Posted: April 24th, 2025

music and labor 0ct 6 2024

Task 1 Write on readings from Weeks 5 and 6. The paper should be a FULL single-spaced page analysis of the readings. I HAVE ATTACHED THE READING BELOW-

Task 2 In this I’m asking you to start analyzing some sources you’ve found. These are not necessarily the only or the best sources you’re reviewing, but they should get you thinking about your research topic and questions more deeply.

Download, fill out, and then upload the attached document.

Task 3 SUBJECT- SEMINAR 114G

task 4 forensic science

task 5 Rhetorical Analysis Peer Review Sheet

PART I:

HEROES

Bass, Amy. In the Game : Race, Identity, and Sports in the Twentieth Century, Palgrave Macmillan, 2005.
ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=307887.
Created from umboston on 2018-08-20 14:29:19.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

ONE

“Richie” Allen, Whitey’s Ways,
and Me: A Political Education

in the 1960s

Matthew Frye Jacobson

I wouldn’t say that I hate Whitey, but deep down in my heart, I just can’t
stand Whitey’s ways, man.

—Dick Allen, Ebony, 1970

“Disrespect” would be a euphemism. Dick Allen was unanimously re-
named “Richie” in 1960 by a white press wholly indifferent to the young
ballplayer’s protestations that everyone from his mother on down had al-
ways called him “Dick.” Later, when Allen finally did insist upon his right-
ful name after several years of patiently accepting what he thought a
vaguely racist diminutive, the press variously ignored his request, spitefully
granted it (“Dick ‘Don’t Call Me Richie’ Allen”), or—worse—depicted the
“name-change” as an emblem of Allen’s unstable character (as in: “in mid-
career he became, adamantly, ‘Dick.’” Sports Illustrated referred to this as
Allen’s “first name sensitivity.”)1 Fans in Philadelphia delighted in throw-
ing objects at Allen—pennies, chicken bones, batteries, bolts, half pints—
and when he took to wearing a batting helmet in the field, the press

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

20 MATTHEW FRYE JACOBSON

intimated that he needed the protection because he was bad with a glove.
Allen twice appeared on the cover of Sports Illustrated: once in 1970 under
the heading “Baseball in Turmoil” (a reference to Curt Flood’s challenge to
baseball’s reserve clause, but Allen was the sport’s better poster boy for
“turmoil”), and once in 1972, smoking what remains the only cigarette in
the history of SI covers.

Nor has Allen’s treatment mellowed over the years. The current entry
for Allen on BaseballLibrary.com (“The Stories behind the Stats”) begins
this way: “Talented, controversial, charming, and abusive, Allen put in 15
major league seasons, hitting prodigious homers and paying prodigious
fines. He was praised as a money player and condemned as a loafer.” The
site does duly note Allen’s Rookie of the Year season in 1964 and his MVP
season in 1972; but its overall flavor tends fairly decisively toward “loafer”
rather than “money player.” (The account of his stellar rookie season opens
on the odd—but for Allen, familiar—note, “He made 41 errors at third
base. . . .”)2 Total Baseball, the baseball encyclopedia, ranks Allen as the
eighty-eighth best player of all time in an entry that begins, “Dick Allen
feuded with writers, fans, managers, and teammates, earned many suspen-
sions and behaved and fielded erratically.”3

In American political life, the phrase “Black Power” will always bring
to mind Stokley Carmichael, H. Rap Brown, Huey Newton, Bobby Seale,
the Black Panther Party, and other black radicals who came to prominence
in the latter half of the 1960s. In the too-clever parlance of ’60s- and ’70s-
era baseball writing, however, its appropriation conjured figures like Hank
Aaron, Willie Mays, Willie McCovey, Frank Robinson, and Richie Allen—
the 1.5 generation of baseball’s integration after Jackie Robinson had bro-
ken the color bar, black sluggers whose speed and playing style and might
were transforming the national pastime. (Absent its black stars, Hank
Aaron points out, the National League’s stand-out player of the 1960s
would have been Ron Santo.)4

But the two meanings of “black power” were not unrelated, as Dick
Allen’s career demonstrates perhaps better than most. The social drama of
the Civil Rights movement constituted the inescapable context within
which black ballplayers of this generation were understood and measured
in the white media—most often, if tacitly, located along an imagined polit-
ical spectrum of “good” and “bad” Negroes (Willie Mays at one end of the
spectrum, Richie Allen, Bob Gibson, and Dock Ellis at the other). “If
[Allen] had been white,” writes Gibson, “he would have been considered
merely a free spirit. As a black man who did as he pleased and guarded his
privacy, he was instead regarded as a trouble-maker.”5 It is only in the con-
text of the wider political and social world of the 1960s, not of the club-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

21“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

house and diamond, that one can comprehend Allen’s becoming “a dart-
board for the press,” in Pirate outfielder Willie Stargell’s phrase.6

Thus the sports page served as a site of oblique but significant social
commentary on the racial questions of the day (indeed it was in relation to
the sports page that whites seem to have first acknowledged and accepted
that there might even be such a thing as a “white press”). It is not just that
the world of Orval Faubus, Martin Luther King, Jr., Strom Thurmond, and
Malcolm X supplied the cues for writing about a figure like Richie Allen, but
also, contrariwise, that commentary on the likes of Allen—or Muhammad
Ali or Cookie Gilchrist or Lew Alcindor—was by its very nature a genre of
political writing whose significations reached beyond the diamond, the ring,
or the gridiron, to the roiling racial world of a nation in unrest.

By the time Allen’s autobiography appeared in 1989, vernacular polit-
ical discourse was better equipped to deal with the experience of someone
“enormously talented and black in a game run by white owners, executives,
and managers,” as one reviewer put it.7 Across the arc of his career in
Philadelphia, however, from 1964 to 1969, the political truths of the sports
world were grasped and analyzed chiefly by athletes and writers on the
black side of the color line, and only very occasionally by a white com-
mentator like Robert Lipsyte or Jack Olsen. Most often, black analyses of
how race mattered—along with black protestations that race did matter—
were simply folded into white power’s already-scripted tale of the “bad
Negro,” as when Cookie Gilchrist mounted a boycott of the AFL’s 1964
All-Star Game in Jim Crow New Orleans, when Tommie Smith and John
Carlos raised their gloved fists on the dais in Mexico City in 1968, or when
Dick Allen or Frank Robinson raised the issue of Major League Baseball’s
racist hiring practices. Bad boys all. By suggesting that race had anything
to do with his image as “the bad boy of baseball,” in other words, a figure
like Allen could only prove himself the “bad boy of baseball.”

This essay is not primarily about Dick Allen, but—quite deliberately—
about Richie Allen, a creation of the white press, a negative icon of the
Civil Rights era, “just about the premier bad boy in sports.”8 It is also
about Richie Allen as a persona who—against the odds, one has to con-
clude—became a positive icon to me, a white kid growing up in the subur-
ban setting of Boulder, Colorado. The sports pages of this era constituted
my political education. I was six years old and just beginning to pay atten-
tion to baseball during Allen’s phenomenal rookie year. If “black power”
signified anything to me at age nine, around the time when the term entered
political parlance, it signified Allen’s towering home run to straightaway
center in the All-Star Game in Anaheim. But by age ten, always hungry for
another story, another AP wire photo, another stat on Allen, I could not
help but notice that most of what I found was some brand of vilification.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

22 MATTHEW FRYE JACOBSON

My fourth-grade teacher Miss Harms could lecture on Reverend King and
the freedom struggle; but what I learned about the injustices and the slan-
ders of racism, I learned mostly by following Richie Allen in the Denver
Post, waiting in vain for someone to write something good. (“Richie played
with fire in his eyes, always,” says Orlando Cepeda. “Never read that in no
newspaper.”9)

Reflecting on the odd oasis of adulation that his own fame provided
him amid a wider, uglier world of racism, harassment, and danger, Bob
Gibson once told baseball writer Roger Angell, “It’s nice to get attention
and favors . . . but I can never forget the fact that if I were an ordinary
black person I’d be in the shithouse, like millions of others.”10 Allen never
did quite get out, even despite his talent and his fame and the awed respect
he earned inside the lines. Here, in what stands as both a historical and a
personal reflection, I seek to discover what that might say about politics
and sport in the 1960s, and also to recover what it did mean to one white
fan, thousands of miles and many worlds away from the Philadelphia shit-
house called Connie Mack Stadium.

1. Philadelphia

“No baseball season in my fifteen-year career had the highs and lows of
’64,” wrote Allen in his autobiography, Crash. “The Temps said it best
baby, I was a ball of confusion.”11 Allen was the National League Rookie
of the Year, hitting .318 with 201 hits, 29 home runs, and 91 RBIs. He also
had 38 doubles and 13 triples, a single-season combination that the likes of
Mays, Aaron, Roberto Clemente, and Pete Rose never matched. Or Jackie
Robinson, for that matter. (Joe DiMaggio bested it back in 1936, with 44
doubles and 15 triples). But Phillies fans found ways to sour on him
nonetheless, many blaming him for the team’s spectacular September
freefall that cost them what had seemed a sure pennant. Fans’ merciless
booing became so common at Connie Mack Stadium in ensuing years that
by the end of his tenure in Philadelphia, Allen had taken to scratching mes-
sages during the game—such as the word “boo”—in the infield dirt with
his spikes.12

Jackie Robinson and the magical date of 1947 seem to have long
passed by the time Allen cracked the majors, but the key to his bitter expe-
rience in the 1960s lies precisely in how little had happened in the inter-
vening years. When one thinks of baseball’s falling racial barriers, the
players who come to mind in addition to Robinson are people like Larry
Doby, Roy Campanella, and Monte Irvin, a generation born in the teens
and twenties, who came of age in the forties and played in the Negro

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

23“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

Leagues before entering the newly integrated majors directly on the heels of
Branch Rickey’s “great experiment” in Brooklyn. The intervening glory
years make it hard enough to recall that Willie Mays and Hank Aaron
played their first pro ball in the Negro Leagues (Mays with the Birmingham
Black Barons, Aaron with the Indianapolis Clowns); but even the players
slightly younger than they—players with no Negro League experience at
all—spent the early part of their careers in a baseball environment no less
white and no less hostile than Jackie Robinson’s Ebbets Field.

Hank Aaron himself refers to them as “second generation black play-
ers,” though 1.5 generation would be more accurate—Willie McCovey,
Billy Williams, Bill White, Orlando Cepeda, Bob Gibson, Curt Flood, Lou
Brock. Though associated with the 1960s and a baseball era far removed
from the Jackie Robinson moment, “most of them came through the minor
leagues in the 1950s, and almost all of them had their own horror sto-
ries.”13 In October 1964, David Halberstam writes of this generation,

If they were not the black players of the pioneer generation, they had come
up right behind them: most had grown up in ghettos, and their way into the
big leagues had been difficult, often through a still-segregated minor-league
system. This obstacle course remained the foundation of big-league baseball,
and it was rife with prejudice. Playing on minor-league teams in tiny South-
ern towns meant the crowds—even the home crowds—were usually hostile.
Worse, most of their fellow players were rural country white boys, who,
more often than not, seemed to accept the local mores.14

“I didn’t know anything about racism or bigotry until I went into pro-
fessional baseball in 1953,” writes Frank Robinson, who grew up in West
Oakland and whose initiation in the taunts of “Nigger, go back to Africa”
came in Sally League towns like Augusta, Macon, and Savannah.15 As
Dock Ellis—ten years younger still than Robinson—put it, “You learn more
than baseball in the minor leagues.” For his own part, Ellis recalls going
into the stands in a game against the Geneva Senators, swinging a leaded
bat at a fan who had called him Stepin Fetchit, or standing defiantly on the
mound, middle finger extended to a hostile crowd, after striking out the last
batter in a game in Wilson, North Carolina.16

Such incidents—Aaron’s racial “horror stories”—punctuate the biogra-
phies of virtually every player of the 1.5 generation. Bill White spent 1953
as the only black player in the Class-B Carolina League, serving, in Hal-
berstam’s words, as “a kind of beacon to local rednecks, who would come
out to the ballpark and, for a tiny amount of money, yell at this one young
black player, who symbolized to them a world beginning to change.” He
sometimes carried a bat with him as he left the clubhouse, according to Bob

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

24 MATTHEW FRYE JACOBSON

Gibson, in order “to get through the hostile crowds that stood between him
and the team bus.”17 Aaron and Wes Covington broke the color barrier up
north in Eau Claire, Wisconsin (Aaron: “We didn’t exactly blend in”; Cov-
ington: “I felt like a sideshow freak”) before Aaron was sent to the Jack-
sonville Braves to break the color line in the Sally League.18 The president
of the Sally League, Dick Butler, later claimed to have “followed Jack-
sonville and sat in the stands to keep a lookout. You were never sure what
was going to happen. Those people had awfully strong feelings about what
was going on.”19 John Roseboro endured taunts of “chocolate drop” in
Sheboygan; Felipe Alou was barred from the Evangeline League because of
Louisiana segregation statutes (and shipped instead to the more hospitable
Cocoa Indians of the Florida State League, “a class D menagerie”).20 In
Fayette, North Carolina, Curt Flood “heard spluttering gasps, ‘There’s a
goddamned nigger son-of-a-bitch playing ball with those white boys! I’m
leaving’”; and in Greensboro, Leon Wagner faced an armed fan by the out-
field fence, issuing a warning, “Nigger, I’m going to fill you with shot if you
catch one ball out there.” “What kind of country is this?” Vic Power
wanted to know, upon confronting racial mores so different from those that
obtained in his native Puerto Rico.21

Even after they had safely reached the majors, far from the redneck
sneers of the Sally League circuit, most of the 1.5 generation had to ne-
gotiate the southern racial climate and the segregated facilities of Florida
sites like Bradenton, Vero Beach, Clearwater, or Tampa during the
months of spring training. Most also had to deal with some element of
segregation in their team’s travel, lodging, rooming, or eating arrange-
ments in cities like St. Louis and Cincinnati during the regular season;
many, like Reggie Smith, had epithets and more dangerous objects hurled
at them at one time or another, even by the “fans” in their home ball-
parks. Some joined major league teams that were themselves deeply di-
vided by race. Gibson and White broke into the majors playing for an
overtly racist manager named Solly Hemus: “either he disliked us deeply
or he genuinely believed that the only way to motivate us was with in-
sults,” remembers Gibson. During one clubhouse meeting, in the presence
of the full team, Hemus referred to an opposing pitcher as a “nigger.” Or-
lando Cepeda, for his part, attributes the perennial also-ran fortunes of
the Giants during the early ’60s to the breakdown of team feeling along
ethnoracial lines. (Among other things, though his lineup featured
Cepeda, all three Alou brothers, Jose Pagan, and Juan Marichal, manager
Alvin Dark tried to ban the Spanish language in the clubhouse. Dark—
who, ironically, had grown up in Lake Charles, Louisiana, the very town
that barred Felipe Alou—also openly questioned the “mental alertness”
of his “Negro and Spanish-speaking players.”)22

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

25“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

Dick Allen drew a cruel hand, even by the standards of such a deck:
after brief stints in Elmira (New York), Magic Valley (Utah), and
Williamsport (Pennsylvania), in 1963 and at the age of only 20, Allen
landed with the Arkansas Travelers, the Phillies’ AAA team whose home
park was in Little Rock (of Central High fame) and whose lineup had, to
that point, been white only. (As Lou Brock, who had been born there, liked
to say, Arkansas was indeed “the land of opportunity”—at the very first
opportunity he had gotten the hell out.23) “When I arrived at the park,”
Allen recalls, “ . . . there were people marching around with signs. One
said, DON’T NEGRO-IZE BASEBALL. Another, NIGGER GO
HOME. . . . Here, in my mind, I thought Jackie Robinson had Negro-ized
baseball sixteen years earlier.” As if to underscore the militant whiteness of
this white world, the season’s festivities began with the ceremonial throw-
ing out of the first pitch by Governor Orval Faubus. Afterward Allen found
a note on the windshield of his car: “DON’T COME BACK AGAIN, NIG-
GER.” “There might be something more terrifying than being black and
holding a note that says ‘Nigger’ in an empty parking lot in Little Rock,
Arkansas, in 1963,” Allen comments, “but if there is, it hasn’t crossed my
path yet.” That AAA season was filled with this sort of menace and dan-
ger; and it was also exceptionally isolating, as off the field Allen was re-
moved from the rest of the team by the maze of segregationist civic codes
and social rituals of pre-Civil Rights Act Little Rock.24

This was perhaps the beginning of bad blood between Allen and both
the Phillies’ white officialdom and Philadelphia’s white press. For one thing,
Allen felt that he was ready for the majors already (his nine spring-training
home runs in 1963 seemed to argue in his favor), and he saw himself as a
sacrificial lamb to the organization’s imperative to desegregate its farm sys-
tem. This might have been workable if, for another thing, the Phillies had
handled Allen’s situation with some of the forethought and sensitivity that
the Dodgers had shown Jackie Robinson. But the organization was quite
calloused in its general disinterest in Allen’s Arkansas experience. As Ebony
wryly noted in 1970, “During [the] 1963 season with Philadelphia’s minor
league team in Little Rock, . . . he complained about racial injustice (Philly
writers say they found no prejudice there).”25 Most telling, perhaps, was
Arkansas manager Frank Lucchesi’s nonchalance toward the social burden
that Allen was made to carry that season: “Richie was upset one night be-
cause one person said, ‘Come on, Chocolate Drop, hit one out. . . . That’s
not in taste but the fan didn’t realize it. They say worse things to white
ballplayers. Richie is sensitive and he is self-centered.”26

And so, one might have thought, the trip north to Philadelphia the fol-
lowing year would be an improvement. But Philadelphia baseball had a
fairly spectacular history of racism of its own: though Connie Mack had

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

26 MATTHEW FRYE JACOBSON

tried to smuggle talented black players into Shibe Park as Italians or Indi-
ans earlier in the century, the Philadelphia stadium—like the Phillies
lineup—remained the most stubbornly anti-integrationist in the National
League. The black press of the 1940s reported that Mack himself was
among the owners “most bitterly” opposed to integration; and according
to historian Bruce Kuklick, when Jackie Robinson joined the Dodgers in
1947, “the cruelest taunts he received at Ebbets Field came from the visit-
ing Phillies. . . .” As for Brooklyn’s visits to Shibe Park, Phillies GM Herb
Pennock pleaded with Branch Rickey not to bring Robinson at all:
“Branch, you can’t bring the nigger here. Philadelphia’s not ready for that
yet.” When Robinson did turn up in Philadelphia, pitchers threw at him,
infielders purposely spiked him, and Phillies players once lined up on the
dugout steps, pointing their bats at him and making gunshot sounds. By the
mid-1950s, the Phillies were the only remaining all-white team in the Na-
tional League; and even after the team finally did integrate, it remained
among the last major league teams to end segregated housing during spring
training.27

Over and above the racialized traditions of Philadelphia baseball, the
city itself was entering a heated and dangerous period in black-white rela-
tions—it was a “racial tinderbox,” as the head of the city’s Urban League
described it.28 In 1964 Allen arrived in a Philadelphia wracked by racial vi-
olence over issues of job discrimination, housing, school segregation, and
police brutality, and in which an aggressive (and aggressively white) former
beat cop named Frank Rizzo was rising rapidly through the ranks toward
the commissioner’s office, which he attained in 1967.29 (Faubus and Rizzo:
two-thirds of some weird, depressing hat trick. Later Allen worked for Al
Campanis.) There had been violent clashes over the integration of Philadel-
phia construction in 1963; and in August 1964, during Allen’s rookie sea-
son, three days of rioting engulfed a 125-block area of Lower North
Philadelphia, one boundary of which was marked by Connie Mack Sta-
dium. Players had to pass through a “police state” to get to the ballpark
during those days. One black resident lamented, “The only thing I regret
about the riot . . . was that we didn’t burn down that goddamn stadium.
They had it surrounded by cops, and we couldn’t get to it. I just wish we
could’ve burned it down and wiped away its history that tells me I’m noth-
ing but a nigger.” Two died and 339 were injured in the rioting.30

Although Philadelphia fans might indeed “boo the losers in an Easter
egg hunt,” as Bob Uecker once quipped, and even white outfielder
Johnny Callison had objects thrown at him, still these fans found a very
special—vitriolic—place in their hearts for the new arrival from the
Arkansas Travelers. Even his Rookie of the Year stats (.318, 29 HR, 91
RBI) were not enough to shield Allen from the tense, racial hatreds of

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

27“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

mid-’60s Philadelphia.31 Fan animosity toward Allen seems a compound
of garden variety racism; scapegoating for the Phillies’ 1964 tailspin;
venting on the larger race questions facing the city; and a misappre-
hending response, as Sports Illustrated noted, to Allen’s expressionless
playing style, which to many whites made him look “arrogant.” (Man-
ager Gene Mauch’s more generous observation of Allen’s demeanor was
that “He doesn’t get way up when things are going good, or way down
when things are going bad. And that’s the best approach to any profes-
sional sport.”) All of which was further fueled by “some of the harshest
press in the city’s sports history.”32

Allen was in fact booed for the first time in the fifth inning of the Phils’
home opener in 1964, and he was booed plenty as the Phillies squandered
their six and a half game lead in the final 12 games of that season. But the
mutual bitterness began in earnest the next season, in July 1965, when a
pregame fight between Allen and Philadelphia favorite Frank Thomas re-
sulted in Thomas’ departure from the Phillies.33 The fight, by most ac-
counts, was itself “racial.” Thomas was already well-known among his
teammates for his derisive comments toward Allen, Johnny Briggs, and
other black players. One thing that particularly enraged Allen was when
Thomas would approach a black player, pretending “to offer his hand in a
soul shake,” and then “grab the player’s thumb and bend it back hard.”34

On the day of the fight, Johnny Callison was razzing Thomas for a failed
bunt attempt the night before, but Thomas chose to answer Allen instead
of Callison. He taunted Allen as “Muhammad Clay,” by some accounts,
and “Richie X” by others—taunts that in either case Allen answered with
a pop to the jaw before Thomas broke a cardinal baseball rule by swinging
his bat at Allen and catching him on the shoulder.35

Teammates pried the two apart, but an ineluctable sequence had al-
ready been set in motion: Thomas was immediately sold off to Houston;
Allen was forbidden from discussing the incident under penalty of a $2,000
fine; but Thomas, meanwhile, freely fed his (partisan, sanitized) version to
the press. Manager Gene Mauch, too, made some rather coy remarks to the
press that not only obscured the nature of the incident and Thomas’ part in
it, but also left an impression that the Phillies had unfairly and quite know-
ingly scapegoated the white veteran in deference to Allen’s talent and
youth. It was here, most significantly, that the press began to tag Allen as a
“troublemaker”—an appellation that would provide a convenient media
peg for the rest of his career. “Baseball should never forget the Allen-
Thomas fiasco,” says Bill White. “ . . . When Dick Allen came to the big
leagues, he was a kid in love with the game. Baseball was all that mattered.
After the Thomas incident, the love was taken right out of him. There’s his-
torical significance in how that was handled.”36

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

28 MATTHEW FRYE JACOBSON

The result was that Allen came out looking unjustly favored and
vaguely militant—a ready-made script for many whites, given the city’s
racial climate—and he was directly blamed for the departure of a popu-
lar (white) player. Banners announcing fans’ unambiguous preferences—
such as “We Want Thomas”—bedecked Connie Mack Stadium; Daily
News writer Larry Merchant embarked on an anti-Allen crusade in print;
one fan “sucker punched” Allen; others at the park jeered him as
“darkie” and “monkey” (when he wasn’t hitting game-winning home
runs), and Allen recalls seeing one father pointing at him and teaching his
little boy how to boo. It was soon after, too, that people started to throw
things at Allen, to vandalize his home, and to harass his family. Across the
balance of the 1960s, Allen was “booed mercilessly,” as Newsweek re-
ported, and he received “hate mail . . . so brutal that he now refuses to
open anything that looks like fan mail”; “people smeared paint on his car,
threw rocks and shot BBs through his windows and booed his children on
the street.”37 As the Daily News once reported in 1967, after Allen’s hero-
ics had dispatched the Cubs, “He should have been grinning and content
in the knowledge that his three-run homer in the twelfth inning won a
game for the Phillies. But it is tough to grin when you come to the ball-
park and there are letters calling you ‘Dirty, Black Nigger.’” It was after
this particular game that Allen started speaking openly about wanting out
of Philadelphia.38

The Thomas incident may have marked a turning point for Allen and
the city, but it was scarcely the only factor in that souring relationship. As
Don Malcolm suggests, the “Angry Negro Problem”—a thematic conven-
tion for writing about a certain kind of athlete, from Dick Allen to Gary
Sheffield—derives not only from the fact that “white Americans still are
manifestly uncomfortable with demonstrative black males,” but also, sig-
nificantly, that they are “probably most uncomfortable with the ones who
are making piles of dough.”39 (As for a bit of context on “angry Negroes”:
five weeks after the Thomas incident, the Phillies landed in Los Angeles just
in time to witness the flames of the Watts riot.40)

Dick Allen, emphatically, was not utterly unappreciated by the baseball
world, and this, paradoxically, may have fueled the animosity against him
in some quarters. Philadelphia had signed him for a cool $70,000 bonus,
the largest ever offered a black ballplayer. Later, Allen became the highest-
paid player on the Phillies (and in 1973, upon signing with the White Sox
for a quarter of a million dollars, he was to become the highest-paid player
in Major League history to that point). “His salary has risen faster than
anyone’s ever did before,” remarked Newsweek in 1968. “ . . . And his
popularity has plummeted just as fast.”41 In the calculus of Philadelphia
race relations—and of the nation’s—these two developments may have

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

29“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

been intimately entwined. It is not just a case of a Negro’s earnings demol-
ishing the white presumption of what would be fitting; it is also a matter of
social demeanor—the white insistence upon “appropriate” black gratitude,
which is to say a bit of the old-fashioned, hat-in-hand bowing and scrap-
ing. But as Sports Illustrated commented, on the contrary, Allen was “the
first black man . . . to assert himself in baseball with something like the un-
accommodating force of Muhammad Ali in boxing, Kareem Abdul-Jabbar
in basketball, and Jim Brown in football.”42

As the economics of the game shifted in the late 1960s, too, there was
the volatile matter of the sheer power attaching to a player’s contract: many
among the white press and white fandom were troubled that the Phillies or-
ganization found even Allen’s white managers (first Mauch, and then Bob
Skinner) more readily expendable than this black star, impetuous though he
was. As Jim Bouton had it in Ball Four, “There is a pecking order in the
major leagues which goes like this: owner, general manager, superstar, man-
ager, established player, coaches, traveling secretary, trainer, clubhouse
man, marginal player.”43 Black superstar over white manager—this was a
problem for many white fans in the 1960s. And while much discussion of
race in baseball has focused on the suspicious paucity of black managers
and team executives, the “problem” of the black super star—the tension
between the racial hierarchy of the culture and the natural pecking order of
the team—has been the cause of much devilment as well.

Within this alchemic mingling of circumstance, ideology, personality,
and history, the media developed an iron framework for reporting on
Allen’s career both on and off the field: Allen was militant, a malcontent, a
troublemaker, a black radical. Allen was not entirely blameless for the vol-
ume of available copy, it should be noted; but the “bad boy of baseball”
label did create a media peg for stories that might have attracted no atten-
tion at all in the case of other players, black or white. (Indeed, the shock
and scandal of a book like Bouton’s Ball Four in 1970—what Bowie Kuhn
called Bouton’s “grave disservice” to the game—was precisely its demon-
stration that the game was made up pretty much exclusively of swearing,
hard-drinking, tobacco-addicted, amphetamine-popping, bed-hopping,
window-peeping bad boys.44) But for Allen and seemingly for Allen alone,
a steady litany of well-publicized “transgressions” mounted throughout the
’60s: the Thomas incident in 1965; a freak, off-field hand injury in 1967,
broadly but baselessly presumed to be the result of either a barroom knife-
fight or perhaps a run-in with a lover’s husband; an actual barroom brawl
in 1968 (which, like the Thomas incident, began with a racial slur); and
also in 1968 a few missed days of spring training, an instance of reporting
late to the ballpark, and his benching by Mauch for being “unfit to play”
(Allen’s trouble, Mauch said, was not with “the high fastball,” but rather

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

30 MATTHEW FRYE JACOBSON

“the fast highball”); and in 1969, income tax problems, a missed plane to
St. Louis, and a missed double-header at Shea.45

Where silence on such matters was the journalistic norm in this cook-
ies-and-milk era of sports coverage (Mickey Mantle was not averse to
showing up at the park “unfit to play,” either, for instance, as we later
learned and as the press corps had surely known at the time), Allen’s every
move seemed to generate acres of copy. “You fellas have created an atmos-
phere where people who have never met me, hate me,” he told reporters.
Later he commented, “Even if they gave me an opportunity to tell all of my
side of the story, I wouldn’t take it because I just don’t trust the white press
in general.”46 If Allen was a perpetual story, race and racism were never an
acknowledged part of that story. But the “race neutral” language of the
white press makes for some interesting reading: Allen “marches to a
mournful tune that only he hears, moving with an insolent grace,” for ex-
ample, according to the Philadelphia Daily News; though one might fairly
ask whether it is even possible for a white man, in America’s media cosmos,
to “move with insolent grace.” Further, Richie Allen is “a superstar with a
built-in distaste for discipline” (New York Times); he is “a player of enor-
mous talents and mercurial moods” who is “known less for his awesome
batting power than for his drinking, horseplaying and habitual tardiness”
(Newsweek); “a man who hits a baseball even harder than he hits the bot-
tle,” a “wondrously gifted misanthrope,” the “chain-smoking, hard-
drinking, horseplaying, perpetually late bad boy of the 1960s” (Sports
Illustrated).47 So infamous did Allen’s movements become, that at the All-
Star Game in 1969 President Nixon sent a personal message through
Allen’s teammate Grant Jackson: “You tell Richie Allen to get back on the
job.” By that same year—his last in Philadelphia, as it turned out—Allen
had begun to “wish they’d shut the gates . . . and let us play ball with no
press and no fans.”48

The contrast with the black press could not have been starker. In 1968,
at the height of his most controversial season and amid a thorough raking
in the white media, for instance, a photo gallery in the Afro-American lov-
ingly depicted Allen as a devoted family man (“$85,000 dad plays mom at
Phils’ ballpark. Richie Allen baby-sits with son between Sunday pitches”).49

After the St. Louis trade in 1970, Ebony directly took up the matter of the
white press’s racism, as “the questions continue[d]” regarding Allen: peo-
ple ought to “question the questioners,” the black journal protested. To
question Allen “presupposes that Richie is guilty of all the bad things writ-
ten about him. . . . Most of the people who hate or love Richie do so on the
basis of what they’ve heard or what they’ve read in the white press.” The
whiteness of the press, in this equation, was as inescapably significant as
the blackness of the ballplayer: “Richie Allen is black and he’s proud and

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

31“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

he has the gumption to be a proud, black man in one of America’s most
conservative sports. He sprouts a lush Afro that’s anchored with long and
wide sideburns”—“his natural and long wide sideburns were targets of
white criticism in Philly for six years.” After pointing out that Allen was
known to read the Bible with some regularity, and that one of his infamous
missed games had to do with his son’s illness, Ebony argued that “Richie’s
stands on baseball’s controversial issues and the fact he’s black” were what
marked him as a “radical.” “Basically, he’s just a ‘regular brother,’ hipped
with all the jive-time routines of coolness, arrogance and a happy-go-lucky
attitude.”50

His were, indeed, the Afro and the pork chop sideburns with which
Sports Illustrated would choose to illustrate its cover story on “Baseball in
Turmoil” in the spring of the Allen-Flood-McCarver trade. Although Allen
did hold out for more money from St. Louis, it is true, the “turmoil” had
mostly to do, not with him—“I’ll play anywhere: third, short, anywhere
but Philadelphia”—but with Curt Flood, who had refused to report at all.
The word “turmoil” itself, in fact, came from an exasperated Gussie Busch,
the Cardinals owner: “I can’t understand Curt Flood . . . or the Allen
case . . . we are going through a hell of a turmoil right now.” Though Busch
was having his problems with the Steve Carlton contract, too, the turmoil
seemed to him largely racial, apparently, and also connected to the broader
social currents of 1960s America: “I can’t understand what’s happening
here or on our campuses or in our great country.”51

Flood’s protest was, in fact, “racial,” even if it was Allen who more
looked the part in SI’s estimation. For one thing, Flood was not eager to
go to Philadelphia, “the nation’s northernmost southern city,” as he put it,
“ . . . to succeed Richie Allen in the affections of that organization, its
press and its catcalling missile-hurling audience.”52 And for another, as
many have remarked over the years, given the bondage and emancipation
motifs of the legalities involved, it was perhaps inevitable that a black
ballplayer would be the first to challenge Major League Baseball’s reserve
clause and seek free agency. Flood himself begins his autobiography, The
Way It Is (1970), with an epigraph from his brother Carl: “Pharaoh, you
better let them chillun go, honey.” Later, noting that “the word slavery has
arisen in connection with my lawsuit” (and conceding sardonically that
“the condition of the major-league baseball player is closer to peonage
than to slavery”), Flood appeals to the language of a 1949 court decision
in the case of the Giants’ Danny Gardella: “Only the totalitarian-minded
will believe that high pay excuses virtual slavery.”53 The reserve
clause/slavery analogy was neither casual nor incidental, in Flood’s view:
“Frederick Douglass was a Maryland slave who taught himself to read. ‘If
there is no struggle,’ he once said, ‘there is no progress. Those who profess

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

32 MATTHEW FRYE JACOBSON

to love freedom, and yet deprecate agitation, are men who want crops
without plowing up the ground. . . . Power concedes nothing without a de-
mand. It never did and never will.’”

To see the Curt Flood case in that light is to see its entire meaning.54

Elsewhere, Bob Gibson quoted Flood as likening a franchise owner’s
powers “to a plantation owner, allowing his players to play for him in
the same way that the plantation owner allowed the sharecropper to
work his land while at the same time keeping him deep in debt and con-
stantly beholden.” The slavery analogy was also clearly among the things
that Gibson had in mind when, during the spring of the Flood-Allen
trade, in dark jest he hung a sign above his locker, “Another happy fam-
ily sold.”55

Sportswriter Sandy Grady was tacitly acknowledging the racialized di-
mension of Allen’s experience—not with the reserve clause, necessarily, but
with the hatreds and disparagements of “The City of Brotherly Love”—
when he wrote of St. Louis GM Bing Devine’s having “emancipated”
Allen. (In typical white press fashion, however, he also suggested that
Devine had “emancipated” Philadelphia from Allen.)56 And Allen, for his
part, drew from the same lexicon: “You don’t know how good it feels to
get out of Philadelphia. They treat you like cattle. It was like a form of
slavery. Once you step out of bounds they’ll do everything possible to de-
stroy your soul.” “Skinner once said he could handle me,” Allen later re-
marked, “ . . . Well you don’t handle human beings, you treat them. You
handle horses.”57 Curt Flood might have said that; so might Frederick
Douglass.

Allen headed into a slightly new era upon his departure from Philadel-
phia; fans never again vented the kind of hatred that Allen had seen in Con-
nie Mack Stadium in the 1960s. Lee Vilensky’s beautiful “Ode to Dick
Allen” vividly captures the death grip that Allen and the white racists of
Philadelphia had on one another during those years. Recalling his first ever
visit to Connie Mack Stadium as an eight-year-old in 1965, Vilensky writes
of the “batteries, bottles, paperweights” that were hurled in Allen’s direc-
tion, and the “nigger, nigger, nigger” and “fuckin’ nigger, nigger” that
swirled around the stands.

I guess it was about the seventh inning when Richie came up for his third at
bat. I don’t recall what he had done in his two previous at bats, but the
chanting started anew. “Nigger.” “Big mouth nigger.” “Fuckin’ nigger.” “Go
back to Africa, Nigger.” Yes, someone actually yelled that. . . . [S]uddenly
there was a crack of the bat as Richie Allen crushed a line drive over our
heads. I turned around just in time to watch the ball bounce off a little eave
above the top of the grandstand, then go completely out of the stadium. A
shot of more than five hundred feet in distance. Not a high, arcing, majestic

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

33“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

home run, but a cold, vicious, angry drubbing of the ball. A loud slap. The
power of it scared me. It made people quiet. Took all their air like a punch
to the gut. As Richie touched home plate, the man next to me said to no one
in particular: “Fuckin’ nigger can hit.”58

2. Boulder

Dick Allen and biographer Tim Whitaker stand on the diamond where
decades before the Homestead Grays and the Pittsburgh Crawfords of the
Negro Leagues had played, directly across from the vacant lot where
Allen’s boyhood home once stood. “Imaginary baseball,” says Allen. “It’s
the purest version of the game.”

Allen tugs at his shirt sleeves and pushes his cowboy hat down on top of his
head, mimicking the same routine he went through whenever he stepped to
the plate against major league pitching. He takes a few practice swings with
his imaginary bat.

Between his feet, Allen has formed a pile of stones with his boots.
He picks up one of the stones, tosses it in the air, and takes a swing with

his imaginary bat.
“As a kid, I used to stand right here,” he tells me, “with a broomstick

in my hands. When I played imaginary ball, I was always the Dodgers. I
would bat stones and work my way through the Dodger lineup—Reese, Fu-
rillo, Snider, Hodges—waiting, just waiting, for his turn to come around.”

Allen pauses dramatically, then cups his hands to his mouth. “Now
battting,” he says, imitating the stadium echo of a public address announcer.
“For the Brook-lyn Dod-gers . . . num-ber four-tee-two . . .”

Dick Allen reaches down and picks up another pebble. “The Jackie
Robinson stone,” he says, tossing the pebble in the air and catching it, “was
always the one that broke a window.”59

When I was growing up there must have been millions of us who were
right with Allen on this: that real players played real games in real stadiums
was just a necessary evil so that the much purer game of imaginary base-
ball could take place, in lots and yards across North America, especially in
the pregnant hours after dinner, as dusk edged into darkness. This scene de-
scribes much of my own childhood, though for me the Richie Allen stone
was the window-breaker. (Well, our developing suburban neighborhood
was still rural enough, the distances still great enough, that no windows
were ever really in danger. Besides, I couldn’t hit that well. But one time
when I was about nine, pretending to be Juan Marichal, pitching off the
side of our brick garage and mowing down the hitters 1–2–3 through the

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

34 MATTHEW FRYE JACOBSON

innings—a real gem—in the top of the eighth I couldn’t resist giving up a
home run to Richie Allen. Num-ber fiff-teeen. In my effort to recreate one
of those awesome shots that cause opposing fielders immediately to slack
their bodies and look skyward in resignation, I threw the ball too high
against the wall, breaking the narrow pane of glass that ran the length of
the garage just beneath the awning. Later, when my dad asked me if I knew
anything about the broken window, I came this close to telling him Richie
Allen did it.)

Why Allen would have idolized Jackie Robinson is pretty obvious, but
how did I come to idolize Allen?

I had the 1965 Topps trading card of Allen—the Phillies flag in one cor-
ner, the little Rookie of the Year statuette in the other—but my real intro-
duction to him was a hero-worshipping book for kids, Great Rookies of the
Major Leagues by Jim Brosnan. The chapter on Allen was enough to make
a huge impression on an eight-year-old, but it was not exactly calculated to
do so: for example, it included Philadelphia owner Bob Carpenter’s judg-
ment, “Allen was the worst-looking infielder I ever saw. I thought he’d be
killed by a ground ball.” This piece of baseball hagiography also featured
a four-panel sequence of photographs depicting Allen letting a grounder
pass between his legs. (The caption reads, “Allen’s uncertain fielding some-
times offsets his great hitting. Here he reaches for a sharp grounder,
searches for the ball and then turns to watch it roll into the outfield. A
Braves runner . . . passes Allen to score on the play.”)60

When I was given the book as a gift (in 1966, I believe—the year of its
publication), I adopted Allen as my hero at once. It may have been because
I was enthralled by his appearance: the chapter itself goes into great detail
on his powerful physique, and there is nothing in the photos of Roy Siev-
ers, Herb Score, Frank Robinson, Tom Tresh, or Pete Rose that begins to
compare with the pure poetry of form in some of the Allen photos—I see it
this way even still. Or, it may have been because I identified with his much-
discussed weakness as a fielder, and took special heart in the story of a
player who was able to overcome his own limitations. If I were going to be-
come a major leaguer (and who could doubt it?), my own path to glory
would surely be strewn with similar obstacles, not to mention the qualms
and denunciations of people like Bob Carpenter. Or it may have been that,
as the fat kid with thick glasses whom everyone made fun of, I gravitated
naturally toward the one figure in the book who was clearly being picked
on. (“He . . . turns to watch it roll into the outfield.” It might have been a
few more years before I could articulate this, but even at age eight I felt
some version of hey, what the fuck, man?) Within two years—1968—when
I was three seasons into my Richie Allen worship and Allen himself was get-
ting more and more press for his off-field behavior, I understood exactly

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

35“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

what it was that I was seeing. This was my education in U.S. race relations.
In a 1970s routine about visiting Africa, Richard Pryor talks about

meeting people who are “so black” that it makes you want to say—and
here he drops his voice to an awed whisper—“BLACK.” My neighborhood
growing up was a lot like that, except in white. It was not the militant
whiteness of South Boston (or Connie Mack Stadium); it was not even the
least bit self-conscious. On the contrary, the neighborhood was so white as
to suggest and naturalize the idea that people of color did not exist at all.
Which is just to say, whatever I learned about racialized relations before
going away to college in 1977, I certainly did not learn by firsthand en-
counters. (Nearby Denver, ironically, was the AAA locale where the Min-
nesota Twins banished black players as punishment for dating white
women.61)

There is a longer-term history that is relevant here, because I did grow
up in a liberal household in which civil rights sympathies were never in
question. Since my father is a New York Jew, naturally we used to listen
to Mahalia Jackson every year when we decorated the Christmas tree. He
had grown up in the Bronx in the 1930s, and at age thirteen, the year he
was not bar mitzvahed, he somehow discovered Harlem and jazz. Though
his was probably not the kind of childhood that encouraged much fellow
feeling with “the shvartzes” (to judge from my grandmother’s social out-
look), from those early jam sessions onward, his glimpses of Harlem and
his captivation by the black aesthetic of the jazz scene translated into a
very particular social sensibility—a whole way of perceiving and under-
standing the human virtues and various political categories like “decency.”
This he tried to pass on to us, along with an appreciation for Louis Arm-
strong. My mother, on the other hand, is a white Ohio Methodist, and her
Tipp City upbringing could not have been much less white—or
“WHITE”—than my own. But as theirs was what was called a “mixed
marriage,” both of my parents had some experience with prejudice—their
parents’, for example.

And so, with the Civil Rights movement rumbling in the distance
throughout my childhood, and my parents’ attention to questions of “dif-
ference” and justice remaining fairly salient, racial matters were not as far
removed from my immediate experience as the demographics of my town
would imply. I remember my father trying to explain the logic of King’s
“passive resistance” to me at a time when, as a political philosopher, I was
probably too young for anything beyond “impulsive vengeance.” My sis-
ters and I got the liberal lecture on the stupidities of prejudice on the ride
to Denver to see Guess Who’s Coming to Dinner. A bit later, it became a
point of bedrock principle in our household that of course one would sup-
port the Broncos’ Marlin Briscoe in his bid to become the NFL’s first black

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

36 MATTHEW FRYE JACOBSON

quarterback. (“He’s not all that good,” my best friend’s father said, “he’s
just all that black.” The opinion was offered up too gruffly not to be sus-
pect, even to a ten-year-old.)

But what strikes me in retrospect is how indirect my political education
was, for the most part. Straight talk like the Guess Who’s Coming to Din-
ner lecture was the exception, not the norm, as was my fourth-grade
teacher Miss Harms’ very interesting prediction of racial retribution in the
wake of the King assassination. When I think closely, I recognize that at the
time I did not actually see much—or any—of the Civil Rights imagery that
now occupies my “memory” of the era—Bull Connor’s German Shepherds
and fire hoses, the flames at Ole Miss, even the “I Have a Dream” speech.
The balcony of that Memphis hotel I think I did see for myself on TV in
1968; but most of the rest of it is later documentary footage, not actual
memory.

My teaching has been animated by Stuart Hall’s dictum that social sub-
jects “are unable to speak, to act in one way or another, until they have
been positioned by the work that culture does.” It is culture above all that
outfits us to behave politically in certain ways and not in others—culture is
politics by other means.62 But rarely have I asked the question: If I was just
coming to consciousness during the Civil Rights years, what was I learning
and how was I taking it in? America’s liberal culture was undoubtedly
teaching a lot, though it may not always have been teaching liberality. The
most potent message of Guess Who’s Coming to Dinner, for instance, does
not involve our common humanity across the color line, but rather a nat-
ural submission to the authority of the Great White Father (in this case
Spencer Tracy): ultimately nobody can make a move without his approval.
Shows like Love American Style and Barefoot in the Park taught that black
is indeed beautiful—as long as it’s almost white. The affable Johnny Car-
son taught that candor is hip and that racist stereotypes can be funny—as
when he joked that there could never be a black quarterback because there
were not seven white guys who would turn their backs on him at the line
of scrimmage, “especially during a night game.”

On the other hand, anti-authoritarianism was occupying an increas-
ingly significant place in the dominant culture—I think of Cat Ballou, Bon-
nie and Clyde, Butch Cassidy and the Sundance Kid, Easy Rider, and a host
of other films from my childhood in which bad guys were the good guys
and good guys were the bad guys. Perhaps this strain in the culture outfit-
ted me with a useful skepticism toward the media’s own claims regarding
the badness of the black radical; perhaps it was this strain that equipped me
to sympathize with a bad boy like Richie Allen, doing battle with “the
man” in the white front office and the white press. How far is it from the
unorthodox authoritarianism of The Mod Squad to the unorthodox anti-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

37“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

authoritarianism—consciously “raced” or not—of Richie Allen, Cookie
Gilchrist, the Smothers Brothers, Jim Bouton?

During these years—confusing enough even for many adults, I am sure—
baseball addressed my childhood confusions in a pre-verbal but nonetheless
poetic and incandescent language. (By “baseball,” I mean the whole cos-
mos—the games themselves, the lineups, the sports page, the fan reactions,
the hypnotizing photographs, the piles of adoring books, the Topps cards, the
on- and off-field lore in Sport, Sporting News, Sports Illustrated.) “I can’t say
it was because of the bombs and the Bull Connors that black players tore up
the National League in 1963,” writes Hank Aaron, “but I can’t say it wasn’t
either.”63 On a particularly fierce streak in the summer of 1968, Bob Gibson,
too, writes: “I really can’t say, in retrospect, whether Robert Kennedy’s as-
sassination is what got me going or not. Without a doubt, it was an angry
point in American history for black people—Dr. King’s killing had jolted me;
Kennedy’s infuriated me—and without a doubt, I pitched better angry. I sus-
pect that the control of my slider had more to do with it than anything, but
I can’t completely dismiss the fact that nobody gave me any shit whatsoever
for about two months after Bobby Kennedy died.”64

Aaron and Gibson might rightly have claimed the whole decade for
black dominance, not just the isolated moments of 1963 and 1968. (Take
the offensive statistic for total bases, the most dramatic instance: from 1960
to 1969 white players made it into the National League’s top three exactly
once—Pete Rose was third in 1968. Aaron, Banks, Mays, Cepeda, Robin-
son, Pinson, Allen, Williams, Alou, Clemente, Brock, McCovey, and Perez
account for the other twenty-nine top-three finishes.65) But in any case,
from the suburban picture window of Boulder, Colorado, the ball field and
The Movement read as being intimately connected. “Baseball was socially
relevant,” wrote Curt Flood, “and so was my rebellion against it.”66 This
is a lesson I imbibed fundamentally but wordlessly between 1966 and
1969. The hateful, swirling “nigger, nigger, nigger” that Lee Vilensky heard
in Connie Mack Stadium, and Richie Allen’s cold, angry drubbing of the
ball in response, was a social drama that was integral, if only implicitly so,
to the game-within-the-game of 1960s baseball as I watched it on Game of
the Week every Saturday.

For one thing, while Gibson, Aaron, Allen, and others may have been
playing “angry,” they looked to me, above all else, to be simply serious; and
the regular access that baseball afforded to African American seriousness
was no small thing. The seriousness of King and the historic moment came
across in the chatter and hum of the adult world around me and in head-
lines to stories that I knew vaguely about but did not exactly read. People
like Sidney Poitier and Diahann Carroll also made an impression. But base-
ball occupied my mind 162 days of the year; and unique among the major

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

38 MATTHEW FRYE JACOBSON

sports, baseball games unfolded at a contemplative pace that was well-suited
to conveying the force of an athlete’s character—neither concealing it behind
the armor of the NFL nor blurring it in the flying speed of the NBA.

“Quiet dignity” is almost certainly a racist construction—or at least
a racialized one—as the phrase never appears in connection with white
people, I notice; and it probably dates from a period when “quiet,” from
Negroes, was especially prized in U.S. culture. But nonetheless, something
like “quiet dignity” is a part of what the 1.5 generation of black stars
communicated to me, at once a contrast and an antidote to the vapid
dronings-on of play-by-play announcers like Curt Gowdy and Joe Gara-
giola; and the “dignity” in the equation tended to keep their “quiet” from
coming across as anything like accommodation. The intensity of concen-
tration—the intensity of mind—evident in the expressions and small ritu-
als of Gibson on the mound, Flood or Robinson at the plate, silently but
decisively dismantled any facile cracker assertions about the brutish ca-
pacities of “the Negro.” That Solly Hemus or the white fans in various
Sally League locations had either failed to acknowledge this, or, perhaps,
had not allowed themselves to see it in the first place, just goes to show
how desperate they were.

But if baseball held the power to dislodge the slanders of racism, so did
it have a tendency to generate some slanders of its own—the denigrating
trope of the black athlete’s “natural gift” is only one among many. “Hang-
ing around baseball, as I have been doing,” wrote Donald Hall in the
1970s, “I don’t see racism in management, in coaching, or in the front of-
fice. Reading the newspapers of Detroit and Chicago and Boston and New
York, I see it every day.” The list of the “Most Unpopular Sports Figures,
in the last decade or two,” he points out, “is largely black”—a younger
Muhammad Ali, Duane Thomas, Dick Allen, Alex Johnson.67

This is where Allen was so significant to me, not just as a personal
idol but as a social emblem: the dissonance between what I felt about
Allen and what the press reported about him became so taut as to snap
my youthful ingenuousness, because to me Allen was clearly a figure of
dignity, too, no less than Gibson or Aaron or Brock or Clemente. I was
too young by about one season to catch and appreciate the Frank Thomas
incident and Allen’s initial falling-out with the press; but it was a stunning
and deflating lesson to me when, in 1967, the media so openly questioned
Allen’s “claim” to have injured his hand while pushing his car, and when
in 1968 and 1969, they so openly denounced him—not just as an outlier
(on the order, say, of Jay Johnstone)—but as someone uncontrolled and
uncontrollable, a kind of pre-criminal, when he missed a plane to St.
Louis or showed up late to Shea. In his paean to Allen, “Letters in the
Dirt,” folksinger Chuck Brodsky—another white kid of almost exactly

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

39“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

my vintage—reflects upon the racial dimension, as he saw it, in Allen’s
treatment by the fans and by the press: “He stood a bit outside the lines
which / made him fair game for those times / Richie Allen never kissed /
a white man’s ass.”68 This is precisely the conclusion I came to myself,
sometime around the age of ten.

Hindsight, of course, clarifies some things but hopelessly clouds others.
Knowing what I now do about the 1960s, about racism, about the Move-
ment, and about Allen himself, can I recover with any certainty the Richie
Allen who occupied my imagination in 1970, when the Cardinals’ road
schedule and my family’s summer vacation intersected for a moment at a day
game in San Diego? Can I see my young self any more clearly than I see
Allen? Allen would not answer, or even look up, when I called out to him
from behind the Cardinal dugout after infield practice, but I had not expected
it to go any differently. I bore him no grudge for ignoring me, nor did it di-
minish in the least the magic of seeing him in person. Did I see the situation
as “racial?” Did I see myself white standing there—another white fan, per-
haps, from Allen’s point of view, who might meet his glance with an insult or
an AA battery—another white boy who had been taught by some jeering
peckerwood how to boo? I believe I did, because for one thing, this was one
of the very first times I had ever addressed an African American directly; it is
doubtful that I was unaware of my whiteness and his blackness, notwith-
standing the era’s liberal bromides on the virtue of being colorblind. And for
another thing, even if I did not know his precise thinking on “Whitey’s
ways,” I had figured out some things by watching Allen and his career from
afar. I understood at least dimly the burden in our exchange; and, rightly or
not, in an inarticulate way I felt his rebuff to concern not me, exactly, but the
larger web of relationships ensnaring us both. I had entered history, in other
words, and this was perhaps the first time in my eleven years that I was aware
of it. At least it seems so to me now. (See figures 1.1 and 1.2.)

After the ’60s crested and began to recede, the culture was hungry for
emblems of reconciliation; the Richie Allen narrative was conveniently
pressed into service. Following his bitter years in Philadelphia, and two
years of marked underappreciation in St. Louis and Los Angeles, Allen
landed in a brief dream sequence with the Chicago White Sox. Not only did
he put up the kind of numbers in 1972 that the best of his early years had
promised (.308, 37 HR, 113 RBI), but in Comiskey Park he found a wel-
coming and comfortable home. The difference, according to Allen, was
White Sox manager Chuck Tanner: “He’s from home and he’s like a
brother.” (The two knew each other from the old days in Pennsylvania—
Tanner’s hometown of New Castle is about seven miles from Allen’s
Wampum—and they often called each other “Homey,” which perhaps hints
at Allen’s intended meaning in the phrase “like a brother.”) Tanner thought

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

1.1 Allen heads for the dugout after
infield practice, ignoring my calls
from the stands. Old habits die hard:
note that Allen wears a batting helmet
in the field, even far removed from the
projectiles of Connie Mack Stadium.
Photo: Jerry Jacobson

1.2 “Dick Allen and me in San Diego,
summer 1970. Allen is the distant figure
directly above my left hand. The glasses
make me look like Pirate pitcher Bob
Veale, don’t you think?” Photo: Jerry
Jacobson

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

41“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

Allen “not only the best player in the American League, but the best in the
majors . . . When he’s through with the White Sox, he’s going to walk right
into the Hall of Fame.”69 Tanner thought that Chicagoans ought to build a
monument to Dick Allen.

The manager’s appreciation for Allen transcended baseball by a long
way. “He has a magnetism,” said Tanner, “—like Clark Gable, say, or
Marilyn Monroe.”70 This is an astonishing thing to say: daring to compare
the appeal of a black man to the enchantment of these white icons—and
one of them a beauty queen at that—strikes me as more radical in its way
than anything Allen ever thought up in defiance of Whitey. This is a world,
after all, where black and white ballplayers are rarely compared: even in the
cosmos of sports talk today, Griffey might remind people of Mays, for in-
stance, but certainly not of Mantle; and McGwire is said to have hit “with
Frank Howard-type power.” Orval Faubus could do no better in segregat-
ing our common conceptions of who is “like” whom; and yet Tanner spot-
ted Dick Allen’s similarities to Marilyn Monroe. We probably ought to
build a monument to him.

From Allen’s White Sox years onward, the baseball establishment fell
in love with the story of its own acceptance of Allen, even if it did not
quite learn to love the ballplayer himself as Chuck Tanner did. (He never
did come near the Hall of Fame, for instance.) But Allen “is a man who
marches to his own wry drummer,” reported Sports Illustrated in 1972.
“On the day his teammates were going out on strike, Allen signed his
1972 contract.”71 “His own wry drummer” is a far cry from the portrait
of the trouble-making militant that had predominated in the coverage of
Allen as a Philly. After Chicago, the press began to find something lov-
ably quirky in Allen’s history of unorthodoxies; but more important, the
press seemed to find something laudable in its own warming up to Allen:
it was as if, in embracing Allen, the white sports establishment could at
once prove and celebrate just how far it had come. “He wrote dismissive
notes to his general manager in the base-path dirt with his foot!” com-
mented Sports Illustrated in tones of mock scandal in 1973. “What kind
of man would do a thing like that? And why didn’t anybody think of it
before?”72 Now Allen was “a team player who has bounced around . . .
a mentor to the young, a seasoned veteran whom managements have
seen as a discipline problem. The more you learn about Allen from out-
side sources,” remarked Sports Illustrated, “the more he swims before
you.” Even the press’s conventional disregard for Allen’s point of view
began to shift: as SI now described it, when Allen entered pro ball, “First
thing, his name got changed . . . he did not care to be issued a new name
by an organization.”73 Dick “Don’t Call Me Richie” Allen suddenly
seemed fairly reasonable.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

42 MATTHEW FRYE JACOBSON

America’s favorite Dick Allen story is the one about how he got a
standing ovation when he returned to the Phillies in 1975. Although he
found himself “wondering where all the brothers had gone” as he looked
around the Phillies’ new, suburban ballpark, evidently Allen is fond of this
one, too. “Things had changed,” he wrote, “ . . . blacks were beginning to
run the city. In the old days, I represented a threat to white people in
Philadelphia. I wore my hair in an Afro. I said what was on my mind. I did-
n’t take shit. But now, like the rest of the country, Philadelphia had come
around to accepting that things had changed and were going to keep chang-
ing, like it or not.”74 The movement, had, after all, accomplished some
things; the logic and the accepted idioms of American race consciousness
had shifted significantly; the terms of sports celebrity, too, had changed, un-
orthodoxy taking its place among the new orthodoxies—Jim Bouton, Joe
Namath, Rosey Grier, Steve Carlton, Bill Lee. Perhaps Dick Allen had
merely been a few years ahead of the curve, and there was no depth to the
tragedy of his Philadelphia story after all. Many found it comforting to
think so.

And yet the reconciliation narrative—the Allen/Philadelphia story, and
the national healing for which it is an implied allegory—cannot plow under
all the chicken bones, the bolts, and the batteries that rained onto the field
in those earlier years in Philadelphia, nor can it wipe from memory Allen’s
whimsical sorrow songs, the letters in the dirt. Perhaps this is why the
player who had integrated professional baseball in Orval Faubus’ Arkansas
and who had later distinguished himself as one of the most powerful hitters
in the Major Leagues, expressed elation in 1987—as if finally receiving af-
firmation—when aging Negro star Cool Papa Bell pronounced that he in-
deed would have had what it takes to make it in the Negro Leagues.
Inverting the conventional storyline of baseball aspiration and fulfillment,
a buoyant Allen exclaimed, “He said I could have been one of them. . . . He
said I had power and I could run, the two most important requirements in
Negro League baseball.” Even he recognized the irony in his being “a big
leaguer who felt like he lost out because he never got a chance to play in
the Negro Leagues.”75 This is not to paint Allen as a victim of desegrega-
tion. But his implied daydream about being “one of them,” a Negro League
star, does say a bit about the operations of race in the game, even two
decades after Jackie Robinson had broken down the color barrier. “People
said there was one set of rules for me and another for the rest of the team,”
Allen once said, reflecting on his image as the Phillies’ troublemaker.
“When I was coming up, black players couldn’t stay in the same hotel or
eat in the same places as whites. Two sets of rules? Baseball set the tone.”76

This is the political lesson that Allen’s career had been teaching all along:
desegregation did not come off as advertised.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

43“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

Notes

1. Dick Allen and Tim Whitaker, Crash: The Life and Times of Dick Allen (New
York: Ticknor and Fields, 1989), xvii; “Dick Allen,” in BaseballLibrary.Com/
baseballlibrary/ballplayers/A/Allen_Dick.stm; Sports Illustrated, Sept. 10,
1973, 105; William Kashatus, September Swoon: Richie Allen, the ’64 Phillies,
and Racial Integration (University Park: Pennsylvania State University Press,
2004), 191.

2. “Dick Allen,” in BasballLibrary.Com; Sports Illustrated, March 23, 1970; June
12, 1972.

3. John Thorn, Pete Palmer, Michael Gershman, Total Baseball: The Official En-
cyclopedia of Major League Baseball [Seventh Edition] (Kingston, NY: Total
Sports, 2001), 158.

4. Hank Aaron with Lonnie Wheeler, I Had a Hammer: The Hank Aaron Story
(New York: HarperCollins, 1991), 334–335.

5. Bob Gibson with Lonnie Wheeler, Stranger to the Game: The Autobiography
of Bob Gibson (New York: Viking, 1994), 224.

6. Willie Stargell and Tom Bird, Willie Stargell, an Autobiography (New York:
Harper & Row, 1984), 168.

7. New York Times Book Review, April 23, 1989, Sec. 7, 36–37.
8. Ebony, Oct. 1972, 192.
9. Allen and Whitaker, Crash, 186.

10. Roger Angell, “Distance” [1980], in Game Time (New York: Harcourt, 2003),
208.

11. Allen and Whitaker, Crash, 53.
12. David Wolf, “Let’s Everybody Boo Rich Allen,” Life, Aug. 22, 1969, 50.

Folksinger Chuck Brodsky’s “Letters in the Dirt” is a paean to Allen and his
infield writing. Baseball Ballads, chuckbrodsky.com, 2002.

13. Aaron and Wheeler, I Had a Hammer, 209. See also Jules Tygiel, “Black Ball:
The Integrated Game,” in Extra Bases: Reflections on Jackie Robinson, Race,
and Baseball History (Lincoln: University of Nebraska/Bison, 2002), 104–117.
Tygiel’s Baseball’s Great Experiment: Jackie Robinson and His Legacy [1983]
(New York: Oxford University Press, 1997) remains the standard in the field
on the early period of integrated ball.

14. David Halberstam, October ’64 (New York: Fawcett, 1994), 113.
15. Frank Robinson and Barry Stanback, Extra Innings (New York: McGraw-Hill,

1988), 23, 26.
16. Donald Hall with Dock Ellis, Dock Ellis in the Country of Baseball (New

York: Simon and Schuster, 1976), 123, 128.
17. Halberstam, October ’64, 203; Gibson and Wheeler, Stranger to the Game, 58.
18. Aaron and Wheeler, I Had a Hammer, 55, 56.
19. Aaron and Wheeler, I Had a Hammer, 79.
20. John Roseboro with Bill Libby, Glory Days with the Dodgers and Other

Days with Others (New York: Atheneum, 1978), 54–55; Felipe Alou with
Herm Weiskopf, My Life and Baseball (Waco, TX: Word, 1967), 29. (Even

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

44 MATTHEW FRYE JACOBSON

so, writing in 1967 the highly conservative Alou averred that the urban up-
risings were inspired by communist agitators, 103.)

21. Curt Flood with Richard Carter, The Way It Is (New York: Trident, 1971), 38;
Samuel Regalado, Viva Baseball! Latin Major Leaguers and Their Special
Hunger (Urbana: University of Illinois Press, 1998), 66, 67.

22. Gibson and Wheeler, Stranger to the Game, 52–53; Howard Bryant, Shut Out:
A Story of Race and Baseball in Boston (New York: Routledge, 2002), 92; Hall
and Ellis, Dock Ellis in the Country of Baseball, 134; Orlando Cepeda with
Herb Fagen, Baby Bull: From Hardball to Hard Time and Back (Dallas: Tay-
lor Publishing, 1998), 74–75; Kashatus, September Swoon, 113; Regalado,
Viva Baseball!, 84–87.

23. Halberstam, October ’64, 151.
24. Allen and Whitaker, Crash, 11–14; Kashatus, September Swoon, 45.
25. Ebony, July, 1970, 90.
26. Quoted in Sports Illustrated, Sept. 10, 1973, 111.
27. Bruce Kuklick, To Everything a Season: Shibe Park and Urban Philadelphia,

1909–1976 (Princeton, NJ: Princeton University Press, 1991), 145–148;
Kashatus, September Swoon, 9–37; Bryant, Shut Out, 5; David Faulkner,
Great Time Coming: The Life of Jackie Robinson from Baseball to Birming-
ham (New York: Simon and Schuster, 1995), 163–164; Tom McGrath, “Color
Me Badd,” The Fan, September, 1996, 39.

28. Gerald Early, This Is Where I Came In: Black America in the 1960s (Lincoln:
University of Nebraska/Bison, 2003), 67.

29. Kuklick, To Everything a Season, 155–156, 158; Early, This Is Where I Came
In, 70–71.

30. Kuklick, To Everything a Season, 155–156; Early, This Is Where I Came In,
75–89; Kashatus, September Swoon, 76–80, 111–113.

31. Sports Illustrated, June 1, 1970, 40; Kashatus, September Swoon, 54.
32. Sports Illustrated, Sept. 10, 1973, 111; Kashatus, September Swoon, 82. See

also William Kashatus, “Dick Allen, the Phillies, and Racism,” Nine, Fall
2000, 151. On Allen’s general mistreatment by the press, see Craig Wright,
“Dick Allen: Another View” (originally published in SABR magazine), posted
at www.expressfan.com/dickallenhof/docs/defense .

33. Kashatus, September Swoon, 80.
34. Allen and Whitaker, Crash, 4.
35. See “The Thomas Incident, July 1965” in Kashatus, “Dick Allen, the Phillies,

and Racism,” and Kashatus, September Swoon, 149–157; Sports Illustrated,
Sept 10, 1973, 111; Allen and Whitaker, Crash, 1–10.

36. Allen and Whitaker, Crash, 58–59, 10; Leonard Schechter, “Richie Allen and
the Use of Power,” Sport, July, 1967, 66.

37. Newsweek, July 8, 1968, 52; Sports Illustrated, Sept. 10, 1973, 111; Kashatus,
September Swoon, 155–156.

38. Kashatus, September Swoon, 172.
39. Don Malcolm, “The Angry Negro Problem,” Baseball Primer: Baseball for the

Thinking Fan, www.baseballprimer.com/articles/malcolm_2001–03–05_0.shtml.
40. Kashatus, September Swoon, 160.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

45“RICHIE” ALLEN, WHITEY’S WAYS, AND ME

41. Newsweek, July 8, 1968, 52.
42. Sports Illustrated, Sept. 10, 1973, 107.
43. Jim Bouton, Ball Four [1970] (New York: Wiley, 1990), 393; Kashatus, Sep-

tember Swoon, 189.
44. Bouton, Ball Four, ix.
45. This is the Richie Allen canon. See Allen and Whitaker, Crash, and Kashatus,

September Swoon (Mauch quoted 166). New York Times, Aug. 23, 1968, 79;
July 3, 1969, 35.

46. “Richie Allen is Not All Bad Boy,” New York Times, May 18, 1969; Ebony,
July 1970, 92.

47. Kashatus, September Swoon, 171; New York Times, “Sports of the Times,”
June 25, 1968; Newsweek, May 19, 1975, 58; Newsweek, Aug. 21, 1972, 83;
Sports Illustrated, March 23, 1970, 18; April 29, 1974, 19; July 19, 1999, 19.

48. Bill Conlin, “Richie Is Beautiful. He Don’t Give a Damn for Nobody,” Jock,
January 1970, 88; Sports Illustrated, May 19, 1975, 59.

49. Afro-American, July 13, 1968, 13.
50. Ebony, July, 1970, 89, 90, 92, 93.
51. Sports Illustrated, March 23, 1970, 21.
52. Flood and Carter, The Way It Is, 188.
53. Flood and Carter, The Way It Is, 139.
54. Flood and Carter, The Way It Is, 206; Halberstam, October ’64, 364.
55. Gibson and Wheeler, Stranger to the Game, 219; Sports Illustrated, March 23,

1970, 22.
56. Kuklick, To Everything a Season, 163.
57. See “Oppositional Identity” in Kashatus, “Dick Allen, the Phillies, and

Racism”; Newsweek, Aug. 21, 1972, 84.
58. Lee Vilensky, “Ode to Dick Allen,” Elysian Fields Quarterly: The Baseball Re-

view, Vol. 20, number 3, www.efqreview.com/NewFiles/v20n3/dustofthefields-
two.html.

59. Allen and Whitaker, Crash, 40.
60. Jim Brosnan, Great Rookies of the Major Leagues (New York: Random

House, 1966), 165–167.
61. Roseboro and Libby, Glory Days with the Dodgers, 232.
62. Stuart Hall, “Subjects in History: Making Diasporic Identities,” in Wahneema

Lubiano, ed., The House that Race Built (New York: Vintage, 1998), 291.
63. Aaron and Wheeler, I Had a Hammer, 231.
64. Gibson and Wheeler, Stranger to the Game, 188.
65. Thorn, Palmer, and Gershman, Total Baseball, 2204–2222.
66. Flood and Carter, The Way It Is, 16.
67. Hall and Ellis, Dock Ellis in the Country of Baseball, 177.
68. Brodsky “Letters in the Dirt,” The Baseball Ballads (Weaverville, NC: chuck-

brodsky.com, 2002), track 5.
69. Sports Illustrated, June 5, 1972, 64.
70. Sports Illustrated, April 29, 1974, 20.
71. Sports Illustrated, June 5, 1972, 64.
72. Sports Illustrated, Sept. 10, 1973, 107.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

46 MATTHEW FRYE JACOBSON

73. Sports Illustrated, Sept. 10, 1973, 108, 110.
74. Allen and Whitaker, Crash, 159–160.
75. Allen and Whitaker, Crash, 85.
76. Sports Illustrated, July 19, 1993, 84.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

TWO

In Sports the Best Man Wins

How Joe Louis Whupped Jim Crow

Theresa E. Runstedtler

A single column cannot begin to describe the feeling of the man of color
who watches a brown-skinned boy like Joe Louis, from Alabama, the
most backward State in the Union, fight his way up from the coal mine
and the cotton field through strength of his body and mind.

—Ted Benson, Sunday Worker,
reprinted in Pittsburgh Courier, February 29, 1936

American Hero or Race Man?

On June 22, 1938, when Joe Louis, the Brown Bomber, won a decisive,
first-round knockout in his revenge match against Nazi-promoted Max
Schmeling, white America embraced the black heavyweight champion as
a national hero. Amid increasing reports of Hitler’s imperialistic aggres-
sion and persecution of the Jews, the mainstream white press highlighted
the bout’s worldwide implications, claiming Louis’s triumph as an Amer-
ican victory in the larger fight against fascism. As Heywood Broun of the
New York World-Telegram mused, “One hundred years from now some
historian may theorize, in a footnote at least, that the decline of Nazi

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

48 THERESA E. RUNSTEDTLER

prestige began with the left hook of a former unskilled autoworker.”1 In-
spiring more than just a mere footnote, Louis’s 1938 win expanded into
a celebrated epic of American patriotism and democracy. Brimming with
postwar confidence in 1947, Louis’s close friend, Frank Sinatra, de-
clared: “If I were the government official responsible for the job of mak-
ing the rest of the world understand our national character and the ideals
that motivate us, I would certainly make use of the case history of Joe
Louis.”2

However well-known the narrative of Louis as the quintessential U.S.
citizen became, another story, one that white America and history have
overlooked, meant more to African Americans in the 1930s: Joe Louis as
Race Man. That Louis earned the customary title of “Race Man” was a
mark of high distinction, since this phrase had long been reserved for men
who best exemplified racial progress and leadership in areas like business,
academics, and politics.3 Writing for the New Masses in 1938, a skeptical
Richard Wright derided the Louis-Schmeling fight as “a colorful puppet
show, one of the greatest dramas of make-believe ever witnessed in Amer-
ica.”4 For Wright, the real significance of Louis lay not in his dubious sta-
tus as a national hero, but in his ability to inspire the black masses. Three
years earlier, in September 1935, when Louis garnered a swift victory over
Jewish American Max Baer in front of 90,000 fans at Yankee Stadium,
Wright described the “religious feeling in the air” on Chicago’s South
Side, where over twenty thousand “Negroes poured out of beer taverns,
pool rooms, barber shops, rooming houses and dingy flats and flooded
the streets.” With Louis’s win over Baer “something had ripped loose, ex-
ploded,” claimed Wright, allowing “four centuries of oppression, of frus-
trated hopes, of black bitterness” to rise to the surface. Louis was “a
consciously-felt symbol . . . the concentrated essence of black triumph
over white.”5

Wright was certainly not alone in recognizing Louis’s influence as the
period’s iconic New Negro. African Americans’ limited access to legal and
political channels of protest meant that sports, and in particular boxing,
became one of the preeminent mass media through which they articulated
their conflict with the racial status quo. Until 1947, when Jackie Robin-
son joined baseball’s Major League, boxing was the only professional
sport that allowed whites and blacks to compete in the same arena. More-
over, in this individual sport of hand-to-hand combat, fighters emerged as
contested symbols of race, manhood, and nation among the American
masses. By 1933 Louis was already a fixture in the black press, supplying
African Americans with the cultural ammunition to critique their persis-
tent lack of democratic rights and dignity. Louis graced the front page of
the Chicago Defender more times than any other black figure during the

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

49IN SPORTS THE BEST MAN WINS

Depression, including Ethiopian emperor Haile Selassie.6 Not only did his
life story become the focal point of sports and human-interest sections in
various weekly newspapers, but his pugilistic exploits sparked larger de-
bates about black representation as editorialists evaluated his role in
racial advancement.

As the dawn of the New Negro era symbolized the race’s passage into
“the sunlight of real manhood,” Louis’s well-documented whupping of Jim
Crow provided a public outlet for diverse expressions of black struggle
across the socioeconomic and political spectrum.7 The term “New Negro,”
meaning a progressive, politically savvy African American, initially
emerged from the turn-of-the-century writings of Booker T. Washington.8

However, black participation in World War I in tandem with the Great Mi-
gration of African Americans to northern cities like New York and Chicago
had a radicalizing effect, infusing the New Negro movement with a height-
ened sense of militancy, urgency, and racial pride. In revisiting the Harlem
Renaissance, historians have begun to expand on its traditional interpreta-
tion as a middle-class, bourgeois literary movement to uncover the various
facets of New Negro activism from black theater companies to leftist inter-
nationalism.9 The sport of boxing offered yet another arena in which New
Negroes could express their racial militancy, albeit vicariously, through the
hard punches and prosperous lifestyle of men like Joe Louis. Indeed, the ris-
ing figure of Joe Louis gave the masculine New Negro ideal unprecedented,
mass appeal.

A detailed analysis of Louis’s coming of age in his first major profes-
sional fight against Mussolini’s darling, Primo Carnera, on the eve of the
1935 Italo-Ethiopian conflict, capped off with a suggestive re-reading of his
well-known loss to Max Schmeling in 1936, not only uncovers how dis-
cussions of black manhood dominated both domestic and diasporic resis-
tance strategies, but also helps to explain the historical emergence of the
male sports celebrity as an integral symbol of black success in the twenti-
eth century.10 The Louis-Carnera match takes center stage, since most ac-
counts have tended to downplay its significance as a matter of coincidental
timing in which foreign affairs overlapped with box-office promotion.
However, a close examination of the riotous celebrations Louis inspired,
along with his mass representation in the black and leftist presses, pho-
tographs, fight films, and blues songs, reveals that African Americans ac-
tively fashioned him as a Race Man, using him to fight racism and fascism
on two fronts—at home and abroad.11 Taken from this vantage point, the
Louis story obliges historians to expand their understandings of the New
Negro’s popular dimensions as a cultural conduit through which African
Americans of the 1930s continued to address the interlocking questions of
race, gender, nation, and class.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

50 THERESA E. RUNSTEDTLER

Biography of a Race

Triumphant tales of the young boxer’s rise to fistic fame filled the pages of
black and young communist publications, along with mass-circulated bi-
ographies. Even though each had a differing agenda, they all spun his life
story into a kind of utopian biography of the race. While the sympathetic
white writer Edward Van Every engaged in hyperbole when he claimed the
boxer’s life made “story book tales of fight heroes seem tame,” the popu-
lar depictions of Louis’s struggles from southern sharecropper to northern
migrant to industrial worker to successful boxer must have resonated with
the experiences of many of his African American fans.12 Providing a myth-
ical link that connected an oppressive black rural “past” with the promise
of a prosperous urban future, the young boxer’s personal story defied re-
gional, class, and even generational boundaries to offer an accessible, yet
decidedly masculine vision of collective progress.

According to the composite story that emerged in the black press, Joe
Louis Barrow was born on May 13, 1914, in Lafayette, Alabama, the sev-
enth of eight children in a sharecropping family. In 1926, Louis and his kin
joined the Great Migration to the North, settling in one of Detroit’s black
ghettos. Soon after their arrival, twelve-year-old Louis developed his young
muscles in a part-time job delivering ice to the city’s wealthier citizens.
Trained in cabinetry at the Bronson Vocational School, Louis later worked
at the Ford plant right up until he joined the ranks of professional boxing.13

As the papers revealed, Louis had honed his fighting skills at Detroit’s
Brewster Recreation Center during his teenage years. By the time he won
the national Amateur Athletic Union light heavyweight championship in
April 1934, the youthful pugilist had participated in fifty-four bouts, win-
ning forty-three of them by knockout, thereby garnering the support of the
African American management team of John Roxborough, Julian Black,
and Jack Blackburn. Writers bragged that at twenty-one, Louis was already
two hundred pounds, standing six feet, one and a half inches tall, with fif-
teen-inch biceps.14 Showcasing his muscular physique, groomed hair, and
boyish smile, the black press helped mold him into a statue of strength and
charm that appealed to men, women, and children.

Even the Young Worker, an interracial communist organ, included fre-
quent reports on Louis that tended to cast him as an exemplary African
American worker. As one journalist related, “He was born in the slums of
Birmingham, Ala. When only a mere lad, he carried cakes of ice to eke out
a living. He worked in King Henry Ford’s plant in Detroit. Always on the
fringe of starvation, he learned how to struggle for self-preservation.” Im-
buing Louis with a black labor consciousness, the writer continued, “He
can see that as a worker, he will end up just where he started from, in the

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

51IN SPORTS THE BEST MAN WINS

slums, because of the widespread discrimination that is practiced against
his race.”15 Portraying him as an everyday man with “a chance to cash in
on his skillful dukes,” the Young Worker used Louis to not only advance a
positive image of African Americans to white youth, but also to show black
workers that they did not have to give up their race heroes to join the com-
munist ranks. White and black laborers both could rally around this male
protagonist.

By the time Louis entered the ring against Primo Carnera in June 1935,
his humble beginnings and subsequent climb to international success had
taken on an epic quality, as sympathetic journalists fashioned his biography
into the ultimate story of racial and economic uplift. In an era when images
of bumbling Sambos, feminized male minstrels, and confused primitives
still held currency, Louis’s public personification of forcefulness and fair-
ness, virility and respectability, stylishness and responsibility, resonated
with popular understandings of manhood, civilization, and modernity.
Thus, from the footnotes of the well-known narrative of Louis as American
hero emerges not only the buried history of a black diasporic icon, but also
a larger story about the intersection of gender and resistance in America’s
race wars.

From Uncle Tom to New Negro

Writing in the New York Amsterdam News, editorialist Theophilus Lewis
dubbed Joe Louis a “Boxing Business Man.” Lewis praised him as a model
of mature focus, telling readers, “Joe Louis prefers to be Joe Louis and not
what white people think Joe Louis should be. Professional boxing is his
chosen road to success.” As Lewis continued, “A man’s success is not a
playful matter—it is a serious business. He refuses to pretend it is a pas-
time, a sort of youthful prelude to mature living.”16 Despite the obvious
passion and respect with which Louis’s African American contemporaries
followed his career, sports historiography, much like popular memory, has
tended to overlook black representations of Louis. For the most part,
scholars’ focus on mainstream daily newspaper accounts has skewed their
assessments of him as a moderate and even ineffectual figure of white
cooptation.17 While several historians challenge this “Uncle Tom” cri-
tique, most still emphasize Louis’s contributions as a crossover American
hero, without deconstructing whites’ and blacks’ differing perceptions of
his cultural and political importance.18 Overall, these approaches obscure
the reality that various segments of black America acknowledged and even
lauded Louis’s accomplishments, fashioning him as a gendered expression
of public resistance.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

52 THERESA E. RUNSTEDTLER

Louis’s folk hero status relied, in large part, on his masculine embodi-
ment of the period’s shifting constructions of black identity and advance-
ment. Just ten years earlier, in the opening essay of The New Negro, scholar
Alain Locke had declared that “Uncle Tom and Sambo have passed on,”
and now the “American mind must reckon with a fundamentally changed
Negro.” According to Locke, despite African Americans’ continued exclu-
sion from the rights of full citizenship, they could still “celebrate the at-
tainment of a significant . . . phase of group development, and with it a
spiritual Coming of Age.”19 As Louis rose in the ranks of professional box-
ing alongside this collective rite of passage, racial progress became increas-
ingly conflated with the redemption of black manhood.

African Americans had long deployed masculine constructions of pow-
erful blackness to confront what historian Gail Bederman describes as the
Progressives’ tradition of weaving race and gender into a web of white male
supremacy. According to popular, early-twentieth-century thought, one
could determine a group’s civilization based on their extent of sexual dif-
ferentiation. In keeping with this pseudoscientific doctrine, black men and
women were supposedly identical, while the patriarchal organization of the
“civilized” white race signified that they were not only the furthest along in
the Darwinist chain of evolution, but also uniquely capable of wielding po-
litical authority and exercising the rights of citizenship.20 According to his-
torian Barbara Melosh, the economic difficulties of the Depression helped
to reify this overall paradigm of white male supremacy. Concerns over fam-
ily stability and conflicts over female labor led to the retrenchment of white
patriarchy after the gender subversions of the 1920s such as the passing of
the 19th Amendment for women’s suffrage, the rise of the assertive New
Woman, and the racy culture of the flapper.21 Not surprisingly, as whites
continued to articulate their racial supremacy through an assertion of male
control, many African Americans attempted to prove their equality using
resistance strategies that embraced male dominance.

Even though the African American political and intellectual movements
of the 1930s shared a common focus on promoting the legitimacy of black
manhood, New Negro activists, by no means, agreed on a standardized de-
finition of its cultural, political, and economic terms. Instead, they har-
nessed and shaped gendered discourses to suit not only their differing
philosophical and tactical aims, but also their varied constituents. While es-
tablished organizations like the National Association for the Advancement
of Colored People (NAACP) and Marcus Garvey’s pan-Africanist Universal
Negro Improvement Association (UNIA) had long appropriated the white
Victorian principles of patriarchy, propriety, industry, and thrift as the
foundation for black advancement, Harlem’s up-and-coming cadre of New
Negro writers and poets began to challenge these rigid ideals by exploring

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

53IN SPORTS THE BEST MAN WINS

homosocial bonds and masculine pursuits beyond the realm of bourgeois
domesticity.22 In turn, the public assertion of militant black manhood be-
came a rallying cry for the emerging politics of collective race and class
protest led by groups like the Brotherhood of Sleeping Car Porters (BSCP)
and the Communist Party.23 Whether they worked within the framework of
American democratic ideals, or rejected their hypocrisy, African American
activists of the 1930s used manhood as a mobilizing force.

As different sectors of black society claimed Louis as one of their own,
his public representation came to embody the class and generational ten-
sions surrounding Depression-era articulations of black manhood. On the
one hand, the period’s constructions of black manliness incorporated the
contradictory ideals of savagery and civilization, as metaphors of battle and
physical prowess existed alongside discussions of intelligence, artistry, and
respectability. On the other hand, the New Negro movement also signaled
a nascent shift toward a more modern sense of masculinity grounded less
in middle-class notions of gentility, and expressed through recreational pur-
suits, the conspicuous consumption of mass-marketed commodities, and
the open display of bodily might and sexual virility.24 The popular celebra-
tion of Louis as Race Man connected these gendered imaginings of black-
ness with the spirit of the masses. This was not a solo performance on the
part of Louis, but rather a collective spectacle involving a complex process
of negotiation among his body of black supporters.

However, even as one uncovers Louis’s significance as the quintessen-
tial New Negro of the 1930s, the inherent dangers of a masculinist critique
of racism inevitably rise to the surface. Trapped in a paradox, Louis, his
black fans, and members of the black press challenged white superiority by
engaging the same constructions of patriarchal authority that were simul-
taneously confirming their racial inferiority. Not only did they ultimately
legitimize existing power relations, but their male-centered modes of resis-
tance also pushed black women to the periphery of the struggle.

Boxing’s New Negro Comes of Age

When Louis celebrated his twenty-first birthday on May 13, 1935, the black
press urged his African American fans to pay tribute to his work as “a ster-
ling young fighter, a gentleman and sportsman.” In calling Louis “the finest
type of American manhood,” they granted him two labels that blackness did
not usually allow.25 On the front-page of the Pittsburgh Courier sports sec-
tion, one writer declared, “Joe Louis, you are a man now. . . . [O]nly a step
across the threshold of boyhood, the hopes of a race and the best wishes of
a nation are with you.” Recognizing Louis’s importance as an emblematic

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

54 THERESA E. RUNSTEDTLER

figure through which gender and race coalesced in a narrative of black
progress, the writer warned the young fighter to “live a clean, honest life . . .
and always remember that your very qualities of modesty and manliness are
the things which bring thousands of people to see you fight.”26 In empha-
sizing Louis’s own coming of age as a man, black journalists exposed the
collective focus on questions of black manhood.

In the buildup to his bout against Primo Carnera, the black press pro-
moted Louis’s redemptive and unifying mission in what some were dubbing
the “battle of the century.” With bold optimism, one writer in the Pitts-
burgh Courier maintained that Louis would defend successfully “the ardent
hopes of more than twelve million Americans” when he stepped into the
ring at Yankee Stadium. Another pre-fight feature in the Chicago Defender
named Louis the most “outstanding Race athlete of the past 30 years,” cit-
ing his unprecedented ability to draw black fans to the box office. In the
month preceding the fight, Harlem buzzed with expectant energy as
African Americans of all ages kept Louis as their favorite topic. The New
York Age even noted that “women from all walks of life, some who had
never taken any interest in fights,” prayed for a race victory in the ring.27

As widespread interest in the Louis-Carnera match cut across racial
lines, many African Americans relished the fact that the black fighter’s rise
was revitalizing the entire boxing industry after years of sparse ticket
sales.28 In a bid to bring Louis closer to a title bout, his African American
managers, Roxborough and Black, had formed a pragmatic alliance with
Mike Jacobs, an influential Jewish American promoter. Jacobs held a vir-
tual monopoly of the industry, organizing major heavyweight events in con-
junction with the Hearst Milk Fund for Babies, a New York charity run by
the wife of publishing magnate William Randolph Hearst.29

Even though Louis was already a superstar in the black press, Jacobs
“introduced” the young fighter to white America. A public relations mas-
termind, he hired press agents like black journalist Russell Cowans to crank
out daily media releases for white and black newspapers all over the coun-
try. These reports carefully constructed Louis as the epitome of white mid-
dle-class respectability.30 While this centralized communications scheme
ensured that overlapping portrayals of the “official” Louis appeared in
both presses, a comparison of white and black sources reveals that writers
reinterpreted and reshaped the Louis image along racial lines, often using
manhood as a metalanguage for race.

While most journalists in the mainstream press certainly favored Louis
to win, they were not ready to count out Carnera, even though a streak of
fixed fights and messy dealings with the mob underworld soiled the veteran
boxer’s seven-year record.31 Despite their high praise of Louis’s technical
abilities and well-mannered conduct, many white writers held reservations

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

55IN SPORTS THE BEST MAN WINS

about his physical and mental toughness. Invoking the emasculating stereo-
types of black cowardice, infantilism, and emotionality, they charged that
Louis’s encounter with Carnera would determine if this “beardless” boy
could hold his own against boxing’s big men. After all, in addition to being
eight years Louis’s senior, Carnera stood nearly half a foot taller and out-
weighed Louis by almost 70 pounds. As one writer in the Macon Telegraph
observed, the question of “Can he take it?” was the “one predominant note
of skepticism” among the white, fight-going public.32 Nationally syndi-
cated sports columnist Grantland Rice agreed that if Louis failed to score
an early knockout, the “rugged” Carnera would “outmaul” the boy to win
by decision. Moreover, Rice and many of his colleagues questioned whether
the young fighter would remain poised in the midst of the “terrific bally-
hoo” of what promised to be one of the biggest fight crowds in many
years.33 Casting Louis as the “dusky David” to Carnera’s “Goliath,” white
journalists wondered whether the youthful, black technician possessed the
gritty manhood to defeat the roughhousing Italian Giant.34

As Louis’s rite of passage to boxing manhood, the fight also became a
litmus test for the strength and maturity of the race. However unconvinced
the white press was, black writers supported Louis with great resolve, pre-
dicting an easy knockout in two to five rounds.35 The question of whether
or not Louis could “take it” reportedly drew a loud chuckle from Manager
Roxborough, who bragged that the young fighter had already prevailed in
the face of knockdowns, a fractured knuckle, and even punches to the
jaw.36 Louis’s manly battle against Carnera not only had “colored America
looking to redeem its honors in the fistic world,” but it took on greater im-
plications as a proxy for larger racial conflicts at home and abroad.37

Enlisted for Ethiopia

While Louis prepared for his conquest of Carnera, another race war threat-
ened to erupt across the Atlantic. Benito Mussolini’s imperialistic designs
on Haile Selassie’s Abyssinia weighed on the minds of many African Amer-
icans. From the Courier to the Crisis, articles in the black press kept read-
ers apprised of the latest news on the impending Italo-Ethiopian conflict
during the spring and summer of 1935. While mainstream publications
tended to bury the reports of Abyssinia, the black press featured them
prominently, often as front-page news. They carried not only current, but
historical accounts of Ethiopia, along with human-interest stories on Se-
lassie, his family, and the plight of the Abyssinian soldiers.

Ethiopia was the last independent nation on the African continent and its
potential takeover had grave implications for struggles of black autonomy

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

56 THERESA E. RUNSTEDTLER

and equality throughout the world. In particular, the perception of a parallel
between Italian fascism and United States racism served to provoke strong,
public African American reactions to the looming invasion.38 Moreover,
when the League of Nations failed to come to the aid of the African country,
it further emphasized the racial dimensions of the conflict, as self-interested,
white governments turned a deaf ear to the pleas of their colored counter-
part.39 Given the depressed economic conditions in northern black commu-
nities like Harlem and the continued terror of Jim Crow in the South, African
Americans recognized the close connections between their plight and that of
their Ethiopian brothers. As poet Langston Hughes declared:

Ethiopia, Lift your night-dark face,
Abyssinian Son of Sheba’s Race!
. . .
May all Africa arise
With blazing eyes and night-dark face
In answer to the call of Sheba’s race:
Ethiopia’s free!
Be like me,
All of Africa,
Arise and be free!40

Out of the crucible of modern colonialism and fascism emerged a growing
sense of black diasporic consciousness.

Many black fans saw the upcoming Louis-Carnera fight as an apt mi-
crocosm of the pending match up between Il Duce and Selassie. In the
major black weeklies, stories and photos of Louis’s training regimen, his
victory, and the subsequent celebrations ran side-by-side with reports of the
Abyssinian crisis and pictures of the Ethiopian emperor. Arguably, even
African Americans who did not read the papers must have picked up on the
obvious analogy. Enthusiastic discussions of the Louis-Carnera bout, from
street corners and front porches to local barbershops and beauty salons,
surely touched on the boxer’s symbolic role as he went fist-to-fist with
Mussolini’s Darling. Not only had Louis become a ubiquitous folk hero by
1935, but as historian William R. Scott argues, Italy’s imminent invasion
stimulated an unprecedented period of black American militancy and group
protest. From Los Angeles to New York, the black masses organized
Abyssinian-defense loans, acts of civil disobedience, huge rallies that at-
tracted thousands of participants, economic boycotts, and even the recruit-
ment of volunteer combat troops.41

Complementing the efforts of grassroots activists, Louis became a pop-
ular outlet for articulations of nascent black nationalism, along with radi-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

57IN SPORTS THE BEST MAN WINS

cal, international critiques of racism. He offered a public embodiment of
the intellectual discussions of the conflict that graced the pages of periodi-
cals like the Crisis, Opportunity, and Marcus Garvey’s Black Man. Various
black groups even met with Louis during his training camp to underscore
the importance of his upcoming fight for black people on the world stage.
Louis recalled, “Now, not only did I have to beat the man, but I had to beat
him for a cause.”42 Enlisted as a fistic soldier in the fight against fascism,
he promised to enact Abyssinia’s struggle for black autonomy in a way that
his legions of African American fans could grasp with a sense of visceral im-
mediacy. In the spectacle of the ring, Louis’s body would perform a utopian
vision of not only the black American body politic, but also that of the
Ethiopian homeland.

Beyond just the basic fact that Louis, a black man, would wage hand-
to-hand combat against an Italian fighter, there were a number of physical
and metaphorical parallels between the real and ring conflicts enabling
African Americans to engage in a gendered critique of domestic racism and
foreign fascism.43 In particular, contemporary black American discourses of
African redemption were suffused with the language of manly battle, inde-
pendence, and honor. To black writers and political figures of the New
Negro era, the colonized continent represented black womanhood, while
the autonomous Abyssinian nation was a decidedly male construct. Writ-
ing to the Negro World, a Garveyite publication, in the lead-up to the an-
nual UNIA convention in 1924, Irene Gaskin exhorted, “Our flag boys [the
African tricolor of red, black, and green] . . . means loyalty to our country
and the protection of our women in our motherland Africa.”44 Labeling
colonized Africa the “motherland,” she placed men at the head of both na-
tion-building and the defense of black womanhood. Since white imperial
justifications often connected a society’s ability to self-govern with its de-
gree of patriarchal order, it is not surprising that African American com-
mentators infused both these battles for racial nationalism with an
overwhelmingly masculine bent.

The conflict between Italy and Ethiopia became anthropomorphized
into a duel between Mussolini and Selassie, as the black press portrayed
Abyssinia’s struggle to remain autonomous as a test of the tiny country’s
racial manhood. At a time when a boxer’s moniker usually had ethnic over-
tones, Louis, dubbed the Brown Bomber, the Ethiopian Exploder, and the
African Avenger, became a natural stand-in for the Abyssinian emperor, and
by extension, black nationhood.45 African American cartoonist Jay Jackson
encapsulated this connection in a clever drawing that showed a much
smaller Louis boxing against a bestial caricature of Carnera in front of
Ethiopian and Italian fans, while a seat reserved for the League of Nations
remained empty in the foreground.46 (See Figure 2.1)

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

58 THERESA E. RUNSTEDTLER

As the celebrated “Crown Prince of Fistiania,” Louis was, in many
ways, the ultimate “Abyssinian Son of Sheba’s Race.”47 While some white
journalists and intellectuals questioned the racial heritage of the light-
skinned Louis and Selassie, writers in the black press embraced both men

2.1 “Ethiopia Shall Stretch Forth,” May 25, 1935. During the lead-up to his match
with the Italian, Primo Carnera, Joe Louis became a natural stand-in for the Ethiopian
emperor, Haile Selassie, and by extension, black nationhood. Used with permission.
Source: Chicago Defender.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

59IN SPORTS THE BEST MAN WINS

as strong Race heroes. White biographer Edward Van Every’s attempts to
connect Louis’s athletic prowess with his tri-racial “blood strain” resonated
with numerous reports in the mainstream dailies that sought to deempha-
size the boxer’s African roots. Although the biographer acknowledged that
Louis “insists . . . the Negro predominates in his blood,” Van Every stressed
the possibility that Louis was “a good part white and more Indian than
African.”48

Flying in the face of such efforts to undermine Louis’s role as Race
Man, African American writers positioned him as the “Black Hope,” argu-
ing that Louis was a “badge of racial prestige . . . in man’s most honored
sphere of endeavor—the noble art of self-defense.”49 Similarly, the black
press showed impressive pictures of the emperor Selassie in his full regalia,
underscoring his links to the ancient kingdom of Cush and claiming him as
the “King of all Negroes everywhere.” One editorial in the Baltimore Afro-
American even maintained that “one glance at . . . [Selassie’s] hair” surely
proved that Ethiopia was a black nation.50 Louis and Selassie’s shared
African roots became a reservoir of strength, and thus, their victories in
manly battle would be victories for the race on both a national and inter-
national scale.

Just as reports conflated Louis with Ethiopia’s emperor, Carnera be-
came the Italian dictator’s sporting deputy. With ethnic epithets like Mus-
solini’s Darling, the Ambling Alp, and the Vast Venetian, Carnera served as
a popular platform for the fascist leader’s chest-beating propaganda. Just
five years earlier in July 1930, when Carnera’s criminal associations had
caught up with him, Il Duce had personally intervened to prevent the
fighter’s deportation from the United States. Moreover, when Carnera won
the world heavyweight title against Jack Sharkey in 1933, Mussolini or-
dered a uniform of the black shirt fascisti for his boxing champion and
posed with Carnera in photos that he sent to newspapers throughout the
world. The fighter even addressed his leader with the fascist salute.51

Paralleling the Louis-Carnera pre-fight publicity, white Americans
wondered whether the tiny Ethiopian nation would survive the onslaught
of Il Duce’s larger, more modernized forces. Despite Italy’s clear military ad-
vantages, an editorial in the Crisis challenged Mussolini’s bravado, claim-
ing that the “last gobble of Africa” would prove to be a “bloody swallow.”
It charged that Il Duce and his army would have to navigate the country’s
treacherous terrain while facing the unpredictable guerrilla strikes of Se-
lassie’s courageous and cunning men.52

In the ring, Louis would have to practice and then engage a similar
guerrilla strategy in order to compensate for the gigantic proportions and
long reach of Mussolini’s Darling. Mapped out by trainer Blackburn and
perfected by Louis, the ingenious battle plan involved breaking down the

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

60 THERESA E. RUNSTEDTLER

Italian’s defensive stance with punishing body shots, and then moving in to
attack Carnera’s head. While Marcus Garvey urged his pan-Africanist
brothers to “act manly, courageously, [and] thoughtfully” in mobilizing for
the crisis that would come with Mussolini’s invasion, the black press high-
lighted Louis’s strict training regimen as another confirmation that he
would prevail. Although Garvey lamented that Abyssinia’s lack of prepara-
tion would only permit a “passionate, enthusiastic, and emotional” re-
sponse to Italy’s attack, the calm and conscientious Louis appeared
well-equipped to conquer Carnera as he slashed his way through a host of
gargantuan sparring mates.53 Intelligence and rational discipline became in-
tegral to Louis’s performance of black nationhood.

Many African American journalists and politicos connected the Louis-
Carnera fight to the gendered debates of savagery versus civilization in the
Italo-Ethiopian conflict. Although Mussolini declared that he sought to
bring progress to the supposedly backward nation of Abyssinia, black in-
tellectuals like James Weldon Johnson questioned the dictator’s rhetoric, ar-
guing that Italy was simply after African loot. Critiquing Mussolini’s
violent designs, Johnson questioned the conventional, Western definition of
civilization, arguing that even though Ethiopians lacked a modern infra-
structure, they were at least civilized in character, with “courage, honesty,
and consideration for the needs of others.”54 Drawing on similar tropes,
Pittsburgh Courier commentator J. A. Rogers compared “Selassie, The
Gentleman, And Mussolini, The Braggart.” Not surprisingly, Rogers used
heavyweight boxing as a metaphor for this larger battle of savagery against
nobility, emphasizing Mussolini’s baseness by equating his “gesturing” and
“clowning” to that of the irreverent black fighter Jack Johnson.55 In this
racial and gendered reversal, Mussolini became the minstrel, as Rogers not
only claimed Ethiopia as a civilized nation, but also referenced Louis’s con-
current role in bringing racial progress to the boxing ring.

Playing on the brutish appearance of Mussolini’s Darling, along with
his reputation for illegal wrestling and holding, the black newspapers’
drawings and photos of the Italian Giant made Carnera appear more beast
than man, while their renderings of Louis retained a lifelike appearance. Al-
though white journalists and cartoonists certainly portrayed Louis in more
humane ways than his predecessor, Jack Johnson, some still tended to de-
pict him using Sambo stereotypes. Paul Gallico’s fight-day column in the
New York Daily News included a thick-lipped, hairy depiction of Louis
chasing after Carnera. Even though Gallico predicted that Carnera would
face a “shy, easily upset man mellow,” the writer also suggested that the an-
imalistic Louis could “go berserk” at any time.56

In contrast, the black press steered away from caricatures of Louis and
quoted him using full sentences. Moreover, while boasting of his strength,

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

61IN SPORTS THE BEST MAN WINS

black journalists also emphasized his kindness and generosity to his mother
and family. In mid-April 1935, many black writers celebrated Louis’s dis-
play of patriarchal responsibility when the fighter used one of his purses to
purchase a fully furnished home for his mother.57 Whereas the Italian Giant
embodied everything that was barbaric and violent about white racism and
fascism, Louis came to exemplify an exalted form of civilized black man-
hood, grounded in a mix of physical prowess and force of character.58 By
more than just a case of coincidental timing, Louis became a gendered
metaphor of black militancy and nationalism that drew on the rhetorical
power of prevailing discourses of manliness and civilization. Even if Se-
lassie had little chance of preventing an Italian takeover, Louis would de-
fend black honor.

The Manly Art of Self-Defense

As Louis fought for Ethiopian independence, he also fought for the dignity
and citizenship rights of African Americans at home. In addition to his sym-
bolic connections to more radical, transnational black activism, he became
the focus of an interrelated debate over questions of black American man-
hood and the state of the race. This discursive battle in the popular media
was an equally significant race war being waged on the African American
home front. While he prepared for his match, black journalists shaped
many of the same gendered critiques associated with the international di-
mensions of his fight into a domestic narrative of black progress.

Black Americans’ disproportionate suffering during the Great Depres-
sion only served to highlight their continued alienation and second-class
citizenship. In the South, Jim Crow segregationists still ruled by legal and
extralegal means, as struggling black sharecroppers and laborers sought to
combat economic exploitation, widespread disfranchisement, and the ter-
ror of lynching.59 Many African Americans left the South in search of safety
and opportunity in the North, but even the Black Mecca of Harlem expe-
rienced police brutality and high unemployment. On March 22, 1935, the
famed New York neighborhood erupted into violence after rumors circu-
lated that the white manager of a local store had beaten and killed a Puerto
Rican boy. Even though several hours later the rumors were discounted,
Harlem’s first-ever race riot continued into the night, as African Americans
expressed their frustrations through mass destruction.60

Against this oppressive backdrop, Louis’s success became the most con-
spicuous argument against the continued exclusion of African Americans
from the benefits of full national citizenship. Black journalists inscribed his
body with the ideals of black manliness and masculinity, and they sculpted

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

62 THERESA E. RUNSTEDTLER

his persona into a cultural vessel in which they poured their hopes and
dreams. As an editorial in Opportunity described, “[t]he picture of a young
Negro boy working in the Ford plant at $5.00 per day . . . who literally
forces his way to a place where he can command a half million dollars
within a single year” appealed to African Americans from “every walk of
life.”61 While establishment uplifters could still embrace Louis for his re-
spectability and productivity, a younger generation of New Negroes lion-
ized him for his style and virility. To them, Louis was not exceptional;
rather, he represented what black America could do with the chance to
compete on level ground. As he climbed his way from the dirt of the cotton
fields to the bright lights of the boxing ring, he linked African Americans
from different classes and vocations in a story of collective progress.

As musicologist Paul Oliver argues, Louis’s heroic climb from the cot-
ton fields of Alabama to boxing fame encapsulated the appealing drama and
seeming invincibility of traditional African American ballad heroes like John
Henry. Indeed, Louis was the only Depression-era athlete that popular blues
artists commemorated in recorded songs.62 As a man who faced the prospect
of punishment alone in the ring, he enacted through sport the same kinds of
struggles confronting many of his fans. Houston singer Joe Pullman’s
recording, entitled “Joe Louis is the Man,” was the first song to honor
Louis’s toppling of Carnera. Although Oliver describes Pullman’s creation as
a “naïve piece of folk poetry,” it captured the essence of Louis as the arche-
typal New Negro. While revering the Bomber as “a battlin’ man,” it also
noted that he was “not a bad dressed guy,” and that even though he was
“makin’ real good money,” it failed to “swell his head.” Just as Pullman cel-
ebrated “powerful Joe” in his performance, the husky-voiced Memphis
Minnie McCoy of Chicago recorded “He’s in the Ring (Doin’ the Same Old
Thing)” as a tribute to Louis’s two-fisted “dynamite.” The mix of Memphis
Minnie’s throaty lyrics, her guitar, and Black Bob’s pounding piano empha-
sized the indestructibility of Louis, who knocked out his opponents with re-
markable consistency to the delight of his poor and working-class fans:

When your people’s goin’ out tonight,
Jes’ goin’ to see Joe Louis fight,
An’ if you ain’t got no money gotta go tomorrow night,
’Cause he’s in the ring doin’ the same ol’ thing.63

As a rallying point for black communities across the nation, the figure of
Louis served to unite the ethereal realm of diasporic politics with the every-
day troubles of African Americans.

Louis received a hero’s welcome from the black community at Grand
Central Station in New York City in the middle of May 1935. As the black

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

63IN SPORTS THE BEST MAN WINS

press included photos of Louis in chic suits enjoying the finer things in life
like driving brand-new cars, he moved beyond his station as prizefighter to
become both celebrity and socialite.64 His bodily display of impeccable
fashion was one of the most integral aspects of his gendered performance
of black pride, since it allowed him to transgress racial norms, moving be-
yond the ubiquitous black identity of poor worker to showcase his wealth
and individuality. One black correspondent praised Louis for looking the
part of fistic champion in “his street togs,” while another carefully itemized
the boxer’s wardrobe of a “dozen suits, nine pairs of shoes, two dozen
shirts, 100 neckties, ten hats, six coats and countless sweaters, zippercoats,
[and] suits of underwear and pyjamas.”65 Likewise, newspaper ads for
Murray’s Pomade, a popular hair straightener, reinforced Louis’s reputa-
tion for being not only a great fighter, but also “one of the best dressed men
in America.” As the text of the advertisement claimed, Louis strived to be
“well-groomed” both in and out of the ring. The company encouraged the
reader to support Louis and to buy their product, since doing both would
enable a man to take on the young boxer’s power and panache in his every-
day life.66 As the consummate New Negro, Louis reinforced his manhood
through his prodigious consumption and street-hip style, offering an opti-
mistic vision of the possibilities of black urban America.

Part politician, part pop idol, and part philanthropist, Louis spent a
busy week in the Big Apple meeting with civic leaders like Mayor Fiorella
LaGuardia, shaking hands with boxing legends like Jack Dempsey, and at-
tending a series of charity benefits. Trading in his trousers for workout
gear four times a day, Louis also starred in a promotional, vaudeville show
at the Harlem Opera House, scoring one of the biggest draws in the his-
tory of the theater. With a kick-line of pretty dancing girls in the back-
ground, he sparred, skipped, and punched the heavy bag to the delight of
packed houses. However, the respite was short-lived. With only a month
left before the Carnera fight, Louis left for his training camp in Pompton
Lakes, New Jersey.67

Black correspondents painted an idyllic picture of the countryside estate
where Louis prepared for battle, emphasizing its connections to old Ameri-
can gentility, while also touting its modern conveniences. Celebrating Louis’s
role as the temporary master of the “Big House,” they cloaked him in a
mantle of both bourgeois respectability and technological efficiency.68 Ac-
cording to local lore, George Washington had slept there, and black writers
claimed that Louis now occupied the same room where the first president
had stayed. Reputedly “one of the most famous fistic training grounds in the
world,” the camp was “[n]estled in a nature-scooped nook of the Ramapo
Mountains,” yet close enough to the city of Patterson to offer all of the
amenities of rural and urban life combined. Although Louis spent most of

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

64 THERESA E. RUNSTEDTLER

his days working out, in his few moments of leisure time he supposedly en-
joyed freshwater fishing, boating, golfing, and even horseback riding.69 The
training camp itself became an expression of not only Louis’s nobility and
modernity, but also the dignity and advancement of his people.

As the first fighter to ever rent the entire grounds for the exclusive use
of his training camp, Louis ruled as lord of the estate. He retained a six-
teen-man, African American entourage that included an eighteen-year-old,
personal valet and the “expert dietician” Frank Sutton, a former restaura-
teur. In particular, Sutton, who had once served Booker T. Washington, be-
came a popular figure in the black press reports from Pompton Lakes.
Referencing the “nutritionist,” black writers presented detailed accounts of
Louis’s disciplined, “two-meal-a-day diet,” countering white reports of the
fighter’s supposed penchant for ice cream and tendency to overeat.70

Editorials in the black press insisted that African American fighters no
longer needed to seek out white assistance to get ahead. Louis reputedly re-
jected the possibility of white patronage, saying that he would “hang up the
gloves for good” if Roxborough and Black sold any part of his earnings. By
this time Jacobs certainly provided much of Louis’s financial backing, but
black reports tended to downplay the white promoter’s role, while empha-
sizing the influence of his black managers. Roxborough, Black, and Black-
burn’s tactical abilities at the negotiating table and at ringside formed an
important plotline in the story of Louis’s success. In true New Negro form,
Louis and his black “Board of Strategy” were beating white men at their
own enterprise.71

A steady stream of cars and pedestrians traveled to the estate to see
Louis in action. In this seemingly apolitical space, showing support for
Louis enabled his black supporters to publicly express their own status and
worth and to gain vicariously the strength of his fists. By the middle of
June, his sparring workouts had already attracted around 3,200 visitors,
and as the fight drew nearer, writers predicted crowds of 1,000 per work-
out of mostly African American fans from all along the East Coast.72

Alongside regular folk, professionals and celebrities made appearances.
Black newspapers like the New York Age and the Baltimore Afro-American
provided weekly lists of the VIP spectators—judges, sportsmen, entertain-
ers, entrepreneurs, orchestra leaders, morticians, and politicians—who
ranged from local to national elites. Many of those who saw Louis in the
flesh achieved their own form of celebrity as they returned home to trum-
pet his prowess on the street corners and in the bars of their urban com-
munities.73 Attending the Louis camp became, for spectators, an expression
of pride and promise.

As Louis toppled his sparring mates, his African American fans cele-
brated him as a polished, physical specimen of black virility. Louis embod-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

65IN SPORTS THE BEST MAN WINS

ied an undeniable, yet understated sexuality that appealed to the younger
generation of New Negroes without upsetting the traditional conventions
of respectability. Even though the Louis team’s “official” position was that
the fighter did not associate with women, black fans still celebrated his
bodily perfection. As public school teacher Helen Harden recounted in a
letter to the New York Age, many spectators visited the camp “with one
purpose,” and that was “to gaze on the Detroit Bomber.” Harden gushed
that he was simply “lovely to look at. Not a blemish on his saffron hued
skin.” Another black female fan refused to believe the official reports that
claimed Louis would keep women out of his life until he won the world
title, arguing that “Joe is a real man, after all.”74

Although the young boxer obviously appealed to women, many articles
in the white press twisted the Louis party line to unsex and infantilize the
black fighter, claiming that “iceberg” Louis had “no time for women” and
that his only “sweetheart” was his mother.75 Challenging these images, the
black press fashioned him as an idol of masculinity, showing suggestive
photos of Louis washing himself in the shower and gazing at the camera
partially disrobed. While black writers did acknowledge that Louis had no
serious plans for marriage, they also reported that camp intimates swore he
was a “lady-killer.”76 However, concerned with dissociating their fighter
from the negative legacy of Jack Johnson, Louis’s handlers kept the young
man’s sexual escapades with white women, along with his love of speeding
cars and frivolous spending, out of the press.77 In an era when black male
sexuality connoted rape and recklessness, Louis’s carefully constructed bal-
ance of physicality and decency offered a positive model of virile black
manhood.

Despite the more daringly masculine aspects of his persona, Louis still
stood as a paragon of manly productivity in the face of racist, white reports
of his laziness. Even a sympathetic white writer like Van Every betrayed his
prejudice when he claimed that Louis’s trainer had to “force Joe . . . to cut
out his dissipation . . . even if it infringed on his sleep.”78 In refuting these
types of disparaging comments, one journalist in the New York Amsterdam
News declared that “[n]o fighter during the past twenty years has trained
with more earnestness than this Detroit boy.”79

Following the conventions of contemporary boxing manuals, the black
press provided detailed descriptions of Louis’s routine, arguing that his
abilities were not just “natural,” but cultivated.80 With scientific precision
and utmost discipline, Louis arose at six in the morning to run in the moun-
tains, followed by a demanding afternoon of sparring matches, bag punch-
ing, rope skipping, and bending exercises. So important was it to counter
notions of black indolence that one sportswriter even maintained that
Louis was a model of efficiency when he slept, taking “it as seriously as he

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

66 THERESA E. RUNSTEDTLER

does his fighting. No faking, no lost motion.”81 In this way, Louis’s persona
combined the traditional watermarks of gentlemanly respectability with the
rising tide of New Negro masculinity. He became not only the Race Man,
but also an Everyman for the race.

The New Negro and his New Crowd

Just days before the fight-date, the impending Italian invasion of Ethiopia
permeated local politics as the Hearst Milk Fund contemplated canceling the
Louis-Carnera bout for fear that it would inspire race riots. The Hearst an-
nouncement marked the high point in a month-long racial debate over the
potential for black-Italian violence at the match. Pointing to the rioting of
Harlem’s black population in March 1935 and the ongoing furor over the
Abyssinian crisis, white sportswriters Westbrook Pegler and Arthur Brisbane
warned that a boxing match pitting a black American against an Italian
fighter would furnish the fuel for racial unrest in both the stands and streets.
Pegler deemed the bout a “new high in stupid judgment,” while Brisbane
worried that it might inspire “a fight bigger than the scheduled fight.”82

Given Pegler and Brisbane’s predictions, it became clear that not just
Louis’s manhood was on the line in the upcoming match, but also the col-
lective manhood of his African American spectators. The black press re-
sponded with vehemence. Al Monroe of the Chicago Defender recognized
white America’s unease with the sudden rise of the Race Man Louis, whose
burgeoning popularity was “moving ‘out of control.’” He dismissed the
warnings of violence, claiming that his Nordic counterparts had no inten-
tion of writing “the real facts.”83 In turn, while the New York Amsterdam
News claimed that “Negroes today are unlikely to riot over anything less
than deep-seated social injustice and economic exclusion,” they also
warned that “Negroes ARE likely to be forced to defend themselves against
attack by whites who have been stirred by repeated comment on the possi-
bilities of rioting.”84

In late June, when a front-page editorial in the white Newark Ledger
called for a boycott of the fight, the black press upped its ante. The Balti-
more Afro-American claimed that this was a deliberate move to prevent
Louis from advancing to the heavyweight championship, reporting that
blacks and Italians in Newark’s “hill” sections had responded with their
own boycott of the Ledger. Linking it to larger political questions, the
Chicago Defender placed the ultimate blame in Mussolini’s lap, declaring
that the dictator’s shameless use of the Louis-Carnera fight as fodder for
race hatred in the Italian American press had provoked the Ledger boy-
cott.85 Just as Louis’s individual victory would prove his boxing manhood,

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

67IN SPORTS THE BEST MAN WINS

so too would his black fans have a communal chance to prove their matu-
rity and respectability as spectators. Characteristic of the period’s wider
questioning of the merits of bourgeois respectability alongside the rise of
popular strains of more aggressive, mass politics, class tensions surfaced in
this aspect of the pre-fight publicity. Recalling the controversy over
Harlem’s first-ever race riot in March, black journalists understood that
much was at stake. Their arguments were not just defensive, but prescrip-
tive. While Louis’s win would certainly be cause for celebration, it had to
remain civilized. Otherwise, his ultimate strength would remain locked in
his fists, unable to transfer its impact to the larger struggle against racism
and fascism at home and abroad.86

On the morning of June 25, 1935, the Brown Bomber and Mussolini’s
Darling readied themselves “to clash for the synthetic championship of two
continents.”87 Despite the reassurances of the black press, the Hearst Milk
Fund was taking no chances with the possibility of violence, and for the
first time in New York City’s boxing history, a troop of armed police would
surround the ringside at Yankee Stadium as Louis and Carnera fought.
Over 1,000 patrolmen and detectives would also be stationed at strategic
points throughout the arena.88

Since the major radio networks of NBC and CBS refused to air the
match for fear of potential bloodshed across the country, the 100 ticket sell-
ers in the stadium box office had their hands full with a last-minute rush of
spectators.89 For weeks before the fight, several black newspapers had ad-
vertised organized bus trips to the event, along with special railroad rates
and flights that welcomed both men and women.90

Under a sunny, steamy New York sky, most of the nearly 15,000
African Americans on hand to see Louis arrived long before the white spec-
tators with ringside seats. They congregated in the right- and left-field
bleachers as soon as the Yankee Stadium gates opened at five o’clock,
singing, cheering, and performing ad hoc speeches during their two-hour
wait for the preliminary fights. A journalist for the New York Age spoke
with one man who had traveled with his wife all the way from Leland, Mis-
sissippi. The writer could only interpret this cotton buyer and Fisk Univer-
sity graduate’s dedication as an example of “the spirit of enthusiasm and
race pride that urged him and thousands of others from Chillicothe, Kinder
Lots and many other hidden hamlets” across the country to attend the
fight.91 In addition to the lively crowds in the bleachers, black America’s
royalty, from politicians to professionals, and from sportsmen to entertain-
ers like Bill “Bojangles” Robinson and Lena Horne sat closer to the ring.92

By the time of the main event, over 60,000 spectators of all races packed
the stadium, with gate receipts totaling nearly $350,000, a new high for a
nontitular match.93

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

68 THERESA E. RUNSTEDTLER

As ring announcer Hugh Balogh urged, “in the name of American
sportsmanship. . . . [R]egardless of race, creed, or color, let us all say, may
the better man emerge victorious.”94 As the fighters approached each
other, Carnera looked like a massive beast alongside the young David. Yet,
it was Louis, expressionless and calm, who commanded the center of the
ring, while Mussolini’s Darling danced around him. By the end of the first
round, Louis had already drawn blood, cutting the Italian Giant’s lip with
a smashing right to the mouth. Louis continued to explode with hard body
shots, followed by rights and lefts that bruised Carnera’s face. Toward the
end of the fifth, Mussolini’s Darling looked ready to collapse, with blood
streaming down his face, but Louis, still fresh-legged, blasted him with
more head and body combinations. Louis rocked Carnera with a series of
hard rights in the sixth round, sending Mussolini’s Darling to the canvas
three times. As Carnera staggered to his feet Referee Arthur Donovan
called off the fight as Louis hit his target with a cannonade of punches.
The crowd burst into cheers as Louis won by technical knockout, with not
a mark on his face.

Even without the benefit of a radio broadcast, news of Louis’s win trav-
eled quickly. Not too far from the stadium, a phone call conveyed the re-
sult to the estimated 20,000 fans who gathered at the Savoy Ballroom in
Harlem. As the Pittsburgh Courier reported, floods of African Americans
poured into the streets from Seventh to Lenox and 125th to 145th with a
carnival spirit “reminiscent of Marcus Garvey’s best days.” The ravages of
the Depression seemed momentarily suspended as celebrants in the taverns
offered up toasts to Louis, while cars with plates from as far away as the
District of Columbia, Illinois, Maryland, Tennessee, Georgia, and Canada
crawled and honked their way down Seventh Avenue.95

As the black press pointed to the relative order of the post-fight festiv-
ities as confirmation that African Americans were not as uncivilized as Pe-
gler and Brisbane had thought, the behavior of Louis’s fans became another
mark of resistance. As a correspondent for the Journal and Guide asserted,
“Contrary to unfounded anxiety expressed in some quarters, there was no
sign of disorder before, during or after the fight.”96 Yet, the glowing de-
scriptions in the black press appear to have obscured the multiple ways in
which African Americans from different walks of life expressed their sup-
port of Louis.

Articles in the white dailies presented a much more raucous picture of
the post-fight revelry. By reading their accounts intertextually with the
black press reports, one can draw a more nuanced portrait of the vigorous
celebration without much regard to hallowed respectability. One elderly,
black orator named Gill Holton reputedly declared, “It [wa]s the greatest
night Harlem . . . had since the riot.” Officers on foot and horseback, along

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

69IN SPORTS THE BEST MAN WINS

with those driving motorcycles and radio cars, monitored the thousands of
fans that surrounded the packed Savoy Ballroom. Mounted police had to
intervene when members of the crowd stormed the entrance, breaking
down one of the doors and injuring a half-dozen people. When the com-
munity’s honorary mayor, entertainer Bill “Bojangles” Robinson, arrived in
a limousine, he made a cursory speech cautioning the throngs of fans to re-
main calm, but minutes later he, too, joined in the shouting as he moved
down the street. Belligerent youths postured on the hoods of moving cars,
yelling at the tops of their lungs, while children who should have been in
bed pounded ashcans on the streets and compared their flexed biceps.97

Even if Louis’s managers advised him against expressing his jubilation in
the ring, the Brown Bomber’s victory gave his fans an opportunity to ag-
gressively assert their racial pride en masse, in a way that defied conven-
tional racial norms.

The events surrounding the Jersey City Riots of August 1935 paint an
even clearer picture of this sense of militancy. According to a report in the
New York Age, around 100 black and Italian men armed with knives, base-
balls, stones, and other blunt objects engaged in a “free for all” of street
fighting on August 11. A verbal dispute over the impending Italo-Ethiopian
conflict and the related Louis-Carnera bout had apparently sparked a fist-
fight that exploded into a massive brawl, leaving four wounded and lead-
ing to eleven arrests. An emergency squad consisting of radio cars, along
with police on foot with tear gas bombs, managed to quell the unrest.
African Americans claimed that Louis’s recent victories had heightened
white aggression in the district. Yet, according to the whites involved, black
youths had been taunting passers-by, demanding that everyone acknowl-
edge Louis’s superiority. After the initial clash, the hostilities almost resur-
faced the next day, as two bands of white males totaling around ninety
exchanged verbal challenges with a group of African American men.98

More than just an inspiration for the writings of New Negro elites, Louis’s
decisive win sparked an already smoldering sense of militant consciousness
among the African American masses, bringing strong expressions of black
pride to the surface that defied the combined strictures of white racism and
elite decency.

Brown Moses?

In addition to energizing the masses, Louis’s conquering of Carnera ignited
a passionate debate in the black press regarding the proper representation
of the race and what constituted legitimate forms of black progress. His vic-
tory gave writers and intellectuals a symbolic slate on which they attempted

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

70 THERESA E. RUNSTEDTLER

to negotiate and navigate their struggle for manhood rights. For the most
part, black writers never questioned whether Louis had “sold out” to the
white establishment or had shirked his duties to black America.99 Rather,
they argued over whether Louis, as boxer, was a suitable male figurehead
for the future of the race, both nationally and internationally. After all, with
his success in the corporal realm of pugilism, Louis presented somewhat of
a dilemma to the traditional politics of bourgeois uplift. Many black elites
struggled to come to terms with the fact that this popular hero was gaining
unprecedented notoriety and wealth through muscular achievement, rather
than education and erudition. As African Americans endeavored to escape
the reductionist stereotypes of black physicality that consigned most to me-
nial labor, Louis emerged as a gendered wild card with multiple possibili-
ties in the changing game of racial construction.

Some commentators expressed their utter joy over Louis’s manly victory
as a source of racial pride and progress. Dan Burley of the Baltimore Afro-
American dubbed Louis the “Brown Moses of the Prize Ring,” claiming that
through his win over Carnera, Louis had become a national leader in the
way that Moses brought the Israelites out of bondage. Citing the fact that
Texas was now competing for a chance to host a Louis fight, along with
Missouri’s decision to lift its ban of interracial matches, Burley maintained
that Louis was literally knocking out Jim Crow, with his wins being every
bit “as good as electing a Congressman to represent us in Washington.”100

In some respects, Louis could exert physical force and command white
attention in a way that escaped his black political and intellectual counter-
parts. Only in the ring could a black man actually harm a white man with-
out being arrested or lynched. Because of the ostensibly apolitical nature of
Louis’s triumph, many black writers, conscious of its larger symbolic im-
plications, could celebrate it in detail without fear of reprisal. Extensive
photo layouts of the Italian Giant’s boxing demise splashed across the
pages of many black newspapers, presenting multiple pictures of Louis
standing over his conquered foe.101

Even though some African American journalists highlighted Louis’s
mix of muscular prowess and mental acuity, contending that “his cunning
brain work[ed] in accordance with fast and deadly fists,” others cautioned
black Americans not to place their hopes in the individual, physical tri-
umph of Louis.102 While the Crisis understood his importance to the “rank-
and-file,” they advised black America not “to hitch its wagon to a boxer,
or base its judgments of achievement on the size of a black man’s biceps or
the speed and power of his left hook.”103 Moreover, another editorial in the
Baltimore Afro-American claimed that the contributions of intellectual
Race Men like Carter G. Woodson and W. E. B. Du Bois, along with the
legal advances in the anti-lynching campaign, were “worth a dozen suc-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

71IN SPORTS THE BEST MAN WINS

cesses in the prize ring.”104 Regardless of its cathartic value, Louis’s win
had not altered the structures of oppression in America, nor had it blazed
any new paths for racial progress. Placing more weight in the potential of
academic and political tactics for achieving manhood rights, they ques-
tioned the significance of sporting victories.

Falling between these two extremes, some editorialists believed that
even if Louis did not bring institutional changes, he was still an appropri-
ate role model of racial uplift, especially for young boys. While not inclined
to view Louis as “a Moses of the race or as an Economic Hope,” one writer
for the Journal and Guide maintained that the Bomber’s “moderation, tem-
perance, [and] modesty” offered the “real moral in his victory, the most im-
portant thing to be proud of.”105 A few weeks after the Louis-Carnera
bout, the New York Amsterdam News attempted to put these ideals into
action, founding and sponsoring a “Joe Louis Boys Club” that encouraged
youngsters to follow in the footsteps of “America’s model young man.” Ac-
cording to its advertisement, the club’s main purpose was to instil the
young men of the community with Louis’s discipline and competitive
spirit.106 Yet, however much adults wished that young boys would emulate
Louis’s respectability, the teen generation had different reasons for idoliz-
ing the boxer. According to the fieldwork of sociologist E. Franklin Frazier,
black youths from all classes in the 1930s admired Louis for his conspicu-
ous wealth and hip style and drew vicarious satisfaction from his brutaliz-
ing of white opponents.107 To them, Louis was less about uplift and more
about black pride and militancy.

Ultimately, even if the heavyweight emerged as a contested symbol with
little concrete effect on the realities of long breadlines and Mussolini’s im-
perial designs, his win over Carnera still served to shine a critical spotlight
on the struggles and ironies of black life. Both journalists and cartoonists
in the African American press used the gendered images of boxing to for-
mulate political critiques that drew explicit connections between foreign
fascism and domestic racism. The focal point of the Chicago Defender’s
picture page showed a battered Carnera on the mat with a caption that
read, “I’d rather be in Ethiopia.”108 In another particularly poignant, post-
fight drawing, a boxer resembling Louis became a proxy for the Brother-
hood of Sleeping Car Porters, standing victorious over a dazed Carnera
look-alike that had “Pullman Company, Unionism” written across his
chest.109 As a figure that embodied the deep connections between diasporic
and domestic politics, Louis’s victory in the ring had underlined the
hypocrisy and unfairness of not only Mussolini and the League of Nations,
but also white America.

Pointing to the sheer absurdity of it all, another Afro-American edito-
rialist wondered what “secret of mass psychology” turned white humanity

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

72 THERESA E. RUNSTEDTLER

in one part of the nation into a murderous mob, while in another “they
cheer to the echo a little brown boy who pummels the gore out of a big
white man mountain?”110 Louis’s victory over Carnera had exposed the
many-headed beast of white supremacy, while also subjecting it to a cul-
tural barrage of strong black manhood.

Schmeling takes Sampson

Following the Carnera fight, many journalists in the white dailies suddenly
became repositories of advice for Louis, offering cautionary tales of what
could happen if the young fighter let amusement and overconfidence get in
the way of his boxing. Bill Corum of the New York Evening-Journal
warned Louis to stick to his “Ma” and to steer clear of the jazzy night life
in Harlem. In a patronizing, almost race-baiting fashion, the writer coun-
seled: “Don’t get big headed. . . . Behave yourself.” Above all, Corum re-
minded Louis that he was not only a fighter, but a symbol to his race.111

On May 16, 1936, in Lakewood, New Jersey, Louis celebrated his
twenty-second birthday, along with the official opening of his training
camp for the first of his two bouts against Germany’s Max Schmeling. Box-
ing’s dignitaries, from Nat Fleischer of Ring Magazine to World Heavy-
weight Champion Jim J. Braddock, honored the young fighter for his
spectacular achievements over the last year.112 However, with his next
match only a month away, one of the most popular questions in the white
mainstream press was whether or not Louis “could take” the pressures of
his newfound fame. As yet another test of his mettle as Race Man, Louis’s
skirmish with Schmeling would once again become a stand-in for larger
racial conflicts at home and abroad.

As Louis began his preparations, Corum’s foreshadowing of the young
boxer’s potential downfall seemed to be coming true. Over ten pounds
heavier and reputedly more interested in improving his golf game than his
fighting skills, Louis appeared disinterested and sluggish during his initial
practices. Even though Louis was the younger and more talented boxer,
journalists from both presses wondered if his apparent smugness would
cause him to falter. As Lloyd Lewis of the Chicago Daily News contended,
“Joe Louis is the only man who can whip Joe Louis.”113

While some writers in the white dailies continued to infer that Louis’s
listlessness confirmed that blacks could not handle positions above their
usual station, the African American press responded with continued faith
in the abilities and ambition of their New Negro of the manly art. Although
one journalist in the New York American argued that “success and plenty”
were spoiling the former “canebrake baby” turned “million-dollar corpo-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

73IN SPORTS THE BEST MAN WINS

ration,” most reports in the black press tended to take on a more positive
view of Louis’s training efforts by the beginning of June.114

Outside the ring, African American writers celebrated Louis’s new
role as husband and provider for his sophisticated, beautiful bride, Marva
Trotter, thereby appropriating the gender roles of white bourgeois society.
After their wedding in September 1935, the black press seized on the op-
portunity to refute the popular racist image of Louis as a “Mammy’s
boy,” promoting the young couple as black America’s first family. Freed
from the responsibilities of her secretarial job, Mrs. Louis pursued char-
ity work, practiced the piano, visited the beauty salon, and attended par-
ties of New York’s black society. While Marva soon gained her own form
of celebrity, admired by black women for her poise, charm, and fashion
sense, she assured her fans that “Joe’s the boss of our family and he’s al-
ways going to be so.” 115 Even though economic imperatives prevented
most African Americans from fulfilling these patriarchal ideals, journal-
ists shaped Louis and his wife into a public display of healthy black
American family life.

Yet, an underlying critique of Louis’s decision to marry before obtain-
ing the heavyweight title would later come back to haunt Marva after her
husband’s loss to Schmeling. Even before their nuptials, many of Louis’s
black fans made it clear that they thought his managers needed to shield
him from the corrupting influences of women to protect his strength. As
one editorialist in the Baltimore Afro-American argued, “An athlete who
marries is usually no good for a year, trainers say. And this is the reason
managers of Joe Louis will be shooing sweet girls away from their charge
until he is champion.” The temptations of female sexuality were apparently
a dangerous distraction in the field of manly battle, and the editorialist
went on to warn Louis’ handlers not to take any chances “with some
Delilah who might snear [sic] their Sampson.”116

In addition to this sexualized, domestic plotline, the Louis-Schmeling
match up became a metaphorical battle in which African Americans could
combat the theory of Aryan supremacy that stripped the Jews of their rights
in Nazi Germany and kept blacks from achieving equality in the United
States. The African American press had already been reporting the Nazi’s
persecution of the Jews and its links to American racism as early as
1933.117 Arguably, the Jewish question did not acquire the same kind of
popular resonance in the black press in comparison to the Abyssinian cri-
sis, which still continued as a featured news item even in the summer of
1936. However, it was clear that, for some sectors of the black population,
the Louis-Schmeling match had both international and national implica-
tions for the race. Although the suave Schmeling did not have the same sav-
age appeal as Carnera, the black press still invited their readers to make

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

74 THERESA E. RUNSTEDTLER

ethnic comparisons, offering side-by-side photos of the fighters’ physical
weapons, along with listings of their measurements.118

In contrast, white sportswriters generally ignored the international im-
plications of the fight, since Hitler’s persecution of the Jews had not yet be-
come an issue in the mainstream daily press. Even the Nazis had little
interest in promoting their ties to the match, since they assumed that
Schmeling would lose.119 In the weeks before the bout, many white Amer-
ican dailies appeared to put aside their national allegiances to promote the
German in articles and pictures. While the text of the Atlanta Constitution
grudgingly argued for Louis’s inevitable victory over Schmeling, the south-
ern paper’s absence of Louis pictures versus its numerous, handsome pho-
tos of the German heavyweight spoke volumes about who they wanted to
win.120 Other white sportswriters were more transparent with their alle-
giances to Schmeling, like Pat Rosa of the New York Post who claimed that
the prideful and industrious German would certainly give Louis the
“Drifter” a run for his money. For Rosa, this test of “mind . . . over mat-
ter” would favor the talents of Schmeling.121 Louis was not the American
hero that he would later become in his rematch against the German in
1938. For many white fans, the upcoming bout was decidedly racial rather
than nationalistic.

Already delayed one day because of rain, the fight took place at Yan-
kee Stadium on the overcast evening of June 19, 1936. The poor weather
coupled with a Jewish boycott of the fight made for a relatively small crowd
of 45,000 spectators. Unlike the cool, lean panther of just a year ago, Louis
looked thicker around the waist, while Schmeling possessed the best
physique of his career. In pre-fight interviews, Schmeling revealed that he
had discovered a weakness in Louis’s supposedly impenetrable defense, and
he intended to exploit it. Throughout the bout as Louis consistently
dropped his left guard when throwing his right, Schmeling hit him with stiff
counterpunches to the jaw. In the fourth round, the German fighter rocked
Louis with a hard right, sending him reeling. Although Louis managed to
stand his ground in the face of many punishing blows, in round twelve
Schmeling smashed him with a right, sending him to his knees against the
ropes. As Louis rose to his feet on the count of four, Schmeling finished him
off with another stiff right. Louis dropped to the canvas and lay prostrate
as if sleeping.122

A shell-shocked black America went into mourning. African American
fans all across the country hung their heads in gloom. Their Race Man had
fallen to the representative of Aryan supremacy. As one report from Louis’s
home base of Detroit described, “It was like a sudden death in the fam-
ily.”123 With black America grieving, the white press quickly threw their
support behind Schmeling, arguing that the so-called Nazi boxer had proved

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

75IN SPORTS THE BEST MAN WINS

“[h]e was too smart for the Negro.” While Grantland Rice exaggerated
when he deemed the fight the “most severe beating in ring history,” the New
York Post presented a pitiful picture of the fallen Louis on his backside, ac-
companied by a headline that reduced him to “Just a Scared and Beaten
Boy.”124 Louis’s loss seemed to confirm black America’s inferiority.

African American fans did not know what to make of their “Super-
man’s” fall from grace. Rumors of doping quickly hit the black press. An-
other particularly vicious example of the post-fight gossip pointed the
finger of blame at Marva, charging that she had distracted Louis before the
match by showing him a recent love letter from her former boyfriend. In
the Black Man, Marcus Garvey maintained that Louis had simply married
too early, reasoning that the young boxer would have won against Schmel-
ing if he were still a single man. For Garvey and many of Louis’s black fans,
the tragic defeat appeared to prove the liability of women in the war of the
races. Their male-centered conceptions of the fight for racial equality
seemed to leave little room for the meaningful participation of women. Ul-
timately, Garvey hoped that Louis had “learned a lesson from the fight, that
when a white man enters the ring in a premier bout with a black man, he
realizes that he has in his hands the destiny of the white race.” Apparently
Louis had not taken his role as Race Man seriously enough.125

On the other hand, many black fans remained supportive of Louis,
pointing to his integrity and respectability even in the face of defeat. In
a letter to the New York Amsterdam News, Sam J. Jones of Brooklyn ar-
gued that Louis had proved his manhood by showing that he could with-
stand prolonged physical punishment. Moreover, Jones suggested that
black America take its lead from Louis in the midst of this crisis because
the young fighter’s denial of the rumors, along with his willingness to
take responsibility for his mistakes, illustrated his true sportsmanship
and dignity.126

While their pillars of racial manhood toppled one by one, with the
Italian conquest of Ethiopia and the continuing problems of the Great
Depression, some journalists in the black press worried about the future
progress of the race. As one post-fight headline in the Chicago Defender
asked, “Haile Selassie First, Now Louis; Who Next?” Louis’s loss against
the German fighter had managed to bring things full circle, intensifying
black Americans’ fears about the implications of the Abyssinian defeat at
the hands of Mussolini. Depicting the instability of racial uplift in the
form of a “Stool of Achievement” lying on its side with two broken legs
labeled “Louis” and “Selassie,” one cartoonist argued, “It can still be re-
paired.”127 In the wake of the Brown Bomber’s defeat, Race Men across
the nation called upon each other to stand up and take charge. (See
image 2.2.)

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

76 THERESA E. RUNSTEDTLER

By the time white America embraced Louis as a national hero with his
famous knockout win in his 1938 rematch against Max Schmeling, black
fans, even outside the United States, had long lauded the boxer as the epit-
ome of black pride and success. In the heart of the Nazi nation, a young
Afro-German man could barely contain his excitement over Louis’s pum-
meling of Schmeling, as he sat surrounded by white patrons in a public bar.
When asked what he thought of the fight, the Louis fan responded, “In
sports, the best man wins.” 128 This subtle, but smug reply incensed some-
one to throw an iron chair at his head. Louis’s victory was more than just
the symbolic overthrow of Nazi fascism; it challenged the masculine foun-
dations of white supremacy. For the young Afro-German, it was not just an
American triumph, but the triumph of a fellow black man connected to him
through a cultural and political identity forged in the transnational crucible
of racist and fascist oppression.

2.2 “It Can Be Repaired, “ June 27, 1936. Joe Louis’s loss to
Germany’s Max Schmeling seemed to underline the uncer-
tainty of racial uplift in the 1930s. With Louis and Haile Se-
lassie defeated, African Americans would have to search
elsewhere for viable Race Men. Used with permission. Source:
Chicago Defender.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

77IN SPORTS THE BEST MAN WINS

Undoubtedly, Louis was neither an uncomplicated hero of American
democracy nor a simple figure of racial cooptation, for the real moral of his
success stands as one of the most important cultural legacies of the New
Negro era. His rise as the preeminent 1930s Race Man points to the pe-
riod’s larger trend toward the engendering of blackness as a male construct.
Despite various class and generational tensions, conceptions of black dig-
nity, black strength, black resistance, and even the imagined black nation
remained intimately connected to the imagined status of black manhood.
From popular culture to academics to political organizations, the “crisis of
black masculinity” moved to the forefront of discussions on racial progress,
with increasingly visible and vocal calls for the “proper affirmation of
black male authority.”129 While political, economic, and social equality re-
mained elusive, the fantastic successes of African American athletes with
the racial integration of U.S. professional leagues in the following decades
meant that sports emerged as the ultimate, public stage for this collective
project in the assertion of black manhood. Moreover, calls for black male
athletes to conform to the bourgeois, patriarchal standards of respectabil-
ity and productivity as “role models” for young African American men,
continues to pervade current discourse on the social significance and re-
sponsibility of black athletes.

Even though the U.S. Army would soon use the figure of Joe Louis to
inspire tolerance among white G.I.’s, African Americans had already laid
claim to him as Race Man and budding patriarch. His model of black mas-
culinity—one that vanquished white men, while leading black women—
stayed with African Americans as they left home to fight Hitler and later
returned to take on Jim Crow again.

Notes

Parts of this article were presented at the 2003 Harvard University Graduate
Conference on “Performing Ethnicity” and at the 2003 Annual Meeting of the
Association for the Study of African American Life and History. The author
would like to thank all of those who graciously helped this piece to evolve over
multiple drafts, including Glenda Gilmore, Seth Fein, Paul Gilroy, Matthew Ja-
cobson, Amy Bass, Jeffrey Sammons, Pamela Grundy, and the members of the
Spring 2002 Yale Research Seminar in American History.

1. Heywood Broun, New York World-Telegram, 1938, qtd. in Chris Mead, Joe
Louis: Black Hero in White America (New York: Charles Scribner’s Sons,
1985), 159.

2. Frank Sinatra, “Foreword,” in Neil Scott, Joe Louis: A Picture Story of his Life
(New York: Greenberg, 1947).

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

78 THERESA E. RUNSTEDTLER

3. St. Clair Drake and Horace Cayton offer a sociological account of the “Race
Man” concept in Black Metropolis: A Study of Life in a Northern City (New
York: Harcourt and Brace, 1945). They argue that this social type developed
as a means for black Americans to resist their second-class status by pointing
to black superiority in particular areas of expertise. In other words, the success
of the Race Man became a metaphor for the success of all African Americans
(390–392). In examining the various facets of Louis’s popular construction as
a 1930s Race Man, this article builds on the gendered critique of twentieth-
century black politics in Hazel V. Carby, Race Men (Cambridge, MA: Harvard
University, 1998). Black feminist scholars like Carby argue against the popular
practice of equating the redemption of black patriarchal manhood with racial
progress, since using the Race Man as the dominant metaphor for black suc-
cess tends to render black women’s roles and struggles, along with the rela-
tionship between racism and sexism, largely invisible. Moreover, this
association of patriarchy with progress has often foreclosed a united front
against the related oppressions of white supremacy and gender inequality.
Please note that I use the terms African American and black or black Ameri-
can interchangeably throughout this article.

4. Richard Wright, “High Tide in Harlem,” New Masses, July 1938.
5. Richard Wright, “Joe Louis Uncovers Dynamite,” New Masses, Oct. 8, 1935,

18–19.
6. Drake and Cayton, Black Metropolis, qtd. in Mead, Joe Louis: Black Hero in

White America, 92.
7. Chicago Whip qtd. in David Levering Lewis, When Harlem was in Vogue

(New York: Penguin, 1979, reprinted 1997), 24.
8. See Booker T. Washington, A New Negro for a New Century: An Up-to-Date

Record of the Upward Struggles of the Negro Race (Chicago: American Pub-
lishing House, 1900).

9. On the Harlem Renaissance as a literary movement see Lewis, When Harlem
Was In Vogue; and Cary Wintz, Black Culture and the Harlem Renaissance
(Houston: Rice University Press, 1988). For research that expands the scope of
the New Negro movement see David Krasner, A Beautiful Pageant : African
American Theatre, Drama, and Performance in the Harlem Renaissance,
1910–1927 (New York: Palgrave Macmillan, 2002); Mark Schneider, We Re-
turn Fighting: The Civil Rights Movement in the Jazz Age (Boston: Northeast-
ern University Press, 2002); and Brent Hayes Edwards, The Practice of
Diaspora: Literature, Translation, and the Rise of Black Internationalism
(Cambridge, MA.: Harvard University Press, 2003).

10. I ground my definition of resistance in the theory of political scientist Jim C.
Scott. See Domination and the Art of Resistance: Hidden Transcripts (New
Haven, CT: Yale University Press, 1990), 8–9, 41. Louis’s victories offered mo-
ments when African Americans’ “hidden transcripts” of grievances could be
brought into public view. Moreover, my discursive deconstruction of the vari-
ety of covert ways that African Americans articulated their notions of black
representation and resistance through Louis’s persona and accomplishments
employs Scott’s overall conception of “infrapolitics” (19).

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

79IN SPORTS THE BEST MAN WINS

11. My analysis draws on historian Penny Von Eschen’s discussion of black dias-
poric activism in the 1930s. See Penny Von Eschen, Race Against Empire:
Black Americans and Anti-Colonialism, 1937–1957 (Ithaca, NY: Cornell Uni-
versity Press, 1997). Louis’s matches against Carnera and Schmeling further
demonstrate the extent to which antifascism and anticolonialism informed
public debates over black identity and politics during the Depression.

12. Edward Van Every, Joe Louis, Man and Super-fighter (New York: Frederick A.
Stokes Co., 1936): book cover. Van Every was the white sports journalist who
gave Louis his first feature break in the daily press. Black newspapers like the
Chicago Defender carried advertisements for Every’s biography (see June 13,
1936, 13).

13. Fans could read about Louis in the 1935 Pittsburgh Courier series, “The Life
Story of Joe Louis, as told to Chester Washington and William G. Nunn.”
Other major black press organs also included regular updates about the
boxer’s life outside of the ring. Also see Van Every, 34, 36, 46.

14. “Here are Details on Weight and Size of Joe Louis,” Pittsburgh Courier, June
8, 1935, section 2, 4.

15. B. Weinstein, “Joe Louis Comes to Town,” Young Worker, June 25, 1935. Also
see “The Real Joe Louis, by his sister Eunice Barrow,” Young Worker, Decem-
ber 24, 1935, 1.

16. Theophilus Lewis, “Boxing Business Man,” New York Amsterdam News, July
6, 1935.

17. Much of the literature depicts Louis as a docile “Uncle Tom” who functioned
as a “race ambassador” to white America. In these treatments, the quiet, gen-
tlemanly Louis pales in comparison to supposedly less conventional boxers like
the flamboyant Jack Johnson and draft resistor Muhammad Ali. See Othello
Harris, “Muhammad Ali and the Revolt of the Black Athlete,” in Muhammad
Ali: The People’s Champ, ed. Elliot Gorn (Chicago: University of Illinois Press,
1995): 56. Also see Harry Edwards, The Revolt of the Black Athlete (New
York: Free Press, 1969); Bill Hawkins, “The White Supremacy Continuum of
Images on Black Men,” Journal of African American Men 3, no. 3 (Winter
1998): 7–18; Othello Harris, “The Role of Sports in the Black Community,”
in African Americans in Sport, ed. Gary A. Sailes (New Brunswick, NJ: Trans-
action Publishers, 1998), 3–14; David K. Wiggins, “The Notion of Double
Consciousness and the Involvement of Black Athletes in American Sport,” in
Ethnicity and Sport in North American History and Culture, eds. George Eisen
and David K. Wiggins (Westport, CT: Greenwood Press, 1994) 133–156; and
Gorn, ed. Muhammad Ali. See Ken Burns, “Unforgivable Blackness: The Rise
and Fall of Jack Johnson.” USA: PBS, 2005; Gail Bederman, Manliness and
Civilization: A Cultural History of Gender and Race in the United States,
1880–1917 (Chicago: University of Chicago Press, 1995) 8–10; and Randy
Roberts, Papa Jack: Jack Johnson and the Era of White Hopes (New York: The
Free Press, 1983). In particular, Ken Burns’ documentary for PBS has brought
the Jack Johnson story to a mass audience on PBS. This biographical film
traces Johnson’s public exploits and the heated controversies they created
within the context of Jim Crow America. In particular, it details Johnson’s

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

80 THERESA E. RUNSTEDTLER

well-publicized marriages to white women, his unapologetic enjoyment of ma-
terial riches from clothes to cars, and his notorious taunting of white oppo-
nents while beating them in the ring.

18. Mead, 156–157. While Mead champions Louis’s contributions to the strug-
gle for racial integration, his project investigates Louis through the eyes of
white sources. For a discussion of state-sanctioned constructions of Joe Louis
in wartime propaganda, see Lauren Rebecca Sklaroff, “Constructing G.I. Joe
Louis: Cultural Solutions to the ‘Negro Problem’ during World War II,”
Journal of American History 89, no.3 (December 2002): 958–983. Also see
Jeffrey Sammons, Beyond the Ring: The Role of Boxing in American Society
(Chicago: University of Illinois Press, 1988): 97–129; Gerald Astor, “And a
Credit to His Race”: The Hard Life and Times of Joseph Louis Barrow,
a.k.a. Joe Louis (New York: E. Dutton, 1974); Jill M. Dupont, “‘The Self in
the Ring, the Self in Society’: Boxing and American Culture from Jack John-
son to Joe Louis,” Ph.D. diss. (Chicago: University of Chicago, 2000); Art
Evans, “Joe Louis as Key Functionary: White Reaction Toward a Black
Champion,” Journal of Black Studies 16, no. 1 (September 1985): 95–111;
William H. Wiggins, “Boxing’s Sambo Twins: Racial Stereotypes in Jack
Johnson and Joe Louis Newspaper Cartoons, 1908–1938,” Journal of Sport
History 15, no. 3 (Winter 1988): 242–254; and Dominic J. Capeci, Jr. and
Martha Wilkerson, “Multifarious Hero: Joe Louis, American Society, and
Race Relations During World Crisis, 1935–1945,” Journal of Sport History
10, no. 3 (Winter 1983): 5–25. Even though several valuable works examine
Louis’s black folk hero status, they still tend to overlook key questions of
gender. See A. O. Edmonds, Joe Louis (Grand Rapids: Wm. B. Eerdmans
Publishing Company, 1973); Lawrence Levine, Black Culture and Black
Consciousness: Afro-American Folk from Slavery to Freedom (New York:
Oxford University Press, 1977); Wilson J. Moses, Black Messiahs and Uncle
Toms: Social and Literary Manipulations of a Religious Myth (University
Park: Pennsylvania State University Press, 1982); Richard Bak, Joe Louis:
The Great Black Hope (Dallas: Taylor Publishing Company, 1996); Donald
McRae, In Black & White: The Untold Story of Joe Louis and Jesse Owens
(London: Scribner, 2002); and Thomas Hietala, Fight of the Century: Jack
Johnson, Joe Louis, and the Struggle for Racial Equality (New York: M. E.
Sharpe, 2002).

19. Alain Locke, ed., The New Negro: Voices of the Harlem Renaissance (New
York: MacMillan, 1925, reprinted 1992): 5, 8, 16.

20. Bederman, Manliness and Civilization, 5, 25.
21. In analyzing New Deal public art and theater, Melosh illustrates the period’s

preference for constructions of rugged, white manhood in opposition to the de-
tested, feminine images of weakness and over-refinement. See Barbara Melosh,
Engendering Culture: Manhood and Womanhood in New Deal Public Art and
Theater (Washington: Smithsonian Press, 1991): 43.

22. See Kevin Gaines, Uplifting the Race: Black Leadership, Politics, and Culture
in the Twentieth Century (Chapel Hill: University of North Carolina Press,
1996), 12–13; and Martin A. Summers, Manliness and its Discontents: The

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

81IN SPORTS THE BEST MAN WINS

Black Middle Class and the Transformation of Masculinity, 1900–1930
(Chapel Hill: University of North Carolina Press, 2004), 8–9.

23. See Beth Tompkins Bates, Pullman Porters and the Rise of Protest Politics in
Black America, 1925–1945 (Chapel Hill: University of North Carolina Press,
2001), 7–12; Robin D. G. Kelley, Race Rebels: Culture, Politics, and the Black
Working Class (New York: Maxwell Macmillan, 1994) 112–114. Bates’s
analysis of A. Philip Randolph’s BSCP places the trade union’s increasingly stri-
dent demands for the “manhood rights” of full citizenship within context of
the larger shift from a “politics of civility” and white patronage to the aggres-
sive, “new-crowd” demonstrations of the 1930s and 1940s. Moreover, as Kel-
ley contends, the Communist International’s 1928 Black Belt thesis of
self-determination offered black radicals a racial platform from which to par-
ticipate in the Party’s masculine vision of militant, international revolution.

24. Summers, Manliness and its Discontents, 151–153.
25. Emphasis added, “Joe Louis Needs Boosters, Not Knockers,” Pittsburgh

Courier, May 11, 1935.
26. “Joe Louis is ‘Three Times Seven,’” Pittsburgh Courier, May 18, 1935, section

2, 4.
27. “Joe Louis–Primo Carnera Fight Holds Spotlight,” Pittsburgh Courier, June

22, 1935, section 2, 4; “Rise of Joe Louis is Biggest Sensation in Sports His-
tory,” Chicago Defender, May 4, 1935; and Lewis E. Dial, “The Sports Dial,”
New York Age, July 6, 1935, 8.

28. In the early 1930s, the sport of boxing was on shaky ground, experiencing its
own kind of depression. With the title changing hands almost yearly in the first
part of the decade, public interest waned. Quickly becoming the sport’s biggest
drawing card, Louis ushered in what some contemporary authors termed the
pugilistic New Deal. See Alexander Johnson, Ten—And Out! The Complete
Story of the Prize Ring in America (New York: Ives Washburn, 1936): 245.

29. For a discussion of Jacobs’ monopoly of fight promotion in the 1930s, see
Richard Bak, Joe Louis: The Great Black Hope, 82–87. Also see Daniel M.
Daniel, The Mike Jacobs Story (New York: Ring Book Shop, Inc., 1949). Al-
though historian Jeffrey Sammons casts Louis’s affiliation with Jacobs as an un-
fortunate loss for Louis and black America, it was necessary for them to align
themselves with Jacobs because the promoter’s influence insured that Louis
would have a chance to challenge for the world title (Beyond the Ring, 98).

30. Mead, 53.
31. Although Carnera was a former world heavyweight champion, by 1935 his

shady associations with gangsters like Al Capone, along with his participation
in what many believed were fixed fights, was common knowledge in the box-
ing world. Moreover, his early days as a carnival sideshow act, in addition to
his freakish size and frequent clumsiness in the ring, made him a kind of laugh-
ing-stock of the profession. For more biographical information on Carnera see
Astor, chapter 7. Also see Clifford Lewis, The Life and Times of Primo Carn-
era (London: Athletic Publications, 1932). Lewis, in conjunction with Carn-
era’s French manager Leon Sée, wrote this biography in defense of Carnera’s
already tarnished image.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

82 THERESA E. RUNSTEDTLER

32. “Joe Louis, Training for Carnera Match, Decides to Become First Beardless
Heavyweight Champion,” Macon Telegraph, June 17, 1935; Van Every, 119,
123. For a contemporary discussion of the common stereotypes of black fight-
ers as “cowardly and unwilling to face punishment,” see Robert Scott McFee,
“The Rise of the Dark Stars,” Vanity Fair, July 1935, 57.

33. “Rice Says ‘Terrific Ballyhoo’ Puts Big Burden on Joe Louis,” Baltimore Sun,
June 25, 1935. Also see Joe Williams, “Negro Star on the Spot, Louis by Early
Kayo, Or—Carnera will Outmaul Him,” New York-World Telegram, June 25,
1935; and Hugh Bradley, “Louis Picked to Win But He Must Start First to Stop
Primo,” June 25, 1935.

34. Sid Mercer, “50,000 to See Fight Tonight,” New York American, June 25, 1935.
35. See Al Monroe, “Fight May Even End in Two if Detroiter Starts Early,”

Chicago Defender, June 22, 1935; Bill Gibson, “Brown Bomber Should Win
before 6th Round,” Baltimore Afro-American, June 15, 1935, 21; Russell J.
Cowans, “Louis in Great Shape, Battle Predicted,” Pittsburgh Courier, June
22, 1935, section 2, 4; and “Louis’s Spar Mate, Six and One-Half Feet Tall,
Gives Carnera 5 Rounds,” California Eagle, June 21, 1935.

36. “Joe Louis Can Take It; His Manager Tells Why,” Chicago Defender, June 22,
1935.

37. “New York Likes Joe Louis,” Claude Barnett Papers, Part I, Series A, Reel 10,
May 20, 1935, 16.

38. William R. Scott, Sons of Sheba’s Race: African-Americans and the Italo-
Ethiopian War, 1935–1941 (Bloomington: Indiana University Press, 1993): 9.
Also see Brenda Gail Plummer, Rising Wind: Black Americans and U.S. For-
eign Policy, 1935–1960 (Chapel Hill: University of North Carolina Press,
1996); Von Eschen, Race Against Empire; Joseph E. Harris, African American
Reactions to War in Ethiopia (Baton Rouge: Louisiana State University Press,
1994); J. Diggins, Mussolini and Fascism: The View from America (Princeton,
NJ: Princeton University Press, 1972); William R. Scott, “Black Nationalism
and the Italo-Ethiopian Conflict, 1934–1936,” Journal of Negro History 63,
no. 2 (April 1978): 118–134. For contemporary explanations of the conflict
and its implications for African Americans see J. A. Rogers, “Italy over
Abyssinia,” Crisis, February 1935, 38–39, 50; Makonnen Haile, “Last Gobble
of Africa,” Crisis, March 1935, 70–71, 90; and George Padmore, “Ethiopia
and World Politics,” Crisis, May 1935, 138–139, 156–157; and Charles H.
Wesley, “The Significance of the Italo-Abyssinian Question,” Opportunity,
May 1935, 148; Marcus Garvey, “Barbarism in America,” Black Man, Octo-
ber 1935, 8. The Abyssinian crisis was arguably the most talked-about story of
foreign fascism for African Americans, as reports on the conflict continued to
appear on the front pages of black newspapers well into 1936.

39. See cartoon entitled “Maybe He Bribed the Guard,” Chicago Defender, March
9, 1935. This cartoon shows an Italian burglar robbing an Ethiopian store-
house of natural resources as a League of Nations security guard looks the
other way. Also see “The League of Nations,” Chicago Defender, March 2,
1935, editorial page; “See Mussolini Forcing a War with Ethiopia: France,
England Join Plot Against Africa,” Chicago Defender, May 25, 1935, 2.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

83IN SPORTS THE BEST MAN WINS

40. Langston Hughes, “Call of Ethiopia,” Opportunity, September 1935, 276.
41. Scott, Sons of Sheba’s Race, 9, 59. For more detailed descriptions of popular

black activism during the Abyssinian crisis, see Scott’s chapters entitled
“Grass-Roots Activism” and “Harlem Mobilization.” For a southern perspec-
tive, see Kelley, “Afric’s Sons with Banner Red” and “This Ain’t Ethiopia, But
It’ll Do” in Race Rebels.

42. Joe Louis, with Edna and Art Rust, Jr., Joe Louis: My Life (New York: Har-
court Brace Jovanovich, 1978): 58.

43. Although journalists in the white, mainstream press also played on the inter-
national implications of the Louis-Carnera fight, they characterized the im-
pending invasion as a wholly foreign affair with no real links to contemporary,
domestic forms of racist fascism in the United States. In their reports, Louis did
not function as a representative of American democracy, but rather, he took on
the role of an Ethiopian auxiliary defending Abyssinia from the ravages of Ital-
ian fascism. See Westbrook Pegler, “Emperor Goes in Training for His ‘Big Boy
Peterson’: Mussolini Takes Leaf Out of Carnera’s Science of Warfare By Se-
lecting Setup For His First Battle,” Birmingham Post, February 16, 1935; “Po-
lice Squads to Guard Louis,” Baltimore Sun, June 25, 1935; and John Lardner,
“Can’t Help Being King, Says Louis: Wins First Real Skirmish Between Men of
Italy and Ethiopia,” Evening Bulletin Philadelphia, June 26, 1935.

44. Irene Gaskin, “Boys Salute the Flag, the Red, Black, Green,” Negro World,
July 5, 1924, 10, qtd. in Summers, 100. On the male-inflected language and
performance of African redemption in Garvey’s UNIA, see Summers, “A Spirit
of Manliness,” in Manliness and its Discontents, 66–110. In Garveyite
rhetoric, the physical space of Africa and the process of redemption both pre-
sented ideal sites for the assertion of black manhood as men took on a mili-
taristic function while women played supporting roles. For the African
Americans’ gendered imaginings of Haiti, see Mary A. Renda, Taking Haiti:
Military Occupation and the Culture of U.S. Imperialism, 1915–1940 (Chapel
Hill: University of North Carolina Press, 2001) 261–288. New Negro artists’
rehabilitation of Haiti as “America’s Africa” involved a re-reading of the Hait-
ian Revolution as a triumphant narrative of black manhood and black pride
through figures like Toussaint L’Ouverture.

45. For a list of Louis’s popular nicknames see Lenwood G. Davis, Joe Louis: A
Bibliography (Westport, CT: Greenwood Press, 1983) 202–203. Also see
Mead, 50–51.

46. Jay Jackson, “Ethiopia Shall Stretch Forth—(Modern Version: His Fist),”
Chicago Defender, May 25, 1935, editorial page.

47. Chester Washington, “Sez Ches,” Pittsburgh Courier, June 1, 1935, section 2,
4; Hughes, “Call of Ethiopia.”

48. Van Every, 24, 26, 27. One pre-fight cartoon even played up the Asian char-
acteristics of Louis’s face, touting him as “more Mongolian than Senegam-
bian.” See Burris Jenkins, Jr., “Brown Study,” New York Evening Journal, June
6, 1935.

49. Dan Burley, “Louis Ready for Baer,” Baltimore Afro-American, April 20,
1935, 21.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

84 THERESA E. RUNSTEDTLER

50. “Just let Italy Try it!” Chicago Defender, June 15, 1935; “One Look at His
Hair,” Baltimore Afro-American, April 6, 1935, 4.

51. New York Times, November 28, 1930, 31; Astor, 95; and The Kings of the
Rings, produced by Jean Labib and T. Celal for HBO Home Video, 1995.

52. “No Snap,” Crisis, March 1935, 81. Also see “Ethiopia Defiant as Italy Plans
to Grab Africa,” Chicago Defender, February 16, 1935; “Ethiopia Has
500,000 Men for Conflict,” Chicago Defender, June 22, 1935, 1–2; “Ethiopia
in Stern Reply to Mussolini,” Chicago Defender, May 11, 1935, 1; and “Look
Out, Italy,” Chicago Defender, June 15, 1935, 1.

53. Compare Marcus Garvey’s treatise on the value of preparedness with respect
to the Abyssinian crisis in “Lest We Forget,” Black Man, Oct. 1935, 4, with
Louis’s various training updates in the black press such as “Couldn’t Take it,
Ace Clark Deserts,” Pittsburgh Courier, June 15, 1935; Russell Cowans,
“Louis in Great Shape, Battle is Predicted,” Pittsburgh Courier, June 22, 1935,
section 2, 4; and Dan Burley, “Louis In Tip-Top Form on Eve of Carnera
Bout,” Baltimore Afro-American, June 22, 1935, 20.

54. Thomas O’Halloran, “Forced Civilization Hit By Educator in Talk on
Ethiopia,” New Jersey Post, October 30, 1935, James Weldon Johnson Scrap-
books, Box 7, Beinecke Library, Yale University, New Haven, CT. For other ex-
amples of the savage versus civilized debate, see Rev. E. A. Abbott, Letter to
the Editor, “Mussolini and Ethiopia,” New York Age, July 20, 1935; and
“‘Civilizing,’ Ethiopia,” New York Age, August 3, 1935, 6.

55. J. A. Rogers, “Selassie, the Gentleman, and Mussolini, the Braggart, Com-
pared: J. A. Rogers Gives Graphic Comparison of Italian and Ethiopian Tac-
tics,” Pittsburgh Courier, August 3, 1935, section 2, 2. Garvey also concurred
with Johnson’s assessment of the Italian aggression, critiquing Mussolini’s
plans to bomb and gas innocent women and children and labeling the dictator,
“the arch-barbarian of our present age.” See Marcus Garvey, “The War,”
Black Man, October 1935, 1. Also see Garvey’s poems, “The Beast of Rome,”
Black Man, October 1935, 4, and “Il Duce—The Brute,” Black Man, July-
August 1936, 6.

56. Paul Gallico, “At it Again,” New York Daily News, June 25, 1935. The white
dailies’ infantilized, Sambo portrayals of Louis continued even after his defeat
of Carnera. See Hoff, “Ink Pot-Pourri,” St. Paul Pioneer Press, July 21, 1935;
Ed Hughes, “Another Case of ‘Bad Hands,’” Brooklyn Daily Eagle, August
17, 1935. For a more thorough discussion of Sambo depictions of black box-
ers, see Wiggins, “Boxing’s Sambo Twins.”

57. See “Joe Louis Purchases Home for his Mother in Detroit,” Chicago Defender,
April 13, 1935, 7; and Julia B. Jones, “How does it feel to be the Mother of the
Next Heavyweight Champ?” Pittsburgh Courier, April 27, 1935, section 1, 11.

58. See “The Stage is Set,” New York Amsterdam News, June 22, 1935, 14. In a
stark inversion of the traditional savage-civilized dichotomy, a picture of the
clean-cut Louis in his defensive crouch stands alongside an enlarged photo of
Carnera’s scowling, teeth-baring mug. For examples of cartoons that follow
these conventions see George Lee, “Sporting Around,” Chicago Defender, May
18, 1935, 15; and Chicago Defender, June 1, 1935, 13.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

85IN SPORTS THE BEST MAN WINS

59. For comprehensive treatments of black activism for economic and citizenship
rights in the South see Robin Kelley, Hammer and Hoe: Alabama Communists
During the Great Depression (Chapel Hill: University of North Carolina Press,
1990); Patricia Sullivan, Days of Hope: Race and Democracy in the New Deal
Era (Chapel Hill: University of North Carolina Press, 1996).

60. Jervis Anderson, This was Harlem: A Cultural Portrait, 1900–1950 (New
York: Farrar Straus Giroux, 1982): 242–244. For contemporary descriptions
of the Harlem Riot, see “Machine Guns Set Up,” Baltimore Afro-American,
March 23, 1935, 1–2; “Harlem Race Riot,” Pittsburgh Courier, March 23,
1935, 1–2; “Blame Radicals for Spreading False Rumors,” Chicago Defender,
March 23, 1935, 1–2.

61. “Joseph Louis Barrow,” Opportunity, October 1935, 295.
62. Paul Oliver, Aspects of the Blues Tradition (New York: Oak Publications,

1970): 149–50.
63. Lyrics qtd. in Oliver, 152 –53.
64. See “New Buick Brings Smile to Joe Louis,” Pittsburgh Courier, May 11,

1935. For the most part, the white dailies only included pictures of a shirtless
Louis in his fighting gear. In comparison to the black press, mainstream white
papers did not print as many photographs of Louis. Even though writers often
made him the centerpiece of their articles, pictures of Louis often failed to ac-
company their words. Instead, the white press tended to showcase more pic-
tures of Louis’s white opponents, even if they were foreigners and underdogs
like Carnera.

65. Al White, “New York Likes Joe Louis,” Claude Barnett Papers, Part I, Series
A, Reel 10, May 20, 1935, 16; and “Louis Called Best-Dressed Heavyweight,”
Baltimore Afro-American, June 15, 1935, 1.

66. Chicago Defender, June 22, 1935, 16. This same ad also appeared in several
other black newspapers. Although some may argue that Louis’s endorsement of
Murrays Pomade is representative of his willingness to ape white culture, his-
torians like Robin Kelley view “the conk as part of a larger process by which
blacks appropriated, transformed, and reinscribed coded oppositional means
onto styles derived from the dominant culture” (Kelley, Race Rebels, 168).

67. Al White, “New York Likes Joe Louis,” Claude Barnett Papers, Part I, Series
A, Reel 10, May 20, 1935, 16; “Defender Cameraman Follows Joe Louis
Around in N.Y.,” Chicago Defender, May 25, 1935, 17; “Joe Louis Captures
New York,” New York Age, May 25, 1935, 15. For a description of the vaude-
ville show, see Louis, Joe Louis: My Life, 54.

68. William G. Nunn, “Courier Writer Paints Word-Picture of Trip to Pompton
Lakes Camp,” June 8, 1935.

69. Bill Gibson, “Hear me talkin’ to ya,” Baltimore Afro-American, June 15,
1935, 20. Also see Russell Cowans, “Room Said to Have Been Used by Geo.
Washington Now Used by Joe Louis,” Baltimore Afro-American, June 8, 1935,
21; Jersey Jones, “Joe Louis’s Training Camp is One of Most Modern and Ideal
Spots in the Metropolitan District,” New York Age, June 8, 1935, 8; Baltimore
Afro-American, June 8, 1935, picture page; Lewis E. Dial, “The Sports Dial,”
New York Age, June 22, 1935.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

86 THERESA E. RUNSTEDTLER

70. See Jones, “Joe Louis’s Training Camp,”8; “Sutton, Who Helped Johnson Be-
fore Title Fight, To Be Dietician In Joe Louis Camp,” Pittsburgh Courier, April
6, 1935, section 2, 5; Chester Washington, “Visiting the Joe Louis Training
Camp,” Pittsburgh Courier, June 8, 1935; “Joe Louis Going Great on 2-Meal
Diet—Sutton,” Pittsburgh Courier, June 15, 1935; and William G. Nunn,
“Courier Writer Paints Word-Picture of Trip to Pompton Lakes Camp,” Pitts-
burgh Courier, June 8, 1935. For descriptions of Louis’s poor eating habits in
contemporary white sources, see Charles Heckelmann, “Eat and Sleep Pas-
times for Bomber Louis,” Brooklyn Daily Eagle, July 17, 1935; and Van Every,
56–57.

71. For articles on Jack Blackburn’s skills as a “mastermind” trainer see “Joe Louis
Going Great as He Trains for Big Bout with ‘Da Preem’,” Pittsburgh Courier,
June 8, 1935; and “Joe and Jack—The Perfect Combination,” Pittsburgh
Courier, July 20, 1935, section 2, 4. For discussions of Roxborough and
Black’s business smarts see “Joe Louis and His Board of Strategy,” Pittsburgh
Courier, March 23, 1935, section 2, 5; “No White Managers,” Baltimore
Afro-American, June 22, 1935, editorial page; and “Joe Louis Wins,” Chicago
Defender, June 29, 1935, editorial page. According to Summers, young, black
radicals of the 1930s spoke out against the traditional avenues of white pa-
tronage, even as they accepted white funds, in order to dissociate themselves
from the prevailing feminized image of the dependent black man (Manliness
and its Discontents, 234–240).

72. Van Every, 127–129.
73. Lewis E. Dial, “The Sports Dial,” New York Age, June 8, 1935, 8; “Many Visi-

tors at Joe Louis’s Camp,” Baltimore Afro-American, June 8, 1935, 16; and
“Johnny Dundee, Claude Hopkins Visit Louis Camp,” Baltimore Afro-American,
June 15, 1935, 20; Joseph Mitchell, “Harlem Argues Itself to Sleep About Joe
Louis and How He’ll Tear the Stadium to Pieces Tonight,” New York World-
Telegram, June 25, 1935.

74. Harden letter qtd. in Lewis E. Dial, “The Sports Dial,” New York Age, June
22, 1935; Julia B. Jones, “How Does it Feel to Be the Mother of the Next
Heavyweight Champ?” Pittsburgh Courier, April 27, 1935, section 1, 11.

75. Wilbur Wood, “Louis Iceberg in Ring or Out: Bomber Abhors Flattery and
Flatterers and Girls Don’t Interest Him,” New York Sun, August 12, 1935;
“Mother is Louis’s Only Sweetheart,” Buffalo Evening News, July 16, 1935;
Jack Miley, “Naw, I ain’t got no girl ‘cause I ain’t got no time for women,” San
Francisco Chronicle, June 27, 1935.

76. “Famed Bomber Ready,” New York Amsterdam News, June 22, 1935, 14;
“Hear me talkin’ to ya,” Bill Gibson, Baltimore Afro-American, June 15,
1935, 20; Doc Morris, “Following Joe Louis,” Chicago Defender, June 15,
35; and “Live Clean Life, Louis Advises Ring Hopefuls,” Baltimore Afro-
American, June 15, 1935.

77. Most biographers have pointed to an apocryphal list of Roxborough’s rules of
etiquette for the young fighter printed in many white and black papers to
demonstrate Louis’s dissociation from Johnson. According to the list, the
Bomber was never to have his picture taken with white women; he was never

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

87IN SPORTS THE BEST MAN WINS

to go to a nightclub alone; he would not participate in soft or fixed fights; he
was never to gloat over his opponents; he was to keep “deadpan” in front of
the cameras; and he was to live and fight clean (Mead, 52; Sammons, 98). Yet,
as Louis himself admitted in his various autobiographies, he often did things
that were in direct violation of Roxborough’s “list.” However, the Jacobs pub-
licity machine kept these aspects of his character out of the public eye.

78. Van Every, 56–57. After the fight, many articles continued to describe Louis as
lazy and sleepy. See John Lardner, “Joe Louis Sleeps and Sleeps But He’s
Happy, Family Says So,” New York Post, June 27, 1935; Margaret Garrahan,
“Fame Doesn’t Bother Giant Killer Louis: Joe Just Sleeps and Eats as Rest of
World is Agog Over Win,” Birmingham News, June 28, 1935; Henry
McLemore, “Joe Louis May be a Whirlwind Killer Inside Ring Ropes, but Out
of Them He is World’s Laziest Man,” Wilkesbarre Times-Leader, July 2, 1935;
and Charles Heckelmann, “Eat and Sleep Pastimes for Bomber Louis,” Brook-
lyn Daily Eagle, July 17, 1935.

79. “Famed Brown Bomber Ready,” New York Amsterdam News, June 22, 1935,
14.

80. For examples of contemporary boxing manuals see Nat Fleischer, Scientific
Blocking and Hitting and Other Methods of Defense (New York: C. J.
O’Brien, 1935); Boxing: A Guide to the Manly Art of Self Defense (New York:
American Sports Publishing Company, 1929); and Tommy Burns, Scientific
Boxing and Self Defence (London: Athletic Publications, 1927). These books
teach the reader how to be a skilful boxer rather than a brutish brawler.

81. Russell Cowans, “News from the Joe Louis Camp,” Pittsburgh Courier, June
1, 1935, section 2, 4; Lewis E. Dial, “The Sports Dial,” New York Age, June
8, 1935, 8; Gibson, “Hear me talkin’ to ya,”20; and “Live Clean Life.” In ad-
dition to these articles, many photos showed Louis in various stages of his
training day. See “Joe Louis at Work,” Chicago Defender, June 1, 1935, 15;
and “Defender Scribe Does Road Work with Louis,” Chicago Defender, June
15, 1935.

82. Westbrook Pegler, “Fair Enough: Plan to Stage Italian-Negro Prizefight at Very
Door of Embittered Harlem is Called New High in Stupid Judgment,” New
York Sun, 1935. Brisbane article qtd. in Mead 58.

83. Al Monroe, “Speaking of Sports,” Chicago Defender, May 25, 1935, 15. Also
see “Pegler Inspires Race Riot,” Chicago Defender, May 25, 1935; “Do They
Want Trouble?” New York Amsterdam News, June 15, 1935; and “Columnist
Spoofs Rumor of Trouble at Louis-Carnera Go,” Journal and Guide, June 1,
1935, 14.

84. “Do They Want Trouble?” New York Amsterdam News, June 15, 1935.
85. “Louis-Carnera Fight Boycott is Sought by Daily,” Baltimore Afro-American,

June 22, 1935, 1; and “Uses Papers to Separate Groups Here,” Chicago De-
fender, June 22, 1935, 1.

86. The Harlem Riot had exposed an existing class divide in terms of appropriate
race representation and activism. In the aftermath, establishment uplifters ex-
pressed their disapproval of the riot and attempted to distance themselves from
“the mob.” The editors of Opportunity claimed that “the mob does not and

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

88 THERESA E. RUNSTEDTLER

cannot reason,” and that it drew its sanction from the underworld of the “ir-
responsible soap box orator and the street corner agitator.” Thus, more civi-
lized black leadership needed to “direct the aspirations of the Negro into
peaceful channels” of protest. See “The Harlem Riot,” Opportunity, April
1935, 102. For a general discussion of these political tensions, see Gaines, Up-
lifting the Race, 246–251.

87. Chester Washington, “Louis Favored to Win by Knockout,” Pittsburgh
Courier, June 22, 1935.

88. Jack Miley, “Riot Guns Ready at Primo-Louis Fight,” New York Daily News,
June 26, 1935.

89. The box-office stat is from Miley, “Riot Guns Ready at Primo-Louis Fight.”
Also see “Prejudice Kept Joe Louis-Carnera Fight Off the Air,” Indianapolis
Crusader, July 6, 1935. This was the first major fight in years that had failed
to get national airplay, and the Indianapolis Crusader argued that networks’
actions exposed their racial prejudice. Bowing to popular demand, the Michi-
gan Network comprised of several stations managed to put the fight on air,
aided by the sponsorship of Detroit’s Stroh Brewery. A couple other Detroit
stations also aired telegraphic reports of the fight.

90. For advertisements for organized trips see the Chicago Defender, June 8 and
June 22, 1935; and the Baltimore Afro-American, June 8, 1935.

91. “Distinguished Gathering Throngs Stadium for Heavyweight Battle,” New
York Times, June 26, 1935, 24; Dan Parker, “Fans on Hand Early,” New York
Daily Mirror, June 26, 1935; “Singing, Happy Negroes Jam Bleachers To See
Ring Idol Continue Win String,” Boston Herald, June 26, 1935; and “Louis-
Carnera fight drew sport fans from all over country; Gross receipts were
$328,655.44,” New York Age, July 6, 1935.

92. “Stars of Stage, Screen, Mingle with the Masses,” Pittsburgh Courier, June 29,
1935, section 1, 4; and “List of Those at Big Bout Amazes,” Pittsburgh
Courier, June 29, 1935, section 1, 4.

93. “Louis-Carnera fight drew sport fans from all over country.” The fight also
broke the record for newspaper coverage, with hundreds of journalists from
both presses on hand.

94. Mead, 59. The following description of the fight is based on my viewing of the
fight film acquired from private collector Ken Noltheimer of Ringwise, Inc.,
along with contemporary white and black press reports, and secondary
sources. See New York Daily Mirror, June 26, 1935; “Al Monroe in Vivid
Story of Big Fight,” Chicago Defender, June 29, 1935, 14; and “‘Ches’ Gives
The Courier Readers Ringside Story,” Pittsburgh Courier, June 29, 1935, sec-
tion 1, 4.; and Mead, 59–61.

95. Floyd J. Calvin, “Harlem Goes ‘Mad With Joy’ as Joe Louis Chops Down
Giant Opponent,” Pittsburgh Courier, June 29, 1935, 1; and Astor, 102. Sim-
ilar scenes played out across the country. In Macon, Georgia, a throng of
6,000, with an estimated 3,500 blacks, congregated in front of the press offices
of the Telegraph to hear regular updates of the fight. In Detroit, thousands of
supporters reportedly converged on the Joe Louis headquarters at St. Antoine
and Beacon Streets, and in Chicago, around 10,000 fans blocked traffic out-

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

89IN SPORTS THE BEST MAN WINS

side the offices of the Chicago Defender until the wee hours of the morning.
See Bobby Norris, “White or Black—He’s Dynamite,” Macon Telegraph, June
27, 1935; “Detroit Fans Believe Baer Gave up Title to Evade Joe Louis,” Jour-
nal and Guide, June 29, 1935, 14; and “10,000 Hear Defender Broadcast of
Fight,” Chicago Defender, June 29, 1935, 1.

96. Bernard Young, Jr., “Conquest of Italian Foe is Complete,” Journal and Guide,
June 29, 1935, 2. Also see “Brisbane and Pegler,” Chicago Defender, July 6,
1935.

97. Joseph Mitchell, “Harlem is Wild About Joe Louis, Don’t Folks Here Sleep?
He Asks,” New York World-Telegram, June 27, 1935; Archer Winsten,
“There’s only Joy in Harlem as Joe Louis is Acclaimed”; Joseph Harrington,
“Many Injured Celebrating Victory”; and “Harlem Celebrates,” Chicago
Daily Tribune, June 27, 1935. This scene is also supported by pictures and de-
scriptions contained in the documentary I Remember Harlem, Schomburg
Center for Black Culture, New York Public Library, Audiovisual Division,
New York, NY.

98. “Race Riot Quelled in Jersey City,” New York Age, August 17, 1935, 1, 11.
99. There is one article that warned Louis “not to get too broad in [his] sympa-

thies” and therefore, neglect the special needs of his people and his special
obligation to black America. However, this piece was the exception, rather
than the rule. See Gordon B. Hancock, “A Letter to Joe Louis,” Journal and
Guide, July 13, 1935, 6.

100. Dan Burley, “Calls Joe Louis Worth Vice President or Congressman,” Balti-
more Afro-American, July 6, 1935, 16.

101. See “How Louis Smashed Primo’s Defense,” Pittsburgh Courier, July 6, 1935;
“The Scene as Joe Louis Smashed his Way to Victory Over Giant Carnera,”
Pittsburgh Courier, June 29, 1935, picture page; “Through the Magic of the
Speed Camera the Guide Gives you a Louis-Carnera Ringside Seat,” Journal
and Guide, July 4, 1935, 14; and “‘David Anoints Goliath’ with Barrage of
Bruising Leather,” Chicago Defender, July 6, 1935, 14.

102. Bill Nunn, “Perfect Fighting Machine,” Pittsburgh Courier, June 29, 1935.
Also see “Celebrities Praise Louis for Victory,” Chicago Defender, June 29,
1935, 7.

103. “Joe Louis and Jesse Owens,” Crisis, August 1935, 241.
104. “How Proud Should We Be of Joe Louis’s Victory?” Baltimore Afro-American,

July 6, 1935, 4. Also see “Three of a Kind,” New York Amsterdam News, July
6, 1935, editorial page.

105. “The Moral in Joe Louis’ Victory,” Journal and Guide, July 13, 1935, 6.
106. “Joe Louis Boys Club,” New York Amsterdam News, July 14, 1935, 4.
107. E. Franklin Frazier, Negro Youth at the Crossways (New York: Schocken

Books, 1940, 1967), 174–185.
108. “Graphic Story of Louis-Carnera Fight Told in Pictures,” Chicago Defender,

June 29, 1935, 13.
109. Chase, “Another Joe Louis,” New York Amsterdam News, July 13, 1935, 12.

Also see “Front Page,” Chicago Defender, July 6, 1935. In this cartoon, the
artist has a black man with the words “you and me” on his back reading a

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

90 THERESA E. RUNSTEDTLER

number of front page headlines like “Joe Louis Wins,” “Jesse Owens Sets New
Records,” “Haile Selassie Defies Italy,” while gruesome caricatures of the
Brain Trusters, Huey Long, Mussolini, and Hitler complain that the black men
have stolen their space. Once again, race progress, and even the racial subject
is male.

110. William N. Jones, “Day by Day,” Baltimore Afro-American, July 6, 1935, 4;
and Ralph Matthews, “Watching the Big Parade,” Baltimore Afro-American,
July 6, 1935, 4.

111. Bill Corum, “Stick to your ‘Ma,’ Joe,” New York Evening-Journal, June 27,
1935. Also see “Risko Warns Louis Against Overconfidence with Max,” New
York American, June 16, 1936.

112. Fred Van Ness, “Louis Celebrates 22nd Birthday; Cuts Cake and Gets Gold
Belt,” New York Times, May 14, 1936.

113. Lloyd Lewis, Chicago Daily News, June 17, 1936. For other examples from
the white press that discuss Louis’s poor showing at Lakewood, see “Louis
Listless in Sparring with Mates,” New York American, May 27, 1936; and
Hype Igoe, “Bomber Can’t Resist Lure of Golf Course,” New York Evening
Journal, June 4, 1936. For examples in the black press, see Al Monroe,
“Bomber Fails to Slay ’Em in Workouts, Chicago Defender, May 30, 1936; and
Ralph Matthews, “Joe’s Camp Upset,” Baltimore Afro-American, June 6,
1936.

114. James Cannon, “Fame and Riches May Bring About Louis’ Downfall,” New
York American, June 2, 1936. For examples of positive reports in the black
press, see “Brown Bomber Back in his Stride,” Chicago Defender, June 6,
1936; “Joe Louis Impressive in Camp Workout Sunday,” Baltimore Afro-
American, June 13, 1936.

115. Thelma Berlack-Boozer, “Joe’s Always To Be the Boss of the Family,” New
York Amsterdam News, June 20, 1936. For examples of Marva’s exposure in
the black press, see “Sunday Workout Shows Look Like Social Affair,” Pitts-
burgh Courier, June 6, 1936, section 1, 9; “Harlem Elite Deluge Marva Louis
with Favor!” Pittsburgh Courier, June 20, 1936, section 1, 9; “The Bomber’s
Bride,” New York Times, June 19, 1936. For examples of white press reports
that depicted Louis as a young boy under the disciplinary control of his
“Mammy,” see the series of articles that ran in the New York Daily Mirror in
the early part of July 1935: “Joe’s Mammy Sees Lesson in Poverty,” July 6,
1935; “Joe’s Behavior Mother’s Care,” July 7, 1935; “Joe in Church Sunday
Under Mother’s Care,” July 8, 1935; “Mother Warns Joe of Sugar-Mouths,”
July 8, 1935; “Joe’s Mother O.K.’s Fights,” July 12, 1935; and “Mother Con-
fident Joe Will Be Champ,” July 13, 1935.

116. “Keeping the Girls Away from Joe,” Baltimore Afro-American, July 13, 1935.
For descriptions of Louis’s appeal with the ladies see the series of articles in the
Baltimore Afro-American from July 13 to August 24, 1935, that described his
fan mail and the various incidents in which mobs of women rushed him for his
autograph.

117. Several secondary sources offer general analyses of the connections between
Jim Crow in the South and Nazi Germany. See Glenda Gilmore, “An Ethiop

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

91IN SPORTS THE BEST MAN WINS

Among the Aryans: African Americans and Fascism, 1930 to 1939,” unpub-
lished manuscript; Stefan Kouhl, The Nazi Connection: Eugenics, American
Racism, and German National Socialism (New York: Oxford University Press,
1994); and Seth Forman, Blacks in the Jewish Mind: A Crisis of Liberalism
(New York, New York University Press, 1998). For some contemporary dis-
cussions of the connections see Rabbi Stephen S. Wise, “Parallel Between Hit-
lerism and the Persecution of the Negroes in America,” Crisis, May 1934,
127–129; Jacob J. Weinstein, “The Jew and the Negro,” Crisis, June 1934,
178–179 (part 2 in July 1934 issue); “Stop Lynching Negroes is Nazi Retort to
American Critics,” Pittsburgh Courier, August 10, 1935, section 1, 3; and
“American Nazis Quite as Bestial as Their German Brothers,” Baltimore Afro-
American, August 24, 1935, 6.

118. See “Powerhouses of Heavyweights Compared,” and “Fighting Eyes Show De-
termination of Heavyweight Fighters,” Pittsburgh Courier, June 20, 1936.

119. Mead, 92. See “Schmeling’s Departure for the U.S. Practically Ignored in Ger-
many,” New York Times, April 16, 1936; and “Hitler Still Frowns on Max
Fighting Joe Louis in U.S.,” New York American, May 19, 1936.

120. See “It’s All Part of Day’s Work—In Busy Schmeling’s Camp,” Atlanta Consti-
tution, June 14, 1936; “Can He Stop the Bomber?” Atlanta Constitution, June
18, 1936; “Mapping Out Maxie’s Battle Plans,” Atlanta Constitution, June 19,
1936. This same trend was characteristic of other papers like the St. Louis
Daily Globe Democrat and the New York Daily News.

121. Pat Rosa, “Stolid Uhlan’s Pride and Ideals May Halt Joe Louis the ‘Drifter’,”
New York Post, June 13, 1936. Several reports also praised Schmeling for his
hard work at training camp. See Bill Farnsworth, “Industrious Max Changes
Style for Louis Bout,” New York Evening Journal, May 13, 1936; Mary
Knight, “Girl Reporter Discovers Civilized Fight Camp,” Dayton Herald, June
13, 1936.

122. I base the above description of the fight on my viewing of the fight film ac-
quired from private collector Ken Noltheimer of Ringwise, Inc., along with
contemporary white and black press reports.

123. “Detroit, Harlem in Gloom as Idol Collapses,” Detroit Evening Times, June
20, 1936.

124. Fred Digby, “Max in Sensational Win!” New Orleans Morning Tribute, June
20, 1936; “Just a Scared and Beaten Boy,” New York Post, June 20, 1936;
“Schmeling Knocks Out Louis in Twelfth Round; Most Severe Beating in Ring
History, Says Rice,” Atlanta Constitution, June 20, 1936, 1.

125. See “Continued Probe of Rumors That Bomber was Doped,” Chicago De-
fender, June 20, 1936, 1; “Louis Not Doped; Love Rift Spiked,” Baltimore
Afro-American, June 27, 1936; and Marcus Garvey, “The World As It Is,”
Black Man, July/August 1936, 19–20.

126. Sam J. Jones, “He Can Take It,” New York Amsterdam News, June 27, 1936,
12.

127. Chicago Defender, June 27, 1936, 19.
128. Interview with Gupha Voss recalling her father’s story of the second Louis-

Schmeling fight qtd. in Clarence Lusane, Hitler’s Black Victims: The Historical
Bass, Amy. In the Game : Race, Identity, and Sports in the Twentieth Century, Palgrave Macmillan, 2005.
ProQuest Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=307887.
Created from umboston on 2018-08-20 14:29:19.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

92 THERESA E. RUNSTEDTLER

Experiences of Afro-Germans, European Blacks, Africans, and African Amer-
icans in the Nazi Era (New York: Routledge, 2002), 215. There are other ex-
amples of international attention from people of color. See “What the People
Think,” Pittsburgh Courier, December 28, 1935, section 2, 4, for a congratu-
latory letter from “the colored young people of Costa Rica.” For a reference
to purported fan mail from India, see “Fans Advise Joe Louis on Marriage,”
Baltimore Afro-American, July 20, 1935, 2. Also see “Joe Louis Beats Brad-
dock and Is World Champion,” The Bantu World, South Africa, June 26,
1937, 1.

129. Philip Brian Harper, Are We Not Men? Masculine Anxiety and the Problem of
African American Identity (New York: Oxford University Press, 1996), x.

C
op

yr
ig

ht
©

2
00

5.
P

al
gr

av
e

M
ac

m
ill

an
. A

ll
rig

ht
s

re
se

rv
ed

Supportive Writing Assignment #2

Prepping for the Lit Review

In this assignment I’m asking you to start analyzing some sources you’ve found so far. These are not necessarily the only or the best sources you’re reviewing, but they should get you thinking about your research topic and questions more deeply.

EXAMPLE:

Title/ Author

“Greenwashing and Bluewashing in Black Friday-Related
Sustainable Fashion Marketing on Instagram” by Sailer, Wilfing, & Strauss

Publication Source (eg Journal of XYZ)

Sustainability, https://doi.org/10.3390/su14031494

What question is it asking?

Do fashion ads purporting sustainable practices actually increase buyer receptiveness, and are the brands advertising this way actually engaging in sustainable practices? Are the Black Friday ads specifically flagrant in their abuse of these tropes? Are they particularly effective?

What is its argument?

Green/Bluewashing has a number of repeated tropes and language signifiers. Ultimately, the findings show the ads are appealing to those who view sustainable behavior favorably, but not so for those whose behavior is “actually” sustainable.

What does it not address?

“Cause-related marketing” has seen an increase in prevalence, but do its users follow through? The article takes for granted that these practices are happening, it is not an investigation into specific cases.

How does it differ from other writings on this topic?

Other research on this topic primarily investigates the credibility of these marketing claims. Few studies limit themselves to Black Friday-specific advertising. It does not, for example, investigate the effects of other high-profile advertisements, such as Super Bowl ads.

Fill out the following tables for three of your sources so far.

Title/ Author

Publication Source (eg Journal of XYZ)

What question is it asking?

What is its argument?

What does it not address?

How does it differ from other writings on this topic?

Title/ Author

Publication Source (eg Journal of XYZ)

What question is it asking?

What is its argument?

What does it not address?

How does it differ from other writings on this topic?

Title/ Author

Publication Source (eg Journal of XYZ)

What question is it asking?

What is its argument?

What does it not address?

How does it differ from other writings on this topic?

SUBJECT- SEMINAR 114G

This week’s response is going to take a more creative approach. Instead of writing an essay, you will be creating a handbook (or part of a handbook) or guide directed towards detectives with the aim of informing them how to avoid procuring a false confession.

You may do this by:

-looking at common mistakes and advising how to avoid them

-using case studies from the textbook or podcast

-convincing them that while they may not believe in false confessions, they happen more often than we would like them to

– something more creative!

You may format this however you’d like- brochure,podcast, presentation, essay, etc. You may use images or other multimodal methods of communicating your points.

I do want you to have
at least 500 words and to show me those analysis skills! You got this.

Note that these 145 words of instruction don’t count 🙂

DUE BY 1ST OCT 2024

We know that forensic science is one of the main ways defense builds cases against the accused.

AND we know that forensic science is still evolving. On top of that, we know that forensic experts are expensive and not always easily afforded by the accused. With all that in mind, write a 1.5-2 page response in which you evaluate the role of forensics in the courtroom. Do the benefits outway the losses? Are there ways to make it more equitable?

Please use your text and “Death by Fire” as your sources and include at least one specific reference. You may refer back to your quick responses. Begin on the page below.

Constraints: 1.5-2 pages, double spaced, one inch margins

I HAVE ATTACHED THE TEXTBOOK AND DEATH BY FIRE (SOURCE FOR THE ASSIGNMENT) BELOW –

In the scene when Steve first meets King, we see Steve taking photos and videos of his friends

when King takes notice of him doing that while playing basketball. King tells Steve to come over

and photograph him. While talking to him King mentions a movie involving photography and

mentions the clicking camera sound with the motion of a finger gun inching closer to Steve’s

face while making clicking noises, making Steve look nervous, but trying not to show it. Then

King asks Steve if he has seen a real gun before and asks how he expects to pursue photography

as a career like in his favorite movie if he’s never even seen a gun flash before and then says

that’s why Steve should be ‘rolling’ with him, insinuating that gun flash is something he’d see

often if he was friends with King. This scene alone immediately foreshadows that getting in with

King would bring Steve trouble in the future and he was really not the type of person Steve

should have been spending his time with, plus King probably had something in negative

connotation planned. This scene alone shows us how easily involved Steve got with the wrong

crowd and how one interaction led him being involved in such a heinous crime. We can also see

how casual it was for King to bring someone into that kind of life and how easily someone might

have brought King into it as well. Overall just how easily the ‘streets’ can change an innocent

person into a criminal. So although the court system didn’t take it easy on Steve, he was

unfortunately there when the crime took place and there is no denying that. And though they

hardly tried to hear Steve’s side of the story, it is really being on the wrong side of town and one

interaction with the wrong person that changed the trajectory of his life.

Try it together: “Death by Fire,” Charcoal Accelerant

The first inspectors find a can of charcoal accelerant under the porch. In the trial, they

have a fire expert point to this as proof that Todd used this to set the fire. However, when

the second fire expert looks at the evidence years later, he says, “ Of course, there was

charcoal accelerate, it was right next to a barbecue.”

Sam: It shows that there was a faulty investigation and the inspectors did not thoroughly

examine the scene or use their common sense to tell that there was a grill next to the can. And

they also had mistaken the burn.

Ashley: Thus, we witness how forensic testimony can clash between experts. The difference

between the first inspectors (the firemen) and the second fire expert is based on their occupations

and knowledge on the subject. While they’re both considered experts in the eyes of the court, the

second fire expert makes the case that the first inspectors misconstrued pieces of evidence (such

as the charcoal accelerate, and the combustion patterns) that ultimately put an innocent man to

death.

Brody: In the film, it is explained that, “firemen are not scientists.” This shows how firemen are

not arson experts. They mistake the burn patterns to be a pour, but it was just a certain type of

fire that burned that way. Testimony among experts in forensic sciences can cause breaks in

evidence, and thus, punishing or even killing an innocent man.

Reina: This shows how we can’t entirely trust what these ‘experts’ tell us are facts. The

forensics of fire was still not established during Todd’s case but they still insisted saying that the

fire was not an accident and using the proof of the fire pattern as an example of it being a pour.

But, as the science of fire began to be more stable, these actual scientists later learned how the

fire was an accident and the fire pattern was caused by the accident. It also explains why Todd’s

feet were not burned from trying to save his kids, it was because the fire had started from the

wall and not from the floor

Kay: This highlights the necessity of cross-examinations for all cases. Had a second or

even third fire expert been added to the case, evidence portraying Todd’s innocence

would’ve come to light earlier, potentially saving Todd’s life. Incorrect analysis in cases

leads to prolonging of trials, furthered punishments, and gilded prosecutions. Ultimately,

this could’ve been avoided through peer-reviewing.

Max: This example proves how experts aren’t always 100% correct and they often make
mistakes that turn innocent people guilty. It also shows that “experts” aren’t always really
experts because in this case it was firefighters that were the “experts” that made the mistake.
Experts can also have contrasting opinions which will make a case confusing.

Ava: The differing levels of expertice caused differing opinions and evidence. I think that this
should be avoided in a court because in order to fairly convict a suspect you need facts not
guesses. With varying levels of knowledge about the fire and charcoal accelerant there came
confusion which messed with the idea of whether or not Todd was guilty. With confusion like this
or to this extent, you start to wonder who really are the “experts” and can they be trusted?

Evan Stockdale:
This proves that those firemen, not fire scientists mistaked all 20 indicators of Arson in that
house because they were not looking to clear Willinghams name. The firemen wanted to prove
he was guilty. A thorough inspection and examination of the crime was not performed and
shows how forensic evidence can be misconstrued and spun in a way that favors the defendant.
A barbeque was not mentioned in the initial report.

Kevin: This example shows that the evidence provided wasn’t thoroughly investigated and
analyzed possibly painting the defendant guilty. The “experts” revision of the evidence shows
some sense of fault within the case as he may have felt off about the outcome. Also, this is a
key aspect to take into consideration so it is shocking that this finding was made years later.

Brayden: This shows that the firemen who were there that day weren’t qualified to make a
scientific analysis on the cause of the fire as well as the patterns shown from the fire because
they don’t have the knowledge and expertise in that area to make that conclusion. The

firefighters looked with just their eyes but in reality it took fire scientists to actually look into it and
find the true cause of how this could have started.

Molly: The example shows that even though people are considered experts, what they are
saying is not always true. This can prove that in court the defendants can be wrongfully
convicted. This is important because it was proven in this case and finding out the truth later on.

Ronon: Based on the summary, it shows that the first fire expert didn’t seem to thoroughly
search for or think about other factors that could’ve possibly pointed out that the fire could’ve
been an accident. It might show that the first fire expert could’ve either been somewhat
inexperienced or biased against Todd as it would be a bit hard to believe in this day and age for
someone investigating a case to completely rule out factors without having a full understanding
of the situation or even having all the evidence; the second fire expert proves this as he points
out that there was a barbeque nearby, which the first fire expert failed to realize or even
mention.

Saim: This paragraph shows how evidence can be shaped by context. The first expert uses its
presence to suggest Todd set the fire, while the second point shows it was near a barbecue
offering a non criminal explanation. This highlights the risk of forensic science, where evidence
may be selectively framed to fit a narrative.

Johansi→ This illustrates that the first fireman who saw the charcoal accelerant as

definite proof of arson, didn’t look at the barbeque that was next to the accelerant. The

second fire expert suggested a more legitimate excuse (the bbq) instead of criminal

intent. This also shows how evidence can be interpreted differently based on who is

doing the analysis, like their history, experience, and biases they may have. Evidence

can be shaped by the perspectives of those analyzing it.

Saneika: This example shows how evidence can be twisted to go against a person in court.
Although there was physical evidence of an accelerator can, that is not enough evidence to
show that it was directly involved in the fire. The first fire “experts” did not state the fact that
there was a barbecue nearby in which the accelerator could have been used. These firefighters
were clearly looking for evidence to prove Todd guilty and not evidence for the cause of the fire.

Myles: This is a perfect example showing us that not all professional evidence can be trusted
when it comes to investigating these cases. Especially if you have people that are from the
same town that hates you come over and investigate, things are not always going to look good
for you. The evidence was against him from the start. It’s safer to have multiple sources come
and help especially when the evidence is not strong enough to convict somebody. The second
investigators we more experienced than the first one’s but the first investigators could have also
been biased.

Taksh: This talks about an incident in relation to law enforcement or an investigation, it likely has

to do with a fire that was possibly started on for a reason.

Under the porch, the first inspectors find a can of charcoal accelerant. This seems to be

evidence that Todd, a suspect, may have used the accelerant to start the fire.

To back up this argument during the trial, a fire expert was called around. This expert is

probably going to provide evidence that the accelerant’s presence fits in with burning or

deliberate fire starting, which could put Todd in trouble.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest Ebook
Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Convicting the Innocent

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Convicting
the Innocent

WHERE CRIMINAL PROSECUTIONS

GO WRONG

Brandon L. Garrett

Harvard University Press

Cambridge, Massachusetts • London, En gland

2011

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Printed in the United States of America

Library of Congress Cataloging- in- Publication Data

Garrett, Brandon L., 1975–
Convicting the innocent : where criminal prosecutions go wrong / Brandon L. Garrett.

p. cm.
Includes bibliographical references and index.

ISBN 978- 0- 674- 05870- 5 (alk. paper)
1. Judicial error— United States. 2. Evidence, Criminal— United States.

3. Post- conviction remedies— United States. I. Title.
KF9756.G37 2011

345.73’064—dc22 2010037172

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

To the exonerees whose stories are told here

and to the innocent people who have not yet been exonerated

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contents

1. Introduction 1

2. Contaminated Confessions 14

3. Eyewitness Misidentifi cations 45

4. Flawed Forensics 84

5. Trial by Liar 118

6. Innocence on Trial 145

7. Judging Innocence 178

8. Exoneration 213

9. Reforming the Criminal Justice System 241

Appendix 277

Notes 291

Ac know ledg ments 351

Index 355

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

In October 1993, Ronald Jones sat on death row in Illinois waiting
to be executed. He had been sentenced to death for a gruesome rape

and murder in Chicago. Jones clung to one last request— for a DNA test,
which he claimed would prove his innocence. His lawyers off ered to pay
the $3,000 that it would cost to do the test. At the time, only a handful of
people had ever proven their innocence using postconviction DNA test-
ing. The prosecutors opposed testing, arguing that it would make no
diff erence. Indeed, there appeared to be overwhelming evidence of Ron-
ald Jones’s guilt. Cook County circuit judge John Morrissey agreed and
angrily denied the motion, exclaiming, “What issue could possibly be
resolved by DNA testing?”1

Eight years before, in March 1985, the victim, a twenty- eight- year- old
mother of three, was out dancing late with her sister on the South Side of
Chicago. She was hungry and decided to get food a few blocks from her
home at Harold’s Chicken Shack. She ran into a friend on the street. As
they talked, a panhandler approached; people in the neighborhood had
nicknamed him “Bumpy,” because of his severe acne. “Bumpy” asked the
friend for fi fty cents. She gave him fi fty cents and all three parted ways.

Several hours later, the victim was found half- naked and dead in a
nearby alley behind the abandoned Crest Hotel. She had been stabbed

c h a p t e r 1

Introduction

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

2 Introduction

many times and beaten. Ronald Jones, who was familiar to the police
because he was a suspect in a sexual assault case that was never brought
to a trial, was arrested. He had a severe acne problem and was known as
Bumpy. He may have in fact been the man the friend saw that night. He
was “a homeless, alcoholic panhandler” with an IQ of about 80.2

After Jones was arrested, he was placed in a small police interrogation
room with walls bare save a sheet of paper listing the Miranda warnings.
During an eight- hour- long interrogation, he confessed. Jones did not just
say, “I did it.” He made far more damning admissions, signing a written
statement that included a series of details that only the killer could have
known. The victim was assaulted in a room inside the vacant Crest Ho-
tel, where police found a large pool of blood and some of her clothing. In
his statement, Jones said that on the morning of the crime, he was walk-
ing “by the Crest,” saw the victim, and assaulted her in a room inside.
Police analysts detected semen in the victim’s vagina. Jones said that
they had sex. The pathologist testifi ed that the victim had injuries from
trying to fend off blows. Jones said they were “wrestling and tussling.”
The victim had been stabbed four times. Jones said he lost his temper
and “cut her a few times” with a knife. Police had found a trail of blood
leading out of a window that had no glass, into the alley where they found
the victim’s body. Jones knew there was “an alley” by the hotel and said
he came and left through an open “side window.”3 It was unlikely that
anyone could coincidentally guess so many details that matched the
crime scene.

The lead detective testifi ed at trial that he brought Jones to the crime
scene, where Jones off ered more details. Jones “showed us the room”
and “showed us where the struggle took place and where she was actu-
ally stabbed.” He accurately described the victim’s appearance.4 Jones
supposedly off ered all of these crime scene details without any prompt-
ing. Those details sealed his fate.

Ronald Jones’s confession was not the only evidence against him.
Forensic evidence also linked Jones to the crime. DNA testing was at-
tempted on the semen evidence, but the results were said to be inconclu-
sive. At the time of the trial, in 1989, DNA technology was brand- new
and could only be conducted in cases with large quantities of biological
material, so conventional A-B- O blood- typing was performed. At trial,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Introduction � 3

the forensic analyst explained that 52% of the population could have
been the source of the semen and that Ronald Jones’s blood type placed
him in that group.

At the fi ve- day trial, Jones took the witness stand and recanted his
confession. In his closing argument, the prosecutor told the jury to con-
sider that Jones was a “twice- convicted felon.” He added, “Please don’t
be fooled by this man’s quiet demeanor in this courtroom and on the
witness stand. The only two eyes that witnessed the brutal rape and
murder . . . are in this courtroom, looking at you right now.”5 The jury
convicted Jones and sentenced him to death.

Jones appealed and lost. He argued that his confession was coerced
and said procedural errors infected his trial. The Illinois Supreme Court
denied his petition, as did the U.S. Supreme Court. Then the trial judge
denied his request for DNA testing.

But at the eleventh hour, Ronald Jones’s luck began to change. In 1997,
the Illinois Supreme Court reversed the trial judge and granted his request
for DNA testing. The DNA profi le on the sperm did not match Jones. The
DNA also did not match the victim’s fi ancé, with whom she had been liv-
ing at the time of the murder. It belonged to another man, who remains at
large.6 Jones’s conviction was vacated. But prosecutors waited until 1999
to drop the charges. Governor George H. Ryan pardoned Jones in 2000.7
He had spent more than thirteen years behind bars.

DNA testing saved Ronald Jones’s life. Jones later commented, “Had
it not been for DNA, who knows about me?”8 He likely would have been
executed.

What went wrong in Ronald Jones’s case? Why did he confess to a
crime he did not commit? How did he confess in such detail? Why did
the blood evidence appear to modestly support the State’s case? The
answers appear in the rec ords from Jones’s trial.

The transcripts of the criminal trial reveal a troubling story. Ronald
Jones signed the written confession statement only after enduring hours
of interrogation. On the witness stand at trial, Jones testifi ed that a de-
tective had handcuff ed him to the wall and hit him in the head again and
again with a long black object, because he refused to confess. Jones said
that a second detective then entered the room and said, “No, don’t hit
him, because he might bruise.” That detective instead pummeled him

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

4 Introduction

with his fi sts in a fl urry of blows to the midsection.9 The defense had
argued prior to trial that, based on this police misconduct, the confes-
sion should be suppressed. The detectives both denied Jones had been
struck. The lead detective denied using any interrogation techniques at
all. He testifi ed, “I sit down, I interview people, I talk to people. That’s
all I do, sir.” The judge ruled that the confession should be admitted at
trial, explaining, “I do not feel that there was any coercion or any undue
infl uence used upon the defendant.”10

Even if the confession was physically coerced as Jones described at
trial, it still raises a puzzle. Now that we know Jones was innocent, one
wonders how he could have known so much detailed inside information
about the crime. At trial, Jones explained that when police took him to
the crime scene they had walked through how the crime happened. The
detective “was telling me blood stains on the fl oor and diff erent clothing
that was found inside the abandoned building,” and that the victim “was
killed with a knife, and she was stabbed, three or four times.”11 It ap-
pears that Jones repeated the specifi c details about the crime in his con-
fession statement not because he was there, but because the police told
him exactly what to say.

The forensic evidence at Jones’s trial was also fl awed. Although the
prosecutor told the jury in his closing statement that “physical evidence
does not lie,” in fact, the forensic analyst had grossly misstated the sci-
ence.12 Jones’s blood type was the most common type. He was a Type O.
However, he was also a nonsecretor, meaning that his body fl uids did
not reveal his blood type. Only 20% of the population are nonsecretors.
The victim was a Type A secretor, as are about 32% of the population.
The vaginal swabs collected from the victim’s body matched her type
and had Type A substances on them. The analyst testifi ed that the per-
centage of males who could have been the source for the semen was the
percentage of nonsecretors added to the percentage of Type A secretors,
which would add up to about half the population.13

The analyst was wrong. A competent analyst would have explained
that any man could have been the rapist. The analyst had found nothing
inconsistent with the victim’s Type A. This raised a problem that was
common at the time, called the problem of “masking.” Substances from
the victim could “mask” any material present from the rapist. The evi-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Introduction � 5

dence from this crime scene was totally inconclusive. Nothing at all could
be said about the blood type of the rapist.

The 250 Exonerees

In retrospect, Ronald Jones’s case provides a stunning example of how
our system can convict the innocent. If his case were the only case like
this, we might call it a tragic accident, but nothing more. But his case is
far from unique. Since DNA testing became available in the late 1980s,
more than 250 innocent people have been exonerated by postconviction
DNA testing.

Who were these innocent people? The fi rst 250 DNA exonerees were
convicted chiefl y of rape, in 68% of the cases (171), with 9% convicted of
murder (22), 21% convicted of both murder and rape (52), and 2% con-
victed of other crimes like robbery (5).14 Seventeen were sentenced to
death. Eighty were sentenced to life in prison. They served an average of
thirteen years in prison. These people were typically in their twenties
when they were convicted. Twenty- four were juveniles. All but four were
male. At least eigh teen were mentally disabled. Far more DNA exonerees
were minorities (70%) than is typical among the already racially skewed
populations of rape and murder convicts. Of the 250 exonerees, 155 were
black, 20 Latino, 74 white, and 1 Asian.15

DNA testing did more— it also identifi ed the guilty. In 45% of the 250
postconviction DNA exonerations (112 cases), the test results identifi ed
the culprit. This most often occurred through a “cold hit” or a match in
growing law enforcement DNA data banks. The damage caused by these
wrongful convictions extends far beyond the suff ering of the innocent.
Dozens of criminals continued to commit rapes and murders for years
until DNA testing identifi ed them.

Before the invention of DNA testing, the problem of convicting the
innocent remained largely out of sight. Many doubted that a wrongful
conviction could ever occur. Justice Sandra Day O’Connor touted how
“our society has a high degree of confi dence in its criminal trials, in no
small part because the Constitution off ers unparalleled protections against
convicting the innocent.”16 Judge Learned Hand famously called “the
ghost of the innocent man convicted” an “unreal dream.” Prosecutors

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

6 Introduction

have from time to time claimed infallibility, announcing, “Innocent men
are never convicted.”17 Others acknowledged that human error is inevi-
table, but doubted that convicts could ever convincingly prove their in-
nocence. Scholars spoke of “the dark fi gure of innocence,” because so
little was known about wrongful convictions.18

DNA exonerations have changed the face of criminal justice in the
United States by revealing that wrongful convictions do occur and, in
the pro cess, altering how judges, lawyers, legislators, the public, and
scholars perceive the system’s accuracy. This sea change came about
because of the hard work of visionary lawyers, journalists, and students
who suspected that the criminal justice system was not as infallible as
many believed. Barry Scheck and Peter Neufeld, two well- known de-
fense lawyers, founded the pioneering Innocence Project at Cardozo
Law School in the early 1990s, which helped to free many of the fi rst 250
exonerees. I fi rst met several of these exonerees when, as a rookie lawyer,
I worked for Scheck and Neufeld representing innocent people who
sued to get compensation for their years behind bars. Over the years,
lawyers, journalists, and others established an “innocence network,”
including clinics at dozens of law schools, designed to locate innocence
cases. Today, DNA exonerations have occurred throughout the United
States, in thirty- three states and the District of Columbia.19 Public dis-
trust of the criminal justice system has increased, and pop u lar tele vi sion
shows, books, movies, and plays have dramatized the stories of the wrong-
fully convicted.20 We now know that the “ghost of the innocent man”
spoken of by Judge Learned Hand is no “unreal dream,” but a nightmar-
ish reality.

What Went Wrong

What we have not been able to know, however, is whether there are sys-
temic failures that cause wrongful convictions. Now that there have been
so many DNA exonerations, we have a large body of errors to study. Did
the fi rst 250 DNA exonerations result from unfortunate but nevertheless
unusual circumstances? Or were these errors the result of entrenched
practices that criminal courts rely upon every day? Are there similarities
among these exonerees’ cases? What can we learn from them?

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Introduction � 7

This book is the fi rst to answer these questions by taking an in- depth
look at what happened to these innocent people. Collecting the raw
materials was a challenge. Although scholars have surveyed jurors and
judges using detailed questionnaires, no one has studied a set of crimi-
nal trial transcripts to assess what evidence was presented, much less
studied the criminal trials of the exonerated.21 One reason is the diffi culty
and expense of locating trial rec ords. These voluminous rec ords must
often be pulled from storage in court archives or requested from the
court reporters. I was able to overcome these diffi culties with the help of
numerous librarians and research assistants. For each of the fi rst 250
DNA exonerees, I contacted defense lawyers, court clerks, court report-
ers, prosecutors, and innocence projects around the country. I located
documents ranging from confession statements to judicial opinions and,
most important, transcripts of exonerees’ original trials. I obtained 88%
of their trial transcripts, or 207 of the 234 exonerees convicted at a trial.22
I also obtained hearing transcripts and other rec ords in thirteen of six-
teen cases where exonerees had no trial but instead pleaded guilty. In the
remaining cases, the rec ords had been sealed, destroyed, or lost.

When I began to assemble this wealth of information, I had a single
goal: to fi nd out what went wrong. When I analyzed the trial rec ords, I
found that the exonerees’ cases were not idiosyncratic. The same prob-
lems occurred again and again. Like Ronald Jones, almost all of the other
exonerees who falsely confessed had contaminated confession statements.
Most other forensic analysis at these trials off ered invalid and fl awed con-
clusions. As troubling as it was, Ronald Jones’s case looked typical among
these exonerees: his case fi t a pattern of corrupted evidence, shoddy inves-
tigative practices, unsound science, and poor lawyering.

These trials call into question the “unparalleled protections against
convicting the innocent” that the Constitution supposedly aff ords. The
system places great trust in the jury as the fact fi nder. When the Supreme
Court declined to recognize a right under the Constitution for convicts to
claim their innocence, it reasoned, “the trial is the paramount event for
determining the guilt or innocence of the defendant.”23 Yet at a trial, few
criminal procedure rules try to ensure that the jury hears accurate evi-
dence. To be sure, celebrated constitutional rights, such as the require-
ment that jurors fi nd guilt beyond a reasonable doubt and that indigent

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

8 Introduction

defendants receive lawyers, provide crucial bulwarks against miscarriages
of justice. But those rights and a welter of others the Court has recognized,
like the Miranda warnings, the exclusionary rule, and the right to con-
front witnesses, are procedural rules that the State must follow to prevent
a conviction from being overturned. Few rules, however, regulate accu-
racy rather than procedures. Such matters are typically committed to the
discretion of the trial judge.

Exonerations provide new insights into how criminal prosecutions
can go wrong. We do not know, and cannot ever know, how many other
innocent people have languished behind bars. Yet there is no reason to
think that these 250 are the only ones who were wrongly convicted be-
cause of the same types of errors by police, prosecutors, defense lawyers,
judges, jurors, and forensic scientists. The same unsound but routine
methods may have contaminated countless other confessions, eyewitness
identifi cations, forensic analysis, in for mant testimony, and defenses. Each
chapter in this book poses a diff erent question to analyze an aspect of
what went wrong in these exonerees’ cases.

Why did innocent people confess in such detail to crimes they did not
commit? Chapter 2 begins with the case of Jeff rey Deskovic, who, like
Ronald Jones, supposedly said much more than “I did it.” Forty of these
250 exonerees (16%) confessed to crimes they did not commit. I expected
to fi nd some false confessions among these exonerees, since several such
cases have become well known and because psychologists have studied
how police pressure can coerce a false confession. I was surprised to dis-
cover, however, that all but two of those exonerees reportedly confessed
to details about the crime that only the killer or rapist could have known.
Those specifi c facts must have been improperly disclosed to exonerees,
most likely by police. In Deskovic’s case, it is not surprising that he suc-
cumbed to police pressure. He was sixteen years old and had to be com-
mitted to a mental hospital after a breakdown during his last interroga-
tion session. More than half of the exonerees who falsely confessed were
juveniles or mentally disabled. Criminal procedure played a role: the Su-
preme Court long ago abandoned reviewing the reliability of confessions,
focusing instead on whether police coerced the admission of guilt.

Why did victims and other eyewitnesses testify that they were certain that
they saw innocent people commit crimes? Chapter 3 begins with the case of

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Introduction � 9

Habib Abdal, who, like 76% (190) of these 250 exonerees, was misidenti-
fi ed by an eyewitness. At trial, the victim pointed to Abdal and said she
was absolutely certain that he was the rapist— but she was wrong. I ex-
pected eyewitnesses to have been confi dent by the time of the trial. I never
expected to fi nd that in most of these cases, eyewitnesses admitted at trial
that earlier they were not so certain and that police had used unreliable
and suggestive procedures. In Abdal’s case, police showed the victim a
single photograph of him and repeatedly pressured her to identify him.
Initially the victim described a culprit that looked very diff erent. When
she fi rst saw Abdal at the police station, she said he was not the rapist.
Most other witnesses had similarly expressed uncertainty, or identifi ed
another person, or described an attacker who looked very diff erent. Abdal
was one of seventy- four black exonerees who were misidentifi ed by white
eyewitnesses— studies have long shown how cross- racial identifi cations
are especially error prone. Once again, criminal procedure off ered little
guidance; the Supreme Court de cades ago adopted a very fl exible “reli-
ability” standard that allows police to use highly suggestive procedures.

Why didn’t forensic science show at trial that these people were inno-
cent? Chapter 4 begins with the case of Gary Dotson. At Dotson’s trial,
an analyst gave the jury inaccurate statistics about blood types, just like
in Ronald Jones’s case. He also gave exaggerated testimony about hair
evidence. Most exonerees had forensic evidence analysis in their cases.
Although cutting- edge DNA technology later set them free, at trial the
forensics were fl awed. Reading these trials, I was confronted with a pa-
rade of invalid forensics by 61%, or 93, of 153 analysts called by the pros-
ecution. Analysts concocted false probabilities. Analysts asserted, with
no scientifi c support, that bite marks or hairs came from the defendant.
Analysts used unreliable techniques. Nor is DNA testing infallible;
DNA testing errors contributed to the wrongful convictions of three ex-
onerees. All of this invalid and unreliable forensic testimony was one-
sided. Only a handful of defense lawyers obtained experts to counter the
fl awed forensics. The Supreme Court now ostensibly requires that ex-
pert testimony be both reliable and valid, but judges to this day have
largely disregarded those requirements in criminal cases.

Why did in for mants testify against innocent people? Chapter 5 begins
with the case of David Gray, in which his jail cellmate became a star

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

10 Introduction

witness for the prosecution. In for mants played an unsavory supporting
role in 21% of these trials. Most were jail house in for mants, and they did
not just lie, as I expected. Instead, they did something more pernicious—
they claimed to have overheard telling details about the crime. In Gray’s
case, the prosecutor admitted that the in for mant was a known liar but
argued that the jury should still believe him. After all, the in for mant
claimed that Gray told him detailed facts about the crime: he ripped the
victim’s phone off the wall; he wore fl ashy wine- colored shoes. In for mants
then denied they received any deals, although prosecutors in fact rewarded
many with light sentences for telling such lies. Judges do not meaningfully
limit or regulate the testimony of these untrustworthy in for mants.

Why didn’t defense lawyers prevent convictions of their innocent cli-
ents? Chapter 6 begins with Earl Washington Jr.’s trial, which lasted just
two days— even though it was a death penalty case. Imagining what I
would do if I were innocent and on trial, I pictured these innocent peo-
ple vigorously litigating their innocence. Washington, like more than
half of these exonerees, took the stand to protest his innocence. Aside
from their own words, though, all that most could do was present a weak
alibi. Their alibis were usually family members who could be accused of
bias. Washington’s lawyer had no experience with death penalty cases.
He called just two witnesses and presented only forty minutes of evi-
dence. In his brief closing statement, he told the jury to do their job to
decide if Washington did it. Afterward, the jury took only fi fty minutes
to convict an innocent man. De cades later, lawyers also uncovered how
the police and prosecutors had concealed evidence of Washington’s in-
nocence from the defense. Although the Supreme Court has emphasized
how the “crucible of meaningful adversarial testing” protects defen-
dants, judges have long failed to ensure that defendants receive adequate
repre sen ta tion or access to evidence of their innocence at trial.24

Why didn’t appeals or habeas corpus review set innocent people free?
Chapter 7 turns to the lengthy appeals and habeas proceedings that
these innocent people pursued. One might expect that judges would
look for serious mistakes like the ones in these exonerees’ cases. How-
ever, judges only reluctantly review the evidence after a conviction. After
all, most constitutional rights relate to procedure and not accuracy. Most
exonerees did not try to challenge fl awed trial evidence, and when they

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Introduction � 11

did they almost always failed. Judges typically refused to grant a new
trial, and some were so sure that these people were guilty that they
called the evidence of guilt “overwhelming.”

Why did it take so long for innocent people to be exonerated? Chapter 8
begins with the case of Frank Lee Smith, who died of cancer on death
row after fi ghting for years to obtain DNA testing, which exonerated him
only posthumously. Most of the exonerees fought for years to obtain ac-
cess to DNA testing and an exoneration. The word exoneration refers to
an offi cial decision to reverse a conviction based on new evidence of in-
nocence. An exoneration occurs if the judge, after hearing the new evi-
dence of innocence, vacates the conviction and there is no retrial, or
there is an acquittal at a new trial, or if a governor grants a pardon.25 The
250 exonerees spent an average of thirteen years in prison. It took longer,
an average of fi fteen years, for them to be exonerated. Judges and prose-
cutors sometimes opposed requests for DNA testing. Judges sometimes
initially refused to exonerate these people even after DNA tests provided
powerful proof of their innocence.

Why didn’t criminal justice systems respond to exonerations? Chap-
ter 9 begins by telling how North Carolina responded to exonerations by
creating an Innocence Commission to review potential innocence cases
and by passing laws to reform interrogations and eyewitness identifi ca-
tion procedures. But most jurisdictions have not responded at all, much
less so comprehensively. In this last chapter, I revisit the questions raised
in each of the prior chapters and suggest a set of solutions. State and lo-
cal law enforcement are increasingly adopting reforms, such as recorded
interrogations, sound eyewitness identifi cation procedures, and audits
of forensic work, each of which can help to prevent future wrongful
convictions.

The Tip of an Iceberg

Do cases like Ronald Jones’s tell us anything about the vast numbers of
criminal cases pro cessed in the United States each year? These 250 exon-
erations are just the tip of an iceberg. The submerged bulk of that iceberg
lurks ominously out of view. The most crucial question about these exon-
erations cannot be answered. We do not know and we cannot know how

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

12 Introduction

many other innocent people languish in our prisons. They remain invis-
ible. One of the most haunting features of these exonerations is that so
many were discovered by chance. Most convicts who seek postconviction
DNA testing cannot get it. Some jurisdictions still deny convicts access to
DNA testing that could prove innocence. In our fragmented criminal
justice system, exonerations hinge on cooperation of local police, prose-
cutors, and judges. Most police in the 1980s did not routinely save evi-
dence after a conviction. In Dallas, Texas, where biological evidence was
preserved, we have a large set of nineteen DNA exonerations.26 In nearby
Houston, Texas, the county clerk destroyed evidence whenever its store-
rooms got full, so there have been just six DNA exonerations, despite a
major scandal that led to the shutdown of the Houston crime lab.27 Fur-
ther, for most crimes, there is no useful evidence collected at a crime
scene that can be tested using DNA. If DNA is a “truth machine,” it tells
us only about a sliver of very serious convictions, most for rape, chiefl y
from the 1980s.28

For all of those reasons, we will never know how large the iceberg is—
but we have good reasons to worry that intolerable numbers of innocent
people have been convicted. A federal inquiry conducted in the mid-
1990s, when police fi rst began to send their samples for DNA testing,
found that 25% of “primary suspects” were cleared by DNA before any
trial.29

The trial rec ords in these cases give us every reason to believe that
most of the police, prosecutors, forensic analysts, defense lawyers,
judges, and jurors acted in good faith. Few likely realized they were tar-
geting the innocent. They may have suff ered from the everyday phenom-
enon of cognitive bias, meaning that they unconsciously discounted evi-
dence of innocence because it was inconsistent with their prior view of
the case, or they were motivated to think of themselves as people who
pursued only the guilty. That makes these cases all the more troubling.
If there is no reason to think that anyone acted diff erently in these cases
as compared to others, then similar errors may have convicted countless
other innocent people and led to the guilty going free.

Rather than blame individual people for their mistakes in these cases,
my study of these wrongful convictions suggests that there are systemic
problems calling for reforms. We should safeguard the accuracy of cru-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Introduction � 13

cial pieces of evidence like eyewitness identifi cations, confessions, fo-
rensics, and in for mant testimony. These 250 exonerations open a unique
window on the underside of our criminal justice system. The chapters
that follow will each take a hard look through that window to better un-
derstand why criminal prosecutions can go wrong— and how we can avoid
convicting the innocent.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

A fifteen- year- old girl’s body was found by a dirt path in a
heavily wooded area of a park in downtown Peekskill, New York,

near where she lived. She was attacked while walking in the park, listen-
ing to a New Kids on the Block cassette tape. She had been severely
beaten with a blunt object. The medical examiner concluded she was
raped.

Jeff rey Deskovic was seventeen years old and he was a classmate of
the victim— both were sophomores at Peekskill High School. Deskovic
attended her wake and funeral and appeared distraught. He told the po-
lice he was eager to help them solve the crime.1 The fi rst time they took
him to police headquarters, a detective recalled asking him if he wanted
to talk to his mother, and he said no. He was taken to an offi ce and given
the warnings required by the Supreme Court’s decision in Miranda v.
Arizona. Reading from a card, the detective read, “You have the right to
remain silent and refuse to answer any questions. Anything you do say
may be used against you in a court of law . . . You have a right to a lawyer
before speaking to me and to remain silent.” The detective asked, “Do
you understand each of these rights I have explained to you?” Deskovic
said he did and then signed the card to indicate his willingness to an-
swer questions without an attorney.2

c h a p t e r 2

Contaminated Confessions

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 15

Deskovic spoke to the police many times over several weeks. He was
interrogated for hours over multiple sessions. In one session, “police had
a tape recorder, but turned it on and off , only recording 35 minutes.”
The tape recorder was on when Deskovic was read his Miranda rights,
but the tape was not running when the detectives aggressively confronted
him.3

Deskovic never said that he had raped or murdered the victim. But he
reportedly off ered other details. The detective explained that “he had
begun telling me things that he could not possibly have known had he
not, in fact, been the person responsible.”4 During another discussion,
Deskovic “supposedly drew an accurate diagram of the crime scene,”
depicting details from “three discrete crime scenes” which were not ever
made public. He had marked the map with notations such as “Struggle,”
“Raped there,” and “Body found.”5 He had also told the detectives that
he “sometimes hears voices.”6

Particularly telling was the fi nal interrogation, which began with a
polygraph examination. The detective explained at trial that he did not
conduct the polygraph examination to actually test Deskovic’s truthful-
ness, but instead to “[g]et the confession.”7 After the polygraph exam, the
detective told Deskovic that he had failed the test, in an eff ort to place
more pressure on him. The interrogation ended with Deskovic curled up
on the fl oor of the polygraph room in a fetal position and crying.

Before collapsing, he reportedly said that although “when he fi rst be-
gan coming in he thought he was helping to get the guy responsible,” he
had gradually realized “that the guy was him.”8 He reportedly volun-
teered that the victim suff ered a blow to the temple and that he tore her
clothes. He said he struggled with her and held his hand over her mouth,
maybe for “a little too long.” He said that he ripped off her bra. The vic-
tim’s torn bra had been found at the crime scene. He knew that the vic-
tim had taken photos on her way to the park and that she had dropped
her keys. Finally, Deskovic reportedly said he had “hit her in the back of
the head with a Gatorade bottle that was lying on the path.” After hear-
ing this, the police later claimed that they returned to the crime scene,
where they found a Gatorade bottle cap.9

The day after his last interrogation and arrest, Deskovic was declared
“suicidal.” After he was hospitalized for months, the judge ordered him

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

16 Contaminated Confessions

examined and found him competent to stand trial.10 Before trial, the
judge also held a hearing to evaluate whether the confession statements
were “voluntary” as required by U.S. Supreme Court and state court
rulings. Police had used methods shifting from “passive” tactics eliciting
information to “active” confrontations. The judge found these tactics
lawful and the confession voluntary and therefore admissible, so the jury
could hear about it at trial.11

Yet at trial the State had a problem. Although the State’s theory was
that Deskovic had raped and murdered the victim alone, the FBI labora-
tory had conducted DNA tests before trial. Deskovic was not the source
of the semen found in swabs taken from the victim’s body. At trial, the
district attorney asked the jury to simply ignore that powerful scientifi c
evidence. The district attorney speculated that perhaps the victim was
“sexually active” and “romantically linked to somebody else” with whom
she engaged in sexual relations shortly before her rape and murder. After
all, “she grew up in the eighties.”12 There was no evidence she was sexu-
ally active and no investigation was conducted to test this conjecture, ei-
ther by the prosecution or by the defense.

The trial transcripts highlight how central Deskovic’s alleged admis-
sions were to the State’s case. The confession was the only evidence con-
necting Deskovic to the crime. The district attorney emphasized in clos-
ing argument that Deskovic had been given the Miranda warnings
“numerous” times. He said there were “no deceptive tactics utilized,”
“no threats of violence. Handcuff s placed on him? Guns drawn? Beat-
ings? None of this.”13 The district attorney then turned to the accuracy of
Deskovic’s statements, noting that after he told police about the Gatorade
bottle “it was found there” (although just a cap was found) and that this
was a heavy weapon, and “not a small little bottle.”14 The detectives “did
not disclose any of their observations or any of the evidence they recov-
ered from Jeff rey nor for that matter, to anyone else they interviewed.”
They kept their work nonpublic “for the simple reason . . . that [if a sus-
pect] revealed certain intimate details that only the true killer would
know, having said those, and be arrested could not then say, ‘Hey, they
were fed to me by the police, I heard them as rumors, I used my common
sense, and it’s simply theories.’ ”

The district attorney told the jury to reject the implication that Des-
kovic was fed facts, stating, “Ladies and gentlemen, it doesn’t wash in

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 17

this case, it just doesn’t wash.” He reviewed each of the facts, asking the
jury, “How would Jeff rey know this” unless he was the killer?15 The
facts that the detectives said Deskovic had volunteered trumped the
hard science— the DNA evidence. The jury convicted Deskovic of rape
and murder.

At sentencing, Deskovic told the judge, “I didn’t do anything. I’ve al-
ready had a year of my life taken from me for something I didn’t do.” The
judge seemed troubled: he agreed that “maybe [he was] innocent” but
the “jury ha[d] spoken.”16 The judge imposed the minimum sentences
of fi fteen years to life. Deskovic then fi led an appeal, which the appellate
court denied, pointing to “overwhelming evidence of the defendant’s
guilt in the form of the defendant’s own multiple inculpatory state-
ments.”17 Deskovic ultimately served sixteen years in prison.

In 2006, new DNA testing again excluded Deskovic, but also did
more: the tests matched the profi le of Steven Cunningham, a murder
convict who subsequently confessed and pleaded guilty. He had com-
mitted another murder while Deskovic was in prison. Cunningham later
said that if he had known that another person had been wrongly con-
victed, he would have come forward, commenting, “Nobody should
have to do time for something they didn’t do.”18

Now that we know Deskovic is innocent, how do we explain how he
reportedly knew so many “intimate details” about the murder? Police
could have made up the entire set of statements, but there is no evidence
that occurred in Deskovic’s case. Instead, police likely contaminated the
confession by leaking facts or feeding them to him. The district attor-
ney’s postexoneration inquiry concluded:

Given Deskovic’s innocence, two scenarios are possible: either the

police (deliberately or inadvertently) communicated this informa-

tion directly to Deskovic or their questioning at the high school and

elsewhere caused this supposedly secret information to be widely

known throughout the community.

What are the chances that an innocent person, and a frightened youth at
that, could have possibly guessed so many details? Given the level of
specifi city said to have been provided by Deskovic, it is most likely that
offi cers told facts to him. Whether the offi cers did so intentionally or

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

18 Contaminated Confessions

inadvertently, in hindsight we know that they provided an inaccurate
account of who said what during the interrogations.

Deskovic has since commented, “Believing in the criminal justice
system and being fearful for myself, I told them what they wanted to
hear.”19 Psychologists call such a confession a “coerced- compliant” con-
fession. In a coerced- compliant confession, the pressure police apply
during the interrogation may not be illegal, and it may come from tactics
that judges have approved. The suspect complies with law enforcement
pressure and confesses chiefl y to obtain a gain, such as “being allowed
to go home, bringing a lengthy interrogation to an end, or avoiding
physical injury.”20 Deskovic is now suing for civil rights violations,
which his complaint calls a “veritable perfect storm of misconduct.” His
lawsuit alleges that police disclosed facts to him.21

The Puzzle of False Confessions

False confessions present a puzzle: why do innocent people confess in
such detail to crimes they did not commit? For de cades, commentators
doubted that a suspect would falsely confess. For example, Professor
John Henry Wigmore wrote in his 1923 evidence treatise that false con-
fessions were “scarcely conceivable” and “of the rarest occurrence.”22
This traditional understanding of confessions changed dramatically in
recent years, as DNA testing exonerated people who had falsely con-
fessed. We now know that false confessions may not be “of the rarest
occurrence.” While we do not know how often false confessions occur,
there is a new awareness among scholars, legislators, judges, prosecu-
tors, police departments, and the public that innocent people can falsely
confess, often due to psychological pressure placed upon them during
police interrogations.23 Studies increasingly examine the psychological
techniques that can cause people to falsely confess.24

I take a diff erent approach, by focusing not on the psychology of false
confessions but on their content. Forty of the fi rst 250 DNA exoneration
cases (16%) involved a false confession. I wondered what people who we
now know are innocent reportedly said when they confessed. I used the
trial transcripts to fi nd out what was said during interrogations and how
the confessions were described and litigated at trial. When I began this

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 19

pro cess, I expected to see confessions without much information. An
innocent person might be able to say “I did it,” but obviously could not
say what exactly he did, since he was not there at the crime scene. I knew
it was possible that a confession could be contaminated if police prompted
the suspect on how the crime happened, and I thought that I might fi nd a
handful of cases where this had happened. Scholars have noted that “on
occasion, police are suspected of feeding details of a crime to a compliant
suspect,” and have described several well- known examples.25 However,
no one has previously studied a set of known false confessions to analyze
what facts must have been disclosed to the innocent suspects.26

To my great surprise, when I analyzed these case materials I found
that not just a few, but almost all, of these exonerees’ confessions were
contaminated. I sought out trial materials and court rec ords for all forty
exonerees who falsely confessed and I was able to obtain them for all
forty.27 I also located the text of written confession statements for most of
these exonerees. All of those rec ords provided a rich source of mate-
rial.28 All but two of the forty exonerees studied told police much more
than just “I did it.” Instead, police said that these innocent people gave
rich, detailed, and accurate information about the crime, including what
police described as “inside information” that only the true culprit could
have known.29

Contaminated False Confessions

Confessions are incredibly powerful evidence at a trial. As the Supreme
Court has put it, “A confession is like no other evidence. Indeed, ‘the
defendant’s own confession is probably the most probative and damag-
ing evidence that can be admitted against him . . . confessions have pro-
found impact on the jury, so much so that we may justifi ably doubt its
ability to put them out of mind even if told to do so.’ ”30 These confes-
sions were particularly powerful because of their detail.

The Deskovic case shows how false confessions do not happen quickly
or by happenstance. As law professor Samuel Gross puts it, “false con-
fessions don’t come cheap.”31 Confession statements are carefully con-
structed during an interrogation and then reconstructed during any
criminal trial that follows. The Constitution requires that an admission

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

20 Contaminated Confessions

of guilt was “voluntary” under the circumstances. However, the power
of a confession derives from the lengthy narrative describing what hap-
pened, and not from the bare statement that “I did it.” During the
“confession- making phase” of an interrogation, police ask about details
regarding how the crime occurred.32 Police are trained to carefully test
the suspect’s knowledge of the crime, by assessing whether the suspect
can freely volunteer specifi c details that only the true culprit could know.

When police interrogated them, each exoneree supposedly gave self-
incriminating statements and admissions of guilt to police, though some,
like Deskovic, did not confess to all of the charged acts. In thirty- eight
cases, the detectives testifi ed that these defendants did far more than say
“I did it.” According to the detectives, the suspects had “guilty” or “in-
side” knowledge. For example, in exoneree Robert Miller’s case, the
prosecutor said, “He supplied detail after detail after detail after detail.
And details that only but the killer could have known.”33 Only two exon-
erees relayed no specifi c information in their confessions. Travis Hayes
could not say much of anything about the crime, but he was still con-
victed, even though the jury heard that DNA testing on the clothes the
attacker wore had excluded him and his codefendant. The second, Fred-
die Peacock, was mentally ill, and he just told his interrogator, “I did it,
I did it,” but he was unable to provide any details at all about what he
supposedly did, even after the detective told Peacock some of those de-
tails, including who the victim was.

In all of the other 95% of the false confessions studied (thirty- eight of
forty cases), detectives claimed that the suspects volunteered key details
about the crime, including facts that matched the crime scene evidence,
or scientifi c evidence, or accounts by the victim. Police offi cers went far-
ther and also claimed they assiduously avoided contaminating the con-
fession by not asking leading questions, but rather allowing the suspect
to volunteer each of the crucial facts.

Although defense lawyers tried to challenge almost all of these confes-
sions, judges found the confessions to have been voluntarily made and
appropriate for the jury to hear. The nonpublic facts contained in con-
fession statements then became the centerpiece of the State’s case. The
facts were typically the focus of the State’s closing arguments to the jury.
Even after DNA testing excluded these people, judges sometimes ini-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 21

tially denied relief, relying on the seeming reliability of these confes-
sions. These false confessions were so persuasive, detailed, and believ-
able that judges repeatedly upheld the convictions during appeals and
habeas review. After years passed, these innocent people had no option
but to seek the DNA testing that fi nally proved their confessions false.

These forty false confessions are unique and unusual. These forty
examples cannot tell us how many suspects falsely confess. Nor are these
cases representative of other confessions, much less other false confes-
sions.34 Many who confess do so in short interrogations.35 In contrast,
almost all of these interrogations were prolonged aff airs, lasting many
hours or even days. Fourteen of these exonerees were mentally retarded,
three were mentally ill, and thirteen were juveniles. Just as in Deskovic’s
case, in 65% of the false confession cases (twenty- six of forty cases),
DNA testing not only excluded the exoneree but also confi rmed the guilt
of others, some of whom themselves later confessed and pleaded guilty.36
The forty false confessions were also concentrated in cases involving
both a rape and a murder. Of the forty who falsely confessed, twenty- fi ve
cases were rape- murder cases, three were murder cases, while twelve
were rape cases.37 Seven exonerees who falsely confessed were sentenced
to death. In contrast, most DNA exonerees, 68%, were not convicted of
murder, but were convicted of rape (171 of 250 cases). As chapters that
follow will describe, most of the rape cases involved a victim who identi-
fi ed the assailant. Cases with false confessions were typically murder
cases where there was no eyewitness, police had few leads, and it was a
higher priority for police to secure a confession.

All of those features make these cases highly atypical, but also espe-
cially interesting. They came to light not because of a challenge to the
confession, but due to the in de pen dent development of DNA technology.
These false confessions also allow one to study the problem of confession
contamination, since one can be confi dent in retrospect that these inno-
cent people could not have themselves known the specifi c details of
crimes they did not commit.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

22 Contaminated Confessions

Law Enforcement Practices

How are police trained to interrogate suspects, and what went wrong in
these cases? Well- trained police receive instruction on the complex psy-
chological tactics used to elicit confessions from suspects. Just as in the
Deskovic case, a suspect who is identifi ed as a potentially important
source of information is asked to participate in an interview. The sus-
pect is placed in a small interrogation room and given the Miranda
warnings. In the initial stages of an interrogation, detectives try to make
the suspect feel at ease. They ask open- ended questions. Many suspects
readily confess at this point. Should that not occur, police may directly
confront the suspect with their belief that he committed the crime. If
that fails, police may use more elaborate and heavy- handed tactics.

The leading manual on police interrogations, originally written by
Fred Inbau and John Reid and now in its fourth edition, provides a set of
protocols branded the “Reid Technique.” The Reid Technique instructs
police to apply pressure but at the same time suggest that the suspect has
something to gain by confessing. Those techniques are called “maximi-
zation” and “minimization.” Police engage in storytelling and off er the
suspect a series of alternative narratives. They try to get the suspect to
initially agree to having committed legally excusable or less reprehensi-
ble acts. For example, using one common alternative narrative, police
may try to get the suspect to admit having attacked the victim, but only
in self- defense.38 Police may also imply that they might grant leniency if
the suspect confesses or impose consequences if the suspect does not.
Such techniques were used in several of these exonerees’ interrogations,
ranging from “Mutt and Jeff ” or “Good Cop, Bad Cop” techniques, to
going so far as to threaten the death penalty.

Police can also sometimes use deceptions or fabrications by telling
suspects, falsely, that forensic evidence proved their guilt or a collabora-
tor has confessed and implicated them.39 Police used a “false evidence
ploy” in several exonerees’ cases. David Vasquez confessed after he was
told that his fi ngerprints were found at the scene.40 Similarly, in the Cen-
tral Park Jogger case, the detective told Yusef Salaam: “we have fi nger-
prints on the jogger’s pants or her jogging shorts, they’re satin, they’re a
very smooth surface. We have been able to get fi ngerprints off of them.”

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 23

Telling suspects that they failed a polygraph, perhaps falsely, may have
played a role in eight of these false confessions, in which polygraph tests
were used.41 As in Jeff rey Deskovic’s case, judges have found such psy-
chological tactics permissible.42

Although they can use a range of psychologically coercive and decep-
tive tactics, police have long been trained to never contaminate a confes-
sion by feeding or leaking crucial facts. The Reid manual is emphatic on
this point. During the confession- making stage of an interrogation, po-
lice may engage the suspect in a pro cess of joint storytelling, but the nar-
ratives they employ to develop what happened have important limits. If
the suspect is told how the crime happened, then the police cannot ever
again properly test the suspect’s actual knowledge of how the crime hap-
pened. Police have long known that suspects may admit to crimes that
they did not commit, for a range of reasons, including mental illness,
desire for attention, and to protect loved ones. For that reason, police
are trained to ask open questions, like “What happened next?” Leading
questions are not to be asked, at least not as to crucial corroborated de-
tails. The Reid manual advises that “[w]hat should be sought particu-
larly are facts that would only be known by the guilty person.” Further,
police are trained not to leak such facts. Police black out key facts so that
the press and the public do not learn those facts during an investigation.
By avoiding contamination of the confession, the offi cer can at trial “con-
fi dently refute” any suggestion that those facts were fed to the suspect.43

Corroborated and Nonpublic Facts

If police are trained to avoid disclosing key facts, what happened in these
exonerees’ cases? In all but two of them, police claimed that the defen-
dant had off ered a litany of details, which we now know these innocent
people could not have known. The police did “confi dently refute” at trial
that they disclosed those facts, making the confessions appear unassail-
able. Police told the jury that the defendant freely off ered information that
only the perpetrator could have known. If police had told the jury that the
defendant had reluctantly nodded “yes” to a series of leading questions,
then the confession would not be particularly credible; it would be obvi-
ous that it had been fed by the police.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

24 Contaminated Confessions

Cases like the Deskovic case, involving unusual, specifi c, or numer-
ous details, raise troubling questions. In such cases, it is highly unlikely
that a person could have made such an elaborate and detailed “unlucky
guess.”44 Similarly, in exoneree Dennis Brown’s case, the sergeant who
interrogated him testifi ed as follows:

Q. You’re stating under oath you did not know what the victim had

on that night, is that correct? You did not know the color of the

couch?

A. No, sir.

Q. You did not know which arm she was grabbed by?

A. No sir, I did not.

Q. And that the defendant confessed to the rape of [the victim],

correct?

A. Yes.

Q. And he gave you specifi cs as to that rape?

A. Yes, sir.

Q. And he told you about the house?

A. Yes, sir.

Q. And he told you what color the couch was?

A. Yes, sir.

Q. And he told you how he committed the rape?

A. Yes, sir.45

The sergeant not only testifi ed that Brown knew details regarding the
crime, down to the color of the victim’s couch, but that the sergeant
himself did not know those crime scene details. The clear implication
was not just that he did not feed those facts to Brown, but that it was
impossible for him to have fed those facts to Brown. Now that we know
with the benefi t of postconviction DNA testing that Brown is innocent,
it is quite likely that police did disclose those specifi c facts. It is un-
likely that innocent suspects could reconstruct the crimes out of whole
cloth from their own imagination. It is unlikely that police improperly
leaked all of the crucial details of the crimes to the public, which the
innocent suspects heard and then accurately recounted during their
interrogations.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 25

Douglas Warney’s case provides still another example of a confession
involving a litany of detailed nonpublic facts concerning the crime. War-
ney was said to have told police numerous facts, including that the vic-
tim was wearing a nightshirt, the victim was cooking chicken, the victim
was missing money from his wallet, the murder weapon was a twelve-
inch- long serrated knife that was kept in the kitchen, the victim was
stabbed multiple times, the victim owned a pink ring and gold cross, a
tissue used as a ban dage was covered with blood, and there was a porno-
graphic tape in the victim’s tele vi sion.46 When asked under oath, “did
you suggest any answers to him,” the sergeant said that he did not. The
prosecutor argued in closing that the reliability of Warney’s confession
was corroborated by each of these facts:

The Defendant says he’s cooking dinner, and he’s par tic u lar about

it, cooking chicken . . . Now, who could possibly know these things

if you hadn’t been inside that house, inside the kitchen? You heard

the Defendant say that he took money . . . You know the wallet was

found upstairs, empty, near the closet . . . You will see photographs

of it . . . You heard the Defendant say that he stabbed [the victim]

with a knife taken from the kitchen . . . that was the knife that they

kept in the house. Where did they keep it? They kept it in a drawer

under the crockpot where the chicken was cooking. Now, who

would know the chicken was cooking? A person who got that knife

and used it against [the victim], the killer. The Defendant described

the knife as being twelve inches, with ridges . . . it was thirteen

inches with the serrated blade.47

Notice how the prosecutor speaks of how the jury “heard the Defendant
say” this and that fact. The jury never heard such a thing from Warney.
They heard an offi cer read a typed confession statement that Warney had
signed. They heard the sergeant describe the interrogation pro cess. When
the jury “heard” Warney testify at trial, he recanted his confession state-
ment and proclaimed his innocence.

Years later, Warney was exonerated by DNA testing (the tests also in-
culpated another man who subsequently confessed). When interrogated
he had “a history of delusions, an eighth- grade education and advanced

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

26 Contaminated Confessions

AIDS.”48 Warney has since maintained that the sergeant had told him all
of those telling details described at trial.

Not only can individual confessions be contaminated, but false confes-
sions can have a “multiplying” eff ect, in which additional innocent peo-
ple are drawn into an investigation, and a group of confessions is con-
taminated. Seventeen exonerees not only falsely confessed to their own
guilt, but they falsely implicated others who did not confess, eleven of
whom were later also exonerated by postconviction DNA testing.49 One
example is the “Beatrice Six” case. A gruesome rape and murder of an
el der ly woman had remained unsolved in the small town of Beatrice, Ne-
braska. Eventually, four people confessed to participating in the murder.
All four pleaded guilty, along with a fi fth who pleaded guilty but did not
confess.

A sixth person, Joseph White, was the only one in the group who re-
fused to plead guilty; he requested an attorney during his interrogation
and did not confess. In return for a lighter sentence, three of those who
pleaded guilty testifi ed against White at his trial: James Dean, Debra
Shelden, and Ada JoAnn Taylor. Each openly admitted that police had
suggested facts to them and that before speaking to police, they could
not remember much of what had occurred the night of the crime. The
contamination of the confessions was no secret. Taylor testifi ed as fol-
lows at a deposition, which was read into the record at trial:

Q. Can you actually separate today what you remember from the

night this happened and what was suggested to you to help you

remember what happened that night?

A. No. It would almost be impossible to separate.

Q. So what ever statements you have made recently, I take it, are not

from your memory but from suggestions that have helped you

remember?

A. There has been parts from my memory as well as the suggestions.

Q. Tell me what parts you actually remember that you don’t have

that you didn’t have suggested to you.

A. Oh, God.

Q. Is there anything?

A. Not that I can remember off hand . . . 50

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 27

At trial, Taylor said that police “somewhat” suggested facts to her, and
helped her to remember much, but not “all of the information.” She ex-
plained: “Well, I have a tendency to believe all offi cers.” Taylor admitted
that police had showed her a video of the crime scene and had given her
the statements of the other defendants to read.51 She also testifi ed that
she was diagnosed with a “personality disorder” and had problems with
her memory, though she also took credit for having some mental telepa-
thy capabilities.52

Taylor admitted that police told her par tic u lar facts. For example, she
said that police suggested a particularly idiosyncratic fact, an explanation
for the presence of a ripped half of a fi ve- dollar bill at the crime scene. She
testifi ed that Joseph White had “a trick that he does with a $5 bill” where
he would rip it in half, and recalled asking him after the murder what he
had ripped, and he had said a “fi ve,” meaning a fi ve- dollar bill. When
asked to explain the trick, she said, “I’ve never really understood it. I
know he pulls a $5 bill out and he does something with it and he ends up
with a ripped $5 bill. And he usually tosses part of it away.” However, on
cross- examination, she admitted that the deputy sheriff originally told
her about the fi ve- dollar bill; when he fi rst asked about the money trick,
she told him that White would make pictures with the money; and fi nally
the deputy told her that White would tear the fi ve- dollar bill.53

Although it was clear at trial that the confessions were contaminated,
the jury chose to credit Taylor, Shelden, and Dean’s statements. There
was little other evidence in the case; the forensics were presented errone-
ously, but were not particularly probative. Joseph White was convicted
and served nineteen years in prison until DNA tests led to his exonera-
tion in 2008, followed by exonerations of the fi ve others.54

Expert evidence also corroborated facts in false confessions. Ronald
Jones, whose case was discussed in Chapter 1, reportedly confessed to
stabbing the victim “a few times.” This fact from his confession statement
was corroborated by the medical examiner’s opinion that the victim had
been stabbed four times.55 Ronald Williamson was said to have described
wrapping a cord around the victim’s neck to strangle her.56 At trial, the
medical examiner testifi ed that the victim was in fact strangled.57

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

28 Contaminated Confessions

Denying Disclosing Facts

We now know that in many of these cases, police contaminated the con-
fessions by disclosing facts to the suspects. But why didn’t this come to
light in the cases that went to trial? The answer is that at trial, the police
denied ever having disclosed those key facts. The trial transcripts show
that in twenty- seven of the thirty confession cases that went to a trial, the
police offi cers testifying under oath at trial denied that they had dis-
closed facts to the suspect, or they described the suspect having volun-
teered the central facts during the interrogation. I do not know whether
these police offi cers were aware that their testimony was false. Some of
these offi cers may have been malicious. However, many of them may
have believed they were interrogating a guilty person. Offi cers may con-
taminate a confession unintentionally (or intentionally), out of a belief
that the suspect is guilty and is a danger to the public. Later they may
not recall that they disclosed facts during a complex series of questions
posed during a long interrogation. Washington, D.C., detective James
Trainum has described how he unintentionally secured a false confession
and did not initially realize what had happened. Later he determined he
had actually disclosed key facts to the suspect, which he confi rmed when
he carefully reviewed the videotape of the entire interrogation.58 Police
did not record the entire interrogation for any of these exonerees, so we
can never be certain precisely how these people came to falsely confess in
such a convincing way.

In Nathaniel Hatchett’s case, police unequivocally denied that they
disclosed facts to him:

Q. Did you ever supply the Defendant with details, specifi c details

of the off ense so that he would be able to recite them back to

you when and if he decided to give you a statement about his

knowledge and involvement with these crimes?

A. I didn’t.

Q. You say you didn’t, so I will ask the next question: Did you hear

anyone else or see anyone else provide him with the kind of

details that he eventually later gave you demonstrating his knowl-

edge and involvement in this crime?

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 29

A. No. As a matter of fact, as lead investigator I was the only one

privy to such details at this point.59

Earl Washington Jr.’s case provides another example where the deni-
als of disclosing facts were crucial to the State’s case. Washington falsely
confessed to a rape and murder in Culpeper, Virginia. He came within
nine days of execution and was in prison for eigh teen years before fi nally
being exonerated by DNA testing. A long string of state and federal
courts denied his appeals and postconviction petitions, each citing the
reliability of his confession. Although he was borderline mentally re-
tarded, the Fourth Circuit, for example, emphasized that: “Washington
had supplied without prompting details of the crime that were corrobo-
rated by evidence taken from the scene and by the observations of those
investigating the [victim’s] apartment.”60

Two offi cers told prosecutors and then testifi ed at trial that Washing-
ton was the one who, without prompting, identifi ed a shirt as his own. It
was a distinctive shirt, its existence had not been made public, and it
was found in the rear bureau of the victim’s bedroom many months after
the murder. The typed statement read as follows:

Offi cer 1: Did you leave any of your clothing in the apartment?

Washington: My shirt.

Offi cer 1: The shirt that has been shown you, it is the one you left in

apartment?

Washington: Yes sir.

Offi cer 1: How do you know it is yours?

Washington: That is the shirt I wore.

Offi cer 1: What makes it stand out?

Washington: A patch had been removed from the top of the pocket.

Offi cer 1: Why did you leave the shirt in the apartment.

Washington: It had blood on it and I didn’t want to wear it back out.

Offi cer 2: Where did you put it when you left?

Washington: Laid it on top of dresser drawer in bedroom.61

This statement was powerful for several reasons. Most of this con-
fession statement consisted of offi cers asking leading questions with

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

30 Contaminated Confessions

Washington— recall he was mentally retarded— providing very brief an-
swers. His most frequent response was “Yes sir.” However, in that cru-
cial passage, Washington gave uncharacteristically detailed answers. He
reportedly off ered that he left his shirt— yet the police had not made
public that a shirt was found at the crime scene. Further, he told police
about an identifying characteristic making the shirt unusual: the torn- off
patch. He knew precisely where the shirt had been left, in a dresser
drawer in the bedroom. And most remarkable, he said that he left it there
because it had blood on it. Yet the shirt that the offi cers showed Wash-
ington no longer had blood on it; the stains had been cut out for forensic
analysis.62

The prosecutor emphasized in closing arguments that the police were
not “lying” and “didn’t suggest to him” how the crime had been com-
mitted, but that Washington knew exactly how the crime had been com-
mitted. The prosecutor ended by discussing the shirt, and noting that
Washington knew “the patch was missing over the left top pocket.” The
prosecutor continued, “Now, how does somebody make all that up, un-
less they were actually there and actually did it? I would submit to you
that there can’t be any question in your mind about it, the fact that this
happened and the fact that Earl Washington Ju nior did it.”

Now that we know Earl Washington Jr. did not commit the crime, but
rather another man later identifi ed through a DNA database who has
now pleaded guilty, there are limited explanations for how Washington
could have uttered those remarks about the shirt, together with other
details about the crime. This was precisely the issue in a civil rights law-
suit brought by Washington after his exoneration. During that lawsuit,
Washington’s lawyers discovered that nearly ten years after the convic-
tion and near the time that Virginia’s governor was considering a clem-
ency petition based on DNA testing, the second offi cer expressed real
doubts about the interrogation. He admitted that facts were likely dis-
closed, telling the Virginia attorney general that either he or his col-
league “must have mentioned the shirt to Washington” and that “his
testimony in the record did not accurately refl ect that the shirt had fi rst
been mentioned by the police.”63 In 2006, a federal jury found that the
second offi cer fabricated the confession and violated Washington’s con-
stitutional rights by recklessly claiming that Washington volunteered

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 31

crucial nonpublic facts. The jury awarded Washington $2.25 million in
damages.64

The high- profi le “Central Park Jogger” case, in which fi ve youths
falsely confessed to brutally assaulting and raping a jogger in Central
Park and were all exonerated by postconviction DNA testing years later,
also involved a striking detail concerning a shirt. The prosecutor empha-
sized in closing arguments that one of the youths, Antron McCray, knew
information that only the assailant could have known:

You heard in that video Antron McCray was asked about what [the

victim] was wearing and he describes she was wearing a white

shirt . . . You saw the photograph of what that shirt looked like.

There is no way that he knew that that shirt was white unless he saw

it before it became soaked with blood and mud. I submit to you that

Antron McCray describes details and describes them in a way that

makes you know . . . beyond reasonable doubt that he was present,

that he helped other people rape her, and that he helped other peo-

ple beat her and that he left her there to die.65

In twenty- two cases, prosecutors emphasized in their closing argu-
ments that the relevant facts were nonpublic or corroborated by crime
scene evidence. For example, in Bruce Godschalk’s case, the prosecution
took the position that “Well, if he were guessing, he was guessing pretty
darn good.”66 The prosecutor, in an incredulous tone, then told the jury
that it was a “mathematically [sic] impossibility” that Godschalk could
have guessed correctly on so many nonpublic facts regarding how the
crime was committed. In Robert Miller’s case, the prosecutor empha-
sized that “[h]e described the details . . . details that only the killer could
have known.”67 In response to the defense’s suggestion that he could have
guessed such details, the prosecutor said, “Are you kidding? Are you kid-
ding?” and added, “He supplied detail after detail after detail after de-
tail.” In the Rolando Cruz and Alejandro Hernandez trial, the prosecutor
closed by telling the jury that Cruz told offi cers the victim “was anally
assaulted. There’s no way to know that information. He knows it because
he was there.”68

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

32 Contaminated Confessions

Recorded False Interrogations

What makes the contamination of these confessions initially appear
surprising is that many of the interrogations were recorded; in twenty-
three of the forty cases (58%) interrogations were partially recorded.
Fourteen were audio recorded and nine had video. Additional interro-
gations were at some point partially transcribed by a stenographer. In
all of these cases, however, only part of the interrogation, often a fi nal
confession statement, was recorded. There was no recording of what came
before.

In some cases, police disclosed facts to the suspect even during the
recorded part of the interrogation. For example, during David Allen
Jones’s recorded interrogation, when he did not recall the location of a
crime, police reminded him that they had earlier showed him photos of
the crime scene, asking, “You remember yesterday we showed you that
picture” and that it was “by the water fountain” and “you remember that
gate we showed you right there,” fi nally eliciting only a response from
Jones that was transcribed as “This right here (Untranslatable).”69

We do not know what transpired during the unrecorded portions of
the interrogation. Law professor Steven Drizin notes that it is “not un-
common” for police to conduct an initial interview in which they “use a
gamut of techniques” to secure admissions, but do not tape that inter-
view. Rather, police tape a second interview only once the admissions
have been secured.70 Thus, in Nicholas Yarris’s case, law enforcement
emphasized that the two crucial facts he volunteered, that the victim was
raped and that her car had a brown landau roof, were off ered in “addi-
tional conversation that was not part of the normal statement,” and there-
fore not on tape or transcribed.71

Chris Ochoa has since his exoneration described blatant abuse of re-
cording, in which the Austin Police Department detective asked him
leading questions about the crime. None of that questioning was on the
tape. By Ochoa’s recounting, the detective stopped the tape each time
that he “came to a detail” and “it was wrong.” The offi cers would “get
mad,” and show him photographs of the crime scene and the autopsy or
tell him the answers, and then “start it, stop it, till they got the details. It
took a long time.”72

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 33

Other interrogations were not recorded, in some cases despite a depart-
ment practice of doing so. In Dennis Brown’s case, for example, police
testifi ed that interrogations were normally recorded, but police “chose not
to turn it on” in his case because the defendant “just started talking.”73

Some of the conversations that may have contaminated these confes-
sions could have occurred at the crime scene itself. In thirteen cases,
police brought the exonerees to the crime scene. Each reportedly identi-
fi ed in person features of the crime scenes that were nonpublic and cor-
roborated by the investigation. Such visits were not recorded, so we do
not know whether police in fact tested exonerees’ knowledge of how the
crime occurred at the scene, or whether they instead familiarized exon-
erees with the crime scene. Ronald Jones, whose case was discussed in
Chapter 1, testifi ed that the detective took him to the crime scene and
told him a series of details about the crime.74 Earl Washington Jr. led
police to locations all around Culpeper, Virginia, having had no idea
where the victim was murdered.75 Even after being driven by the victim’s
building several times, he did not identify it; when asked to point to her
building in the apartment complex he pointed to “the exact opposite
end” of the complex. Only when the offi cer pointed to her apartment
and asked if that was it did he say “yes.”

Inconsistent Facts

We have heard about the detailed facts that these innocent people likely
repeated during their interrogations. What do innocent suspects say when
they are not prompted by police? In most of these exonerees’ cases,
when given a chance to volunteer information, they got the facts totally
wrong, since they did not know anything about the crime aside from
what they were told.76 In at least 75% of these cases (thirty of forty cases),
the exoneree supplied facts during the interrogation that were inconsis-
tent with the known facts in the case.77 These inconsistencies should
have been a real warning sign that the confessions could be false.

Indeed, police may have been so frustrated that these innocent sus-
pects were not giving them the answers they wanted that they resorted to
feeding the facts rather than asking proper open- ended questions. Earl
Washington Jr.’s case provides an example of this strategy. At his trial,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

34 Contaminated Confessions

police testifi ed about the separate stages of the interrogation. While in
the second stage of his interrogation he supposedly gave accurate details
about the crime, in the initial stage, whenever he was not asked a leading
question, he got the answers wrong. When he was asked what race the
victim was, for example, he told police that the victim was black. That
was wrong; she was white.78 He described stabbing the victim a few
times— she had in fact been stabbed dozens of times. He described the
victim as short when she was tall. He said he “didn’t see” anyone else in
the apartment, yet the victim’s two children were there.79 It was clear he
knew nothing about the crime, but, ever compliant, he would eventually
agree with what ever leading question was posed to him, politely saying,
“Yes, sir.” In fact, the police knew that if they were per sis tent, Washing-
ton would falsely confess. He had already falsely confessed to four other
crimes. He confessed to every crime he was asked about, but in the other
cases the victims told police he was not the assailant, or the confessions
were deemed totally inconsistent with how the crimes occurred.80

Prosecutors had to explain inconsistencies to the jury by downplay-
ing them and emphasizing instead the powerful nonpublic facts that
each was said to have volunteered. Byron Halsey was also mentally
disabled and highly compliant with police interrogators. Halsey made
multiple incorrect guesses as to the manner in which a horrifi c rape and
murder of two children occurred. When asked how he put nails into
one victim’s head, the offi cer recounted, “His initial response was crow-
bar. He grinned at us.” “Then he came up and he said, ‘Hammer.’ He
grinned again.” “Then he mentioned chair. He grinned again.” The
perpetrator apparently actually used a brick. Following that interchange,
the offi cer recalled that “he fi nally did tell us what he used to drive the
nails into [the victim’s] head.” The prosecutor asked, “Any time was it
suggested to the defendant what the answer should be?” and the offi cer
answered, “No.” In his closing statement, the prosecutor not only de-
nied that any facts were disclosed but argued that the inconsistencies in
Halsey’s confession were part of a master plan to confuse the jury: “You
know why, ladies and gentlemen, because he thinks he’s just given a
sworn statement that’s going to let him off .” The prosecutor added,
“he’s trying to mislead . . . he’s trying to lie his way out of this in that
confession.”81

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 35

In a few cases, police may have inadvertently suggested mistaken facts
due to their incomplete knowledge about the crime scene evidence, which
turned out to be inaccurate. For example, in a recorded interrogation,
Bruce Godschalk described entering the kitchen window of one victim’s
apartment. There was a problem. That apartment had no kitchen win-
dow. The detective claimed at trial that Godschalk had later changed his
statement to say he had entered through a spare bedroom.82

Police often stopped investigating once they obtained a confession.
They then failed to investigate major inconsistencies between the con-
fession and the other evidence. In Earl Washington Jr.’s case, not only
were the facts that he did volunteer inconsistent with the evidence, but
police refused to test forensic evidence that might have shed light on his
guilt or innocence. Recall that the shirt with the torn pocket was a key
fact in his confession statement. Hairs were found in the pocket of that
shirt, and early in the case, fi ve other suspects’ hairs were compared and
they were excluded. After Washington confessed, police instructed the
state crime laboratory not to compare his hairs to the ones found in the
pocket of that shirt.83

In other cases forensic testing was not done because police did not
want to call the confession into question. In Lafonso Rollins’s case, the
crime lab technician had found that blood- typing excluded Rollins and
had asked his supervisors to send the evidence to the FBI for more so-
phisticated DNA testing. As the technician later testifi ed, “his request
was refused because police said Rollins confessed.”84

In eight cases, the suspects had confessed, but DNA tests conducted
at the time had already excluded them, providing powerful evidence of
innocence. Yet police and prosecutors continued to insist that the defen-
dants had committed the crimes. Judges and juries presumably believed
the confessions and disregarded the DNA evidence. Those eight people
were all exonerated only after additional DNA tests not only excluded
them but also inculpated other persons.85

For example, in Deskovic’s case, the police apparently stopped inves-
tigating the case after they secured the confession. Although the State’s
theory was that Deskovic raped and murdered the victim alone, recall
that DNA results at the time of trial excluded him. The prosecutor had
speculated that perhaps the victim had a boyfriend, and perhaps the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

36 Contaminated Confessions

boyfriend’s semen and not the murderer’s accounted for the test results.
There was no attempt to investigate the theory that the victim was sexu-
ally active, most likely because there was no such boyfriend. An inquiry
conducted after Deskovic’s exoneration concluded, “There is no evi-
dence, for example, that much was done to locate the ‘boyfriend’ who
was the supposed source of semen or even to document [the victim’s]
movements in the 24 hours before her death.”86

Nathaniel Hatchett had a bench trial, before a judge and not a jury.
The victim testifi ed she had been raped by a single man, and DNA test-
ing at the time of the trial excluded Hatchett. Yet the judge explained
that “in light of the overwhelming evidence that the Court has . . . the
Court does not fi nd that the laboratory analysis is a fact which would
lead to a verdict of acquittal.” The judge emphasized that despite power-
ful DNA evidence, “in this case there is an abundance of corroboration
for the statements made by Mr. Hatchett to the police after his arrest,
about what happened during the assault.” The judge had no theory to
explain away the DNA test results that excluded Hatchett, but said in-
stead that he found the confession “to be of overwhelming importance in
determining the outcome of the trial.”87

Judicial Review of False Confessions

Once these false confessions were obtained, why didn’t judges and juries
reject them? Almost all of these exonerees’ defense attorneys did chal-
lenge the confessions before trial, and in each case the judge rejected the
challenge. The only exonerees who we know did not challenge their
confession at trial were those who pleaded guilty and had no trial, and
one, Rolando Cruz, whose lawyers did not challenge his confession as
involuntary but instead argued that police had completely made up his
alleged statements regarding a dream or vision of the crime.88 Of the
twenty- nine exonerees who went to trial and whose available rec ords
indicate whether a challenge to the admission of the confession was
made, twenty- eight (97%) made a challenge. All were unsuccessful.

The most obvious reason why defense lawyers’ eff orts to challenge
the confessions failed is that the confessions seemed so accurate. After
all, these defendants supposedly knew facts that “only the killer could

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 37

have known.” Nor could they try to point out inconsistencies and ask the
judge to scrutinize the reliability of these confessions. The U.S. Su-
preme Court has held that unreliability is irrelevant to the question
whether a confession statement is suffi ciently voluntary to be admitted
at trial.89 Judges are also very reluctant to rule that a confession cannot
be heard at trial. The legal standard that judges use to review confes-
sions focuses on the voluntariness of the bare admission of guilt, and
judges will usually only exclude confessions in rare cases of fl agrant
police misconduct. Confession statements are regulated by the U.S. Con-
stitution in two ways: the Miranda warnings and the requirement of
voluntariness.

The Supreme Court developed the well- known Miranda v. Arizona
protections to shield suspects from coercion and avoid diffi cult volun-
tariness questions by providing notice in the interrogation room that
suspects have certain rights.90 If the Miranda procedures had not been
followed, these exonerees might have had straightforward challenges to
their confessions. However, all of these exonerees waived their Miranda
rights, just as the vast majority of suspects do.91 Twenty- two of the exon-
erees signed waiver forms and eight waived their rights on video. We
cannot know how many more exonerees who, like Joseph White, did ask
for a lawyer and cut their interrogation short, might have also falsely
confessed if their interrogations had continued.

Perhaps the most remarkable of the rulings concerning Miranda warn-
ings was at the trial of Eddie Joe Lloyd. The most obvious evidence of his
mental illness and suggestibility was that Lloyd had been interviewed
while he was involuntarily committed in a mental hospital, with a pre-
liminary diagnosis of bipolar aff ective disorder. The judge ruled that his
statements were voluntary, noting, “[a]fter he is advised of his rights, it
doesn’t make a diff erence what form it takes.”92

The Supreme Court’s voluntariness standard regulates interrogations
more broadly, by examining the “totality of the circumstances” to assess
whether the confession was coerced, focusing on “the characteristics of
the accused and the details of the interrogation.”93 The prosecution has
the burden of showing that no undue coercion was applied.94 Trial judges
conduct hearings before trial to examine the question whether the confes-
sion was voluntary and should be admitted.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

38 Contaminated Confessions

Judges ruled that these exonerees’ confessions were voluntary, but of-
ten despite substantial evidence to the contrary. For example, one factor
the Court has long taken into account when analyzing voluntariness is
“the youth of the accused.” Yet 33% of the exonerees who confessed were
juveniles (thirteen of forty cases; fi ve were in the Central Park Jogger
case). A second factor is “low intelligence” of the accused. Studies have
long shown that the mentally disabled tend to be submissive and eager to
please authority fi gures, and vulnerable to stress and pressure.95 Yet at
least 43% of the exonerees who falsely confessed (seventeen of forty cases)
were mentally disabled or ill, many obviously so. A third factor is “the
lack of any advice to the accused of his constitutional rights.” All of these
exonerees waived Miranda rights prior to confessing.

A fourth factor is the “length of detention” together with “the re-
peated and prolonged nature of the questioning.” Most of these exoner-
ees endured quite lengthy interrogations. Typically, John Kogut was
told “you’re not going anywhere until we get the truth.”96 Only 10% of
the exonerees for whom materials were obtained (four of forty cases)
were interrogated for less than three hours. The other exonerees were
interrogated for far longer, typically in multiple interrogations over a
period of days, or interrogations lasting for more than a day with inter-
ruptions only for meals and sleep. Jerry Townsend was interrogated for
thirty to forty hours over the course of a week.97 Robert Miller was inter-
rogated for thirteen to fi fteen hours just during the portions of his inter-
rogation that were recorded.98 Yet again, judges found such confessions
to be voluntary.

An additional factor includes “use of physical punishment such as the
deprivation of food or sleep.” While less common among these cases,
some exonerees did allege, whether accurately or not we often do not
know, that police used physical force on them during the interrogation.
Marcellius Bradford later claimed that police beat him. Dennis Brown
claimed that the offi cer fi rst unbuckled his gun belt, then “put a knife on
me.”99 Ronald Jones, whose case was discussed in the Introduction, de-
scribed that a detective handcuff ed him to the wall and hit him in the
head several times with a long black object.100 According to Jones, a sec-
ond offi cer then entered the room and “was surprised” at “the treatment”
“and told him, he said no, don’t hit him, because he might bruise.” That

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 39

offi cer proceeded with hitting him in the “midsection with his fi st” when
he would not confess. Paula Gray, who was seventeen and borderline
mentally retarded, was kept at hotels with seven police offi cers for two
days as a “protected custody witness” regarding a double murder investi-
gation.101 During her custody she made a statement inculpating herself
and four other innocent people, in a case that became known as the “Ford
Heights Four” case. Gray testifi ed that she was asked, “Did they empha-
size what would happen if you did not tell this story?” and answered,
“That they would kill me.” Police denied engaging in any such conduct.

Judges also consider the demeanor of the defendant during the inter-
rogation to shed light on whether the confession was voluntary. The State
admitted that David Vasquez gave a statement that “at times” was “virtu-
ally incomprehensible.”102 Byron Halsey, who had a “sixth- grade educa-
tion and severe learning disabilities,”103 by the detective’s own admission
gave a statement that was full of “gibberish” in which Halsey was “saying
one syllable words, free fl owing like a water fall” and at some times he fell
into a “trance” state in which he “just stopped and looked out into
blankness.”104

Perhaps judges, when they ruled that those confessions were volun-
tary, relied on the fact that most exonerees signed written statements that
contained boilerplate language on voluntariness. Thus, Lafonso Rollins
signed a statement repeating that “he has been treated well” and “he has
eaten pizza and a sandwich and has had coff ee and cola to drink while he
has been at the police station. Lafonso Rollins states that he has not been
threatened in any way, nor has he been made any promises for this state-
ment.”105 Similarly, Bruce Godschalk admitted at trial that he had been
asked on tape by the detectives, “How have we treated you?” and he an-
swered, “Very well.”106

Why did judges look past all of the signs of coercion in so many exon-
erees’ cases? Not only are judges reluctant to exclude a confession, which
is central to the prosecution’s case, but they may be especially reluctant
to exclude a confession that appears so powerfully corroborated by de-
tailed facts. For example, in Douglas Warney’s case, though he alleged
coercion, the judge relied heavily on how he initially approached police
to talk about the crime: “This was information that he volunteered to sub-
mit himself to present, and he followed through on that.”107 Though the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

40 Contaminated Confessions

Supreme Court has ruled out reliability as a reason to exclude a confes-
sion, judges focused on the apparent reliability of these confessions when
they admitted these confessions and found them voluntary. In Bruce
Godschalk’s case, the judge emphasized that he “had given information
to the detectives which was not released to the general public,” though
the judge did not say which specifi c facts he found convincing.108 Simi-
larly, in the trial of Paula Gray, who was seventeen years old and border-
line mentally retarded, the judge noted, “Incidentally, the defendant tes-
tifi ed with skill, with knowledge, explicitly, extremely clear, made her
points well and all it means to me is whether she’s in twelfth grade or
what ever her educational level is she’s a very intelligent person. That’s my
judgment and those are my fi ndings and my decision.”109

Use of Experts

Jurors may have a hard time imagining that a person would falsely con-
fess. A defendant may try to use an expert to explain to the jury that false
confessions can actually happen, and that the defendant was psychologi-
cally vulnerable to police suggestion or coercion. However, judges often
deny indigent defendants the funds to hire such experts or they refuse
to allow such expert testimony. Only 8% of the exonerees (three of forty
cases) had testimony by defense experts such as psychologists or psy-
chiatrists concerning their confessions. The three were David Allen
Jones, Jerry Townsend, and David Vasquez. In Vasquez’s case, the pre-
trial hearing transcripts are not in the court fi le; all that the fi le indicates
is the judge’s view that there was “confl icting evidence given by the psy-
chiatrists.”110 Jones retained a psychologist who testifi ed that he found
Jones “mildly mentally retarded.”111

At Jerry Townsend’s suppression hearing, there was a lopsided battle
of the experts; surprisingly, given a typical lack of defense resources, the
contest was weighted in the defense’s favor. Jerry Frank Townsend was
mentally retarded and he confessed to every unsolved murder he was
asked about, including about twenty murders, most in Florida. His law-
yer later noted that he “was capable of confessing to just about any-
thing,” and the police “wanted to clean out the cold case fi les on him.”112
The defense called seven diff erent clinical psychologists and psychia-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 41

trists as experts. All of these doctors agreed Townsend could not readily
understand the Miranda warnings, lacked capacity to be tried, and did
not confess voluntarily. For example, one expert described how Townsend
was retarded with an IQ of 59 to 61 and the mental development of a six- to
nine- year- old.113 Other experts testifi ed that he was “easily led” and “highly
suggestible.” Townsend not only lacked capacity and had a mental age of a
seven- to nine- year- old, but he “confabulate[d] all the time,” engaging in
“unconscious making up of material to fi ll in for what one doesn’t know,”
like a person who is “senile.” Townsend spoke like a “madman” and, in
medical terms, was “intellectually limited and functioning at a psychotic
level.”

The prosecution called only one expert witness, who only had a mas-
ter’s degree in psychology (he was not a doctor).114 He mea sured Townsend’s
“social adaptive” ability as nineteen years, although Townsend’s IQ was
“signifi cantly below” average. He testifi ed that Townsend would under-
stand his Miranda rights.115 The State also called one of the detectives
who had interrogated Townsend, who gave his nonexpert opinion that
Townsend had a “very severe speech impediment,” but had “street
smarts” and “made it clear that he understood his rights.” The prosecu-
tor then elicited lengthy testimony concerning the “specifi c facts” that
Townsend supposedly knew about how the crimes were committed
“which we have been able to, of course, corroborate.”

In the end, the judge found this battery of defense expert testimony to
be “testimony of con ve nience” and found that “the credibility of it is
rather low and the believability of it just as low.” The judge ruled: “I fi nd
that Mr. Townsend was suffi ciently societally, if I may use that word,
and functionally intelligent to know— to know his Miranda rights and to
signifi cantly and suffi ciently waive them . . .” He added that Townsend
“had so much knowledge of street parlance.” He found it a closer ques-
tion whether Townsend confessed involuntarily, but denied relief, not-
ing, “You know, it would be good if all confessions were perfect.” The
judge explained, “I have yet to see a perfect confession, but I’ve not been
in the criminal area that long.”116

Townsend called the same experts to testify at trial. The State argued
that Townsend’s actions indicated the work of a methodical serial killer,
and he was convicted. Townsend served twenty- two years in prison

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

42 Contaminated Confessions

before a Fort Lauderdale police reinvestigation led to DNA testing that
cleared him and inculpated serial killer Eddie Lee Mosley (the DNA
testing also linked Mosley to a murder for which Frank Lee Smith was
falsely convicted).117

Other exonerees, had they obtained experts, might have presented the
jury with research on causes of false confessions, or evidence that they
lacked capacity or otherwise were susceptible to coercion or suggestion.
In the Deskovic case, for example, the district attorney’s inquiry found
that “the defense did not attempt to introduce psychiatric evidence that
might have persuaded jurors that Deskovic was particularly vulnerable to
the police tactics employed against him and that those tactics induced a
false confession. In the absence of such evidence, the defense attack on
the statements seemed scattershot and unfocused.”118

Reforming Interrogations

Ronald Jones’s conviction and the years he spent battling for his free-
dom, with judges denying relief at every turn, now begin to come into
focus. In Chapter 1, I described how not only was his trial short and his
defense tepid, but every step of the way, judges relied on the facts in his
confession to argue that he could not possibly be innocent. Ronald
Jones, Jeff rey Deskovic, Earl Washington Jr., and all of the others I have
discussed in this chapter each gave detailed false confessions that viv-
idly show how a false confession may involve not just coercion but the
ability to convince an innocent suspect to develop a crime narrative.
When that narrative was contaminated by the disclosure of key facts,
absent DNA tests, the criminal system could not untangle what trans-
pired. Social scientists and legal scholars have long recommended that
judges evaluate the reliability of entire interrogations rather than simply
focus on Miranda warnings or voluntariness.119 However, even robust
reliability review faces another problem. As these cases show, to the ex-
tent that facts were disclosed to the suspect, confessions appear uncan-
nily reliable.

As we will see throughout this book, evidence can be contaminated in
many ways. Police can leave physical evidence exposed to the elements
so that it degrades. They can tell an eyewitness that one of the people in

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Contaminated Confessions � 43

the lineup has a prior conviction, which would tend to encourage the
identifi cation. Contamination can be very hard to detect or undo. Once
police tell a suspect details about the crime, one can never know for sure
whether the suspect truly knew those details or if he learned them from
the police.

Contamination of confessions can be discouraged. The easiest way to
avoid it is to require that police record interrogations from start to fi nish.
As Chapter 9 will discuss, this reform has been adopted in eigh teen
states, the District of Columbia, and hundreds of police departments. A
lengthy and complex interrogation, the centerpiece of a criminal investi-
gation, should not be undocumented and then recounted based on bi-
ased or fallible human memory. Indeed, police have embraced recording
interrogations so they can show a judge that the confession was reliable
and voluntary.

Recording entire interrogations is an important step but is not enough.
If the recording shows that the confession was contaminated, and prose-
cutors nevertheless pursue the case, the judge should step in to assess
whether the confession should be excluded as unreliable. David Vasquez’s
case provides a cautionary tale. Vasquez was borderline mentally re-
tarded, and he was interrogated by the police after becoming a suspect.
The police had determined “that the bindings used to secure [the vic-
tim’s] hands had been cut from the venetian blinds in the sunroom. The
noose employed for her execution had been cut from a length of rope
wrapped around a carpet in her basement.”120 Part of the interrogation
was recorded. It was obvious from the recording that Vasquez had no
idea what was used to bind and murder the victim:

Det. 1: Did she tell you to tie her hands behind her back?

Vasquez: Ah, if she did, I did.

Det. 2: Whatcha use?

Vasquez: The ropes?

Det. 2: No, not the ropes. Whatcha use?

Vasquez: Only my belt.

Det. 2: No, not your belt . . . Remember being out in the sunroom,

the room that sits out to the back of the house? . . . and what

did you cut down? To use?

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

44 Contaminated Confessions

Vasquez: That, uh, clothesline?

Det. 2: No, it wasn’t a clothesline, it was something like a clothes-

line. What was it? By the window? Think about the Venetian

blinds, David. Remember cutting the Venetian blind cords?

Vasquez: Ah, it’s the same as rope?

Det. 2: Yeah.

Det. 1: Okay, now tell us how it went, David— tell us how you did it.

Vasquez: She told me to grab the knife, and, and, stab her, that’s all.

Det. 2: (voice raised) David, no, David.

Vasquez: If it did happen, and I did it, and my fi ngerprints were

on it . . .

Det. 2: (slamming his hand on the table and yelling) You hung her!

Vasquez: What?

Det. 2: You hung her!

Vasquez: Okay, so I hung her.

The feeding of facts to David Vasquez was blatant— and it was caught on
tape. Nevertheless, the judge denied the motion to suppress the confes-
sion. Facing the death penalty, Vasquez then chose to plead guilty rather
than risk a trial. He was exonerated by postconviction DNA testing fi ve
years later.121

These contaminated false confessions vividly illustrate why what goes
on in the interrogation room should not remain undocumented, unregu-
lated, unreviewed, and as a result, shrouded in darkness. Recording can
bring interrogation practices into the sunlight. Although contamination
is an insidious problem that may be hard to eradicate, recordings can
validate professional interrogations and make review by judges far more
eff ective. Reforming criminal procedure to safeguard the reliability of
interrogations can help us better separate true from false confessions.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

A woman was bird- watching with her husband in a park in Buf-
falo, New York, in May 1982. As the sun set, she was getting cold

and decided to return to the car alone. But as she walked down a jogging
path, a man approached. He suddenly ran up and jumped out from be-
hind, and said, “Scared you, didn’t I?” She caught only a glimpse of him
before he grabbed her, choking her neck and blindfolding her with a
bandana. He dragged her into a thicket, raped her, and left.1 She re-
ported the rape to the Buff alo Police Department that eve ning and gave
them a description of the attacker. Although blindfolded, she “could feel
a close cropped beard and moustache and fairly thick lips.” She saw only
“dark skin but rather reddish undertones,” and that he had “a space be-
tween his upper front teeth.” The man was between 5′8″ and 5′10″.

More than four and a half months passed, and she heard nothing from
the police. Finally, in September, the police called her to the police sta-
tion and asked whether she could “make an identifi cation,” from “a black
and white photo, a line- up photo.” She told the offi cer that she was not
sure if she “would be able to identify him again.”2

She was shown a standard photo array with six photographs, one of
which was a photo of a man named Habib Abdal. She was not able to
identify her assailant. She told the offi cer that there were three men in the

c h a p t e r 3

Eyewitness Misidentifi cations

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

46 Eyewitness Misidentifi cations

array who looked somewhat similar to her attacker. At trial, she testifi ed
that she tentatively chose Abdal. She said: “I then chose one man that
looked closest to my description that I had in my head that I knew of the
assailant.” This should have immediately indicated a serious problem.
She could not identify Abdal, but said his photo looked the “closest.”
She had in fact told police that he looked diff erent from the culprit. So-
cial science research has shown that witnesses should be told that the
culprit might not be in the lineup, because otherwise the witness may
comparison shop and pick whoever looks most like the attacker.3

Police next asked if it would help if she saw a live lineup. The police
did not conduct a lineup with other men who resembled Habib Abdal.
Instead, they took her to a room to look through one- way glass at just
two men— Abdal and a white man. In eff ect, they used a showup, a pro-
cedure where a lone suspect is presented to the witness. Even that was
not enough to convince her. For about fi fteen minutes, the police made
suggestions to her that she identify Abdal. They told her that he was in
custody because he was involved in another crime, or in another rape.
The police then put overt pressure on her to make an identifi cation. The
victim recalled all of this pressure and testifi ed about it at trial: “And the
more pressure they put on,” she said, “the less I would listen to them and
the longer I took in making an identifi cation, because this was a man’s
life, I didn’t want to just make an identifi cation to make it easy for every-
one.”4 She testifi ed that the police had a “specifi c mind set” and that “they
wanted” her to identify Abdal. She told the jury, though, that she made a
conscious eff ort to resist. She told police that “there were two diff erences
still” and she could not identify Abdal. They decided to then take her to
look at Abdal up close—“I was able to look eye to eye” with him. Even
then, she was not able to identify him. She heard him speak, and that did
not help change her mind; “he used diff erent voices.”

Finally, an offi cer said, “Wait a minute, I have an older photo,” and
brought her a black- and- white mug shot of Abdal. She “immediately
identifi ed the man in the photo.” That photo had been taken in 1978, fi ve
years before the assault. Yet the victim said when she saw that photo, she
“had no hesitation whatsoever” and was “positively sure.”5

Her certainty increased at trial and other aspects of her account also
changed. At trial, she testifi ed that her attacker spoke to her and she

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 47

looked at him directly for about thirty seconds before he assaulted her. In
her original statement she said he spoke from behind her and she barely
saw him before he blindfolded her. Her original statement described only
being able to feel facial hair; she could not describe it.

Where the victim had seen diff erences between Habib Abdal and her
attacker, at trial she found similarities. She testifi ed, “It’s the same pro-
truding eyes, but, slightly bulging. It’s the same yellowing whites. The
same slightly hooded eyelid. The same oval face that comes to the same
slight narrowing pointing of the chin. It’s the same width of the shoul-
ders I felt. It’s the same coloration . . .” However, in her descriptions to
police she had never described those “hooded eyelids.” She testifi ed
that the defendant had the same dark gums or spaces between his teeth.
However, her original description was very specifi c about a single space
between the upper front teeth. She told police the man could be between
twenty- fi ve and forty. However, Abdal had a gray beard and was forty-
three years old. She described an attacker who was between 5′8″ and
5′10″, just a few inches taller than she. However, Abdal was much taller
than that. He was 6′4″.6

Finally, the victim testifi ed at trial that she was completely certain
Habib Abdal was her attacker. She claimed, “I have the same picture in
my mind today as I had then.”7 That statement was false. But it was very
powerful. There was no other evidence linking Abdal to the rape. The
forensic evidence from the rape kit was inconclusive. The hairs found on
the victim did not match Abdal. The jury convicted Abdal based on the
account of this eyewitness whose memory was deeply tainted, the prod-
uct of sustained police eff orts to induce her to make an identifi cation.
There was a picture in her mind after the rape— and at fi rst, her memory
may have been accurate. Her initial descriptions did not resemble Ab-
dal, but rather another man. After the police used suggestive techniques,
though, the picture in her mind changed and her certainty about her
identifi cation of Abdal increased. In the last chapter, we saw how police
can contaminate a confession by suggestion or outright feeding of facts
during an interrogation. Here, we see contamination of a diff erent kind
of evidence: the memory of an eyewitness. In Abdal’s case, to this day,
the true culprit has not been found. Abdal spent sixteen years in prison
before being exonerated by postconviction DNA testing in 1999. We will

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

48 Eyewitness Misidentifi cations

hear more about Abdal’s postconviction ordeal in Chapter 8, which de-
scribes the struggles of these exonerees to obtain DNA testing and an
exoneration.

Suggestive and Unreliable Identifi cations

As in many trials, the key moment in Habib Abdal’s trial occurred when
the eyewitness pointed to him in the courtroom and said she was com-
pletely certain that he was the man who attacked her. As Justice William
Brennan wrote, “[T]here is almost nothing more convincing than a live
human being who takes the stand, points a fi nger at the defendant, and
says ‘That’s the one!’ ”8 Almost all of the eyewitnesses at these innocent
persons’ trials testifi ed that they were certain they had identifi ed the right
person.

Judges expressed disbelief that these eyewitnesses— or any
eyewitnesses— could be mistaken. In Jerry Miller’s trial, the judge said,
“I have never heard that in my life, that a rape victim with a person who is
unmasked says they can’t identify him. I never heard that in my life.”9
Jurors were moved by this eyewitness testimony. In Marcus Lyons’s case,
a juror later described how powerfully it aff ected them that the victim was
“shaking like a leaf” when she identifi ed Lyons at trial.10

The role of mistaken eyewitness identifi cations in these wrongful con-
victions is now well known. Eyewitnesses misidentifi ed 76% of the exon-
erees (190 of 250 cases).11 Nevertheless, when I began studying these cases,
I was worried that I would not be able to say very much about eyewitness
misidentifi cations. After all, we do not often have rec ords of what trans-
pired during the identifi cation procedures, because police usually do not
document them. The trial rec ords are often all that we have to go on. To
my surprise, I found that the trial rec ords alone told a troubling story. In
the vast majority of these cases, the seemingly powerful eyewitness testi-
mony was fl awed. I obtained trial materials for 161 or 85% of the 190 exon-
erees (out of 250) who were misidentifi ed by eyewitnesses. Two related
problems recurred: suggestive identifi cation procedures and unreliable
identifi cations.12 In this chapter, I will focus on each problem in turn.

First, eyewitness memory is not just fallible; more important, it is mal-
leable. I did not expect these trials to describe police using suggestive

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 49

procedures. Yet in 78% of these trials (125 of 161 cases), there was evi-
dence that police contaminated the eyewitness identifi cations. Many of
these eyewitnesses were asked to pick out the suspect using suggestive
methods long known to increase the chances of an error. Police either
made remarks that indicated to the eyewitness who should be selected
during a lineup, used showups where the eyewitness was asked whether
a single person was the attacker, or used lineups in ways that made it
obvious whom the eyewitness was supposed to choose because the de-
fendant stood out.

Suggestion is related to the second problem, false certainty. Although
these witnesses were confi dent at trial, their false confi dence may have
been created by what came before. Not only were eyewitnesses in fact
wrong, having identifi ed an innocent person, but at trial the fl aws in
their memory were often quite glaring. I expected to read that these eye-
witnesses were certain at trial that they had identifi ed the right person.
They were. I did not expect, however, to read testimony by witnesses at
trial admitting that they earlier had trouble identifying the defendants.
Even if there had been problems, eyewitnesses might not recall their
hesitance at the time they fi rst identifi ed the defendant. Yet in 57% of
these trial transcripts (92 of 161 cases), the witnesses reported they had
not been certain at the time of the earlier identifi cations. Witnesses said
that they had been unsure when they fi rst identifi ed the defendant, or
they had identifi ed other people, or they had trouble making an identifi –
cation because they had not seen the culprit’s face. While not my focus
here, almost all cases also involved defendants who looked very much
unlike the person the victim initially described, with glaring diff erences
in their height, weight, or hair, or in other telling physical features like
tattoos or gold teeth or scars. Witnesses also admitted they could not get
a good look at an attacker because it was dark or the attacker hid or or-
dered them not to look. As the months passed after the crime, their mem-
ory had also faded.

This is not to say that police necessarily engaged in purposeful mis-
conduct. Police likely believed they had a guilty suspect. They may have
had good reasons to place the exoneree in a lineup. Some of the proce-
dures they used might seem like a good idea, if one were not trained
about the nature of eyewitness memory. For example, police may have

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

50 Eyewitness Misidentifi cations

conducted repeat lineups because they wanted to be sure they had the
right person, and not to cement an error. They may have tried to comfort
the victim by mentioning that they had already arrested a suspect, not
realizing that doing so aff ects the reliability of the identifi cation that fol-
lows. Moreover, as we will see, lenient criminal procedure rules do not
require police to do anything diff erent than what they did in many of
these cases. With judges taking a hands- off approach, most police de-
partments have few procedures and little formal training on eyewitness
identifi cations, despite the importance of eyewitness identifi cations in so
many criminal cases.

Yet for de cades social scientists have warned of the fragility and mal-
leability of eyewitness memory and the ways that commonly used police
practices create unnecessary risks of tragic errors. As Professor Gary
Wells, a psychologist who helped pioneer the study of eyewitness mem-
ory, puts it, “eyewitness testimony is among the least reliable forms of
evidence and yet persuasive to juries.”13 Eyewitness misidentifi cations
are common. Data suggest that in actual police lineups, witnesses choose
known innocent “fi llers” an average of 30% of the time.14 Upward of
75,000 suspects are identifi ed each year by eyewitnesses in the United
States.15 These exonerations raise disturbing questions about how ac-
curate those eyewitnesses may be.

Eyewitnesses and Victims

Who were the eyewitnesses who misidentifi ed innocent people? I was
surprised to discover that 36% (68 of 190 exonerees) were identifi ed by
multiple eyewitnesses, some by as many as three or four or fi ve.16 At least
290 eyewitnesses misidentifi ed the 190 exonerees who had eyewitness
evidence in their cases. Of those, 73% (213 of 290) were victims, typically
victims of a rape. In cases involving multiple eyewitnesses, prosecutors
emphasized that the evidence was particularly strong. For example, in
Cody Davis’s case, the prosecutor told the jury that “in this case you
have two witnesses, again, that made the identifi cation of Cody Davis,
the defendant, on their own, in separate locations, at separate times they
did that.”17 One wonders how separate individuals could have all made
the same mistake. In some cases, the exoneree may have become a sus-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 51

pect because his appearance resembled the perpetrator; eyewitnesses
could have made the same, but understandable, mistake.18 However,
poor police procedures could also have tainted all of the witnesses.
Studies suggest that one eyewitness can be infl uenced by the identifi ca-
tion of another.19 For example, in Richard Alexander’s case, as in several
others, the eyewitnesses all viewed photo arrays together in the same
room.

Of the exonerees convicted based on an eyewitness identifi cation, the
vast majority, 84%, were convicted of rape (159 of 190 cases).20 In 93% of
the cases of an exoneree convicted of rape (159 of 171 cases), a victim tes-
tifi ed as an eyewitness at trial. In contrast, in the fewer murder cases that
had eyewitnesses, the eyewitnesses were typically not victims, but rather
people who said they saw the culprit near the crime scene.21 Rape cases
might not have gone forward had the victim not been able to identify the
defendant; the victim identifi cations were often crucial to closing the
case.

Much of this chapter focuses on testimony by victims of rape. Only
fi fteen of the exoneree rape cases had identifi cations by acquaintances
(in contrast, most reported rapes involve acquaintances).22 These vic-
tims were called on to identify total strangers. Moreover, victims saw
these strangers during rapes in which they often feared for their lives.
High levels of stress strongly aff ect the ability to accurately identify a
person. While most of these victims were women, women do not have
more trouble identifying people as eyewitnesses; the evidence is quite
to the contrary.23 In addition, most of these identifi cations were cross-
racial, with a white victim and a black defendant, unlike most rape pros-
ecutions, which are same- race. As I will discuss later in this chapter, that
may be part of the reason why so many witnesses misidentifi ed the
culprits.

To protect the privacy of victims, their names are not used, except in
a few instances in which the victims themselves have spoken extensively
about their cases. Criminal investigations and trials can be extremely
diffi cult for the victims and their families, who must revisit violent
events. Wrongful convictions severely impact victims and their families
in new ways. They learn that the true culprit was not convicted, but re-
mained or still remains at large. Sometimes victims played a role in the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

52 Eyewitness Misidentifi cations

error, often very much inadvertently, but as I will explain, police could
have prevented some errors by more carefully preserving the victim’s
memory.

Police Procedures

How do police determine whether an eyewitness can identify a culprit?
Police use a range of techniques to test an eyewitness’s memory. If a sus-
pect is found shortly after the crime, police may present that suspect to
the eyewitness directly. Such a one- on- one procedure, called a “showup,”
is inherently suggestive. Police may use such a procedure only in the
hours immediately following an incident, in order to quickly identify the
perpetrator or rule out the suspect and continue their investigation.24 If
police do not quickly locate a suspect, they may show a witness books or
computerized collections of mug shots. If that also fails, police may ask
the witness to work with a police sketch artist or with a computer pro-
gram to generate a composite image. That composite may be used in
wanted postings to try to apprehend a suspect.

In at least forty- six exonerees’ cases, eyewitnesses initially worked with
a sketch artist or a police offi cer using an “Identikit” or computer pro-
gram to make a composite image of the attacker. Studies suggest compos-
ites are highly problematic. People have far more diffi culty giving an ac-
curate verbal description of individual facial features than recognizing an
entire face.25 In Allen Coco’s trial, the victim said that only “after the
composite” did she begin “looking and putting pieces together myself of
what I remembered of that person’s face.”26 The pro cess of creating the
composite may have distorted her memory.

If police eventually locate a suspect, they conduct an identifi cation
procedure to test the eyewitness’s memory. In a live lineup, a suspect
stands in a row of “fi ller” individuals and the witness looks at the group
from behind one- way glass. In the past few de cades, police have mostly
stopped using live lineups because it is so time- consuming to fi nd people
who look similar to a suspect. Instead, they use photo arrays, typically
using a set of six photos (a “six- pack”). Among the 161 exonerees studied,
118 were identifi ed in a photo array, 61 in a lineup, 53 in a showup, and 46
from composite images prepared by witnesses. Most witnesses viewed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 53

more than one type of procedure, and viewing multiple procedures may
have reinforced false identifi cations.27

Although most of these crimes and police investigations occurred in
the 1980s, some were more recent. In any case, police practices have not
signifi cantly changed in most departments. Police departments have
detailed procedures, manuals, and training on a host of subjects, rang-
ing from traffi c stops to use of force. Yet many still do not have any writ-
ten procedures or formal training on how to conduct lineups or photo
arrays. For example, in James Ochoa’s case, an offi cer saw nothing wrong
with conducting a showup six hours after an incident, rather than using
a photo array. At a preliminary hearing, he admitted that he had never
been trained on how to create a photo array. Ochoa’s lawyer sarcasti-
cally asked, “Do you think that at any time between now and the
jury trial you will have a chance to refresh yourself on proper lineup
procedure?”28

The Supreme Court’s Due Pro cess Test

If techniques like showups are so prone to error, why are police still al-
lowed to use them? In Manson v. Brathwaite (a case involving a prison
custodian named Manson, not the famous serial killer), the U.S. Supreme
Court noted the dangers of suggestive identifi cation procedures.29 The
Court had long recognized “[t]he vagaries of eyewitness identifi cation,”
where “the annals of criminal law are rife with instances of mistaken
identifi cation.”30 Accordingly, in Manson the Court affi rmed that the Due
Pro cess Clause of the Constitution embraces a right to be free from unduly
suggestive eyewitness identifi cation procedures, such as showing the eye-
witness a single photograph of the suspect, or telling the eyewitness whom
to identify in a lineup.

However, the Court in Manson added a caveat that undercut the power
of that holding. Even if the police engage in suggestive procedures so po-
tentially suggestive that they violate due pro cess, the identifi cation may
still be admitted at trial if it is otherwise “reliable.” The Court’s fl exible
multifactor test for what counts as “reliable” provides almost no guidance
to police or judges. Even the aggressive use of suggestion by the police in
Habib Abdal’s case could be excused if the judge found the identifi cation

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

54 Eyewitness Misidentifi cations

“reliable” because the witnesses seemed certain by the time of the trial.
In practice, the test leaves it to the defense lawyers to bring out the weak-
ness of the identifi cation during cross- examination, and to the jury to
decide if the eyewitness is reliable. In the exonerees’ cases, defense law-
yers typically did cross- examine eyewitnesses, often aggressively. Yet the
apparent confi dence of these eyewitnesses may have trumped evidence of
suggestion and unreliability. Jurors place great weight on the apparent
confi dence of an eyewitness, although, as I will discuss, confi dence at the
time of trial is very misleading.31

Suggestive Procedures

We know, then, that police have expansive authority to encourage eye-
witnesses to identify the person the police already suspect committed a
crime, and that police are in a position to suggest, sometimes uninten-
tionally and sometimes overtly, which person the eyewitness should
choose. Is that what happened in these exonerees’ cases?

We do not know what happened and what was said at these identifi ca-
tions, apart from what witnesses later recounted at the trial. Police typi-
cally do not document identifi cation procedures, aside from keeping a
copy of the photo array, or perhaps a form the witness signed. In only
four cases was the entire procedure recorded. Experts have long recom-
mended that police record eyewitness statements and identifi cations to
create an accurate record of what eyewitnesses say and how confi dent
they are.32 Defense attorneys were present in only sixteen of these iden-
tifi cation procedures. Defendants have a right to have a lawyer present at
postindictment lineups, but most of these cases involved photo arrays,
not live lineups, and most involved identifi cations conducted well before
an indictment. The defense lawyers could only try to fi nd out what actu-
ally happened at an identifi cation by questioning the eyewitness and the
police who conducted the procedures at trial.

Given how poorly documented eyewitness identifi cations usually are,
I was particularly surprised that even based on what the eyewitnesses
and police said at trial, it was clear these lineups were riddled with ir-
regularities and procedural fl aws— though perhaps no more than is typi-
cal. These exonerees’ trials typically involved procedures long known to

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 55

create risks of suggestion, ranging from showups in 33% of the trials
with eyewitness testimony (53 of 161 trials), to biased lineups in 34% (55),
to suggestive remarks in 27% (44), to hypnosis in 3% (5). While some
cases had more than one type, 78% (125) involved one of those types of
suggestion. The sections that follow will discuss each of those suggestive
procedures in turn.

Showups

One of the most obviously suggestive procedures simply shows an eye-
witness a single photograph or single suspect— a “showup.” Showups
outright tell the eyewitness who the suspect is and for that reason have
been “widely condemned.”33 Some 33% of the trials obtained with eye-
witness misidentifi cations (53 of 161 trials) involved a showup, either
live or by showing a single photo.34

Showups are clearly allowed and justifi ed only shortly after a crime oc-
curs, when a suspect is found near the scene. Under such special circum-
stances, police have an important public safety reason to detain a poten-
tially dangerous person immediately; otherwise they would have to let
him go while they tried to prepare a lineup procedure. Further, police
can ideally rule out innocent persons quickly without an arrest. Only
eleven exonerees were identifi ed in crime scene showups. For example,
Gene Bibbins was identifi ed in a showup fi ve to fi fteen minutes after the
assault. Similarly, Anthony Robinson was identifi ed a half hour after the
assault.

Such cases are noteworthy. Those victims misidentifi ed the defendant
right after the crime when their memory was still fresh, perhaps in part
because showups are inherently suggestive. Unlike lineups, which in-
clude fi llers who can help to protect an innocent suspect from a misiden-
tifi cation, showups provide no alternatives for the victim to choose.35 Not
only is a showup by its nature suggestive, but it can also be conducted
even more suggestively. In exoneree Gene Bibbins’s case, for example, a
showup was made more suggestive where police placed him in a squad
car for the victim to view, after they had just showed her his apartment,
where she identifi ed a radio as the one stolen from her during the
assault.36

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

56 Eyewitness Misidentifi cations

In addition, forty- two exonerees were identifi ed in showups that were
not conducted in the crime scene area right after the crime. Those show-
ups were suggestive and had no justifi cation, because they occurred long
after the crime, when police could and should have easily prepared a
lineup. Six exonerees were identifi ed for the fi rst time in court at a hear-
ing or at trial. Police made no eff ort to use a lineup or photo array to test
the eyewitness’s memory in the months after the incident. Perhaps they
suspected the eyewitness could not identify the attacker if given a choice.

Many of these showups were blatantly suggestive and unnecessary,
yet they were readily admitted at trial. Willie Davidson was identifi ed in
a showup in which police staged a sort of per for mance. The victim had
only seen her assailant in the dark with his face covered by a stocking.
Police repeatedly pulled a stocking over Davidson’s head and off his
head, with the victim present. Each time they asked the victim to iden-
tify him, asking, “is this it, is that it.” Although the victim had not seen
her assailant’s face, after that display, she identifi ed Davidson as the
culprit.37

Similarly, in Alan Crotzger’s case, a showup was used precisely be-
cause an eyewitness could not otherwise identify him. The eyewitness
had failed to identify him in the photo array. The police then took his
photo out of the array and presented it alone in a showup. The eyewit-
ness then identifi ed the photo and at trial said “there was no doubt” that
it was of the attacker. When the defense brought a challenge at trial, the
judge ruled, “I don’t believe [the identifi cation] in any way was tainted.”38

A showup that directly subverted judicial review was conducted in
Neil Miller’s case. His defense attorney was concerned that earlier photo
arrays had been conducted in a suggestive manner. The victim had se-
lected two photos from an array, but was not sure if she could pick either.
The fi rst of the two was a six- year- old photo of Neil Miller taken when he
was only sixteen. The second was of another man. The detective show-
ing her the photos manipulated the situation by instructing her, “if she
had a fi rst impression, that the best thing to do was go with her fi rst im-
pression.” Having picked Neil Miller’s photograph fi rst, the victim then
identifi ed Neil Miller’s photo.39

The defense lawyer scheduled a hearing to argue that a new photo ar-
ray should be conducted. In the court house, just before the hearing was

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 57

to take place, the prosecutor walked the victim past Neil Miller in the
hallway outside. Yet even after being told that her attacker might be in
that hallway, she still was not sure, and just thought he might be her at-
tacker. She was not positive until she followed Miller into the courtroom
(where it was obvious who he was), looked at him again, and said, “This
is him.” Now the hearing to request a new lineup was a pointless exer-
cise. Even if the judge ordered that a new photo array be conducted, due
to both of the prior suggestive procedures the victim would likely again
pick out Neil Miller, and then testify with confi dence before the jury that
he was the person who assaulted her.

Additional eyewitness misidentifi cations, other than from showups,
occurred after witnesses were exposed to single images of the exoneree.
Some witnesses also saw wanted posters or media coverage displaying a
composite drawing of the suspect. In Michael Blair’s case, for example,
all three eyewitnesses had already seen photos of him in the local news-
paper or on tele vi sion before picking his photo from a lineup.40

Stacked Lineups

Although showups may be highly suggestive, a lineup may also be sug-
gestive if it is not set up fairly. At least 34% of the trials obtained with
eyewitness testimony (55 of the 161 trials) were biased, or stacked to make
the suspect stand out. If some of the fi llers in the lineup do not look any-
thing like the description of the culprit, or the suspect, then the lineup is
not a sound test of the eyewitnesses’ memory.41 I say at least 34%, because
problems with the photo array or lineup may not have been brought out at
trial. Once again, it was surprising how often problems with the identifi –
cations were raised at trial.

The case of Marvin Anderson, who was convicted in 1982 of a brutal
rape, involved a clear example. The victim, a white woman, had told
police at the hospital after her attack that she could never forget the face
of her attacker as long as she lived. Her attacker was a light- complexioned
black man with short hair, medium height, and a thin mustache. He had
told the victim that he dated “a white girl.” The investigating offi cer
knew of only one black man in town who lived with a white woman, and
that was Marvin Anderson. The offi cer did not have a photo of Anderson

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

58 Eyewitness Misidentifi cations

on fi le, because he had no prior record. So, the offi cer obtained a photo
ID from his employer. The victim was given a group of photographs to
review. At trial she said she “didn’t notice any diff erence in them.” How-
ever, unlike all of the black- and- white mug shots shown to the victim,
Anderson’s work identifi cation was a color photograph, and also unlike
the others it had his employee number on it. She picked Anderson’s
photo.42

Just an hour after being shown those photos, the victim was then
shown a live lineup and asked “if she could pick out the suspect.” Ander-
son was the only person in the lineup repeated from the photo array.
The people in that lineup were chosen because they looked similar to
Anderson. The victim looked at the lineup twice and picked Anderson
both times.43

Anderson’s trial attorney moved to exclude the pretrial identifi cation,
arguing that it was unduly suggestive. He pointed out that Anderson “is a
dark- skinned person,” and not light complexioned (by the time of trial
the victim now remembered describing her attacker as medium complex-
ioned). He argued that Anderson was the only person shown in both the
photo array and lineup. He did not point out the obvious— that Ander-
son’s was the only color photo in the array. The judge ruled that “there’s
been no showing here that the photographs were irregular or were arranged
in any irregular way or, uh, were presented in any way to, uh, identify a par-
tic u lar person.”44

Another man whose photo was included in the photo array, but who
was not in the lineup, was named John Otis Lincoln.45 He turned out to
be the actual culprit. The victim had failed to identify her actual attacker
in the photo array. After Anderson had served fi fteen years in prison,
DNA testing in 2001 not only exonerated Anderson but inculpated Lin-
coln, who was then convicted. This came as no surprise to Anderson,
who had heard from many in the community that Lincoln committed the
crime. Lincoln had stolen the bicycle used by the assailant a half hour
before the rape. Lincoln even confessed in 1988 at a hearing, but the trial
judge called Lincoln a liar and denied the motion for a new trial.46

The most common type of fl aw in these lineups was to construct them
so that the exoneree stood out. In Ronnie Bullock’s case, the only man in
the six- person lineup with distinctive bumps on his face like the victim

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 59

had described was Bullock.47 In Lonnie Erby’s case, he was the only
person in the array who had facial hair.48 The list goes on and on. Other
lineups included markings identifying the suspect. Thomas Doswell’s
photo had an “R” written on it, signifying he had previously been charged
with rape.49

In at least fourteen cases, the exoneree was the only person repeated
in multiple viewings. Studies have shown that such repetition can ce-
ment a false identifi cation by making a wrong person look more famil-
iar.50 For example, in Larry Fuller’s case, the victim was shown two ar-
rays, and his was the only photo repeated. After the fi rst array, the victim
said Fuller looked “a lot like the guy” but could not identify him. The
second time, although the attacker had no facial hair, while Fuller had a
full beard, the victim explained, “I knew that was the face,” and she
“looked at the picture again and I put my fi nger over the part, the hair,
and then I could identify him.” She did say she was “very slow in identi-
fying him because [she] did not want to identify an innocent man.”51
Had police repeated photos other than Fuller’s, maybe she would not
have done just that.

A voice lineup, nicknamed an “earwitness identifi cation,” may be even
less reliable than eyewitness identifi cations. Forty- four eyewitnesses
identifi ed the defendant not just by sight, but by voice, in a “voice lineup”
where the police had persons speak a phrase that the attacker had uttered.
What research exists on such voice lineups suggests that they may be
similarly susceptible to the infl uence of suggestion.52 Other witnesses
identifi ed voices in the voice equivalent of a showup, as in Habib Abdal’s
case, where the witnesses were asked if they recognized a single suspect’s
voice.

Suggestive Remarks

What ever the type of procedure used, police may contaminate the iden-
tifi cation by telling the eyewitness whom to pick out of a lineup or saying
something to reinforce an identifi cation. In another 27% of the trials in-
volving eyewitness testimony (44 of 161 trials), it came out at trial that
police made suggestive remarks. We do not know how many more cases
involved suggestive remarks not brought out at trial; what is surprising

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

60 Eyewitness Misidentifi cations

was how often the eyewitnesses and offi cers openly described use of sug-
gestion at trial.

Suggestion can occur before the lineup begins. The American
Psychology- Law Society recommends that police instruct a witness that
the suspect might or might not be present in the lineup or array. Other-
wise the witness may assume that the culprit is present and pick the per-
son who looks most like the culprit. The failure to give those admonitions
substantially increases the risk of a misidentifi cation.53 Although some
states and police departments have adopted lineup reforms in response to
these exonerations, as I will discuss in Chapter 9, most police depart-
ments still have no written procedures and no set of standard instruc-
tions. At the time these exonerees were convicted, police might have ca-
sually shown the witness photos and just asked if the suspect was there.
In only fourteen trials was there evidence that the witness was specifi cally
told that the suspect might not be in the lineup.54 The danger is that the
witnesses may logically assume that the police brought them to the sta-
tion for a reason— because the police arrested the culprit. Indeed, many
of these witnesses said they assumed they had been asked to come to the
police station to look at the culprit.

In some cases, police further biased the eyewitnesses by telling them
beforehand that they had apprehended a suspect whom they had placed in
the lineup.55 For example, in Robert Clark’s case, police told the victim
that not only did they have a suspect but that he was caught driving her
car. In Richard Johnson’s case, the offi cer called the victim and said, “you
need to come in for a line- up, we think we have him.”56 In Robert McClen-
don’s case, the victim recalled at trial that “they just told me to pick out the
person who done that to me.”57 In many trials the issue was not raised, and
since almost none of these lineup procedures were documented, we do not
know whether police gave biasing instructions to the witnesses in even
more cases. Again, what is surprising is how often police or the victim did
recall and describe biasing instructions at trial. Police may have made
such remarks to victims to comfort them and to encourage them to come
in to see a lineup that they might have been quite afraid to look at.

In some cases, police made remarks that were perhaps not so well in-
tentioned. A few eyewitnesses even recalled police outright telling— or
demanding— that a par tic u lar person be identifi ed, and eyewitnesses, or

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 61

even the police, openly described this police misconduct at the trial in
their testimony. Habib Abdal’s case is an example of aggressive police
suggestion; recall how the victim said that police kept pressuring her to
identify the suspect. There were others. In Alejandro Dominguez’s case,
police told the victim, “watch the one sitting there. Tell me if that is the
one.”58

In Richard Johnson’s case, the victim looked at his photo in an array
for about fi ve minutes but was not certain; “I said I wanted to see him, I
thought that from a picture it was hard to tell.” The victim testifi ed at
trial: “After a while passed and I hadn’t said anything, they said, you
know, you have been staring at that photograph a long time and I said I
think it’s him.”59 In Gilbert Alejandro’s case, before the lineup the victim
was told that the suspect was named Gilbert Alejandro, whose family she
knew, so she knew what he looked like. The lineup was also apparently
designed so Alejandro would stand out; she testifi ed that the others in the
lineup were “fatter than him.”60

In Larry Mayes’s case, the police offi cer testifi ed that the victim looked
at the live lineup and “She said, ‘No. 5.’ ” This was a problem: No. 5 was
a fi ller. The offi cer recalled, “So I grabbed her and I said, ‘Come on in the
room, and shut the door and look at everyone and see if he is in here.’ ”
The victim then “studied all fi ve subjects and made the statement, ‘No. 4.
That’s him. It looks like him. Yes, that is him.’ I asked her, ‘Are you posi-
tive? . . .’ And she said, ‘Defi nitely, yes.’ ”61 Larry Mayes was No. 4. Not
only was the victim pressured to disregard her initial nonidentifi cation,
but years later the victim revealed that police had hypnotized her in an
eff ort to enhance her memory.

In fi ve exonerees’ cases, eyewitnesses had been hypnotized prior to
identifying the defendant.62 Lesly Jean had his conviction reversed be-
cause the fact that hypnosis was done was concealed from the defense,
just as in the Mayes case; indeed, the victim may have fi rst identifi ed him
while under hypnosis.63 Hypnosis was initially developed as a therapeu-
tic technique, but by the 1970s law enforcement began to use hypnosis to
“recover” or “refresh” memories of forgotten traumatic events. However,
studies show that hypnotism does not increase the accuracy of eyewit-
ness testimony and in fact can make a witness far more susceptible to
suggestion.64

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

62 Eyewitness Misidentifi cations

Police also made statements after an identifi cation that would tend to
enhance the victims’ confi dence. After the victim had identifi ed Larry
Fuller, police told her that Fuller had previously been in prison for armed
robbery.65 In Anthony Green’s case, the attacker had told the victim that
his name was Tony, and after she identifi ed Green’s photo, police said,
“this guy’s name is Anthony, Tony.”66 In Ronald Cotton’s case, the victim
later recounted that “When I picked him out in the physical lineup and I
walked out of the room, they looked at me and said, ‘That’s the same guy,’
I mean, ‘That’s the one you picked out in the photo.’ For me that was a
huge amount of relief.”67

Several cases highlight how untrained police can unintentionally re-
inforce a mistaken identifi cation. In Thomas McGowan’s case, the vic-
tim identifi ed his photo from an array. The victim explained that she
slowly reviewed each of the photos, looking at them for “quite a long
time.” Reviewing one photo for the second time, she told the offi cer, “I
think this is him.” She recalled that he “said I have to be sure, and when
I picked it up the second time, I said, ‘I know this is him.’ ”68 While the
offi cer may have been trying to be cautious, telling the victim that she
had to “be sure” encouraged her to express more certainty. If the offi cer
had instead just asked how certain she was, he could have obtained
more objective information. Similarly, in Larry Johnson’s case, the vic-
tim was asked after she identifi ed him whether she was certain. She
said, “Yes,” and the offi cer told her that she “had to be a hundred percent
sure.”69

In none of these cases did anyone claim the police gave any nonverbal
cues to the eyewitness. A lineup administrator who knows which person
is the suspect can shake her head, nod, point, or slow down or speed up
the procedure to indicate the suspect. Such cues can be unintentional
and unconscious. We do not know from reading these trial transcripts if
other types of suggestion occurred.

The Reliability Test

In many of these cases, then, police engaged in the kind of suggestion
that the Supreme Court has said is impermissible, although nevertheless
excusable if there is enough evidence that the identifi cation is “reliable.”

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 63

A close look at the trial materials also sheds some light on why judges
may have decided that these identifi cations passed muster at the time of
trial. Judges can rule that an eyewitness identifi cation may not be pre-
sented to the jury, if they decide that it was both tainted by suggestion
and unreliable. However, the test that they use to evaluate reliability is
deeply fl awed and has been much criticized by social scientists. The
Supreme Court’s test in Manson asks the judge to look at fi ve factors to
assess reliability, in no par tic u lar order of importance: (1) the eyewit-
ness’s certainty; (2) the opportunity to view the perpetrator; (3) the de-
gree of attention; (4) the accuracy of the description the witness gave;
(5) the time between the incident and identifi cation. The test is so fl exible
as to be toothless, and, it also includes some factors that do not actually
give judges good information about the reliability of eyewitness identifi ca-
tions. I start with eyewitness “certainty,” because the role played by that
factor was most problematic in these cases.70

False Confi dence

The problem with focusing on eyewitness certainty is that the certainty
of eyewitnesses by the time of trial may be completely diff erent from
their certainty at the time their memory is most reliable— when they fi rst
identifi ed the defendant. These exonerees’ cases illustrate that problem
vividly. Almost all of the eyewitnesses in these trials expressed complete
confi dence at trial that they had identifi ed the attacker. The eyewitness
in Steven Avery’s case testifi ed, “There is absolutely no question in my
mind.”71 In Thomas Doswell’s case, the victim testifi ed, “This is the
man or it is his twin brother” and “That is one face I will never forget.”72
She rated, on a scale from one to ten, her certainty as ten. However, she
described her attacker having a full beard, which Doswell did not have
when picked up three hours after the attack. In Dean Cage’s case, the
victim was “a hundred percent sure.”73 One hundred percent was a
modest fi gure. In Willie Otis “Pete” Williams’s case, the victim said she
was “one hundred and twenty” percent sure.74 The victim in Donald
Wayne Good’s trial gave a chilling depiction of what it was like to see
him in the lineup. She said: “Instantly, I knew it was him. Instantly. And
my whole body reacted. I just fell, I collapsed. I grabbed my husband. If

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

64 Eyewitness Misidentifi cations

he hadn’t been there, I would have been on the fl oor.” She said, “My
God, it’s him.”75

When did these eyewitnesses become so confi dent? I assumed they
would recall having always been sure of their identifi cation. I did not
expect to learn that many of these witnesses were actually not so sure
when they fi rst identifi ed the defendant, long before the trial. Police of-
ten do not document eyewitness certainty at the time of an identifi cation,
yet their certainty at that time is most indicative of the accuracy of the
identifi cation.76 In 57% of the trials with eyewitness testimony (91 of 161
trials), the witnesses had earlier not been certain at all, a glaring sign that
the identifi cation was not reliable. In 40% of the trials with eyewitness
testimony (64), the trial transcripts revealed how an eyewitness did not
initially identify the defendant, but rather a fi ller, another suspect, or no
one at all. In 21% of the trials with eyewitness testimony (34), an eyewit-
ness admitted initial uncertainty about the identifi cation. In 9% of the
trials with eyewitness testimony (15), an eyewitness reported not having
seen the culprit’s face at all. Some cases had more than one type of evi-
dence of eyewitness uncertainty.

It is powerful to see a person point to the defendant in court, explain-
ing that his memory is like a photograph, and that man is the one. But
human memory is not like a photograph; it is dynamic and fragile. While
jurors are particularly impressed with confi dent eyewitnesses, certainty
is one of the most malleable features of human memory. Social science
research has for de cades shown how suggestion or poor procedures can
generate false and infl ated confi dence.

Suggestion and certainty are related. Only 12% of the cases reviewed
(20 of 161 cases) had no evidence of police suggestion or of clear unreli-
ability involving prior uncertainty. The other 88% involved evidence at
trial of suggestive procedures, evidence of clear unreliability, or both (141
of 161 cases). And still more cases may have involved suggestion or unreli-
ability that the trial transcripts do not reveal.

These eyewitnesses’ certainty may have increased with each identifi –
cation procedure used, if police conducted not one but multiple identifi –
cations. In addition, not only identifi cation procedures but routine prep-
aration for trial increases an eyewitness’s certainty. A prosecutor will
rehearse each of the questions, walking the eyewitness through every-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 65

thing that happened, everything she saw, and everything she told police.
The eyewitness would then see the defendant again in court, perhaps at
pretrial hearings, and then fi nally at trial, sitting at the bench with his
lawyer. At trial, the prosecutors typically followed the same script, dis-
cussing chronologically everything that happened, and culminating with
the courtroom drama of the eyewitness pointing to the defendant from
the stand— and then at the very end of the questioning, asking the eyewit-
ness if she is certain.

Every step in the criminal pro cess that comes before that question,
“how certain are you that the man you have pointed to is the man you
saw,” enhances the certainty of the eyewitness. The Manson test’s focus
is wrong. Instead, the way to prevent contamination of the early and
most reliable impression in the eyewitness’s mind is to take a clear state-
ment from the eyewitness immediately after the identifi cation proce-
dure, to fi nd out how confi dent he was then. That is what social scien-
tists have long recommended, to preserve eyewitness memory just as
police carefully preserve trace evidence from a crime scene to ensure its
integrity.

Although the most crucial and accurate information is how certain the
eyewitness was at the time of the fi rst identifi cation, at that time most of
these eyewitnesses were not sure.77 For example, Calvin Johnson was ar-
rested in 1983 for two rapes in College Park, Georgia. Both victims where
white, while Johnson was black. The fi rst rape victim picked Johnson’s
photo out of a photo array, but when she observed a live lineup, she
picked a fi ller. The second victim did not identify Johnson from the
photo array, saying she “wasn’t sure,” but did identify him at the live
lineup. The photo used in the array showed him as clean- shaven. How-
ever, at the time of the crime, Johnson had a full beard and mustache, as
his boss and others testifi ed. Yet both victims had described their at-
tacker as clean- shaven, or at most as having had some stubble.78 Those
inconsistencies should have made the victims uncertain, and in fact at the
time of the initial procedures, they were not certain.

Yet at trial both victims identifi ed him. One explained that the reason
she didn’t identify Johnson when looking at the lineup was that “I just
refused to let myself look at him anymore, and I just picked another per-
son.” She said, “I picked the wrong one, and I knew it,” thus admitting

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

66 Eyewitness Misidentifi cations

that she had falsely accused an innocent person.79 At trial, however, she
said she was certain: “I know I’m right. I know that’s him.”

The second victim, against Johnson’s lawyer’s objection, had been
allowed to hear the fi rst victim’s testimony at the preliminary hearing,
and the second victim admitted that after that hearing she remembered
more details, including a ridged sweater the attacker wore. She now tes-
tifi ed she was certain that Johnson was her attacker, saying, “It’s the man
with the beard sitting right there” and “you’ll get your time.” After leav-
ing the stand, she lunged at him, yelling out: “You stupid bastard.”80

In these exonerees’ cases, we now know that the accurate response,
when viewing these lineups, was “none of the above”— except in three
remarkable cases. In John Jerome White’s case, the victim picked White
out from a photo array. She had been told by the detective that “he had
caught somebody” and she should come see a live lineup, where she again
identifi ed White. The seventy- four- year- old victim said she was “almost
positive” he was her attacker, even though there was very little light dur-
ing the attack and she was not wearing her prescription eyeglasses.
White’s lawyer argued that she was initially uncertain, but became more
certain as police showed her White’s photo, explaining, “Once she saw
that picture though, once that poison seed was planted it began to grow
and she became more positive as to the statements from then on . . . by
the time she got into the court, she was absolutely positive that was the
man.”81

DNA testing not only excluded White and exonerated him after
twenty- two and a half years in prison, but it confi rmed the guilt of an-
other man. That man, James Edward Parham, was coincidentally stand-
ing in the lineup in 1979, when the victim picked White. He was not a
suspect but just happened to be in the county jail at the same time and
was placed in the lineup. He had raped another woman in Meriwether
County six years later and eventually pleaded guilty to the crime for
which White was convicted.82 Parham looked very diff erent from White.
Parham was older, was much stockier, and had a round face, just as the
victim had initially described. Yet having already identifi ed White’s
photo and then been told by the detective that “he had caught some-
body,” she may have looked for White in the lineup and never carefully
looked at the actual perpetrator who was standing right there before her.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 67

We heard how the victim in Marvin Anderson’s case had failed to
identify the actual perpetrator, and there was also a third case like that.
Jennifer Thompson has described how she misidentifi ed Ronald Cot-
ton; she said, “I was certain and I was wrong.” She was sure Cotton was
the attacker even when at a postconviction hearing she saw Bobby Poole,
whose guilt was later confi rmed through postconviction DNA testing.
When she was asked if she had ever seen Poole before, she said, “I have
never seen him in my life. I have no idea who he is.”83

Unless the witness happens to mention doing so at trial, we may not
know if the witness had previously picked out the wrong person or had
been uncertain, since police do not typically document how certain eye-
witnesses are when they fi rst make an identifi cation. In 21% of the cases
(34 of 161 cases), the eyewitness’s initial uncertainty did emerge at trial.
In Bruce Godschalk’s case, one of the victims testifi ed at trial that she
was very certain: to the question “Are you positive in this identifi ca-
tion?” she answered, “Absolutely.” Yet at the lineup, she was unsure and
had to look three separate times before making an identifi cation. Police
departments often do not keep rec ords of such repeat viewings, but data
suggest that eyewitnesses make more errors each time they need another
viewing to make their identifi cation.84 In Anthony Capozzi’s case, a vic-
tim had identifi ed a fi ller in one lineup, but testifi ed at trial that she was
now sure that Capozzi was the attacker. She explained that she had felt
“rushed” and therefore identifi ed the fi ller.85 In 40% of the trials with
eyewitness testimony (64 of 161 trials), the eyewitness did not initially
identify the defendant but instead selected a fi ller, another suspect, or no
one at all.

Similarly, one of the eyewitnesses in Kirk Bloodsworth’s case had ini-
tially told the police offi cer that she never got a good look at the person she
saw with the murder victim. At trial, though, she repeated her identifi ca-
tion, stating that she saw him for forty- fi ve seconds, and that questioning
her identifi cation “really upset” her because “I’m trying to be so strong
and hold onto this case as much as I can. You don’t know how many doc-
tors I have seen to try to forget this, you know.”86

In Terry Chalmers’s case police conducted several lineups, and since
they documented each, we can see the victim’s confi dence changing over
time. At the fi rst photo array, the victim was shown a book of mug shots

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

68 Eyewitness Misidentifi cations

as well as a photo array. She said that two of them “could be” the at-
tacker. At a second photo array, she said that Chalmers was “very close”
to the person who raped her. She asked if he was the same person that
had been shown to her in the fi rst photo array, and police told her that “it
was the same person.” This reinforcement then prompted her to ask to
see him again in a live lineup. At a fi nal lineup, which was transcribed
with counsel present (and objecting to the fact that many in the lineup
did not resemble Chalmers), the victim picked Chalmers within “twenty
seconds of the viewing” and he was arrested.87

Although many of these eyewitnesses were uncertain initially, by the
time of the trial, almost all were positive they had identifi ed the right per-
son. In only four cases were the eyewitnesses not sure at trial that they
had identifi ed the right person.88 Jerry Miller had an eyewitness testify
that she was not sure but was willing to say that it “looks like him.”89 In
Jimmy Ray Bromgard’s case, the victim testifi ed that at the lineup, “when
I fi rst saw him I wasn’t too sure.” At trial, she testifi ed that she was “not
too sure.”90 Yet the offi cer who conducted the lineup off ered the jury a
diff erent account: “My impression she was quite sure, very sure.”91

Discrepancies in Descriptions

A second factor in the Manson “reliability” test, which does not in fact
provide very useful information about reliability, is the degree to which
the eyewitnesses’ initial description of the culprit matches the defen-
dant’s appearance. This factor seems at fi rst blush more objective than
the others. However, an innocent person could be placed in a lineup be-
cause police felt that he matched the description. More important, peo-
ple are far better at recognizing entire faces than accurately describing
par tic u lar features.92

Since police often did not carefully record initial descriptions of the
attacker, we cannot always tell if the descriptions shifted over time to
resemble the defendant. I was surprised at how often it emerged at trial
that exonerees did not resemble the initial descriptions of the culprit. In
unusual cases like Willie Jackson’s, a misidentifi cation was understand-
able: the true culprit was his brother, who looked like him. However, in
62% (100 of 161 cases), the exonerees looked very diff erent from the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 69

initial description. Witnesses had described culprits with major diff er-
ences from the exonerees in hair, facial hair, height, weight, scars, tat-
toos, piercings, teeth, and eyes.93

An extreme example of such discrepancies is Ricardo Rachell’s case,
in which the victim and another eyewitness both identifi ed Rachell but
had described an attacker who could speak clearly. Rachell had been
seriously injured by a shotgun accident years before that disfi gured his
face and allowed him to speak only with great diffi culty. At trial, Rachell
took the stand in his own defense and demonstrated his signifi cant
speech impediment. Nevertheless, Rachell was convicted and sentenced
to forty years in prison. Indeed, in thirteen cases, eyewitnesses failed to
describe prominent scars. Dean Cage had distinctive scars on his face,
which the victim never mentioned. Anthony Capozzi had a prominent
scar on his forehead and a face pitted from acne, which three witnesses
did not describe.

Eight exonerees had distinctive teeth. James Giles’s lawyer argued:
“Now, what is unusual to me is that he has got two gold teeth, two
prominent gold teeth. No mention of that. Yes, I think it is mistaken
identifi cation.”94 Carlos Lavernia had a silver front tooth with a star in
it, as well as numerous tattoos, none of which the victim described.
Eyewitnesses in forty- seven cases inaccurately described obvious facial
hair (or lack thereof ). In Steven Avery’s case, the victim said the at-
tacker had brown eyes, but Avery had blue eyes. She admitted: “I origi-
nally said brown eyes . . . When I picked out his photo in my hospital
room as I handed it to the Sheriff I commented, “He’s got blue eyes, I
was mistaken.”95 In sixty- seven cases, eyewitnesses did not accurately
describe the exoneree’s hair, getting the color or hairstyle completely
wrong. In still more cases, they failed to describe height or weight with
any accuracy. In eight cases, they did not accurately describe the com-
plexion of the attacker.

If judges took the Manson test seriously, perhaps they would have ex-
cluded some of these identifi cations based on discrepancies. However,
when defense lawyers tried to raise these discrepancies at trial, they had
no success. Judges or jurors may have intuitively understood that this
Manson factor is not a useful gauge of reliability. People can sometimes
accurately recognize a face, even if it is hard to accurately describe the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

70 Eyewitness Misidentifi cations

details of the facial features. Social scientists have found no connection
between ability to accurately describe a person and the accuracy of an
identifi cation. The discrepancies could instead be a symptom of the
suggestive procedures used in so many of these cases. Exonerees did
not look like initial descriptions of the culprit because they were the
wrong men. Yet despite those diff erences in appearance, they were picked
anyway.

Opportunity to View

The remaining Manson factors each focus on how good a look the eye-
witness had. Unlike the previous factors discussed, these do provide
judges with useful information about reliability. Yet in many of these
cases, eyewitnesses had a poor view of the culprit, but judges still admit-
ted the identifi cations.96 The next Manson factor asks what opportunity
to view the eyewitness had. While police often do not document how
good a look the eyewitness had, and by the time of the trial, eyewitness
accounts may be aff ected by suggestion, I was surprised at how many
cases involved poor opportunities to view the culprit.

In most of these cases, the assailant ordered the victims not to look at
them or covered their eyes, even threatening to kill the victims if they
looked at their faces. The culprits often kept their faces hidden from view
or wore masks. In general, most rapes occur at night and in a home.97 In
only twenty- one of these exonerees’ cases, an eyewitness described excel-
lent lighting conditions, while in most the lighting was poor. Only slightly
more than half of the cases involved a witness who reported seeing the
off ender’s entire body.

The victim in Dennis Brown’s case testifi ed that she “was memoriz-
ing his face, his eyes, his arms.” But when asked, “Could you see his
face?” she admitted she only “saw his complete face twice when his
mask dropped for like a— long enough— as soon as it dropped, he pulled
it back up.”98 In Gilbert Alejandro’s case, the victim could only see his
eyes and nose. In Fredric Saecker’s case the victim saw only the attack-
er’s build or outline.99 The victim in Clark McMillan’s case identifi ed
him at trial, but admitted that it was so dark in the woods where she was
assaulted that she could not even see blood on a cut on her own fi nger.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 71

Victims often saw attackers in the dark or in poor lighting, and under
terrifying circumstances in which they feared they would be killed.

Obviously, better police procedures cannot change the fact that violent
crimes commonly occur under conditions where the eyewitnesses do not
see the attacker for long, cannot see well, are focused on a weapon, and
experience great stress. All of those features can reduce accuracy of iden-
tifi cations.100 Social scientists call such factors “estimator variables,” as
opposed to “system variables” that improved police procedures can
change.101

Degree of Attention

The Manson factor regarding degree of attention is closely related to op-
portunity to view. At trial, the victims most often said that they only had
a chance to look at the attacker for a matter of seconds. Many of these
assailants disguised their own face, covered the victim’s face, remained
in the dark, or ordered the victim not to look at them. Fewer of the cases
involved long viewings of the attacker for many minutes.102 The actual
degree of attention of these eyewitnesses was likely far less than the short
viewings they recalled. Reports at a trial occurring many months later
are to be distrusted. Studies suggest that eyewitnesses overestimate the
amount of time they see an event, particularly a stressful event, and they
also overestimate how much of a person’s face they could observe.103 As
the victim testifi ed in Willie Jackson’s case, “It could have been a couple
of minutes, and it could have been a couple of hours because to me it
seemed forever. And I don’t want to exaggerate . . .”104

Suggestion aff ects memory of duration, as with everything else that
eyewitnesses recall. For example, witnesses are especially likely to say that
they had an adequate or excellent view of a culprit if police make remarks
that confi rm their identifi cation, like “Good, you identifi ed the suspect.”105
Recall how the victim in Habib Abdal’s case initially told police she was
not looking at her attacker at all when he spoke to her, but later, after police
pressured her to identify Abdal, she said she saw his face for thirty to forty-
fi ve seconds.

Most of these eyewitnesses were victims under conditions of extraor-
dinary stress, during brutal sexual assaults, often with the attacker holding

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

72 Eyewitness Misidentifi cations

a weapon. Understandably, victims will fi rst focus on a weapon, and not
the attacker’s face; as the victim in one exonerees’ case recalled, “I was
focused on the gun until he was physically on top of me.”106 Studies have
found that while moderate stress may improve concentration, “a witness
in a high stress situation is more likely to be an unreliable witness.”107
However, prosecutors argued and jurors may have believed the oppo-
site, that stress makes one concentrate better on the face of an attacker.

The Passage of Time

Finally, the Manson test includes as a factor the passage of time between
the incident and the fi rst identifi cation procedure. This is a useful factor,
because accuracy greatly decreases over time. Studies suggest this pro-
cess begins to occur within hours after an event.108 Many of these identi-
fi cations were made long after the crime. Recall that the victim misiden-
tifi ed Habib Abdal four and a half months after the assault. Only a handful
of cases, including the ten crime scene showups, included identifi ca-
tions the same day as the attack. Most of these eyewitness identifi cations
occurred weeks or months after the crime.109 Trials occurred even later,
typically more than a year after the crime. If judges took this factor seri-
ously, as well as the opportunity to view and degree of attention factors,
they might have excluded many of these identifi cations. The problem
with the Manson test is that it is so fl exible, and worse, it includes un-
sound factors. Judges may have put those other factors to the side and
focused on eyewitness certainty at the time of trial, a factor that is mis-
leading and does not predict accuracy.

Cross- Racial Identifi cations

There are other factors that are not included among the fi ve factors that
judges use, but that should be very important to judges when they evalu-
ate eyewitness identifi cations. Social scientists have long found an
“other- race eff ect,” in which the likelihood of misidentifi cation is higher
when an identifi cation is cross- racial.110 Judges should more carefully
examine cross- racial identifi cations, but in the past they did not, al-
though this is slowly changing, as I will discuss in Chapter 9. One expla-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 73

nation for why so many DNA exonerees were minorities, out of propor-
tion even to their overrepre sen ta tion among rape and murder convicts, is
that so many of these cases involved cross- racial identifi cations. At least
49% of the exonerees identifi ed by eyewitnesses had a cross- racial iden-
tifi cation (93 of 190 cases). In 71 of those cases, white women misidenti-
fi ed black men.111

Racial disparity is glaring in these exonerees’ cases. Many more
DNA exonerees were minorities (70%) than is typical even among aver-
age and already racially skewed populations of rape and murder con-
victs. Most striking, 75% of the exonerees who were convicted of rape
were black or Latino, while studies indicate that only approximately
30% to 40% of all rape convicts are minorities.112 Why were so many of
these exonerees minorities? Is being black a risk factor for being wrongly
convicted?

Those are not easy questions to answer, since several factors may
explain the racial disparity. One explanation may have to do with the
race of the victims, and not just the race of the defendants. The vast
majority of the victims in the 250 exonerees’ cases were white 72% (180),
while 24% of the victims were minorities (43 black, 15 Latino, and 1
Asian), with no information available in the remaining cases (11). In
par tic u lar, the cases of 49% of the exonerees convicted of rape (84 of 171
cases) involved black or Latino off enders and white victims. In contrast,
most sexual off enses, almost 90%, are committed by off enders of the
same race as the victim.113 There is a long history of discrimination sur-
rounding charges of rape involving white women and black men. There
is also some evidence that prosecutors may pursue more serious charges
in cases involving white victims and black men. There is evidence that
white jurors may be less sympathetic to black defendants, and may be
more likely to convict black defendants in crimes involving white vic-
tims. And more generally, these exonerees’ cases may refl ect racial dis-
parity in the criminal justice system. Black males are incarcerated at
many times the rate of white males, and minorities make up the major-
ity of prison inmates at any given time.114 Finally, a related explanation
is that, as social scientists have shown, eyewitnesses are more likely to
make mistakes in cross- racial identifi cations. Once an identifi cation is
made, it is more likely that the defendant will be convicted, even if the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

74 Eyewitness Misidentifi cations

eyewitness was initially unsure. Therefore, one might expect to fi nd many
cross- racial identifi cation cases in a group of innocent convicts.115

One victim described her diffi culty making a cross- racial identifi ca-
tion in Thomas McGowan’s case; she said that when she was looking at
the lineup, “I didn’t pick anyone out, but I showed the police offi cer a
person who looked similar to the defendant, because I didn’t know how
to describe a black person to the police.”116 The defense lawyer did not
follow up on that statement. In contrast, a victim at Patrick Waller’s trial
explained that when she viewed the lineup, “I did not want just some
African- American male to be punished. I wanted the individual who has
done this to be punished, and I was going to be sure that I had that indi-
vidual before I did anything.”117

Larry Johnson’s lawyer argued that the cross- racial identifi cation may
have contributed to a mistake, and the prosecutor responded by telling
the jury, “Black men as you can see among yourselves are all very diff er-
ent.”118 One exoneree sought to introduce an expert to testify about the
risk of a cross- racial misidentifi cation. In Perry Mitchell’s case, an ex-
pert began to testify about his work on cross- racial identifi cations, but
before he could say what the research showed, the judge sustained a State
objection and held the testimony was not admissible under South Caro-
lina law.119 Anthony Hicks also raised the issue, and the judge held, “We
have a black man, and there’s a record, at least some record in this coun-
try of mistakes about black men who have been imprisoned after convic-
tion, misidentifi cations, and we have had at least one in this community.
I hear you.” The judge found, though, that there was not “anything in
this record that suggests that because Mr. Hicks is a black man he is
here.”120

The only exoneree to have success arguing that cross- racial identifi –
cations pose additional risks of error was McKinley Cromedy. As I will
describe in Chapter 9, the New Jersey Supreme Court later struck down
Cromedy’s conviction and required judges to instruct jurors that cross-
racial identifi cations may be more prone to error.121

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 75

Child Witnesses

A second factor that the Manson test does not explicitly take account of
is the age of an eyewitness. In 12% of the cases with eyewitnesses (22 of
190 cases), the eyewitnesses were juveniles or children. All states have
rules regarding the competency of child witnesses, although many courts
assess competency quite deferentially. Social science research has found
that children often accurately answer open- ended questions, but when
children are asked leading questions, they can easily incorporate false
information and are highly vulnerable to suggestion.122

In Leonard McSherry’s case, the victim was six years old, and seven
at the time of trial. She had been abducted from her home for ten hours
and raped. The only other eyewitness was her fi ve- year- old brother, who
saw her walk away with a stranger and get into his car. Minutes later, he
told their mother that his sister got into a green truck with a stranger. At
trial, a police offi cer testifi ed and described how they used suggestive
techniques that should be avoided, especially with child witnesses. For
example, the offi cer indicated they asked the victim leading questions:
“We were trying— the child was under a lot of strain and stress, so we
were not asking her direct questions like you are asking. They were more
kind of trying to help her go along with what we were trying to empha-
size.” Similarly, at trial and on the stand, where the victim was asked an
open- ended question, she was unresponsive, but when the question was
changed to a leading question, she could answer yes or no.123

The victim was interviewed two days after the incident. At trial, the
offi cer admitted that her testimony at the time seemed very unreliable.
“We knew she was confused and did not put too much credibility to
anything she was saying.” A month later, the victim was asked to draw a
picture of the house, but what she drew was confusing and not useful.124
The defendant was arrested and the police claimed that the next day, the
victim gave a very detailed description of a house and also picked the
house from a photo array. The offi cer testifi ed that she described a tall
dresser with a TV on it, a photograph of the perpetrator, a bed with a
blue sheet cover, three doors in the bedroom, and a large round mirror—
and the offi cer testifi ed that she told them all of this before the police had
entered the house.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

76 Eyewitness Misidentifi cations

The judge gave great weight to that testimony. When denying a mo-
tion for a new trial, he noted, “I remember that little girl’s testimony, and
that bedspread stands out in my mind. I don’t think this was an I.D.
case. Her testimony weighed very heavily in how I felt of the rightness of
the jury’s verdict.”125 Yet there were inconsistencies. At the lineup, she
picked a fi ller, though supposedly after leaving she told police that al-
though she wrote down number six, she had meant to write down num-
ber three, which was McSherry. The defense lawyer argued: “Ladies
and gentlemen, the bottom line is, if you pick two people out of a lineup,
that is a misidentifi cation.” The lawyer added, “She saw some other guy.
She saw the green truck guy.”126

As it turns out, the defense lawyer was right. The police offi cer and
the victim’s mother both testifi ed that she initially described a green
truck but later insisted that the car was a station wagon. At trial, the po-
lice claimed that the victim picked the defendant’s car, a yellow station
wagon, from an array of photos. However, when asked at trial what color
the car was, she answered, “Green.” Similarly, her brother said he saw a
green truck. Before the alleged descriptions of the house and only two
days after the abduction, the victim took police to an apartment complex
and led them upstairs to a door. She said the attacker had taken her to a
place with the numbers zero and one on the door. The police discounted
her initial descriptions when they focused on McSherry.

It is possible the offi cers had fed the victim those detailed facts about
the inside of McSherry’s house— which we now know she never saw. Or
perhaps the police offi cers unintentionally led the victim to make these
descriptions through the use of leading questions. The offi cer admitted
they reinterviewed her, after conducting a search of the house, to try to
fi nd out more details.127 In a civil rights case McSherry later fi led, it
emerged that police also interrogated McSherry regarding the details of
the house he lived in after they arrested him. Only later did they inter-
view the victim. In 2002, she stated in a deposition that the police gave a
false account at trial. She was impatient when police interviewed her,
she did not actually identify a yellow car or McSherry’s house, and she
never described the inside of McSherry’s house to the police.128

Who was the “green truck guy”? When McSherry was exonerated,
the DNA matched that of George Valdespino, who was serving a life

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 77

sentence in California state prison at the time of McSherry’s release.
Valdespino had been arrested in Costa Mesa one week after the abduc-
tion, and was charged at that time with kidnapping and molesting a four-
year- old girl. In December 2001, Valdespino admitted in a taped confes-
sion that in 1988 he had kidnapped a girl in the Long Beach area while
driving a green Ford Ranchero and had taken her to a motel room (with
numbers on the door).

Judging Eyewitness Identifi cations

In view of the evidence of police suggestion and unreliability in these
misidentifi cations, the question arises: why did judges admit them at
trial? Judges are highly reluctant to suppress eyewitness identifi cations;
after all, in many of these cases, the eyewitness was the rape victim, and
without her testimony, there might be no case at all. In fi fty- eight cases,
we have some information about how the trial judge applied the Manson
factors, because the trial rec ords include rulings on a motion to suppress
the eyewitness identifi cation. As discussed, the Manson test is so defer-
ential that all of these tainted and unreliable identifi cations were admit-
ted. The judges that discussed the identifi cation in any detail typically
emphasized the witnesses’ certainty at trial, the factor that is the most
misleading. Still worse, in some cases, judges discounted evidence that
should have been of far greater concern than certainty at the time of
trial— the eyewitnesses’ earlier uncertainty. In Ulysses Charles’s case, one
victim did not identify anyone in the lineup, telling the prosecutor, “I
think or I know that it was No. 4, but I was afraid to say so because it had
been so long”— the lineup was conducted three years after the off ense. At
trial, though, she identifi ed Charles, saying that earlier she “was scared.”
The judge ruled, “To specify that a witness must . . . make a declaration
within a short period of time of the lineup is not to take into account hu-
man traits, human frailties.”129

Larry Holdren had an identical twin brother, which presented a spe-
cial problem. In a pretrial hearing the defense discussed the photo array,
and said that of the fi ve people in the lineup, three had the wrong color
hair (black instead of blonde), and the other two were of the defendant—
and of his identical twin brother, whose photo was wrinkled and washed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

78 Eyewitness Misidentifi cations

out. Larry Holdren’s photo was the only one in the lineup that resem-
bled the victim’s description. Worse, during the assault the victim was
not wearing her glasses and there was not much light. The judge ruled,
“This is not a very good photo array . . . there is a suggestive aspect of
the photo array,” but emphasized “without hesitation the victim selected
one of two people, in any event, who had remarkable similarity in their
appearance . . . and I think that that tends to take the edge off of the
suggestiveness.”130

Kirk Bloodsworth, the fi rst person exonerated from death row, and
convicted based on testimony by fi ve eyewitnesses, tried to call an ex-
pert to testify regarding social science research on eyewitness identifi ca-
tions, but the judge denied the request, fi nding that “the evidence would
tend to confuse or mislead the jury.” The Mary land Supreme Court
agreed, in a decision that is still the law in Mary land.131 Ronald Cotton
also had a motion for an eyewitness expert denied. We will see in the
next chapter how judges routinely allowed prosecution experts to render
unscientifi c forensic testimony. In contrast, judges refused defense law-
yers’ requests to present social science evidence on eyewitness error. Just
three exonerees had eyewitness experts testify.132

The law regulating eyewitness identifi cations is even more lenient
than I have described so far. Even a defense “victory,” the suppression of
an identifi cation, is usually at best only partial and temporary. A pretrial
identifi cation could be suppressed, but an in- court identifi cation before
the jury may still be allowed. Neil Miller’s case was an example of a rare
case where the judge excluded an identifi cation because it was con-
ducted in an unconstitutionally suggestive manner and was unreliable.
Recall how the police conducted an illegal showup on the morning of a
hearing scheduled to examine the identifi cation in the case. However,
judges bend over backward to let the eyewitness still identify the defen-
dant at trial, ruling that the courtroom identifi cation has an “in de pen-
dent source,” based on the witness’s original view of the perpetrator.133
Thus, the jury in Neil Miller’s case still heard about her identifi cations
of Miller before the illegal showup. They also saw her identify him in
court and say she was positive he was the attacker. Similarly, in Carlos
Lavernia’s case, the victim admitted that none of the lineup fi llers re-
motely matched her description, and the judge suppressed her prior

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 79

identifi cation— but she identifi ed Lavernia in court and said she was
“absolutely positive” he was her attacker.134 Given the fl aws in the Man-
son approach and the casual way that judges apply that fl exible test, it
was no wonder that police so often used suggestive eyewitness identifi –
cation procedures— judges let them do it.

Closing Arguments and Jury Instructions

Once judges found these eyewitness identifi cations admissible, they
were typically the central evidence at trial. Both the prosecution and the
defense typically focused their eff orts on the identifi cations. As one
would expect, the defense emphasized irregularities in the identifi cation
procedures and any evidence of unreliability. For example, in William
Harris’s case, the defense argued that only suggestion could explain why
an initially unsure eyewitness could now be so certain, arguing: “there’s
too many inconsistencies, too many problems with the correct identifi –
cation made early as compared, as it was molded later by all of these po-
lice offi cers and later on into the positive identifi cation, ‘the most con-
vincing I.D. ever.’ ”135

In contrast, prosecutors emphasized eyewitness certainty. For ex-
ample, in Carlos Lavernia’s case, the prosecution argued in closing,
“Do you think a woman under those circumstances gets any surer?
Have you ever seen anyone surer, any more positive about who the per-
son was?”136 Prosecutors also downplayed inconsistencies in victim
descriptions. The prosecutor at Anthony Hicks’s trial mocked the
defense:

[The victim] gives you all of this that matches him, and because she

doesn’t initially see the cleft in his chin, [defense counsel] would

have you believe, you know “He’s the innocent victim of the Madi-

son Police Department! We’ve got the wrong guy!”137

Since defense lawyers were rarely present at the identifi cation proce-
dures and the procedures were not documented, the defense could do
little but point to such inconsistencies— but prosecutors could respond
by pointing to the witnesses’ certainty.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

80 Eyewitness Misidentifi cations

Or prosecutors could dwell on the seriousness of the crime. In Perry
Mitchell’s case, the prosecutor argued: “Some Supreme Court Justice
said eyewitness identifi cation is untrustworthy. Well, I bet that Supreme
Court Justice never had a knife pulled up to him and drug him over in
the woods. I think he would change his mind then.”138

In some exonerees’ trials, the judge’s instructions to the jury were
transcribed. Judges sometimes gave detailed instructions. For example,
in Alan Newton’s trial, the judge told the jury, “Identifi cation testimony
is an expression of belief or impression by the witness. Its value depends
on the opportunity the witness had to observe the off ender at the time of
the off ense and to make a reliable identifi cation later.”139 The judge then
explained factors to consider when weighing such testimony. The case
was also unusual in that the jury sent back questions concerning the
eyewitness testimony; they asked to reread the victim’s testimony regard-
ing how good a look she had at her attacker. In Nicholas Yarris’s case,
two witnesses were initially unsure about their identifi cations but were
positive at trial. The judge instructed the jury that “an individual can
make a mistake as to identifi cation even when the individual is trying to
tell the truth,” and then asked the jury to evaluate the reliability of the
identifi cations.140 Beyond such instructions telling jurors to assess the
credibility of eyewitnesses and to consider certain factors, the judges did
no more. None explained to the jury, for example, relevant social science
research regarding identifi cation procedures or possible eff ects of police
suggestion.

Reforming Eyewitness Identifi cation Procedures

There is every reason to think that these eyewitness misidentifi cations
may be part of a larger problem that continues to this day. Observers
have called eyewitness misidentifi cations the “leading cause” of wrong-
ful convictions.141 There is a reason why there were so many eyewitness
misidentifi cations among the DNA exonerations. DNA testing is partic-
ularly helpful in rape cases, where there is biological material collected
from the victim that can be tested. In turn, many rape cases involve eye-
witnesses. In rape cases, the case will not easily go forward unless the
victim can identify the culprit. However, very few suspects are charged

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 81

with rape or murder and fewer are convicted.142 Other crimes that many
more people are convicted of each year, such as robbery, also commonly
involve eyewitness identifi cations, but usually do not have evidence that
can be tested using DNA. We simply do not know how common eyewit-
ness misidentifi cations are for those other types of crimes.

Still more worrisome, the types of suggestive and unreliable eyewit-
ness identifi cation procedures used in these exonerees’ cases have been in
wide use for a long time. Police did not begin to adopt any written eyewit-
ness identifi cation procedures at all until recently, much less sound pro-
cedures and training. Archival studies of police identifi cation procedures
have found that suggestive procedures, such as showups, are used fre-
quently and appear to bias eyewitnesses.143 Nor have very many judges or
prosecutors actively sought to improve the quality of eyewitness identifi –
cations, although this is changing, as Chapter 9 will describe.144

For more than thirty years, beginning with pioneering work by psy-
chology professors Elizabeth Loftus and Gary Wells in the mid- 1970s,
social scientists have used hundreds of laboratory and fi eld experiments
to study eyewitness memory. A vast body of work has uncovered fl aws in
eyewitness identifi cation procedures and the Supreme Court’s Manson
approach.145 Courts admit identifi cations despite highly suggestive pro-
cedures based on factors that do not correspond to reliability, such as
certainty, which may be the product of suggestion. Police may intention-
ally suggest a suspect to a witness, but they may also give inadvertent
cues that can corrupt an eyewitness’s memory.

There is one straightforward way to ensure that police do not engage
in suggestive conduct during a lineup procedure. That is to use a double-
blind procedure, in which the administering offi cer does not know
which person in the lineup or array is the suspect, so he cannot infl uence
the witness, intentionally or not. The eyewitness must also be told that
the police offi cer does not know which person in the lineup is the sus-
pect, so the eyewitness does not perceive any behavior from the offi cer as
a cue, correctly or not. Eyewitness researchers have long recommended
requiring double- blind lineups as the most crucial reform of all.

It should be no surprise that in these exonerees’ cases, double- blind
procedures were not used. Only one of these eyewitnesses, in Dennis
Brown’s case, misidentifi ed a defendant in a double- blind procedure.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

82 Eyewitness Misidentifi cations

However, his lineup had another problem: the victim was not told the
suspect might not be present. Brown was poor, seventeen years old, and
he had been off ered thirty dollars by a police offi cer to be a “volunteer
fi ll- in” in a lineup. He was unlucky to have been in the wrong lineup at
the wrong time; the victim identifi ed him, probably to the surprise of the
police.146

Adopting double- blind lineups is cheap, and more police departments
have now adopted this reform.147 All that police need to do is bring in
another offi cer to administer the lineup. Increasingly, police rely on com-
puterized arrays using databases of photos with automated instructions
to witnesses. There is also a clever low- tech way to make a double- blind
array that some states recommend: simply place each photo in a folder
and shuffl e the stack of folders, so that the witnesses can look inside the
folders to view the photos, but the offi cer cannot see which photo the wit-
ness is looking at.148

Other reforms are also inexpensive but important. The lineup itself
should be constructed so that the suspect does not stand out. Lineups
in which all of the suspects are presented at the same time, in person or
in a photo array, may run the risk of “comparison shopping.”149 The
eyewitness must be told that the suspect may not be in the lineup. Offi –
cers must document the descriptions and certainty of eyewitnesses.150
As described, almost none of these identifi cations used those best
practices.

The judge at Habib Abdal’s trial told the jury, “Guilt or innocence is
in your hands and in your hands alone.” He then told them more than
many judges do. He instructed the jurors to “exercise extreme caution,”
and try to assess whether the victim’s identifi cation was accurate or was
made “by virtue of any suggestions or prompting.”151 Despite strong evi-
dence of an unreliable identifi cation made under sustained police pres-
sure, the jury deliberated for about fi ve hours and then convicted Abdal.
Exonerees’ cases provide a series of powerful examples of how common
but fl awed and suggestive police procedures can contaminate the mem-
ory of an eyewitness. Judges, not just police, should also be attentive to
risks of error. They could conduct a more searching inquiry than under
the Manson test, to ensure that police use sound procedures or face con-
sequences, such as pointed instructions to jurors or perhaps even the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Eyewitness Misidentifi cations � 83

exclusion of a tainted identifi cation. Reformed eyewitness identifi cation
procedures are fi nally starting to be adopted, as I will discuss in Chapter
9, and they are important. DNA exonerations show how jurors can be
convinced that “seeing is believing” even when the eyewitness on the
stand points to an innocent man.152

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

In July 1977, police in Homewood, Illinois, came across a young
woman, disheveled and walking by the side of the road. She told po-

lice she had been raped. Following the standard practice, police took her
to a hospital for examination. Medics combed her pubis for any loose
hairs, they used cotton swabs to sample fl uid from her vagina, and they
preserved her stained underwear. They assembled this evidence in a
standard collection box, called a rape kit, and sent it to a lab for analysis.

Meanwhile, police worked with the victim to prepare a composite
drawing of the attacker— a white man with long hair. Since they had no
suspect, police showed her books of mug shots. She selected one, a mug
shot of Gary Dotson, whom she later identifi ed in a lineup. Dotson’s
appearance diff ered from the mug shot in an obvious way— he had a
large mustache. But this case is not at all like the cases discussed in the
last chapter, in which suggestive or unreliable identifi cation proce-
dures may have led to a misidentifi cation. This victim later admitted
she made up her injuries and her accusation of Dotson as part of an
elaborate ruse to conceal from her parents consensual intercourse with
her boyfriend.1

At Dotson’s trial in 1979, the victim, as expected, pointed him out in
the courtroom and identifi ed him as her attacker. Only one other witness

c h a p t e r 4

Flawed Forensics

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 85

testifi ed for the State— a forensic analyst.2 The term forensic science
means the use of science to help answer legal questions. In criminal
cases, forensic methods are frequently used to analyze crime scene evi-
dence. Most forensic analysts in the United States work for the police, at
law enforcement crime laboratories. This analyst worked for the Illinois
Department of Law Enforcement. When he testifi ed in Dotson’s case, he
began by claiming, falsely, that he had done “graduate work” at the Uni-
versity of California at Berkeley. He had in fact “only attended a two- day
extension course there.”3 Unfortunately, his testimony did not improve
in its accuracy.

He next discussed his analysis of hairs collected in the rape kit. In
1977, before DNA testing had been invented, most of what forensic scien-
tists did was to compare objects from crime scenes: hairs, fi bers, fi nger-
prints, bite marks, tool marks, or bullet casings. Using a microscope, they
looked at the objects side by side. The analyst would then reach a sub-
jective conclusion about whether the sets of objects looked alike or not.
A conclusion that an object is defi nitely diff erent is called an “exclu-
sion,” since the object is found not to be part of the group of similar ob-
jects. An exclusion can be defi nitive. If the defendant’s hair is blond but
the hairs from the crime scene are brown, then the analyst would say
that the defendant is “excluded”— he could not be the source of brown
hairs. An analyst may also fi nd an “inclusion,” if the object is included,
or is part of a group. If the defendant has brown hair and the crime scene
hairs are also brown, then the defendant is included in the group of peo-
ple who could have left those hairs— but that inclusion may not say much.
After all, the group of people that have brown hair is enormous. Finally,
an analyst may also fi nd that all of the evidence is inconclusive. For ex-
ample, hairs from the crime scene might be in such poor condition that
they are unsuitable for analysis.

This analyst told the jury that he found “several pubic hairs” that
were not like the victim’s but “were microscopically consistent” with
Dotson and “could have been originated from the same source.” What
did any of that mean? Did the hairs just look “similar” to those of Dot-
son? Dotson was included, but in how large a group of people? Could
just a hundred other people have the same type of hair? Could a million
other people have the same type of hair? Or were these microscopic

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

86 Flawed Forensics

characteristics so unusual that only Dotson could have been the
perpetrator?

The analyst could not have answered those questions, even if the de-
fense lawyer had tried to ask them. The analyst’s scientifi c- sounding
terminology may have impressed the jury, but the jury might have
thought diff erently had they known that there was no scientifi c defi ni-
tion of what it meant for someone to conclude that hairs are “consistent”
with each other. All it meant was that the analyst looked at the hairs and
thought they looked alike. To this day, no one has done any study of
what portions of the population share par tic u lar microscopic hair char-
acteristics. It could be that many thousands or millions of people share
some type of hair characteristic.

The technique of hair comparison is an unreliable method. What
does it mean for a method to be unreliable? A method is unreliable if it
does not produce consistent or accurate results. Hair comparison con-
sists in an analyst’s own personal opinions. Two analysts could (and do)
disagree about whether hairs seem to them to be “consistent” or “simi-
lar.” Nor is there any defi nition of what those terms mean. Studies have
found very high error rates in hair comparison. Scholars have called on
courts to limit the use of such unreliable methods.4 Nevertheless, judges
continue to allow jurors to hear about such unreliable evidence, even in
serious criminal cases like Dotson’s.

The same analyst also used a second forensic method that was reli-
able. ABO blood- typing, which is called serology, was used widely in
sexual assault cases in the 1970s and 1980s. A person’s blood type,
whether A, B, AB, or O, does not change during his or her lifetime. Fre-
quencies of those blood types were derived from well- established and
scientifi cally valid databases. Most people are called “secretors,” be-
cause their blood type can be detected not only in their blood but also in
their other bodily fl uids, including semen. Unlike hair comparison, se-
rology was a reliable method. An analyst could use serology to defi ni-
tively exclude someone, not based on a subjective opinion but because
the person’s blood type did not match the material tested. An analyst
could also fi nd that a person was included, and could explain what that
meant by telling the jury how a person fell within a percentage of the
population that shares a blood type. However, many millions of people
might also share that blood type.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 87

At Gary Dotson’s trial, the analyst testifi ed that he analyzed the vagi-
nal swabs collected at the hospital. He testifi ed that Dotson was Type B,
and that so was the source of the semen. Type B is a type shared by only
11% of Caucasians. This appeared to be suggestive, although certainly
not conclusive, evidence of Dotson’s guilt. It narrowed down the group
of possible culprits to 11% of the population.

Unfortunately, the analyst’s conclusions were unsound. The analyst
never told the jury that the victim was also Type B and that her fl uids
were mixed in the sample. The Type B substances observed might not
have come from the rapist at all. The Type B substances could have come
entirely from the victim. In fact, they most likely did, since it was quite
common that the material from the victim would overwhelm, or “mask,”
any material from the semen. Based on the testing technology at the time,
nothing more could be done with the evidence. The source of the semen
could have had any blood type at all. The semen could have belonged to
any man in the world. It was false and highly misleading to tell the jury
that only 11% of the population (including Gary Dotson) could have been
the source and to imply that 89% of the population was excluded.

Although serology was a reliable method, the conclusion that the ana-
lyst drew from the evidence was a false one. I call such testimony “in-
valid,” because the term validity refers to whether claims or inferences are
supported by the evidence. In Gary Dotson’s case, the analyst went on to
off er the jury a second invalid conclusion about the blood evidence. He
not only claimed to include Dotson, by falsely saying that Dotson was part
of a group of Type B people that could have been the source of the semen,
but he failed to exclude Dotson.

He turned to analysis of the victim’s underwear. Here he was con-
fronted with a problem. Type A material was found in stains in several
places on the victim’s underwear. Since both the victim and Dotson had
Type B blood, the A material could not have come from either of them.
The analyst testifi ed: “Unfortunately for us there are lots of materials;
dust, wood, leather, certain kinds of clothes, diff erent cloth materials,
detergents in materials” that could explain the presence of the Type A
substances.5 He speculated that contamination somehow created the
Type A substances.

This testimony was simply false. The analyst should have told the
jury that the A substances could not have come from the victim or Dotson

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

88 Flawed Forensics

and that Dotson was excluded. The analyst’s testimony refused to rec-
ognize evidence of innocence. In his eff ort to disregard clear evidence of
innocence, the analyst was willing to misstate the fundamental basis of
blood- typing. As pioneering forensic scientist Edward Blake explained
in his report on the case years later, the analyst could test other portions
of the stains to examine whether contamination had really occurred. If it
were actually the case that contamination could never be ruled out, and
blood substances could spontaneously appear from contact with every-
day materials, then “ABO typing of biological samples” would have al-
ways been an “inherently unreliable” type of analysis.6

Dotson’s defense lawyer never questioned or objected to any of this
wildly inaccurate testimony. Not having obtained an expert, Dotson’s
public defender may not have realized that the hair comparison method
was unreliable or that the analyst’s conclusions about the serology were
invalid.

During the closing arguments, the prosecutor exaggerated the sci-
ence even more, claiming that the analyst had “matched” Dotson’s hairs
to those found on the victim. The analyst had not gone that far, but had
only said that the hairs were “similar” and “could have” come from Dot-
son. The term match, used by the prosecutor, implied that Dotson’s
hairs were identical to those found on the victim. The defense lawyer
did object to that statement. However, the judge denied the motion,
thereby permitting the suggestion to the jury that perhaps the hairs re-
ally did match.

The jury convicted Dotson of rape in 1979. He was sentenced to
twenty- fi ve to fi fty years in prison. The victim subsequently recanted.
She admitted that she fabricated her accusation to conceal from her par-
ents that she had consensual sex with her boyfriend at the time. How-
ever, after three- day televised clemency hearings in 1985, Illinois gover-
nor James Thompson denied Dotson a pardon despite the victim’s
recantation.7

DNA testing was Dotson’s last hope. Edward Blake would later per-
form testing showing both the innocence and guilt of many convicts.
Blake found that the DNA results excluded Dotson as the source for
the male ge ne tic profi le, and that the victim’s boyfriend was included.
The governor still refused to pardon Dotson, but based on the DNA

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 89

results, the trial court vacated his conviction and he was exonerated in
1989.8

Unreliable and Invalid Forensics

The pervasive use of forensic science in exonerees’ cases provides a
study in contrasts. One story to tell about science is triumphant— DNA
technology developed by innovative scientists set these innocent people
free, literally saving some of their lives. Yet in the Introduction, I posed
the question: Why didn’t forensic science show that these people were
innocent at trial? Some forensic science did tend to show innocence at
trial. Most of the forensics did not. Indeed, forensic science played a role
in most of these exonerees’ original convictions. Exoneree Roy Brown
aptly said: “Junk science sent me to prison, but real science proved my
innocence.”9

I focus on the exonerees who had a trial, and whose trial rec ords
could be located, because by reading these trials one can see, as in Dot-
son’s case, what the forensic analysts actually said on the witness stand.
Forensic evidence was present in 74%, or 185, of the DNA exonerees’
cases. Some of those had no trial, or no forensic evidence was presented
by the prosecution, but for most, 169 exonerees, analysts testifi ed for the
prosecution at trial. Of those, I was able to locate 153 trials where ana-
lysts testifi ed for the prosecution.10 The forensic testimony was fairly
evenly distributed across the diff erent crimes that exonerees were con-
victed of, and it was the second most common type of evidence, after the
eyewitness evidence discussed in the last chapter.11

The prevalence of forensics did not surprise me. After all, most of
the cases, like Dotson’s, were rape cases in which potentially important
forensic evidence was collected in a rape kit. I expected to fi nd that fo-
rensic analysts provided the best evidence they could, but that they
were limited by the technology available at the time. Instead, when
reading these exonerees’ trials, I found that even by the standards of the
1980s, when most of these trials took place, forensic analysts should
have known that the evidence they presented was unsound. There were
two recurring types of problems with the forensics: reliability and
validity.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

90 Flawed Forensics

First, many of these exonerees were convicted based on forensic meth-
ods that were unreliable. Serology was the most common technique
used, in 116 cases, and DNA testing had been conducted at the time in
20 cases. While those are both reliable techniques, other methods that
were used were not reliable. Gary Dotson was just one of 75 exonerees who
had testimony at trial concerning microscopic hair comparison. Other
methods were used, including fi ngerprint comparison (20 cases), bite
mark comparison (7 cases), shoe print comparison (6 cases), and voice
comparison (1 case). When analysts use each of those methods, they make
a detailed but subjective comparison of objects from the crime scene with
those from the defendant. Each method is, in the words of a landmark
report by the National Academy of Sciences (NAS), “supported by little
rigorous systematic research to validate . . . basic premises and tech-
niques.”12 Each is based on an opinion that two objects are “similar” or
“consistent” or that they “match.” Although scientists and scholars have
criticized many of those forensic methods due to errors,13 evidence of
unreliability, vague terms used for conclusions, and lack of scientifi c
rigor, most are still in wide use today.14

Second, many of the cases involved invalid conclusions drawn from
the evidence. In 61% of the trials where a forensic analyst testifi ed for the
prosecution, the analyst gave invalid testimony. The analysts did this
across a wide range of forensic methods, ranging from serology, in which
58% of the testimony was invalid (67 of 116 trials); to hair comparison, in
which 39% was invalid (29 of 75 trials); to bite mark comparison, in which
71% was invalid (5 of 7 trials); to shoe print comparison, in which 17% was
invalid (one of six trials); to fi ngerprint comparison, in which 5% was in-
valid (1 of 20 trials). As in Gary Dotson’s case, in which the analyst mis-
stated the meaning of the ABO blood- typing, even if the underlying
method is reliable, the jury can be misled if the analyst exaggerates the
evidence. Not only did analysts typically reach invalid conclusions about
serology, a reliable method, but of cases with DNA testing, 17% had in-
valid testimony (3 of 18 trials). All of this invalid testimony had some-
thing in common. All of it made the forensic evidence seem like stronger
evidence of guilt than it really was.

Invalid and unreliable forensics were so prominent in these cases that
they raise more troubling questions about the use of forensics in crimi-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 91

nal cases more generally. How can analysts testify in such a patently un-
scientifi c way without any consequences? As judges have long recognized,
forensic evidence can play an important role in criminal trials. Juries may
give special weight to testimony by forensic scientists. The Supreme Court
has cautioned that “[e]xpert evidence can be both powerful and quite mis-
leading because of the diffi culty in evaluating it,” and judges talk about the
“talismanic signifi cance” that jurors may attribute to expert evidence.15
That is why judges must conduct a “gatekeeping” inquiry, to ask whether
a method is reliable and valid, before allowing an expert to take the wit-
ness stand. The Supreme Court’s decision in Daubert v. Merrell Dow
Pharmaceuticals, Inc. instructs judges to ask whether an expert’s method
has a valid and reliable scientifi c foundation.16 And that is what judges
routinely do in civil cases where money damages are in dispute.

However, there is a reason why so many experts used unreliable meth-
ods and gave invalid conclusions in these exonerees’ cases, and why such
testimony is still allowed. Even unreliable methods— or worse, methods
rejected by the scientifi c community— have long been permitted in crim-
inal courtrooms. Daubert had not yet been decided in the 1980s when
most of these exonerees’ trials took place. Nor have matters changed
much since that time. Daubert has not been meaningfully applied in
criminal cases where life or liberty is at stake. Once judges decide, with-
out looking very carefully, that the methods are acceptable, they do
nothing if experts provide exaggerated conclusions on the witness stand.
Judges do not examine the validity of the conclusions that forensic ana-
lysts reach.17 Why have judges ignored basic requirements that expert
testimony be reliable and valid?18 Judges assumed that forensic analysts
adhered to scientifi c standards. In the rare event that one crossed the
line, then judges assumed that cross- examination by the defense lawyer
would illuminate the inaccuracy and help the jury to understand the
truth.19 As one crime lab director conceded, judges “have given us a free
ride.”20

Who Are Forensic Analysts?

In civil cases involving disputes about money, there is a phrase, the
“battle of the experts,” referring to confl icts between expert witnesses

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

92 Flawed Forensics

testifying for opposite sides about scientifi c or technical matters. There
is rarely any battle of the experts in a criminal case. Instead, the pre sen-
ta tion of forensic evidence is almost entirely one- sided. Almost all of the
analysts testify for the prosecutors. Almost all of them work for the po-
lice, at state or local law enforcement crime laboratories.21 There are over
350 crime laboratories in the United States. Law enforcement agencies
operate the vast majority as state or regional laboratories, though some
are operated by local governments in large metropolitan areas.

These analysts, therefore, report to law enforcement and see them-
selves as working for law enforcement. An extreme example from several
exonerees’ cases was Fred Zain in West Virgina, who asked to be ad-
dressed as “Trooper,” and who wore a police uniform and gun even
though his job was to supervise a crime lab. Studies have suggested fo-
rensic analysts may be subtly or sometimes not so subtly biased by their
role as analysts working for police and prosecutors.22 Scientists design
experiments that are blind, so that not all of those involved know certain
aspects of the experiment, to prevent the infl uence of bias (conscious or
not) on their work. In contrast, forensic analysts do not do their work
blind. They receive information about the crimes being investigated that
may have nothing to do with the forensic analysis being conducted, such
as the fact that a suspect confessed, or the prior criminal record of the
suspect. As the U.S. Supreme Court has put it, a forensic scientist may
not be “neutral” because an analyst “responding to a request from a law
enforcement offi cial may feel pressure— or have an incentive— to alter the
evidence in a manner favorable to the prosecution.”23

Some analysts, such as Joyce Gilchrist of the Oklahoma State Police,
have since been the subject of searching audits and investigations, but
most have not.24 Gilchrist testifi ed in four exonerees’ trials. She hid evi-
dence, altered reports, and off ered conclusions on the stand that hairs
were unique and certain to have come from the defendant. U.S. senator
Orrin Hatch, commenting on the need to provide new resources for fo-
rensic sciences, referred to the Gilchrist scandal, noting, “[W]e are all
troubled by allegations that mistakes by a police chemist in Oklahoma
helped send innocent people to prison. This isolated situation should
not be used unfairly to indict the thousands of forensic scientists who
perform their work professionally and responsibly.”25 While I agree that

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 93

we should not indict an entire group of people based on the bad acts of
one person, the sad truth is that the invalid testimony in DNA exonera-
tion cases did not involve just a few “bad apples.”26 Almost all of the ana-
lysts who delivered invalid testimony did so in one trial each. These
cases included invalid testimony by 81 forensic analysts employed by
54 laboratories, practices, or hospitals from 28 states.27

One cannot know what these analysts were thinking at the time, or
what their state of mind was, based on the trial materials. Some may have
simply been inexperienced or poorly trained. They may have been sus-
ceptible to cognitive bias based on their role working for law enforcement
and based on the other information that they were given about the cases
they worked on. One also wonders how these analysts testifi ed in their
other cases. Each one of these analysts could have testifi ed in many cases
each year. For example, the next section presents invalid testimony by an
analyst in exoneree Neil Miller’s case. He said that he had testifi ed “at
least a thousand times.”

Other analysts were themselves supervisors. Indeed, Gilchrist was a
supervisor who ran the Oklahoma City serology unit and established
their new DNA lab; prosecutors commended her work and wrote her let-
ters of appreciation. Later audits uncovered problems with missing evi-
dence and mishandling of evidence in the lab, and cited the fact that the
offi ce had “no operating, procedure, or safety manuals.”28 One wonders
about her training and supervision of others. If analysts lacked adequate
training or supervision, then one wonders about their testimony in other
cases. One wonders about testimony by their colleagues, with whom they
would have been in close contact; most crime laboratories do not employ
more than a dozen analysts.

The next sections describe the types of unreliable and invalid foren-
sics that convicted these innocent people, beginning with serology, which
was used most often in exonerees’ cases, and then the techniques used
less frequently.

Serology

Conventional serology, using ABO blood- typing, was the most common
type of forensic analysis in these exonerees’ trials, and it was present in

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

94 Flawed Forensics

116 trials. There was invalid testimony in 58% of the cases involving se-
rology (67 of the 116 trials). Unlike some other disciplines, serology was
grounded in sound science and empirically validated population data— in
other words, it was a reliable method. This makes it all the more disturb-
ing that the analysts in so many of these trials either did not understand
or exaggerated the clear boundaries of the science when they off ered
their conclusions. The most common single scientifi c error made during
all of these trials involved the same straightforward problem of “mask-
ing” that arose in Dotson’s trial. Fifty- one trials involved that same basic
error, in which the serology evidence was in fact inconclusive because it
was consistent with the victim, but that was not what the analyst told the
jury.

A telling example is in the case of Neil Miller, who was wrongly con-
victed of a rape. The victim was examined in a hospital after reporting
the crime and a vaginal swab was collected. Such swab stains are mixed;
they may include the victim’s fl uids and semen from the assailant. Mixed
stains presented a serious challenge. The victim’s own ge ne tic markers
could overwhelm or “mask” any markers from semen. It could be impos-
sible to detect the blood type of the assailant.29

In Miller’s criminal trial, the forensic analyst from the Boston Police
Department testifi ed that the substances he found “had to be deposited
by a Group O individual,” and “[a]pproximately forty- fi ve percent of the
population are Group O individuals.” Yet the victim was also Type O,
so all of the substances could have come from the victim. The analyst
had no information at all about the rapist. The entire male population
was included and none was excluded.

That testimony was particularly remarkable, because of a coincidence
(or maybe not a coincidence at all): the same analyst testifi ed in the trial of
another DNA exoneree, Marvin Mitchell. The evidence in Mitchell’s case
also raised the problem of masking, and there the same analyst did explain
the problem to the jury— where doing so assisted the prosecution. He ex-
plained why Mitchell, who had Type A blood, was not excluded by a test
of the vaginal swab consistent with the victim, a Type O. He testifi ed:

Mr. Mitchell could not be excluded. No secretor could be excluded

from depositing that stain because the stain may have been too di-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 95

luted or graded [sic] to pick up Mr. Mitchell’s blood type. So I can-

not exclude him, but I cannot say that I found the A blood group

type. In other words, again no secretor can be totally excluded from

the stain.

Even in Mitchell’s case, he did not accurately explain the evidence. He
erroneously implied that as a secretor, a person whose blood type can be
detected in his body fl uids, Mitchell was in a subset of a population that
could be the source. He never told the jury that no person, whether a
secretor or not, could be excluded.

A range of additional errors were made in serology cases. For exam-
ple, in several exonerees’ cases, analysts wrongly divided frequencies in
half. In Perry Mitchell’s case, the semen came from a Type O secretor,
and O secretors comprise about 35% of the population. The serologist
divided the accurate statistic in half and testifi ed that only 17.5% of men
could have contributed the semen and thus 82.5% of the relevant popu-
lation was excluded. However, the population statistics on blood types
are identical for both sexes; 35% of men are O secretors and 35% of
women are O secretors. There was no valid reason to divide that statistic
in half.

Microscopic Hair Comparison

Forensic hair microscopy, the type of hair comparison conducted in Gary
Dotson’s case, was the second most common type of analysis in these ex-
onerees’ trials. Seventy- fi ve exonerees had such testimony. In all but fi f-
teen trials, the testimony was used to show the guilt of the defendant. In
twenty- nine cases, the testimony was invalid. Forensic hair microscopy
involves the side- by- side comparison under a microscope of head and
pubic hairs found at a crime scene with dozens of head and pubic hairs
plucked and combed from the victim and suspects. Hair examination has
long been important in police investigations, because hairs are so com-
monly and readily transferred to skin or clothing.

However, the method of forensic hair comparison is unreliable. While
we might have some intuitions about what percentage of the population
has diff erent hair colors, the microscopic characteristics that analysts

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

96 Flawed Forensics

compare cannot be observed with the naked eye. Unlike in serology,
where we know how many people have Type A or Type O blood, no
scientist has studied how often hairs have par tic u lar microscopic char-
acteristics, like smooth versus rough cuticles, or large versus small pig-
ment granules. As the NAS report put it, “[n]o scientifi cally accepted
statistics exist about the frequency with which par tic u lar characteristics
of hair are distributed in the population.”30 And unlike in serology,
where two analysts should not disagree whether tests show that someone
has Type A blood, hair analysts can disagree about whether sets of hairs
share microscopic characteristics. Profi ciency tests of hair examiners
dating back to the 1970s have found very high error rates ranging from
28 to 68%.31

There is no agreed- upon set of characteristics that analysts look for
when they compare hairs. The NAS report explained, “[t]here appear to
be no uniform standards on the numbers of features on which hairs must
agree before an examiner may declare a ‘match.’ ” Indeed, the reason
why analysts look at large sets of hairs, and not just single hairs, is that
there is often wide variation in the characteristics of the hairs on a single
person’s body. The NAS report, using strong language, stated that any
eff ort to link a defendant to hair evidence has “no scientifi c support.”32

One of the few judges to intervene and suggest that there is a problem
with such unreliable analysis was the federal judge reviewing Ronald Wil-
liamson’s habeas petition. In 1995, not knowing Williamson was innocent,
the judge took an important stand against faulty forensic testimony and
ruled that testimony fi nding hairs to be “consistent” was erroneous, and
that hair comparison testimony was too error prone, “scientifi cally unreli-
able,” and too grounded in the experts’ “subjective opinion” to be admis-
sible.33 That decision postponed his execution and paved the way for the
DNA testing that exonerated Williamson in 1999. However, other courts
have not followed suit to bar or limit the use of such unreliable hair
comparisons.

These exonerees’ cases also provide a large body of examples of just
how unreliable hair comparison can be in practice. We now know the
hairs from these crime scenes were not in fact from these innocent defen-
dants. In quite a few trials, the errors were particularly egregious, be-
cause the analyst falsely concluded that not just one hair, but a large set

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 97

of hairs, all came from the wrong man. For example, Jimmy Ray Brom-
gard was exonerated by DNA testing after fi fteen years in prison in Mon-
tana. The evidence at his trial consisted of an identifi cation by the vic-
tim, who was one of a few eyewitnesses to testify she was not sure of her
identifi cation, and the hair comparison. Arnold Melnikoff , then the di-
rector of the Montana State Crime Laboratory, had compared more than
thirty hairs and found that they all “matched” Bromgard’s hairs.34

Not only is hair comparison an unreliable method, but analysts also
often reached conclusions that were invalid, in an eff ort to make their
conclusions sound more powerful. Twenty- nine exonerees had invalid
testimony concerning hair comparison. For example, Melnikoff ’s testi-
mony was not only grossly in error but also invalid. He testifi ed that be-
cause both Bromgard’s pubic and head hairs were a “match” with those
from the crime scene:

Well there are actually two mutually exclusive events because they

come from diff erent areas of the body, and their characteristics are

not necessarily the same. So if you fi nd both head and pubic hair

there you have one chance in a hundred for the head hair matching a

par tic u lar individual and one chance in a hundred for the pubic

hair. If you fi nd both it’s a multiplying eff ect, it would be one chance

in 10,000.

He testifi ed similarly in other cases, such as exoneree Chester Bauer’s
case, stating:

To have them both match, it would be the multiplication of both fac-

tors so as an approximately using that 1 out of 100, you come out

with a number like 1 chance in 10,000. Multiply 100 100. It becomes

a very highly improbable situation . . .35

There is no scientifi c research about what percentage of the population
shares microscopic hair characteristics. The analyst made up probabili-
ties based on his own personal estimates. He then multiplied those con-
cocted probabilities by the types of hairs that were “matched,” as if each
represented in de pen dent events. The probabilities of two events can

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

98 Flawed Forensics

only be multiplied if the events are statistically in de pen dent, and the
outcome of one event does not infl uence the outcome of the other. There
is no research on the question of statistical in de pen dence for head and
pubic hair. It may be that a person’s head and pubic hairs tend to share
microscopic characteristics.

Another example of invalid testimony, where the analyst tried to dress
up hair comparison as sound science, is Timothy Durham’s case. There
the analyst testifi ed that the par tic u lar reddish- yellow hue of both his
hair and the crime scene hair were only found in “about 5 percent of the
population.” The analyst did not provide scientifi c support for that sta-
tistic, nor could one do so.36

During William Gregory’s trial, an analyst from the Kentucky State
Police testifi ed that the hairs “more than likely” belonged to Gregory. In
part this was based on a fi nding of what she called “ovoid bodies” in the
hairs, which she said were “kind of an unusual characteristic.” The ana-
lyst explained:

A. I told you, there is no statistics on this. I can tell you this is the

fi rst time I have ever had a negroid origin hair that has not had

a medulla in it.

Q. What percentage of people have ovoid bodies in them?

A. This is probably the fi rst time I have ever seen an ovoid body in

a human hair. I have seen them in cattle hair before.

The analyst added that while “hairs from brothers and sisters” might
share some characteristics, “you wouldn’t see that kind of an overlap in
two people you would just pick off the street.” Finally, although she testi-
fi ed that the hairs had an “unusual” characteristic, she admitted “there
is no statistics [sic] on this.”37

In two exonerees’ cases, the analyst went even farther and simply con-
cluded that the defendant’s hairs were found at the crime scene. In Cur-
tis McCarty’s case, Joyce Gilchrist ended her testimony with an explo-
sive statement. Based on examining McCarty’s hairs and hairs from the
crime scene, she concluded that “he was in fact there.”38 Similarly, in
Larry Peterson’s case, the analyst found that all of the questioned hairs
examined were identifi ed as either “belonging” to the victim or to Peter-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 99

son.39 No research remotely supports making such strong conclusions
based on hair evidence.

Some of the analysts in these exonerees’ cases did not reach invalid
conclusions. They admitted that there was no research on how many
people might also share the same hair characteristics as the defendant.
However, in thirty- one cases, although the testimony was valid, the ana-
lysts used vague and potentially misleading terminology. They said that
the hairs were “consistent” or “similar” or they were a “match.” They
could not explain what those terms meant.

Most were following the standard practice at the time. At a 1985 sym-
posium convened by the FBI, the community of hair comparison ana-
lysts adopted several standards. They discouraged use of the term
match, because it might imply a comparison as detailed as a fi ngerprint
“match.” Instead, they decided that the strongest statement that could
be made is that hairs are “consistent” with the defendant’s or “could
have” come from the defendant. They advised that an examiner should
not make “any statements about the probability that a specifi c hair
could have come from someone other than the person to which it was
associated.”40

While the eff ort to provide standards for hair comparisons was com-
mendable, even a statement that hairs are “consistent” or “similar” is
confusing and can be misleading. After all, the jury may think that “con-
sistent” means a perfect “match.” The jury may think that the character-
istics making the hairs “consistent” are rare. The analyst testifying in
the Central Park Jogger case was asked whether thousands or millions of
people could share the same hair characteristics. He explained, accu-
rately, that “I really can’t give you a number because there hasn’t been
any research along those lines.”41 The analyst cannot explain how rele-
vant any similarities are, precisely because he cannot say if ten or a thou-
sand or a million other people might also have similar hairs.

It is hard to understand how judges can so routinely allow forensic
analysts to give such vague and unreliable testimony. The NAS report
emphasized “the problem with using imprecise reporting terminology
such as ‘associated with,’ which is not clearly defi ned and which can be
misunderstood to imply individualization.”42 This problem extends far
beyond hair comparisons (which are themselves conducted less often

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

100 Flawed Forensics

today, since mitochondrial DNA testing can sometimes be conducted
on hair evidence). Analysts conducting fi ber comparison, tool mark
comparison, bullet comparison, tire tread comparison, and a host of
other methods still in wide use all lack any standard terminology and
use similarly vague language to characterize conclusions. Since the meth-
ods are unreliable and are not grounded in sound scientifi c research, it
is hard for an analyst to say anything about what it actually means to
reach a conclusion that, in her opinion, two objects are “similar” or
“consistent.”

DNA Testing

Prosecutors complain that jurors are exposed to tele vi sion portrayals in
shows like CSI: Crime Scene Investigation and they think that forensic
analysts are infallible and capable of using technology to solve any crime.
Prosecutors say they see a “CSI eff ect,” where jurors will not convict un-
less there is strong forensic evidence of guilt. Yet in these exonerees’
cases, the reverse often occurred, and jurors convicted in spite of strong
forensic evidence of innocence. Twenty exonerees had DNA testing at
the time of their conviction, and sixteen of them had DNA tests that
showed innocence, while four had DNA tests that appeared to show their
guilt. In one case, the exoneree pleaded guilty despite DNA test results
that excluded him. The other fi fteen exonerees had testimony at trial
about DNA test results that excluded them, but despite that powerful
DNA evidence of innocence, the jurors or— in some cases without juries—
the judges still convicted.43

Why did they convict these innocent people? Eight of them had con-
fessed; the jurors or judges may have credited the confession over the
DNA. They may have credited thinly supported accounts, as in Jeff rey
Deskovic’s case discussed in Chapter 2, of how the semen could have
come from another person. In fi ve cases, defense witnesses presented
the DNA results that excluded the defendant. Although in some cases,
the DNA analysts worked for state crime laboratories and routinely testi-
fi ed for the prosecution, the prosecutors nevertheless tried to object to
the testimony to keep the jury from hearing about the DNA evidence of
innocence.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 101

DNA testing is not immune from invalid testimony, and three exoner-
ees, Gilbert Alejandro, Chad Heins, and Josiah Sutton, all had invalid
prosecution testimony describing DNA test results. Gilbert Alejandro’s
case involved invalid DNA testimony by the infamous Fred Zain, who
testifi ed in Bexar County, Texas, that he had obtained DNA test results
inculpating Alejandro. He told the jury “they could only have originated
from him.”44 That testimony was invalid. Although he told the jury that
“the beauty of DNA testing is that it can give you a hundred percent cer-
tainty,” that is not the case. DNA testing is based on statistics about the
percentage of the population that shares each ge ne tic characteristic that
is tested. The analyst must calculate what percentage of the population
could be expected to randomly match each characteristic in a ge ne tic
profi le (those characteristics are par tic u lar locations on the DNA strand
that are called “loci”; thirteen loci are now typically tested). The num-
bers may be very high, because DNA testing focuses on ge ne tic charac-
teristics that vary widely. Analysts can sometimes conclude that one per-
son in many millions or billions could be expected to randomly have the
same DNA profi le. However, it is invalid to conclude that the DNA could
only have come from the defendant, rather than present the relevant sta-
tistic. In Alejandro’s case, an internal inquiry concluded that Zain did
worse— he had in fact not fi nished doing the DNA tests in the case. Later
tests excluded Alejandro.

Another exoneree trial involving DNA evidence illustrates inaccurate
uses of statistics. In the Josiah Sutton case, the victim was raped by two
men in the back seat of her car. Semen was present in the vaginal swab
and on the stain removed from the back seat where the rape occurred.
The Houston Police Department Crime Laboratory analyst presented
highly misleading DNA results.45 The analyst provided invalid testi-
mony as Zain did in Alejandro’s case, testifying, “No other two persons
will have the same DNA except in the case of— of identical twins.” The
jurors were not told what percentage of the population would be expected
to randomly match the par tic u lar DNA profi le. Instead, the jury was left
with the misimpression that the DNA evidence uniquely identifi ed
Sutton.

The analyst left something even more important out of her testimony
and reports. The lab notes actually showed that Sutton was not the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

102 Flawed Forensics

source of one of the stains; he was excluded.46 Five years later, new
DNA tests conclusively established Sutton’s innocence, and he was
freed.

Another DNA case, Timothy Durham’s, involved a serious laboratory
error and an error in the interpretation of the DNA test results.47 Errors
persist, even though DNA testing is now state of the art. These cases
provide a vivid reminder that even evidence as reliable as DNA evi-
dence, if presented in an invalid way in the courtroom or misinterpreted
in the laboratory, can cause a wrongful conviction.48 Innocence Project
cofound er Peter Neufeld commented that although DNA testing can be
“a truth machine,” nevertheless “any machine when it gets in the hands
of human beings can be manipulated or abused.”49

Bite Mark Comparison

Bite mark comparisons provide perhaps the most notorious of all the fo-
rensic techniques used in these exonerees’ cases. In seven cases exoner-
ees were convicted based on bite mark comparisons, and in fi ve of those
cases the testimony was invalid. In the other two cases, vague terms like
“consistent” and “similar” were used. That should come as no surprise.
Forensic odontology, or bite mark comparison, is a discipline with no
objective criteria at all. The discipline of “forensic odontology” involves
interpretation of lacerations, abrasions, and bruises of questionable ori-
gin on decomposing skin. Such bite mark work is “based on the assump-
tion that every person’s dentition is unique,” though this assumption has
not been tested.50 But even if true, the uniqueness of teeth is far easier to
identify when comparing pristine wax molds of teeth made in a dentist’s
offi ce. Forensic odontolists admit that skin is a “poor impression me-
dium” for bite marks, and particularly in what they euphemistically call
a “dynamic” biting situation. It can be hard to say whether a few marks
come from a human bite and not some other kind of abrasion related to
decomposition or the fatal attack. Experts regularly disagree not only
about whether a bite mark matches a defendant, but whether a mark
comes from a human bite at all.51

For all of those reasons, the NAS report described how “no scientifi c
studies support” the assumption “that bite marks can demonstrate suf-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 103

fi cient detail for positive identifi cation.” More damning, the report con-
cluded that “[s]ome research is warranted in order to identify the circum-
stances within which the methods of forensic odontology can provide
probative value.”52 Somehow judges appear unconcerned by the scien-
tifi c disrepute of bite analysis, and “no reported case has rejected bite
mark evidence.”53

In the cases of Kennedy Brewer, Ray Krone, Willie Jackson, and Rob-
ert Lee Stinson, odontologists testifi ed they were certain that the defen-
dant left the bite marks.54 Such testimony was erroneous and without
scientifi c support. Ray Krone’s case was particularly troubling, because
he was sentenced to death, and the central evidence at his trial was faulty
bite mark evidence. Two experts concluded that the defendant made the
bite mark on the victim. A forensic odontologist presented the bite mark
evidence at trial, along with a dentist who was just beginning to serve as
the police department’s odontologist. The odontologist presented a
highly infl ammatory and unusual video with images of the dentist hold-
ing molds of Krone’s teeth to the marks on the deceased victim’s body.
The odontologist gave the jury probabilities at trial:

And it turns out that on average a tooth can be in about 150 diff erent

positions, each one of which is easily recognizable. And if you are

looking at a tooth in that kind of detail, then you can see that very

quickly. Just having two teeth, the possibilities of two teeth being in

the same position, it would be 150 times 150, what ever that is. Maybe

1200 or something like that.55

Actually 150 times 150 is 22,500, but those numbers made up on the fl y
merely led to the ultimate conclusion. The odontologist then told the
jury in no uncertain terms that Krone had left the bite marks:

A. That’s as nice a match as we— as we really ever see in a bite mark

case.

Q. By “nice” do you mean accurate?

A. Yes. That was a nonscientifi c term. This is really an excellent

match, and would be held in high regard by forensic odontolo-

gists. Now there’s a wiping action just to show the same thing.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

104 Flawed Forensics

Again, high correlation. I mean, that is— that tooth caused that

injury.

On cross- examination he added, “we are talking about a specifi c indi-
vidual, and these teeth are unique.” The dentist similarly testifi ed, “I say
that there is a match. Okay? I’m saying there’s a defi nite match.”

Not only was that testimony erroneous and unsupported, but the de-
fense never learned that, before trial, police had initially consulted an FBI
odontologist who after examining the bite marks concluded that Krone’s
teeth were very diff erent from the marks. He said, “It could not have been
clearer . . . Ray Krone had two higher teeth than his incisors that would
have marked when he bit. Those weren’t there in the evidence.”56

The same odontologist reappeared in Robert Lee Stinson’s case in
Wisconsin, testifying that there was “no question” that his teeth were a
“match” to the bite marks and that the evidence was overwhelming. He
agreed with the “very good work- up” of a colleague, who testifi ed the bite
marks “would have to have been made by Robert Lee Stinson.” Stinson
was exonerated by postconviction DNA testing in 2009. A review of the
bite mark analysis in the case later criticized the testimony as having no
scientifi c basis and found that the analysts “should have excluded Robert
Lee Stinson.”57

These cases vividly show how, while “there is no quantitative base for
bitemark analysis,” analysts still off er conclusions that express near cer-
tainty.58 The guidelines of the American Board of Forensic Odontology
(ABFO) long permitted its members to render conclusions expressing
near certainty— they could conclude that a bite mark matches a criminal
defendant to a “reasonable medical certainty” and “high degree of cer-
tainty,” explaining that the intended connotation is a “virtual certainty;
no reasonable or practical possibility that someone else did it.”59 No
scientifi c criteria exist for what observations and analysis could permit
an expert to draw such conclusions.

The guidelines added that while experts may not convey “uncondi-
tional certainty,” they may express “reasonable medical certainty,” and
noted that it was “acceptable to state that there is ‘no doubt in my mind’
or ‘in my opinion, the suspect is the biter’ when such statements are
prompted in testimony.” The guidelines recommended that experts

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 105

should exaggerate their fi ndings in response to questioning. Perhaps in
response to criticism, the ABFO revised its guidelines in 2009 to say:
“Terms assuring unconditional identifi cation of a perpetrator, or with-
out doubt, are not sanctioned as a fi nal conclusion.”60 However, little has
changed. The guidelines still allow invalid conclusions, such as that a
person “beyond a reasonable doubt” made a bite or “more likely than not”
made a bite. And not only are the guidelines faulty, but the underlying
method is still unreliable.

Shoe Print Comparison

One exoneree, Charles Fain, had invalid shoe print testimony at his
trial. The shoe print comparison method is unreliable and it is subjec-
tive. While it can provide useful evidence if the defendant’s shoes do not
match prints from the crime scene, it is not clear what it means for the
defendant’s shoes to look similar to those from the crime scene. After all,
shoes are mass- produced objects that can be expected to leave similar
prints. At Fain’s trial, an FBI agent testifi ed that the make of the shoe
print was consistent with Fain’s, and “It was possible that this shoe
made this impression.”61 Not satisfi ed with his initial cautious, although
vague, conclusion, he added that although it was a common type of boat
shoe sole, wear patterns individualized the print:

Q. Now, did you indicate that the wear characteristics are put there

by a gait of a par tic u lar individual?

A. You would have to have the same characteristic walk as the indi-

vidual who owned those shoes.

No research suggests that the eff ect of gait on the sole of a shoe is unique.
He added that analysts could go even farther to say that wear patterns on
shoes “correspond exactly.”

Unfortunately, that is the case. Other analysts might go farther on the
recommendation of the Scientifi c Working Group on Shoeprint and Tire
Tread Evidence, which off ers the guideline that an examiner can fi nd an
“[i]dentifi cation (defi nite conclusion of identity).”62 The guideline ex-
plains, “This opinion means that the par tic u lar shoe or tire made the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

106 Flawed Forensics

impression to the exclusion of all other shoes or tires.” Thus, the guide-
lines that these analysts use allow them to reach invalid conclusions.

Voice Comparison

David Shawn Pope’s case involved the use of a totally discredited foren-
sic discipline, voice comparison, an unreliable method that should have
disappeared from our courtrooms de cades ago. Analysts use a spectro-
graph, an instrument that generates a visual pattern depicting an audio
recording using lines that represent the frequency and intensity of the
sound wave over time. The NAS issued a report in 1979 concluding that
using voice spectrograph analysis to identify individuals “is not ade-
quately supported by scientifi c data.” The FBI then stopped permitting
its agents to testify using such analysis. However, some judges continued
to allow such testimony.63

In Pope’s case, the victim of a 1985 rape in Garland, Texas, received
several messages on her answering machine shortly after the crime. The
Dallas County police arrested Pope after the victim identifi ed him in a
lineup. The State retained an expert, who testifi ed that he found “10– 15
similar patterns” shared by a recording of Pope’s voice and the recording
from the victim’s answering machine. He concluded that “the original
producer of these was the same individual.”64 The defense also retained
an expert, who cited to the NAS report and testifi ed that studies showed
that spectrography “is totally unsuitable as a tool for identifying voices
with any degree of accuracy.” Nevertheless, Pope was convicted and spent
fi fteen years in prison before DNA testing exonerated him. If judges simply
ignore scientifi c research that utterly discredits a forensic technique, such
erroneous convictions should come as no surprise.

Fingerprint Comparison

While twenty exonerees had prosecution testimony concerning fi nger-
print comparison, in all but one of those cases the analysts testifi ed that
the prints excluded the exoneree. Unlike the others, Stephan Cowans
was convicted in part due to an erroneous fi ngerprint identifi cation. A
Boston police offi cer was shot by a civilian, and during his escape, the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 107

assailant picked up a glass mug and drank from it. The police lifted two
latent fi ngerprints from the mug. After Cowans became a suspect, a Bos-
ton Police analyst compared Cowans’s known ink thumb print to one of
the latent prints and declared a match. After Cowans was exonerated by
postconviction DNA testing conducted on the glass mug with the fi nger-
print and on a baseball cap the perpetrator left at the scene, the district
attorney asked the Massachusetts State Police to reexamine the thumb
print. The State Police declared that Cowans was clearly excluded. The
Boston Police then hired an external auditor to conduct an in de pen dent
investigation. The audit team had four members, all experts in fi nger-
print comparison. The auditors reached the unanimous conclusion that
the analyst had realized at some point prior to trial that Cowans was ex-
cluded, but that he concealed that fact and instead told the jury that the
print matched Cowans’s.65

We have seen how many exonerees had testimony concerning forensic
methods that were unreliable, like hair and bite comparison testimony.
In addition, in 61% of the exonerees’ trials, analysts testifying for the
prosecution reached conclusions that were invalid and overstated the
evidence to make it seem like stronger evidence of guilt than it really
was. There was invalid testimony both by analysts using reliable meth-
ods like serology and DNA, and by analysts using unreliable methods
like hair comparison.

How about the 39% of these cases that did not have invalid forensic
testimony— what were those cases like? In more than half of those cases,
or thirty- one of the sixty cases, all of the forensic evidence presented was
inconclusive (seventeen cases) or tended to show innocence (seventeen
cases) or both. Such testimony obviously did not play a signifi cant role
in the conviction. In 12% of the cases (19 of 153 cases), analysts did not
overstate the evidence by reaching invalid conclusions. Instead, they
used terms like “match” or “similar” or “consistent” that were vague and
potentially misleading.

Next, I turn to additional problems beyond reliability and validity,
including analysts who concealed evidence, made errors in the lab, or
failed to test evidence.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

108 Flawed Forensics

Concealing Forensic Evidence

If there were so many problems apparent just from the testimony of
analysts on the stand at trial, how good was the work they did in the
laboratory? For one, we do not usually know whether analysts con-
cealed forensic evidence that would have shown innocence. In twenty-
two cases, the State either failed to disclose to the defense exculpatory
data, or analysis, or outright fabrication of evidence.66 These are only
those cases that have come to light. We do not know whether other evi-
dence of innocence has been concealed to this day or was destroyed
long ago.

Recall Earl Washington Jr.’s case from Chapter 2; he was a borderline
mentally retarded man who had falsely confessed to a brutal murder. His
lawyer never brought up forensic evidence at his trial, but one reason
was that at least one piece of forensic evidence had been concealed from
him. An analyst working for the Virginia Bureau of Forensic Science had
tested stains on a central piece of evidence, a blue blanket found on the
murdered victim’s bed, and found Transferrin CD, a fairly uncommon
plasma protein that is most often found in African- Americans. The ana-
lyst even ran a second test to double- check the result. The next year,
when Earl Washington Jr. was arrested, they tested his blood and found
he did not possess the unusual Transferrin CD. The State did not give
the defense the report indicating Washington was excluded by that char-
acteristic. Instead, the State gave the defense an “amended” report.
Without having done any new tests, the altered report stated that the re-
sults of the Transferrin CD testing “were inconclusive.”67 The original
lab report came to light de cades later when Washington fi led a civil
rights lawsuit after his exoneration.68

In Roy Brown’s case, the defense presented an expert who concluded
that the bite marks were inconsistent. Among other manifest diff erences,
one mark showed impressions of six teeth from the upper bite. Roy
Brown did not have six upper teeth. He had missing teeth— he had only
four upper teeth in his mouth. The prosecution never disclosed that the
New York State Police had earlier concluded that the marks were clearly
inconsistent with Brown’s teeth. Instead the prosecutor called to the
stand an odontologist who found the bite marks similar to “a reasonable

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 109

degree of dental certainty” and, while noting inconsistencies, called
them “inconsistent, but explainably so in my opinion.”69

In Gene Bibbins’s case, a Baton Rouge City Police analyst testifi ed at
trial that any comparison between Bibbins’s fi ngerprints and a print
found on the window at the crime scene was inconclusive, calling the
prints “unidentifi able. You can’t— they aren’t any— there aren’t any
prints on there that we can use.” When asked, “Did you double- check
your conclusion with the state crime lab?” and “did they have the same
results,” she answered, “Yes, Ma’am.”70 That testimony was false. The
state crime lab’s fi nding and report had excluded Bibbins. Other cases
involved concealment of exculpatory information regarding hair com-
parison. In William Gregory’s case, not only did the analyst testify about
the supposedly unusual characteristics in his hair, but she concealed
that at least one hair was found not to be consistent with Gregory’s
hair.71

Gross Errors in Analysis

We also do not know how many analysts made errors in the laboratory
in exonerees’ cases. Only a few exonerees had the evidence in their
cases retested or reexamined. Scandals involving faulty work at some of
our nation’s preeminent crime laboratories have led to investigations,
audits, and eff orts to provide in de pen dent oversight.72 Some of the
gross errors uncovered in these cases suggest a lack of adequate quality
control. Most crime laboratories do not routinely conduct blind profi –
ciency testing or in de pen dent audits. They chiefl y rely upon peer review
of casework.

In only a few exonerees’ cases was there an audit or an eff ort to exam-
ine whether errors were made in the lab. In some cases, errors happened
to surface when the DNA testing was conducted. For example, in several
exonerees’ cases, the analysts reported that they did not try to do serol-
ogy testing, because there was no semen in the sample to test. Yet when
later examiners reviewed the evidence using the same tool, a simple mi-
croscope, they saw abundant semen. In Larry Peterson’s case, “[a]lthough
the New Jersey State Police Laboratory had reported that there was no
semen in the victim’s rape kit,” the Serological Research Institute, before

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

110 Flawed Forensics

conducting postconviction DNA testing, “identifi ed sperm on her oral,
vaginal, and anal swabs.”73 Other cases involved faulty test results. In
the Ford Heights Four case, a Chicago Police Department analyst re-
ported that Dennis Williams was a Type A secretor, but Edward Blake
retested the evidence postconviction and found that he was in fact a
nonsecretor.74

Failures to Conduct Elimination Testing

The consequence of the faulty forensic analysts in many of these cases
was not only that an innocent person was sent to prison, but that real
perpetrators remained free. Sometimes the real culprits committed ad-
ditional crimes. And in at least some of these cases, the forensic analyst
tested the very person who was later shown by DNA testing to be the
guilty party, and they mistakenly concluded he could not have been the
culprit.

For example, in Ronald Williamson’s case, a supervisor at an Okla-
homa State Laboratory testifi ed unequivocally that he had compared
the hairs of the State’s witness Glen Gore with those at the crime scene.
He testifi ed, “I did direct comparison with the unknown hairs,” and
when asked if any of Gore’s hairs were microscopically consistent with
the questioned hairs, he testifi ed, “No, sir.”75 Later, during William-
son’s appeal, the Tenth Circuit reviewed the lab report, which revealed
that the analyst “did not compare Mr. Gore’s samples with unidentifi ed
hairs.”76 Indeed, the analyst had also testifi ed about a “match” of seven-
teen hairs, including both scalp and pubic hairs. Later it was deter-
mined that none of the hairs belonged to Ronald Williamson, nor to the
other individual charged in the case, Dennis Fritz. Instead, Glen Gore
was shown by postconviction DNA testing to have been the actual
perpetrator.77

Similarly, in the Robert Miller case, Joyce Gilchrist excluded a sus-
pect who was later identifi ed by DNA testing and indicted.78 In the Bea-
trice Six case, the four suspects who confessed had all been excluded by
serology testing, although analysts claimed, without any scientifi c basis,
that some mixture of several of them could account for the evidence.
However, early on in the case, police had focused on another man, a

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 111

convicted rapist who had been seen near the victim’s apartment the
night of the murder. He lived in Oklahoma City, so the Beatrice police
asked the Oklahoma City crime lab to obtain samples from him and test
them. Once again Joyce Gilchrist botched the tests, mistakenly exclud-
ing the suspect. Six people would then be wrongly convicted, and years
later that original suspect was identifi ed by DNA testing as the actual
culprit.79

Also troubling, some experts made clear during their trial testimony
that they only performed the specifi c test requested by police or pros-
ecutors, rather than all of the tests that could shed light on the guilt or
innocence of the defendant. In several exonerees’ cases, DNA tests could
have been conducted at the time, but were not. For example, in Marlon
Pendleton’s case, his lawyer asked to do a DNA testing, but the Chicago
Crime Laboratory analyst represented that there was insuffi cient mate-
rial present to conduct DNA tests, and the judge denied the motion.
Years later an expert tested the evidence, proving innocence and noting
that there was enough material to have conducted testing back in 1992
when Pendleton was convicted.80 In Neil Miller’s case, the prosecutor
said that serology testing on the sheet from the victim’s bed did not
match either the victim’s or Miller’s blood type. However, he explained
that the victim’s roommate had a boyfriend “who sometimes stayed
overnight.”81 The prosecution never had the boyfriend’s blood type
tested, which could have been easily done before trial. Later postconvic-
tion DNA testing revealed that the semen stain was not from the boy-
friend but was from the rapist, who was not Neil Miller.82

Lawyers and Judges

After hearing about all of this unreliable, invalid, and erroneous forensic
analysis, one naturally wonders what role lawyers played in allowing it
to happen. Prosecutors elicited the testimony. Did they help the forensic
analysts prepare and encourage them before trial to testify in an invalid
way? We do not know, but since prosecutors sometimes summarized in
their opening statements how they expected the analysts to testify, they
were certainly sometimes aware of what that testimony would conclude.
In addition, in eigh teen exonerees’ cases, as in Gary Dotson’s case, the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

112 Flawed Forensics

prosecutors themselves exaggerated the science in their closings argu-
ments. The ethical and criminal procedure rules regarding closing state-
ments include “few standards for proper prosecutorial argument,” and
though prosecutors may not misrepresent facts in evidence, they may
make arguments concerning inferences to be drawn from the facts.83
There may be a fi ne line between properly drawing inferences and mis-
stating facts. Judges typically order a new trial only in egregious cases in
which the conduct “so infected the trial with unfairness as to make the
resulting conviction a denial of due pro cess.”84

Many of these exaggerations may have left the jury with a faulty im-
pression of what the science had actually shown. The Illinois courts
granted Steven Linscott a new trial based on a fi nding of egregious
prosecutorial misconduct during closing arguments concerning the
forensic evidence. The appellate court damningly held that the prose-
cutor “simply made up” a claim that was contrary to what the analyst
had said and purported to make the serological evidence far more
powerful. In affi rming, the Illinois Supreme Court noted, “A prosecu-
tor must confi ne his arguments to the evidence and to ‘reasonable in-
ferences’ that follow from it . . . We believe that the prosecutor in the
instant case contravened this fundamental rule.”85 Despite that sharp
rebuke, the court did not recommend any disciplinary action against
the prosecutor.

In all of the other cases, judges provided no remedies. In Drew Whit-
ley’s case, the analyst, a laboratory manager at the Allegheny County
Crime Laboratory, examined a number of very short shaved or cut hair
fragments found on a stocking apparently worn by the perpetrator as a
mask. The analyst, though fi nding similarities, admitted the hair frag-
ments were unsuitable for comparison; “Because these hair fragments
were so small, I could not make the statement that they were micro-
scopically consistent . . .”86 During closings, the prosecutor claimed
that the analyst had said there was “no doubt those hairs came from
Drew Whitley.” In fact, she had specifi cally rejected that conclusion.
When the prosecutor had himself asked her, “You can’t say it belongs to
the defendant,” she had answered, “That is correct.” Despite telling the
jury the opposite of her actual testimony, the prosecutor embellished
further:

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 113

But it’s only when the scientists come in and say, hey, we have a stan-

dard, we know this hair to be of Drew Whitley and they compare it

all microscopically. Exact. No doubt about it. (Pointing.) Him.

The defense attorney asked for a mistrial: “That is absolutely, positively
not the evidence; and that is the most vital part of this whole case; and
for him to say that constitutes prosecutorial misconduct . . . She never
said that [the hairs] came from my client, Your Honor.” The judge
equivocated, stating, “I do recall that she answered my question as she
couldn’t say exactly who those hairs belonged to . . . I don’t know if she
did say it. I don’t recall.” When the prosecutor claimed he did hear such
a statement, and asserted, “It’s the jury’s responsibility to remember
things,” the judge provided a curative instruction, merely telling the jury
to resolve any discrepancy themselves.87

Just as important as the question whether prosecutors encouraged
this testimony is the question why defense lawyers did not eff ectively
challenge it. Defense lawyers rarely made any objections and they
rarely eff ectively cross- examined forensic analysts who provided in-
valid science testimony. Indeed, in forty- seven cases, or half of the
ninety- three cases involving invalid forensic testimony, the defense
lawyers failed to ask any questions at all about the areas in which the
analyst testifi ed erroneously. In at least twelve cases, defense counsel
failed to ask for DNA testing that could have proved their client’s
innocence.

Defense experts testifi ed in only twenty- one of the exonerees’ trials.
Perhaps defense attorneys could not be expected to understand the evi-
dence without access to their own experts. Judges frequently deny the
defense funding for experts in criminal cases, even where the forensics
are central.88 Nor is it easy to fi nd an expert, since most forensic analysts
work for law enforcement at crime labs and they usually will not testify
for the defense. Curtis McCarty did have a defense expert from the Kan-
sas City crime laboratory, who disagreed with the Oklahoma lab analyst
testifying for the prosecution. The Oklahoma lab then complained to
the Kansas City lab, which barred their analyst from testifying again as a
defense expert.89 In criminal cases, the pre sen ta tion of forensic science
testimony is almost always one- sided.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

114 Flawed Forensics

Reforming Forensic Science

Professor Eric Lander, a leading ge ne ticist, commented in 1989 that “fo-
rensic science is virtually unregulated— with the paradoxical result that
clinical laboratories must meet higher standards to be allowed to diag-
nose strep throat than forensic labs must meet to put a defendant on
death row.” Two de cades later, little has changed. Forensic testimony
involving unreliable methods and invalid conclusions are still routinely
allowed in our courts. Nor has DNA technology solved those problems.
Although DNA testing is now widely available in sexual assault cases, in
the overwhelming majority of criminal cases, labs still use methods like
those used in these exonerees’ trials.90

Over half of the 250 exonerees, or 128, had one or more of the prob-
lems discussed in this chapter: invalid, unreliable, concealed, or errone-
ous forensic analysis.91 Of the prosecution’s forensic analysts testifying
at these exonerees’ trials, 61% provided invalid testimony. Still more ex-
onerees had analysts use unreliable methods, and sometimes it came to
light that analysts concealed forensic evidence or made lab errors. This
presents a grim picture of forensics in serious criminal trials.

Although the cases of these exonerees cannot tell us whether unreli-
able and invalid forensic analysis was common at the time of their trials,
there is every reason to think that it was and also that it persists today.
One can readily fi nd cases with the same unreliable and invalid foren-
sics.92 In published opinions, judges discuss invalid testimony just like
that in these DNA exonerees’ trials.93 Despite de cades of scholarship
questioning the reliability of many of these forensic disciplines, judges
routinely permit testimony off ering unreliable methods and invalid
conclusions.

These exonerees’ cases also show how we cannot depend on the
adversary pro cess to correct forensic errors. Few exonerees’ lawyers
cross- examined analysts concerning invalid testimony. Few obtained
their own experts, since judges routinely deny funding for defense ex-
perts. Prosecutors, moreover, sometimes gave the jury false accounts
of the forensic evidence. In the few cases in which the defense chal-
lenged unreliable and invalid forensic science, judges denied their
motions.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 115

Because of these structural defi ciencies in our criminal justice sys-
tem, the responsibility for adopting sound scientifi c practices has rested
with forensic scientists themselves. Yet these wrongful convictions have
typically not led to audits or reform. In Gary Dotson’s case, for exam-
ple, there was no offi cial response to the invalid forensic testimony at
his trial. There was no audit ordered of the Illinois Crime Laboratory.
A series of other Illinois exonerations also involved erroneous or in-
valid forensic testimony. A de cade later, an Illinois Governor’s Com-
mission recommended creation of an in de pen dent crime laboratory,
separate from law enforcement. In 2003, the General Assembly failed
to adopt that reform. The laboratory is now peer accredited but still
has no in de pen dent oversight. As Rob Warden and Locke Bowman of
the Center for Wrongful Convictions pointed out, the City of Chicago
paid $36 million to settle a lawsuit brought by the Ford Heights Four,
who were convicted in part based on invalid forensic testimony and
then exonerated by DNA testing. It costs cities a lot of money when
crime labs are negligent, and that alone justifi es the cost of adopting
reforms.94

The few audits that have been conducted in the wake of exonerations
have uncovered systemic problems. Michael Bromwich led an investiga-
tion of the Houston crime lab, sparked by two DNA exonerations, and
his team uncovered hundreds of cases involving invalid analysis.95 An
NAS report uncovered invalid testimony by FBI analysts who testifi ed for
de cades that bullets had “come from the same box” without empirical
support.96 More should be done to examine the methods and conclusions
of forensic analysts, because the unscientifi c testimony in these exoner-
ees’ trials may be part of a worrisome problem extending far beyond known
wrongful convictions.

There is also reason to be optimistic: the issue of reliability and valid-
ity in forensics has fi nally attracted some attention. The American Bar
Association issued reform principles,97 and judges increasingly scruti-
nize forensic evidence.98 Meanwhile, Congress asked the National Acad-
emy of Sciences (NAS) to examine ways to improve the quality of foren-
sic science.99 Indeed, I began my research on the use of forensics in these
exonerees’ trials after I received a request to look into the matter from the
NAS.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

116 Flawed Forensics

In 2009, the NAS recommended sweeping reforms in a landmark re-
port. It found that, apart from DNA testing, no other forensic discipline
“has been rigorously shown to have the capacity to consistently, and
with a high degree of certainty, demonstrate a connection between evi-
dence and a specifi c individual or source.”100 That is, no other tech-
nique is as reliable. The NAS report described how research must be
done to provide a reliable scientifi c foundation for techniques like hair
comparison and bite mark comparison. That research has not been
done. The NAS concluded, “The bottom line is simple: In a number of
forensic science disciplines, forensic science professionals have yet to
establish either the validity of their approach or the accuracy of their
conclusions.” The NAS called on Congress to fund basic research. The
NAS report also advocated the creation of an entity to establish national
forensic science standards to regulate reports and testimony. The NAS
report cited to wrongful convictions as a reason to adopt reforms, stating
that “[n]ew doubts about the accuracy of some forensic science practices
have intensifi ed with the growing numbers of exonerations resulting
from DNA analysis (and the concomitant realization that guilty parties
sometimes walk free).”101 We will see whether strong federal legislation
emerges.

In the meantime, judges should not permit unreliable or invalid testi-
mony in courtrooms. The U.S. Supreme Court recently highlighted
these problems, with Justice Antonin Scalia writing in Melendez- Diaz v.
Massachusetts that the Sixth Amendment right to confront adverse foren-
sic experts is important in part because concerns have been raised about
the reliability of forensic methods and validity of forensic testimony.102 As
the Court indicated, however, the right to cross- examine a witness is not
enough.103 Judges should not permit experts to testify about unreliable
and unvalidated techniques, and they should not permit experts to exag-
gerate their conclusions.

Scientists, judges, and lawyers must all address the two problems iden-
tifi ed in these exonerees’ cases: the lack of reliability of many forensic
methods, and the invalid conclusions that analysts may reach.104 Scien-
tists can use the opportunity to restore our confi dence in forensic science
by establishing a sound foundation for its pre sen ta tion in the courtroom.
Crime laboratories should establish sound quality controls and in de pen-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Flawed Forensics � 117

dent oversight and auditing. Judges should not allow testimony about
unreliable methods. Lawyers should scrutinize such testimony, push for
in de pen dent experts, and object when analysts cross the line on the
stand. Unless all sides overhaul the practice of forensic science, the same
unreliable techniques and invalid testimony will continue to contribute
to wrongful convictions.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

118

Things were looking better for David Gray. He had been ar-
rested and tried for a rape in Illinois. But at the end of Gray’s trial

in 1977, the jury could not reach a verdict. It was a hung jury. Why?
The prosecution case had started strong. The victim identifi ed Gray

and testifi ed that she was “absolutely positive” that he had attacked her.
The culprit had come once before to look at a motorcycle her son was sell-
ing. A few days later, the same man returned to look at the motorcycle.
This time, he sat on the motorcycle and grabbed the rearview mirror, and
then asked to come inside to use the telephone. There, he brutally raped
and stabbed the victim thirty- three times with a knife from the kitchen.
She said that her attacker had worn very fl ashy disco- era shoes— she de-
scribed them as “Shiny plastic wineish colored shoes, with high heels.”
He also grabbed and ripped the telephone off the wall.1

However, Gray had evidence of his innocence. Police dusted the rear-
view mirror of the motorcycle for fi ngerprints, and they found two
prints— which did not match David Gray or anyone else. The police also
dusted the telephone for fi ngerprints, and the print they found did not
match Gray.2 The forensic analyst found the serology evidence incon-
clusive, but the hairs from the crime scene did not match Gray or the
victim.3 The victim’s neighbor said David Gray looked like someone he

c h a p t e r 5

Trial by Liar

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 119

spoke to a few days before the attack. But he did not identify Gray at a
photo array, saying that Gray looked familiar, but he could not be sure.4

Gray also had a strong alibi. Gray’s employer testifi ed that he was at
work as a janitor at the Salvation Army all day on the day that the victim
had said he had fi rst looked at the motorcycle, as his boss’s time sheet
confi rmed. The day of the crime, he was with a group of relatives through-
out the day: his girlfriend, her mother, and his own parents.5 His girl-
friend noted that Gray never wore fancy shoes with high heels, insisting
that he had just two pairs of shoes: black work boots and suede dress
shoes.6 Maybe that alibi evidence, combined with the other evidence, ex-
plained why the jury remained at an impasse.

After a hung jury, the prosecution can ask for a new trial before a new
jury. This is because although double jeopardy would otherwise prevent
the state from prosecuting someone twice for the same crime, the fi rst
prosecution is considered incomplete due to the hung jury. The prose-
cutor, the assistant state’s attorney for Madison County, Illinois, and one
of the county’s “best- known prosecutors,” decided to pursue a second
trial. He immediately formed a plan. He needed a new star witness.

So the prosecutor paid a visit to the jail.7 He later recalled that he
“talked to other prisoners, who were in jail with David Gray, to ascertain
if they knew anything.” He wanted to fi nd out “if Gray had made any
admissions, concerning the rape case.”8 He found a man in the same cell
as Gray, who was twenty- two years old and in jail for burglary.9 Gray’s
cellmate would become a crucial witness in Gray’s retrial.

David Gray faced a new trial in October 1989. There was not only a new
star witness, but also a new prosecutor. At Gray’s retrial, the new prosecu-
tor was “widely regarded as dean of the area’s prosecutors.”10 Why was
there a new prosecutor? The prosecutor who handled the fi rst trial had
helped to secure the cellmate as a witness against Gray. He likely could not
prosecute the case anymore, because he was now a witness, based on his
own interviews with the cellmate- turned- jailhouse- informant.

The jail house in for mant came to the trial prepared. He had written
several pages of notes from his conversations in the jail cell with David
Gray, which he gave to the prosecutors. At trial, the in for mant testifi ed
that Gray had given a detailed description of the general features of this
brutal crime: “once inside the house, his friend raped the lady, and he

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

120 Trial by Liar

said that after he raped her, he went into the kitchen, and got a knife out
of the drawer, and, after that, he said that he stabbed the lady a lot of
times.” They then “went through the purse and found the checks,” and
then left.11 The presence of a second attacker was contrary to the victim’s
account. The victim gave the culprit all of her money, and off ered him a
blank check from her billfold, but he did not take it.12 However, she was
raped in her home and stabbed many times with a kitchen knife.

Then in a very brief portion of his testimony, the jail house in for mant
described three key facts, as follows:

Q. Okay, did he say anything about a telephone?

A. Oh, yes. He jerked— one of them jerked the phone off the wall.

Q. Okay, did he tell you what color of the shoes he had on?

A. Wine colored.

Those facts were crucial to the prosecution case. First, the in for mant said
he could confi rm the most unusual feature of the victim’s description of
her attacker: the wine- colored shoes he wore. Second, the attacker had
violently ripped the phone off the wall. Third, the in for mant added a new
detail: he said that Gray admitted that he “had some gloves or somethin[g]
on his hands.” This could explain why no fi ngerprints matching Gray
were found on the telephone.13

Next, the in for mant bolstered the uncertain identifi cation by the vic-
tim’s neighbor, who had said that Gray merely looked familiar. The in-
for mant claimed that Gray told him in jail that he had tried to burglarize
the victim’s house a few days before, but “he seen a man next door, and
he talked to the man next door.”14

Leaving no stone unturned, the in for mant fi nally addressed Gray’s
alibi. Gray’s employer and his parents had all testifi ed about his where-
abouts the day of the crime. The in for mant said Gray bragged that “he
could account for everything but one hour, and his folks were gonna take
care of that for him.”15 The in for mant suggested that Gray had admitted
to staging his alibi by encouraging his parents to testify falsely.

These striking details could have come from the prosecutor or police,
but the in for mant and the prosecutor denied that any facts were fed. The
prosecutor from the fi rst trial, testifying as a witness, explained that the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 121

“[o]nly fact that I told him, prior to him telling me anything, was that it
was a rape case.” He agreed that the in for mant’s information would have
had to come from Gray.16 Similarly, the in for mant was asked whether the
prosecutor said “anything about the nature of the facts, or anything
else?” and he answered, “No, he didn’t say nothin.” He was then asked,
“So, he didn’t tell you about the maroon shoes?” He answered, “No, he
did not.”

We do not know if the in for mant “was being told what the facts were”
by a prosecutor or the police. Other inmates in the jail could have also
heard about the evidence at Gray’s fi rst trial and passed on the details by
word of mouth. Gray’s defense attorney argued that the in for mant spoke
to police months after a preliminary hearing, and at that hearing, con-
ducted long before trial, “there was testimony in Court about the shirt
and shoes and the telephone, and it was a well known story around here,
wasn’t it?” The in for mant denied having heard any of those facts, but he
did agree that inmates were generally familiar with things going on at the
court house. He had incentives to try to locate that information. Because
they can obtain substantial leniency by helping prosecutors, in for mants
may engage in aggressive eff orts to turn on other inmates.17

What did this in for mant get in exchange for providing the prosecution
with so much help? The in for mant said he received nothing. He ex-
plained that he had “more or less” testifi ed out of his sense of duty as a
“good citizen.”18 However, he was a repeat jail house in for mant. He was
facing three years in prison and had much to gain from helping the State.
In his closing statement, the prosecutor minced no words, and called the
in for mant “a bum”— but added that “he’s been used in the past, too, and
the information that he has provided has been very reliable, in other situ-
ations. This is not uncommon, in law enforcement . . .”19 The prosecutor
noted that the in for mant had every reason to lie, stating: “[I]f he thought
it would do him any good at all, he would lie . . . He’s a thief, he’s a bur-
glar, he’s a liar, and not only that, but just a few days before he was talked
to, he had been sentenced to three years in the penitentiary, and he was
petrifi ed, afraid to go to the penitentiary.”20

The prosecutor from the fi rst trial who fi rst located the jail house in-
for mant denied he gave any promises of a deal, at least when they fi rst
spoke. For his part, the in for mant said that when he fi rst spoke to the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

122 Trial by Liar

prosecutor that he had already heard Gray admit his guilt.21 The prose-
cutor stated that maybe a vague understanding had been reached with
the in for mant, but that there was no deal. He testifi ed, “I know there was
nothing defi nite said or off ered, but I don’t know whether I may have said
something like, ‘It won’t hurt you,’ or something like that. I just can’t re-
member.”22 A prosecutor is obligated under the Supreme Court’s ruling in
Brady v. Mary land to tell the defense about evidence that could be used to
question the truthfulness of a witness, including promises or deals made
with a witness. However, the prosecutor denied that there was a deal with
the in for mant, so he had nothing to disclose to the defense.

Those denials ring false. Although the in for mant was sentenced to
three years, he served only seven months. He testifi ed at Gray’s trial that
“they told me all I would have to do was six or seven months, what ever
they felt was enough time on the burglary.”23 In fact, he was told that he
would be let out a few weeks after he testifi ed at the Gray trial.24 All of
these substantial benefi ts, though, were supposedly provided in grati-
tude for the help he had provided in other cases, and had nothing to do
with his upcoming testimony in the Gray case. He claimed, “They told
me that I was getting out, no matter what. The told me that I didn’t have
to testify” in the Gray case.25 The judge denied the defense motion to
suppress the in for mant’s testimony.26

The prosecutor acknowledged that the in for mant had good reasons
to lie. However, he did not agree that the testimony was worthless. To
the contrary, he emphasized that the in for mant provided reliable infor-
mation in Gray’s case. How could the prosecutor know that? He claimed
that the in for mant reported admissions from Gray that only the true at-
tacker could have known. During the closing arguments, the prosecutor
argued:

Number one, he had heard David Gray, talking in the jail, when he

was a cellmate with him and bragging about, yeah, he had done it.

He referred to the fact, number one, that he ripped the phone off the

wall. That’s a fact that would have been unknown to any person,

other than police offi cers, members of the State’s Attorney’s Offi ce,

[the defense attorney] or Mr. Gray . . . He had wine shoes, that he

was there, at the time of the commission of the off ense . . .27

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 123

The prosecutor also emphasized that these facts were in some respects
inconsistent— the victim did not describe a second attacker— thus “[i]f
he was being told what the facts were, his story would have been right
down the line with the facts, and it wasn’t. He was hearing the version of
David Gray, and he was not even hearing the accurate facts.” Although
the State admitted the in for mant would “say just about anything, to get
out of prison,”28 those allegedly nonpublic facts strengthened their case.

After all, the in for mant testimony was the only evidence corroborating
the victim’s identifi cation. Recall how the forensic analysis of the fi nger-
prints, semen stains, and hair either was inconclusive or it did not match
David Gray. However, the in for mant’s testimony could explain the lack of
fi ngerprints on the telephone. Gray had strong alibi testimony, but the
in for mant’s testimony could explain that away too.

The jail house in for mant’s carefully crafted lies may have made the dif-
ference between the hung jury at the fi rst trial and the conviction at the
second trial. At his second trial, David Gray was sentenced to sixty years
in prison. He served twenty years before being exonerated by DNA test-
ing in 1999. As for the in for mant, when asked about the case years later,
he said he no longer remembers David Gray having ever admitted his
guilt to him.29

In for mants and Jail house In for mants

In for mants have never been thought to be particularly reliable sources.
At common law, evidence from “persons of infamy” was not allowed at a
trial. Use of in for mants was later defended as a necessary evil. The Su-
preme Court, quoting Judge Learned Hand, has noted that while “[c]
ourts have countenanced the use of informers from time immemorial,” it
is because “in cases of conspiracy, or in other cases when the crime con-
sists of preparing for another crime, it is usually necessary to rely upon
them or upon accomplices because the criminals will almost certainly
proceed covertly.”30 In or ga nized crime or drug prosecutions, reliance
on in for mants is pervasive and sometimes necessary to penetrate a closed
criminal enterprise by encouraging participants to cooperate or “fl ip.”
However, these exonerees were convicted of rapes and murders that
were not the work of a criminal or ga ni za tion. Nor were these in for mant

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

124 Trial by Liar

statements recorded and carefully documented as in a meticulous or ga-
nized crime investigation. Most common in these exonerees’ trials was
testimony by a diff erent kind of insider— jailhouse in for mants. In for mants
who are already in jail and testify against cellmates have long been con-
sidered notoriously unreliable sources.31

When I began looking at the exonerees’ cases, I knew that in for mants
may lie in exchange for a prosecution deal. It would come as no surprise
if that occurred in the exonerees’ cases. I found that 21% of the exonerees
(52 of 250 cases) had in for mant testimony at their trials. Of the 52 that
had in for mants, 28 were jail house in for mants; in addition, 23 had code-
fendant testimony and 15 had testimony by confi dential in for mants or
cooperating witnesses (some had more than one type). In almost all of the
cases, the in for mant reported incriminating statements allegedly made
by the defendant.

When I began reading the transcripts of what these in for mants actually
said at trial, I uncovered a diff erent and more troubling problem. These
in for mants claimed that the exonerees not only admitted guilt but knew
details about the crime, which the prosecutor or police claimed only the
culprit could have known. Just as exoneree confessions were contami-
nated by disclosure of inside information, so were these in for mant state-
ments. Unlike in the interrogation setting, these in for mants came forward
later during an investigation. Often they surfaced close to the time of trial.
They may not have always obtained their information from police, but
they sometimes may have obtained it from prosecutors, or perhaps jail-
house sources.

I was most amazed by how the most aggressive in for mants delivered
“made to order” statements neatly molded to the litigation strategy of the
State. Their testimony included details designed to undermine the de-
fendant’s alibi, address weaknesses in the prosecution’s case, or enhance
prosecution evidence. If the fi ngerprint evidence was inconclusive, the
in for mant testifi ed that the defendant confessed to wearing gloves. If an
eyewitness could not identify the defendant, the in for mant testifi ed that
the defendant shielded his face on purpose. These cases suggest the need
for judges to carefully screen testimony by such in for mants. Few states
require any review of such testimony, nor do they require careful docu-
mentation of contacts with in for mants. One wonders whether jail house

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 125

in for mant testimony in par tic u lar should be permitted at all, but such
testimony certainly deserves rigorous scrutiny.

Jail house In for mants in Murder Cases

Prosecutors frequently use in for mants, and they can be indispensable.
In investigations of criminal enterprises, like or ga nized crime groups or
drug gangs, the only way to obtain inside information is to secure the
cooperation of a person on the inside. As the Ninth Circuit Court of Ap-
peals put it, “Without in for mants, law enforcement authorities would be
unable to penetrate and destroy or ga nized crime syndicates, drug traf-
fi cking cartels, bank frauds, telephone solicitation scams, public corrup-
tion, terrorist gangs, money launderers, espionage rings, and the like.”32
Police and prosecutors use the cooperation of such witnesses to pursue the
“big fi sh” and to break up close- knit criminal enterprises. More broadly,
police depend on the conscientious assistance of community- minded citi-
zens to report crimes they witness or hear about.

The types of cases that these exonerees were convicted of did not in-
volve use of cooperators or in for mants to infi ltrate a criminal enterprise.
These cases, typically murder prosecutions, sometimes relied on tips
from concerned citizens in the community. However, they also used in-
for mants who had more selfi sh motives. The cases with in for mants were
concentrated in the murder prosecutions: thirty- two were in cases that
involved a murder and a rape, nine involved murder, and eleven involved
a rape. The cases involving jail house in for mants were even more dispro-
portionately murder cases. The jail house in for mants claimed to have
overheard the defendant, whom they had never met prior to being incar-
cerated together, confess to a crime. Eigh teen of the twenty- eight cases
involving jail house in for mants involved a murder and a rape, and six in-
volved a murder.

In murder cases, unlike in rape cases, there is often no victim or eye-
witness who can identify the perpetrator. A confession becomes all the
more critical to solving the case, and if the defendant will not confess
during an interrogation, then police may rely on jail house in for mants to
secure admissions of guilt. The in for mant may get a better deal from
prosecutors for helping to solve a more serious murder case. As law

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

126 Trial by Liar

professor Samuel Gross explains, “[I]f the witness is lying to get favors
unrelated to the crime at issue, he will do much better if it is a big case,
which usually means a murder, or better yet, a capital murder.”33

Just four of the 171 exonerees convicted of rape had jail house in for-
mants at trial. The four rape cases were all cases in which the jail house
in for mant’s testimony was used to bolster a victim who had diffi culty
identifying the defendant. Prosecutors knew that David Gray’s case was
weak, because they had a hung jury. In Bruce Godschalk’s case, one
victim was initially uncertain about her identifi cation and the other vic-
tim could not identify him at all. In Kenneth Wyniemko’s case the at-
tacker had worn a stocking mask, and forensic evidence excluded him.
In Wilton Dedge’s case, the victim initially described a much larger and
taller man.

Though in for mants provided false inculpatory information in many
exonerees’ cases, this is not to say that prisoners do not sometimes tell
each other the truth when they talk about their cases. Indeed, Clark Mc-
Millan found evidence of his innocence in jail. When he was fi rst ar-
rested, he spoke to the man in the next cell, David Boyd. McMillan told
him what he had been charged with, and Boyd responded that he should
not worry, because he (Boyd) “was good for that crime.” When McMil-
lan tried to introduce that helpful jail house testimony during habeas
proceedings, he had no luck. Twenty- three years later, DNA testing ex-
cluded McMillan and matched Boyd.34

Jail house in for mant testimony, as the Gray case illustrates, is chiefl y
regulated in two ways. First, under the U.S. Supreme Court’s decision
in Brady v. Mary land, the prosecution must disclose to the defense in-
formation about any promises or deals that could be used to challenge
the in for mant’s truthfulness. As the Supreme Court held in one of its
early due pro cess decisions, Napue v. Illinois, the prosecution cannot sit
by idly while a cooperating witness falsely denies the existence of a prom-
ise by the prosecution to provide leniency in exchange for that testimony.
The prosecutor must correct the record if the in for mant lies about a
deal.35

Second, the Supreme Court in Massiah v. United States held that the
defendant has a right to have a lawyer present during questioning by an
in for mant that is already reporting to the police. That is, if there was a

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 127

formal deal with the prosecution and the in for mant was planted to get
information, the defendant has a right to counsel.36 With a lawyer pres-
ent, an inmate will be much less likely to talk to an in for mant. The Su-
preme Court carved an exception to the Massiah rule: even if the State
does unconstitutionally plant an in for mant in an inmate’s cell, the in for-
mant can still be put on the stand to challenge the truthfulness of the
defendant’s testimony.37 Regardless, police and prosecutors have an in-
centive to keep deals with in for mants highly informal. The deals are
rarely in writing, they are often not negotiated with lawyers present, they
can be subject to change, and they are typically conditioned on continu-
ing cooperation.38

The Supreme Court has done little to ensure the reliability of in for mant
testimony, making clear that it leaves “the veracity of a witness to be tested
by cross- examination.”39 But cross- examination did not uncover the con-
tamination of these in for mants’ statements, just as cross- examination did
not typically bring to light the contamination of confessions, the eff ect of
suggestion on eyewitness memory, or the invalid pre sen ta tion of forensics
in these exonerees’ trials.

In the rest of the chapter, I turn to what these in for mants said. I fi rst
discuss the jail house in for mants, then the testimony of codefendants, and
fi nally, testimony by confi dential in for mants or cooperating witnesses.

Deals with Jail house In for mants

Jail house in for mants were the most common type of in for mant in these
exonerees’ trials. These jail house in for mants invariably testifi ed that
they had no contact with law enforcement until after they heard the in-
criminating information. Only two of the twenty- eight jail house in for-
mants admitted they made deals with the State. The in for mant in John
Kogut’s trial admitted that police promised him “if you tell us, you know,
what you know, we’ll see if we can get you a shorter term in jail.”40 In Wyn-
iemko’s trial, the in for mant admitted receiving a plea deal in exchange for
his testimony, through which he would spend one year in county jail
rather than the fi fteen that he had faced.41

The other jail house in for mants all denied that there was any explicit
promise of a benefi t in exchange for their testimony— though many

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

128 Trial by Liar

acknowledged they did obtain a more favorable outcome in their case for
one reason or another. In Dennis Williams’s trial, the in for mant testi-
fi ed: “There was no promises.”42 In Drew Whitley’s trial, the jail house
in for mant was asked, “Has anybody promised you anything in exchange
for your testimony, sir? A. No, sir.”43 In some cases the in for mants de-
nied a deal, but admitted that they hoped to obtain a benefi t from testify-
ing. In Wilton Dedge’s case, the in for mant denied any deal was struck,
but was asked, “Now, you are hoping for some consideration from your
testimony in this case, are you not?” and responded, “I hope it will look
favorably to the Parole Board, yes sir.”44

In some cases, like David Gray’s, the State’s denials that a benefi t had
been promised ring hollow. At Calvin Washington’s trial, the State in-
troduced testimony by a litany of cooperating witnesses, including two
jail house in for mants. Washington’s lawyer called them “a parade of li-
ars.” He noted that since the forensic evidence excluded Washington, all
of the State’s evidence came from “Rapists. Burglars. Thieves. Habitual
criminals. People who escape from custody. Right down the line. And
they are putting you on the spot by asking you to subject a man to the
possibility of death based on this.”45 One member of that parade met
Washington at the McLennan County Jail. That in for mant was facing
charges of theft and unlawfully carry ing a weapon. He claimed he re-
ceived no deal of any kind. In fact, he said that the offi cer “told me he
would get back with me, and he never did get back with me.”46 The in-
for mant then testifi ed: “Hasn’t nobody made me no off ers.” Yet it
emerged during cross- examination that his two felony convictions were
“dropped . . . down to a misdemeanor.” Though he could have received
life in prison, he instead got “nothing” more than the four months he
had already served. He explained, implausibly, that “I got nothing be-
cause they had no evidence.”47

Similarly, in Steven Barnes’s case, the in for mant received eight months
for several felonies after reporting his cellmate’s statement. Barnes’s de-
fense lawyer incredulously argued to the jury that though the in for mant
denied any deal, “that’s the fi rst time in my life I have ever heard of a guy
with a record that long that didn’t go to State prison.”48

Some of the in for mant’s denials may have been accurate. Prosecutors
would typically not make any fi rm promises before trial for two reasons:

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 129

fi rst, because any such promises would have to be disclosed to the de-
fense and would further undermine the credibility of the in for mant; and
second, because prosecutors may not want to make any fi rm promises
until they actually hear the testimony and receive the cooperation. Pros-
ecutors can get around the rule that they must disclose deals to the de-
fense by leaving any deal vague and waiting to provide leniency until
after the trial is over.49

We do not know how many more of these in for mants did later obtain
leniency. There may be no written record of a deal, journalists may not
have investigated the matter after the exoneration, and one cannot usu-
ally tell from looking at judicial rec ords, because prosecutors in their
discretion may have informally dropped charges or pursued more le-
nient charges or a shorter sentence. In some cases, years later, the exon-
erees’ lawyers managed to uncover the existence of a deal. For example,
the State’s most important new witness at Rolando Cruz’s second trial
was a jail house in for mant. The in for mant testifi ed that Cruz had de-
scribed the victim’s gruesome murder in detail, and that when he read
an article in the “Chicago Lawyer” in which Cruz asserted his inno-
cence, he became angry and contacted the prosecutor. He denied receiv-
ing any promises of leniency from the prosecutor. However, he received
a resentencing after his testimony in the Cruz case. The prosecutor testi-
fi ed on behalf of the in for mant at his resentencing hearing by describing
how he “voluntarily provided testimony in the trial of Rolando Cruz”
but admitting that the in for mant was “a calculating individual” who
wanted to be sure before testifying in the Cruz trial that the prosecutor
would help him at his resentencing hearing. The Illinois Supreme Court
concluded in its opinion reversing Cruz’s conviction that the in for mant
had falsely denied having a deal with prosecutors.50

Similarly, in Ronald Williamson’s case, a jail house in for mant was a
crucial witness, who described how Williamson confessed in detail to a
brutal rape and murder. She was also an in for mant in another murder
case, and denied receiving any leniency or compensation in both trials;
however, she had most of her sentences for three felony convictions sus-
pended and never paid her fi nes. The federal district judge found that
Williamson’s lawyer was in eff ec tive for failing to attack her credibility
based on that evidence of her motivation for testifying.51

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

130 Trial by Liar

Rather than promise leniency, in some of these cases, prosecutors may
have used sticks and not carrots. That is, in for mants may have feared that
should they not cooperate they would face harsher treatment. This would
particularly be true if the in for mant had a possible connection to the
crime. In Calvin Washington’s case, another in for mant later explained in
a subsequent statement to a defense investigator that the police “kept
pressuring me” by saying “that I might be charged with Capital Murder.
I don’t need no capital murder charge— I don’t even know when she died
or what ever.”52 He added that he was worried “since I’m on parole you
know.”53

Jail house in for mants also denied that they were planted in the same
cell or vicinity in order to elicit incriminating information. For example,
in Charles Fain’s case, a man was in the same jail cell in the Canyon
County jail for “four or fi ve days.”54 He later asked the jailer to be moved
to another cell because Fain’s talk of “sexual conduct with little chil-
dren” “just made me sick at my stomach.” It would violate Fain’s right to
a lawyer for police to place the in for mant in that jail cell to interrogate
Fain. Yet this in for mant admitted that an offi cer spoke to him before-
hand and said, “they had a reason” for his being placed in the cell. The
judge looked into the matter, and asked the jailer specifi cally whether he
had been instructed “to plant him in there as a witness,” and the jailer
said no.55 However, the police offi cer testifi ed that the in for mant had of-
fered to gather information, but the offi cer refused, saying it would be
illegally obtained, since the defendant was represented by a lawyer.56
Despite the evidence that the in for mant was put in the cell after off ering
to obtain information about the case, the judge ultimately concluded
that “there is no evidence to establish that this was a covert or subterfuge
on the part of the state to plant [the in for mant] in this cell.”57

Specifi c Facts

Even when jail house in for mants deny they are testifying to get a deal, it
can be obvious to jurors that they are not the most trustworthy or altru-
istic witnesses. What made the testimony of these in for mants powerful
was that they described how these exonerees supposedly told them spe-
cifi c details that only the true perpetrator could have known. All but two

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 131

of these twenty- eight jail house in for mant statements included such spe-
cifi c details concerning the crime. The two cases of Roy Brown and
Steven Barnes involved bare admissions that they committed a murder,
but no further details. In Barnes’s case, for example, the in for mant claimed
to have made a reference to a girl, and said that Barnes replied, “You mean
the one I killed,” but he then corrected himself to say, “I mean the one I am
accused of killing.”58

The other jail house in for mants reported that the exoneree confessed
to details concerning the crime. Bruce Godschalk had falsely con-
fessed to police, and at his trial, the State also claimed that he had con-
fessed to a fellow inmate. The in for mant testifi ed, “And one time he was
mentioning about how he didn’t fi nish the job in the one, you know, and
this is a little embarrassing, and that the lady said . . . She was saying
something like, ‘My boyfriend was coming home.’ And he said— he told
me that he got scared and left.”59 One of the victims had described just
such an encounter.

Others described the murder weapon. During Ronald Williamson’s
trial, the jail house in for mant described Williamson as admitting how
the victim was killed: “He was telling— I guess in the bullpen, the guys
back there— that he— he said he shoved a coke bottle up her ass and her
pan ties down her throat.”

Williamson interrupted: “You are lying. I ain’t never said nothing like
that in my life. I did not kill this girl, and I call you a liar.” Williamson’s
lawyer told him: “Be still.”

The district attorney, meanwhile, needed to correct the record. It was
not a coke bottle, but a ketchup bottle that was found at the crime scene.
The prosecutor asked, “let me ask you about the details you were just re-
lating. As far as your memory goes, are you sure about the objects that he
stated he used. You said coke bottle.” The informant answered, “He said
a coke bottle or catsup bottle or bottle . . .”60 Williamson continued to
interrupt, calling her a liar, and eventually the judge called a recess so he
could calm down.

The entire time that the jail house in for mant was testifying during
John Restivo’s trial, Restivo was “muttering under his breath,” and “call-
ing this person a liar and shaking his head from side to side.”61 The judge
asked counsel to tell him to stop doing so; Restivo commented that he

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

132 Trial by Liar

was sorry, but “When I hear somebody lying through their teeth . . .”62
The in for mant then described Restivo’s confession, that along with two
accomplices, he had raped the victim, his accomplice “strangled her after
with a scarf,” and that they then “dumped the girl by some railroad
tracks.”63 As in the other cases, each of those details was corroborated by
crime scene evidence.

Several in for mants specifi cally denied having learned of the facts from
anyone other than the defendant. In Drew Whitley’s case, the in for mant
claimed, “He said the girl was shot in the back” and “he said he took the
gun and threw it over a hillside near McDonald’s into Kennywood Park.”
The in for mant was asked, “At that time had you any other information
concerning a killing at McDonald’s?” and he answered, “Not until he
came to talk to me.”64

In Charles Fain’s case, the prosecution introduced testimony by two
in for mants. The fi rst claimed he did not even know what Fain had been
arrested for.65 He then testifi ed that Fain confessed to picking up the
victim and was “molesting her,” and “when she got away, she tripped
and fell and she hit her head, her forehead. He said, ‘A knot come up on
her about the size of a golf ball.’ ” This injury was consistent with the
autopsy. He went on to say, “He wasn’t satisfi ed and fi nished with what
he was doing, so he went back and got his satisfaction. Then he took her
and put her in the ditch. He carried her to a ditch and held her head un-
der water.” The victim was indeed found drowned in a ditch. He testi-
fi ed that Fain even drew maps of where the murder took place, that Fain
“wadded them up and was going to fl ush them down the toilet,” but he
rescued the maps, and “picked them up out of the toilet and put them in
my pants so he couldn’t see them.”66 A second in for mant similarly testi-
fi ed that Fain said “he had caught the girl, and had placed her in some
water . . . and then he had left . . .”67

During closing arguments, the prosecutor emphasized that the fi rst
in for mant testifi ed that Fain told him about a “golf ball sized lump that
came on her head,” and that he held her head under water, and noted,
“I believe you have seen the photographs of that bruise on her forehead;
and that he drowned her in the water.”68 The prosecutor added that the
jury should especially credit the fact that there were two in for mants:
“[W]e have the statements made by . . . Not one, but two persons, and

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 133

I believe the old rule is that in the mouth of two or three witnesses shall
everything be established. This defendant committed those crimes.”

Similarly, in Chad Heins’s case, two jail house in for mants testifi ed for the
State. The prosecutor emphasized the specifi c and nonpublic facts that the
in for mants relayed—“What rings true . . . other than the fact they didn’t
get any deal or compensation from testifying? The words they claimed he
used. Remember . . . neither one of them knew anything about the facts of
this defendant’s case.”69

How did this happen, that so many jail house in for mants could report
false admissions that contained such detailed crime scene facts? It is
hard to uncover what actually transpired. Obviously, unlike other wit-
nesses, such as eyewitnesses, jail house in for mants can only come for-
ward after a person is arrested. As a result, they often speak to police af-
ter the investigation is well under way, sometimes close to the time of
trial or even, as in Gray’s case, after a fi rst trial. In Miguel Roman’s case,
the jail house in for mant came forward just as the trial was coming to a
close; the prosecutor moved to reopen the case to allow the in for mant’s
testimony. The later in the case, the more facts about the case have been
uncovered and made public to more people, and the in for mant has more
sources for that information. During an interrogation of a suspect, police
may know what the crime scene looked like and they may have some
initial forensic reports, perhaps eyewitness accounts. But by the time
police speak to a jail house in for mant, they may know much more about
the case. In addition, prosecutors, and not just police, may talk to the
in for mant. One troubling scenario is that the crime details that these in-
for mants repeated could have come from police or prosecutors.

Second, the exonerees may have themselves been the unwitting source.
While these innocent people would likely not have falsely admitted their
guilt to a jail house in for mant, they may have known a fair amount about
their own case by the time they were in jail. They may have talked to their
cellmates about their case and why they hoped to win at trial. They may
have explained why they were innocent and why the police were wrong to
suspect them. A clever in for mant could turn that information around to
fabricate a story that the defendant had admitted his guilt.

There is a third possibility. The in for mants may have learned about a
case from public sources or through a jail house network. In cases like

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

134 Trial by Liar

David Gray’s, where there had already been a hearing or a trial, word may
have gotten around about what evidence was presented in the courtroom.
In Calvin Washington’s case, for example, a witness told the defense that
“the ‘jail house rumors’ were that a certain State’s investigator, who testi-
fi ed at the trial, could help an inmate ‘get out of jail quick’ and that an in-
mate should see him ‘before you see your lawyer.’ ”70

Bolstering the State’s Case

The most enterprising jail house in for mants did not just know specifi c
facts about the crime. They knew the facts that the prosecutors had been
unable to prove any other way. Their statements were neatly tailored to fi t
the prosecution strategy at trial. Their testimony bolstered shaky prose-
cution witnesses, or undermined defense alibi witnesses, or explained
weaknesses in the forensic evidence. The jail house in for mant became a
sort of jack- of- all- trades, able to plug all of the holes in the State’s case.

Bruce Godschalk’s trial provides a case in point. He was tried for the
rape of two separate victims. Neither was wearing her glasses during the
rape. One saw the attacker only in a small mirror and had trouble identi-
fying him from a photo array, although at trial she testifi ed she was “ab-
solutely” certain Godschalk was the rapist. Nor did that victim describe
any facial hair— but Godschalk wore a mustache. A second victim could
not identify Godschalk at all.71 The in for mant testifi ed that Godschalk
had not only confessed to him, but he bragged about how he had made
sure that the victims would not be able to identify him, saying that “they
couldn’t identify him and stuff , because they didn’t get a good look at
him, only in a mirror, right?”72

Wilton Dedge was convicted of rape in Florida, but then his convic-
tion was reversed on appeal and he was granted a new trial. Dedge had a
powerful alibi. Eight coworkers at a garage had testifi ed that he had been
repairing auto transmissions the entire day. His workplace was a long
drive from the town where the victim was raped.

Nor was the case against Dedge very strong. The appeals court noted
that the victim’s identifi cation of him was “equivocal.” She described
her attacker as big and muscular, about six feet tall and 180– 200 pounds.
That was a far cry from how Dedge looked. He was fi ve foot fi ve inches

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 135

and 125 pounds. The State’s other witness was a dog, named “Harass
II.” That police dog supposedly identifi ed Dedge’s scent in a “dog scent
lineup,” in which the dog was given the victim’s sheets from the crime
scene.73 We should not blame the dog— the police department dog han-
dler later had his work in cases around the country discredited. Indeed,
the same dog handler also produced a false dog scent identifi cation in
exoneree William Dillon’s case.74

The prosecutors located a new witness for the second trial: a jail house
in for mant who had met Dedge only once, during a three- hour ride in
a prison transport van. The in for mant claimed that in that short time,
Dedge had confessed in great detail. Dedge admitted to raping a woman
in another town, “off a dirt road off U.S. 1,” which accurately described
where the victim lived. Yet Dedge’s coworkers said he was at work all day.
How could Dedge reach that far- off small town in the middle of the day?

The in for mant described an implausible high- speed adventure. He
claimed that Dedge described fooling his alibi witnesses into thinking
he was at work, but that in fact he had slipped away. “So when they testi-
fi ed about him being there at work, as far as they knew they were telling
the truth because when he did return back to work, he didn’t— no one
seen him come back in, either, he just eased his way back into the thick,
and it was about quitting time, anyhow, the end of the day.” That did not
explain how he could have made his way to a far- away town and back
without them noticing. But he claimed that Dedge “he had the trip . . .
down pat where he could make it in fi fteen minutes.”

How could that be done? Dedge had a Kawasaki motorcycle. And ac-
cording to the in for mant, “he cracked the throttle all the way open there
and he said in a very short time the cable started bounding real bad, and
the cable snapped. And he said he knew he was going way over a hun-
dred and sixty miles an hour at that time.”75 The prosecutor then called
as an expert witness to corroborate the account, a motorcycle racer; he
explained that at speeds of over 126 mph, an odometer can break off .76

The in for mant claimed that Dedge said he had intentionally gotten in
a fi ght that eve ning to establish an alibi; “he had danced with some bik-
er’s old lady, and the biker got mad, they got into a fi ght which caused a
ruckus, and the police came and made a police report, and that further
established his alibi.”77

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

136 Trial by Liar

The in for mant also undermined the defense attempt to challenge the
victim’s identifi cation; she had initially described a large man having at-
tacked her, where Dedge was not large. “[H]e said, that shows you what a
dizzy bitch that she is, that she can’t even get her details right. He said—
that’s when he said, look at me, because I’m only fi ve foot six and a hun-
dred and thirty- fi ve pounds.”78 Adding a fi nal note, painting the defen-
dant as confi dent that his various ruses would succeed, he claimed that
Dedge bragged “he was confi dent he would win his new trial.”79

The prosecutor emphasized in the opening arguments that the in for-
mant had uncovered how Dedge “had basically pulled one over on his
alibi witnesses,” because “he had a Kawasaki 900, that he did a hundred
and sixty getting back, that he broke the speedometer, that it actually
sheared off and quit operating because he was going so fast.”80 Indeed,
almost every prosecution argument was bolstered in one way or another
by the in for mant’s remarkable testimony.

Of course, it all turned out to be an elaborate lie. After years of fi ght-
ing to obtain DNA testing, and after multiple DNA tests that all proved
his innocence, Wilton Dedge was exonerated by DNA testing in 2004,
after serving twenty- two years in prison.

As for the in for mant, he had been facing murder charges, but had also
testifi ed in two codefendants’ cases and in an unrelated murder case. He
pleaded guilty and had 120 years taken off his sentences for testifying
against Dedge and the other inmates.81 The in for mant said at Dedge’s
trial that he had come forward not because he expected any deal, but
because he was so upset by violence against women. But it emerged after
Dedge’s exoneration that the in for mant may have been no stranger to
such violence. He had been charged with raping his stepdaughter, be-
fore Dedge’s trial. The existence of that child rape case had never been
disclosed to Dedge’s lawyers, and it lay dormant for thirty years. The
in for mant was prosecuted only after Dedge was exonerated and the
stepdaughter came forward again. In 2004, the in for mant received life in
prison for the child rape.82

Similarly, in Nicholas Yarris’s case, the jail house in for mant claimed
to have heard a series of details that helped the State with their entire
case. The in for mant testifi ed that Yarris told him, “If I had the chance
again, I never would have killed the girl” but off ered no details concern-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 137

ing the murder itself. However, the in for mant instead reported details
that strongly corroborated the testimony of an eyewitness, the victim’s
coworker.83

That victim’s coworker had testifi ed that she worked at the Tri- State
Mall with the victim. She testifi ed she had seen Yarris seven or eight
times before the victim’s death, and that he had worn “dirty jeans.”84 The
day of the murder she did not see Yarris. She said that Yarris saw her a
few days after the murder and said, “I heard she was raped,” a detail
which the police said had not been made public. (Recall from Chapter 2
that Yarris also reportedly confessed falsely to the police and off ered that
same detail.) However, at the preliminary hearing the eyewitness was not
sure whether Yarris was the man she had seen.

The in for mant testifi ed that Yarris told him “that he was at the mall a
couple of times and there was a girl there that seen him at least twenty-
fi ve times and she wouldn’t be able to identify him, and he don’t see how
they could place him at the mall.”85 The district attorney argued that
Yarris’s admissions in jail must have been accurate, since those details
could not have been known by anyone but Yarris. He argued, “Where
does that come from? Is that in any of the newspaper reports?”86

Yarris had several alibi witnesses at trial, including a neighborhood
grocery store own er and his family. The in for mant also implied that Yar-
ris was asking alibi witnesses to commit perjury for him. And “he asked
me, he says, if one of my alibi witnesses are caught lying, he says, could
they be convicted of perjury.”87

The in for mant also addressed the forensics. The serology did not
conclusively link Yarris to the crime. Yarris was Type B, and any Type B
or Type AB could be the source, including approximately 13% of males.
The in for mant testifi ed that Yarris admitted that the Type B material
did in fact come from him.88

In several other cases the in for mants enhanced the forensic evidence.
For example, recall how in David Gray’s case the in for mant explained the
lack of fi ngerprints on the victim’s telephone by claiming Gray confessed
to wearing gloves. In Kenneth Wyniemko’s case, the in for mant explained
that he had admitted that he “got rid of everything” and had worn latex
gloves, which was used to explain how the forensic evidence had excluded
Wyniemko. Similarly, in Charles Fain’s case, the in for mant addressed the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

138 Trial by Liar

State’s lack of forensic evidence by testifying that “Mr. Fain had stated
that the detectives were going about his case all wrong, and that he had
cleaned his car out numerous times, and he cleaned it out so good that
they would not be able to detect anything in his automobile.” The prose-
cutor emphasized in closing arguments how the in for mant described
Fain’s car cleaning, and how “this was quite a joke with him, because he
had cleaned his car out so many times that the detectives could not get one
piece of evidence on him.”89

Stated Motives for Testifying

Some of these in for mants were repeat in for mants. The in for mant in Da-
vid Gray’s case, who was well known to the prosecutors, is one example.
In Wilton Dedge’s case, the in for mant had also testifi ed in several other
cases. On appeal, Dennis Halstead asked for a new trial, arguing that his
right to counsel was violated since the in for mant who testifi ed against
him was already working for the police; the court rejected his appeal, rul-
ing “the mere fact that the informer in this case had previously acted as
a police in for mant does not establish that he was presently acting as an
agent of the police.”90

In for mants rarely admitted that they were testifying because they
hoped for some gain. Instead, these in for mants often gave high- minded
motives for testifying that belied their likely motives. Some in for mants
claimed they were testifying as public- minded citizens. The in for mant in
Yarris’s case clearly could benefi t from testifying, as he spoke to prosecu-
tors before his sentencing for a crime that might particularly trouble pros-
ecutors: the burglary of a prosecutor’s home (his sentencing was then
postponed until after Yarris’s trial). Yet he said, “I did it because it could
have been my wife or children, okay?”91 In Charles Fain’s case, the in for-
mant stated he was off ered no promises of leniency, or benefi ts of any
kind, explaining, “The reason this was done, and the reason I done it, is
because I have two nine- year- old girls of my own, sir.”92 In Calvin Wash-
ington’s case, an in for mant claimed he came forward on his own, after
talking with Washington during Bible study about “repenting” and “and
being saved,”93 and he added he “came to the State because I have a wife,
I have a mother, and I have a sister.”94 In John Restivo’s case, the in for-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 139

mant testifi ed that he was not a “snitch” but was “doing what I feel is
right.”95

Some in for mants described personal motives for testifying against the
defendant, aside from leniency from the State. The in for mant in Dennis
Williams’s case noted that he believed that the defendants were involved
with an assault on the in for mant’s wife.96

In Dennis Fritz’s case, the in for mant emphasized the dangers that an
in for mant faced to show what personal risk he was taking because he
supposedly wanted to do the right thing. He said “you’ll be labeled as a
snitch, and they’ll give you a hard time. And most of the time, they try to
make punks out of you down there . . . They try to turn you into a ho-
mosexual and try to rape you or something just because you testifi ed for
the State.” He told the jury that he kept a pencil in his sock, fearing he
would be attacked in his sleep.97

These jail house in for mants, as unsavory as they may have appeared
to a jury, and as transparent as their motives to testify may have been,
nevertheless may have been quite eff ective witnesses, because they could
off er admissions that seemed to be so convincingly corroborated by the
crime scene evidence— and in some cases, that carefully bolstered all of
the weak points in the prosecution case. I turn next to codefendant testi-
mony, and then to other confi dential in for mants and cooperating witnesses
who also testifi ed that exonerees had made incriminating statements to
them.

Codefendants

Codefendants, charged with helping to commit the same crime, may be
off ered a deal to cooperate with prosecutors, testify against their accom-
plices, and receive a more favorable sentence. Of the fi fty- two exonerees
that had in for mant testimony, twenty- three had codefendant testimony.

In three trials, the codefendants knew quite a bit about the crime and
could easily relay information about how it was committed— because they
were the actual perpetrators, and were later implicated by postconviction
DNA testing. In each of those cases, the codefendant was in fact guilty
but successfully shifted some of the blame to an innocent man. In Bruce
Nelson’s case, the actual perpetrator, Terrence Moore, had confessed but

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

140 Trial by Liar

also told the police that Nelson was the one who initiated the rape and who
committed the murder.98 In Dana Holland’s case, the perpetrator was his
uncle, who was tried with him and found not guilty by the judge.99 In Ar-
thur Mumphrey’s case, a codefendant who confessed and testifi ed against
Mumphrey in exchange for a reduced sentence later had his guilt confi rmed
by postconviction DNA testing along with Mumphrey’s brother Charles
(who had confessed to police yet was inexplicably not prosecuted).100

In all of the remaining codefendant cases, the codefendants were them-
selves also innocent. As mentioned in Chapter 2, seventeen exonerees
confessed and testifi ed against codefendants, who were subsequently ex-
onerated by DNA testing.101 Several of those high- profi le cases were dis-
cussed in Chapter 2, for example the Central Park Jogger case, in which
fi ve youths variously implicated each other after lengthy interrogations
and were later exonerated by postconviction DNA testing.

In the Ford Heights Four case, Paula Gray had testifi ed that she com-
mitted a brutal double murder along with four others: Kenneth Adams,
Willie Rainge, Verneal Jimerson, and Dennis Williams. A court granted
Jimerson a new trial ten years later, when his lawyers uncovered that
Paula Gray testifi ed falsely when she said she received no leniency from
prosecutors. While she was a frightened, mentally retarded youth who
had also falsely confessed, the prosecutors had the obligation to set the
record straight at trial. Indeed, the State eventually conceded that it
should have disclosed the existence of a deal; “the jury did not know that
the People had agreed to drop the murder charges against Paula if she
testifi ed against defendant, Williams and Rainge.” In the case of Jimer-
son, Gray’s testimony “was the only evidence to link the defendant to
these crimes.”102 Verneal Jimerson is the only exoneree to have had a
conviction reversed based on a claim related to codefendant testimony.

After his new trial was granted, DNA testing exonerated Gray, Jimer-
son, and the other members of the Ford Heights Four. The DNA tests
also implicated the real killers, who had been located by three North-
western University college students working with journalism professor
David Protess. They had found reliable witnesses who did know who
the real murderers were. One witness had even told the police just days
after the crime that he had seen the real murderers commit the crime,
and he gave a detailed statement, but the police never followed up.103

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 141

Cooperating Witnesses

In fi fteen trials, there was testimony by cooperating witnesses who told
police they heard the exoneree admit guilt. These were not all con-
cerned Good Samaritans in the community. Although these witnesses
were not in jail, some hoped to obtain benefi ts from law enforcement not
unlike the jail house in for mants. Some were potential suspects and may
have sought to cooperate to avoid being prosecuted. Some sought re-
ward money. Most described the exonerees having confessed in detail to
the crime.

One of those witnesses was the actual perpetrator, whom DNA test-
ing later implicated as the murderer. At a preliminary hearing for both
Dennis Fritz and Ronald Williamson, who faced the death penalty for a
brutal murder in Oklahoma, the actual perpetrator, Glen Gore, testifi ed
and said that he had seen Williamson dancing with the victim at a bar
where she was last seen the night of her murder. He described how Wil-
liamson had supposedly “kept asking her to dance with him, and she
didn’t want to dance with him.” Gore described the victim looking over
her shoulder while dancing with Williamson, tapping Gore, saying “save
me,” and asking for a dance. Gore’s account of the eve ning ended with
him dancing with the victim— whom he had murdered.104

The other witnesses and in for mants had no way of knowing what hap-
pened at the crime scenes, but the level of specifi city in many of their state-
ments suggests that police may have contaminated the investigation. For
example, Kenneth Waters supposedly told two ex- girlfriends a series of
details, including that the victim was murdered by a stabbing, that she was
el der ly and German, and that money and jewelry were taken. One of those
ex- girlfriends had allegedly been told that she would be charged as an ac-
cessory and would have her children taken away from her if she did not
incriminate Waters.105 William Dillon’s ex- girlfriend also testifi ed against
him, but she recanted weeks after the trial, stating she had testifi ed falsely
because she was threatened with twenty- fi ve years in prison and also dis-
closing that she had sex with the lead police investigator the night he fi rst
interviewed her about the case.106

In Roy Criner’s case, three people told police they heard him admit to
using a screwdriver to threaten a hitchhiker into performing oral sex on

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

142 Trial by Liar

him. These accounts were inconsistent and none resembled the details
of the actual murder. Further, one of the three, Criner’s boss, testifi ed at
trial about this hitchhiker story. However, he told journalists years later
that both he and his wife had made clear to the police that Criner was at
work stacking logs when the crime occurred. It was “physically impos-
sible” for Criner to have committed the crime. He was never asked about
this at trial, and recalled that the police “only wrote down what we said
that was bad for Roy.”107

Thus, although it is often crucial that cooperating witnesses in the
community come forward to help police solve crimes, police can also
exercise undue pressure over them, and in some cases, police may have
disclosed crime scene facts to them.

Reforming the Use of Jail house In for mants

These exonerees’ trials suggest that we should not place undue faith in
the pro cess of a criminal trial to uncover in for mant lies.108 To prevent
wrongful convictions based on false in for mant testimony, the Supreme
Court has relied on “established safeguards of the Anglo- American le-
gal system [that] leave the veracity of a witness to be tested by cross-
examination, and the credibility of his testimony to be determined by a
properly instructed jury.”109 In very few of these cases were deals with
prosecutors or details of law enforcement interviews disclosed. In al-
most all of these cases, though, the in for mant described a detailed— but
false— account of a confession.

Several exonerees’ attorneys asked the judge to keep out the jail house
in for mant testimony at trial, but none succeeded. Perhaps the most arbi-
trary judicial ruling admitting jail house in for mant testimony was in
Drew Whitley’s case. The in for mant was on death row and facing execu-
tion. Yet the judge instructed the attorneys that they could not bring up
the fact that he was sentenced to death or that he was even in jail. The
judge ruled that they could bring up that he was on “court supervision”
and psychiatric care, and had to describe any conversation between them
as if it somehow occurred on the street and not in jail.110 Nor did the judge
let the defense lawyer bring out that at the in for mant’s own trial he had
huge memory blocks and could not remember even confessing to the po-
lice. These memory lapses were highly relevant, since the in for mant did

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Trial by Liar � 143

not come forward until fi ve or six months after the crime occurred. He
explained: “I forgot about it until I read in the paper about it” but later
wrote to the prosecutor because “it was on my conscience.” He admitted
having memory problems, though he said, “I remember what yesterday
was, a week before, two weeks.”111 The jury never heard any of this evi-
dence of the in for mant’s unreliability.

Judges could have instead required prosecutors to disclose the details
of any conversations with in for mants, to shed light on whether there was
in fact a deal, and to shed light on how it was that the in for mant came
to hear details about the crime. It was not surprising, though, that these
judges did not closely examine the testimony by these in for mants. Tradi-
tionally, judges do not require any such examination of in for mant testi-
mony, and even since these wrongful convictions have come to light, very
few jurisdictions have adopted any reforms. California requires that the
jury be instructed on the potential unreliability of jail house in for mant
testimony, while Oklahoma requires “complete disclosure” of any nego-
tiations between prosecutors and the in for mant.112 Illinois, in response to
death row exonerations in cases involving jail house in for mants, now re-
quires that the judge conduct a “hearing to determine whether the testi-
mony of the in for mant is reliable.”113 Only a handful of states have ad-
opted these reforms. Such eff orts are modest improvements. No state
requires videotaping conversations to ensure that case information is not
suggested to the in for mant. Few jurisdictions limit the use of in for mants
to cases where they are necessary and reliable.

Tailor- made statements by jail house in for mants provided powerful
but false evidence of guilt. If in for mant testimony is to be permitted,
then it should be subject to far more searching review. Statements could
be videotaped. Prosecutors should be forbidden from entering posttrial
deals or securing informal benefi ts not disclosed to the defense. Discov-
ery could be enhanced. The prosecution should be required to disclose
prior cooperation by a jail house in for mant. All information concerning
any deals should be formalized and provided to the defense. These trials
illustrate how wrongful convictions can result if in for mant testimony is
freely admitted without meaningful scrutiny or precautions.

William Dillon was exonerated after spending twenty- seven years in
prison. After he was fi nally released, a hearing was held to decide whether
the State of Florida would compensate him, and if so, for how much.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

144 Trial by Liar

There he again met the jail house in for mant who had testifi ed against him
at trial. The in for mant had claimed Dillon reenacted the murder in the
jail house bullpen, showing how he pinned the victim down and punched
him in the face on a deserted beach. Twenty- seven years later, the in for-
mant tearfully apologized. He described how police offi cers sat on either
side of him and coached him. One asked questions, while the other stated
the desired answers. The offi cers wrote a statement with just their words
that the in for mant parroted back. The entire statement was fabricated. For
his contributions to this frame- up, the in for mant avoided prison for a rape
charge. Twenty- seven years later, he fi nally apologized to Dillon, saying,
“I’m very sorry.” They shook hands. He explained that the police “put
everybody on the spot.” Dillon told him, “I understand, believe me.”114

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

145

The small town of Culpeper, Virginia, was shaken by the grue-
some rape and murder of a nineteen- year- old woman in June 1982.

Two of her infant children were present when she was raped and stabbed
thirty- eight times. The bedroom walls and fl oor of her apartment were
smeared and splattered with blood. She collapsed outside screaming
“help me” and “he hurt me.” Police and her husband arrived before the
ambulance. She was able to tell them only that a black man attacked her.
She quickly died. Eyewitnesses saw a black man climbing a fence, and
newspapers ran a composite sketch. For more than a year, police had no
leads. The case was a “top priority” with local and state police. But they
were “fairly frustrated” and called the case “a big puzzle.”1

Earl Washington Jr. was a black, borderline mentally retarded farm-
hand. In May 1983, during a night of drinking and arguing with rela-
tives, he broke into an el der ly neighbor’s house. He hit her with a chair,
ran off with her gun, and shot his brother in the foot. Found crying in
the woods, he was arrested. Police suspected an attempted rape initially
(the neighbor later denied any such thing), and began to question him
about several unsolved rapes. As I described in Chapter 2, Washington
confessed to all of the crimes he was asked about, but in the other cases
the victims told police he was not the assailant, or the confessions were

c h a p t e r 6

Innocence on Trial

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

146 Innocence on Trial

deemed totally inconsistent with how the crimes occurred. Police also
asked him about the crime in Culpeper. During a two- hour interroga-
tion, later followed by an interrogation by state police familiar with the
case, he also confessed to that brutal murder. He signed a typed confes-
sion statement that consisted primarily of him saying “Yes, sir” to what-
ever the police asked.

Now the case moved rapidly. The State announced that it would seek
the death penalty. Death penalty trials, because of their seriousness, are
on average far longer than the typical felony trial. The transcript of this
trial, however, was quite short. Earl Washington Jr. was convicted in
Virginia after a trial lasting less than fi ve hours, and he was then promptly
sentenced to death. He was tried as “Earl Ju nior Washington”— he could
not correctly write his own name. Apparently the authorities did not
write his name correctly either. At a brief pretrial hearing, a state psy-
chologist testifi ed and found him competent to stand trial. After a day of
jury selection, the trial began the morning of January 19, 1984.

The opening statements were very brief. The Virginia Common-
wealth’s prosecutor described the gruesome rape and murder of the vic-
tim. He emphasized that Washington had not only confessed but had told
police “a number of diff erent things that could only have been known by
somebody who actually had committed the off ense.”2 The prosecution
then called fi fteen witnesses. They described the confession in detail, the
injuries suff ered by the victim, and the investigation of the case.

The defense called just two witnesses: Earl Washington Jr. and his
sister. Virginia has statutory limits on fees paid to lawyers for indigent
defendants that are still among the lowest in the country.3 Not satisfi ed
with his public defender, his family hired a private lawyer. However, that
lawyer had never tried a death penalty case, and it showed. The entire
defense case occupied only forty minutes.

Earl Washington Jr. was innocent and on trial for his life. How did he
try to convey his innocence to the jury? On the witness stand, Washing-
ton protested his innocence. He said he did not understand many of the
things written in his confession statement, which, as I described in
Chapter 2, included striking crime scene details about a torn shirt that
had been found in the dresser drawer of the victim’s bedroom. He denied
having said some of the things that police testifi ed he had said during his

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 147

interrogations. He had no strong alibi. He could not remember any par-
tic u lar day in June 1982. After all, he was borderline mentally retarded.
Nor would most people remember a par tic u lar day more than a year and
a half earlier. Washington recalled that he was out of work that month
and said that he was probably “at home cleaning up in the yard” the day
of the crime.4

In the closing statements, the defense had one last chance to make a
compelling case that Washington should not be convicted. Washing-
ton’s lawyer off ered few arguments, and did not discuss many facts ei-
ther. He did not mention any of the glaring inconsistencies in the confes-
sion. For example, Washington did not correctly describe the race of the
victim. When police asked him, “Was this woman white or black,” he
had answered, “Black.” She was white. Washington’s lawyer did little
more than remind the jury that “The primary thing for you to decide is
whether or not this gentleman, this man, Earl Washington did it.” He
never claimed that Washington was an innocent man.5

The next morning, on January 20, 1984, the jury deliberated for only
fi fty minutes before concluding that Washington was guilty. Next, there
was a sentencing phase, a separate portion of the trial where both sides
could present evidence concerning whether Washington deserved a death
sentence. The prosecutor vividly described “the suff ering and the agony”
of the victim, and how “two of her children were right there” during the
gruesome murder, while the third child came home from school and saw
her mother lying in a “pool of blood.” The defense simply presented
Washington’s sister and the state psychologist who had found Washington
competent. In closing, he told the jury that “[h]is life is in your hands.”
The jury quickly sentenced Washington to death.6

The prosecution case rested on the apparent strength of Earl Wash-
ington Jr.’s confession. The obvious defense was that this confession, by
a mentally retarded individual, was unreliable, false, and the product of
police suggestion. His lawyer never made those arguments. The lawyer
did not request funding for an expert to review whether mental retarda-
tion aff ected the reliability or voluntariness of the confession.7

Nor did the defense lawyer understand the forensic evidence. The
semen- stained blanket from the victim’s bed was blood- typed, and that
rudimentary technique had ruled out Washington. The lawyer later said

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

148 Innocence on Trial

that while he saw the forensic reports, he “was not familiar with the sig-
nifi cance” of the analysis. The prosecution may have also engaged in
misconduct. As I noted in Chapter 4, it emerged years later, after Wash-
ington fi led a civil rights lawsuit and obtained access to the prosecution
fi les, that the State had concealed crucial evidence of innocence, includ-
ing forensic evidence, from the defense.8 Although Washington was in-
nocent, the jury heard very little evidence of innocence. Instead, they
heard about a detailed confession. Washington would come within nine
days of execution. I will describe in Chapter 8, which shows how the 250
exonerees obtained their exonerations, a series of miraculous and chance
interventions that later saved Earl Washington Jr.’s life.

The Defense Case

We have heard quite a bit about the prosecution evidence at these exon-
erees’ trials, but not about the general characteristics of the trials, such
as their length, nor the cases that these innocent people’s lawyers mounted
in their defense. When I fi rst began looking at the trial transcripts, I ex-
pected to fi nd that actually innocent defendants would try to assert their
innocence at trial. I was surprised to fi nd that although innocence played
a role in most of these trials, the defenses raised were usually fairly weak.
Most of these cases were not high- profi le trials. Most were short trials for
rape. Since almost all were cases where the rapist was a stranger, a central
issue was identity— who was the stranger who committed the rape? Truth
was on trial. These jurors fi rst heard the prosecution case. Defense law-
yers tried with varying success and skill to undermine the prosecution
evidence that their client was the culprit.

However, what followed the prosecution case was typically a very
short defense case that did little to present the jury with an alternative
account. Three reasons explain why the defense case was usually so
weak. First, the defendants had little ability to locate evidence of their
innocence. The primary investigators of crimes are the police; they work
for the prosecution, and their work may be well under way or completed
by the time that these people even became suspects. Second, it is inher-
ently diffi cult to convincingly prove that one did not commit a crime. It
is not easy to have an alibi. One may not remember what one was doing

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 149

on an unremarkable eve ning many months ago. Indeed, it might be sus-
picious if one did have a vivid recollection of that night. Third, the de-
fense lawyers often did not eff ectively challenge the prosecution case. The
defense called, on average, less than half as many witnesses as the prosecu-
tion. Defense lawyers retained few experts, in comparison to the prosecu-
tion. Most of these innocent people could not aff ord an attorney, much less
expert witnesses, and most had lawyers assigned to them. Lawyers may not
have known that these cases were not ordinary, that their client’s protests
of innocence were actually true. And while we do not know how often
this occurred, in some exonerees’ cases, the police or prosecutors con-
cealed evidence of innocence from the defense lawyers.

In this chapter, I explore each of those challenges that these innocent
people faced in mounting a meaningful defense at trial, even though they
could truthfully proclaim their innocence.

Before Trial

The last chapters have all discussed the trial evidence used by the prose-
cution, such as confessions, eyewitness evidence, forensics, and in for-
mant testimony. The State has a head start in assembling its case, since
long before any criminal trial takes place, the police and prosecutors
conduct an investigation. In contrast, very few of these exonerees had
lawyers at the time that crime was being investigated. Most did not obtain
a lawyer until after they were arrested. A few were found near the crime
scene and identifi ed on the spot by a victim, but most became suspects
after some time had passed. While we do not know precisely how all of
these people fi rst came to the attention of the police, we do know that
some were presented to a victim because, as prior off enders, their photos
were on fi le with the police. Others came to the attention of police be-
cause their behavior was suspicious, in for mants provided their names, or
they came forward to talk to the police about the case and then reportedly
falsely confessed.

By the time defense lawyers were involved, witnesses may have been
harder to locate and the police may have largely closed their investigation.
Indeed, these trials took place an average of about one year and three
months after the crime. That delay may have aff ected the reliability of

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

150 Innocence on Trial

the evidence, particularly the memory of eyewitnesses, which can be
expected to fade over time.

The press also played a role leading up to these trials. Eyewitnesses
sometimes saw news reports about a crime or composite drawings that
other witnesses had prepared, and seeing these images may have con-
taminated their memory and increased their certainty. Likewise, studies
have shown that jurors are far more likely to convict in cases in which
there is pretrial publicity.9 Sensational coverage may highlight gruesome
aspects of the crime, which may bias jurors. Extensive coverage fueling
public outrage may also increase pressure on police to close cases, and
on prosecutors to win them or face the anger of voters when they are up
for reelection. Moving the trial to a diff erent court house when there is
extensive pretrial publicity might help, but judges denied every motion
that these innocent people made to transfer the venue.

Guilty Pleas

Some exonerees had no trial at all; 6% of the exonerees (16 of 250) pleaded
guilty and agreed to have no trial. (Nineteen pleaded guilty, but of those,
three were also convicted at a trial on separate charges).10 While we do
not know all of the reasons those exonerees pleaded guilty, what we do
know suggests that these exonerees had the same problem as those who
did have a trial: they lacked compelling evidence that they were innocent,
while the prosecution had already investigated the case and had seem-
ingly powerful evidence of their guilt.

Ten of the sixteen who pleaded guilty had already confessed. Several
initially litigated their cases, but after the judge ruled that the confession
would be admissible, they decided to plead guilty. Others pleaded guilty
because they had falsely confessed and agreed to testify against other in-
nocent individuals in exchange for a more lenient sentence; their cases
were described in Chapter 5.

As in the cases of the exonerees who did have trials, the defense law-
yers may have played a role. Christopher Ochoa, for example, recalled
his lawyers pressuring him to plead guilty, telling him that his detailed
confession to a rape and murder was so compelling he might receive the
death penalty. They told him there was “no way an innocent person

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 151

would give such a detailed statement.” He later said that although his
lawyers probably “believed [he] was guilty,” he also had the impression
that “it was less work” for them if he would plead guilty. He was off ered
a life sentence for his testimony against his also- innocent codefendant,
Richard Danziger.11

John Dixon had asserted his innocence to police and later regretted
pleading guilty. He asked to retract the plea and asked for DNA tests, ar-
guing they could prove his innocence. The judge denied the requests,
noting, “It is not uncommon once having plead guilty there’s a change of
heart.”12 Anthony Gray also tried to withdraw his guilty plea. He was
mentally disabled, had falsely confessed, and then had accepted a wildly
unfavorable plea agreement. He pleaded guilty to the most serious crimes
charged and agreed to receive the maximum sentence, life in prison,
which could have been imposed at a trial. His lawyer discussed this
agreement with him for less than an hour. The American Bar Association
has documented the problem of poor defendants receiving what have
been nicknamed “meet ’em and plea ’em” defense lawyers.13 The judge
denied Gray’s request to withdraw his guilty plea, and DNA testing exon-
erated him seven years later.14

The small percentage of guilty pleas makes these DNA exonerees very
diff erent from typical criminal defendants, who overwhelmingly plead
guilty. More than 95% of all criminal cases result in a guilty plea. The
numbers are lower for felony rape (83%) and murder (69%) but are still
quite high.15 In this unusual set of cases where we now know that the de-
fendant was actually innocent, the number of guilty pleas may be so low
because of the defendants’ knowledge of their own innocence. In many of
these cases, because of the seriousness of the charge, the prosecutors may
not have always off ered bargains that were palatable to an innocent defen-
dant. Several exonerees later explained that they could not bring them-
selves to plead guilty for a crime that they did not commit. It is one thing
to agree to falsely admit to having committed a crime if the crime is rela-
tively trivial and the sentence is a year in prison, but it is another thing
entirely if the result is life in prison and being known forever as a rapist
or murderer. For example, James Lee Woodard, who served more than
twenty- seven years in prison, had been off ered a plea bargain for just
three years, although he was charged with murdering his girlfriend. He

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

152 Innocence on Trial

told his lawyer at the time, “We go to trial,” and explained that he turned
down the deal because “I know one day that I was gonna have to die, and
I didn’t want to go before God saying I did something that I didn’t do.”16

Perhaps these innocent people would have spent far less time in prison
if they had accepted plea bargains. Indeed, the reason we know about
these par tic u lar innocent people is in part because they refused to accept
plea bargains, and so they received far harsher sentences after a trial. If
James Lee Woodard had served three years in prison, he would have
been released in 1984 and might not have bothered to pursue DNA test-
ing a de cade later when it became possible to do so.

Just as the cases of exonerees show that innocent people can falsely
confess, they also show us that innocent people can plead guilty. Al-
though people who plead guilty swear in court that they understand that
they are admitting their guilt and the consequences of doing so, DNA
testing has shown that innocent people can plead guilty for some of the
same reasons that innocent people can falsely confess. They succumb to
pressure from prosecutors or even from their own defense lawyers. They
believe that pleading guilty is a better option than the severe sentence
they might receive at a trial. They may have previously confessed, they
may be vulnerable or mentally disabled, or they may feel as if they have
little choice but to plead guilty.

However, the cases of these exonerees cannot tell us how many inno-
cent people plead guilty. Just because few of these exonerees pleaded
guilty does not mean that wrongful convictions are less of a problem for
people who plead guilty than for those who have a trial. It may be that
many others were also innocent, but they never asked for DNA testing,
because they understood the system better, they were less principled
about asserting their innocence despite the risks at a trial, or they had
more talented lawyers, and as a result, they obtained favorable plea bar-
gains and served short sentences. We know very little about such cases,
and judges do not look closely at them. After all, when one pleads guilty,
one typically waives the right to fi le an appeal or a habeas petition. Many
states also refuse to allow people who plead guilty to pursue postconvic-
tion DNA testing.17

Our criminal justice system is a plea- bargaining system. Prosecutors
resolve the vast majority of cases using plea bargaining— but these exon-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 153

erations vividly illustrate how not all who plead guilty are in fact guilty.
The plea- bargaining system deserves far more careful scrutiny.

Innocence on Trial

We have good reasons to think that “in criminal jury trials, the evidence
matters.”18 Jurors place great value on the types of evidence common in
these trials: eyewitness testimony, police testimony, confession testimony,
victim testimony, and forensic testimony.19 As described in the previous
chapters, although that evidence now looks quite fl awed, at the time it
may have appeared powerful. The confessions and in for mant testimony
appeared to include details about the crime scene that only the true cul-
prit could have known. The eyewitnesses appeared confi dent about their
identifi cations, which impresses jurors, as I discussed in Chapter 3. The
forensic analysts may have exaggerated the evidence, but as I described in
Chapter 4, the jury rarely heard anything about such fl aws.

Moreover, those pieces of prosecution evidence can all reinforce each
other to make the case appear even stronger. Police sometimes secured
confessions by telling the suspect that forensic evidence tied them to the
crime. An offi cer did that in the Central Park Jogger case, where the fo-
rensic evidence later turned out to be quite weak, and largely excluded
the defendants; however, it was used to produce a seemingly strong con-
fession. Even if the medical examiner could not be sure what the murder
weapon was, if the injuries were consistent with the use of a knife, the
exoneree then confessed to using a knife (as in Ronald Jones’s case); or if
an in for mant claimed the exoneree admitted to using a knife, then the evi-
dence would suddenly fi t together in a seemingly convincing way. The
evidence can cross- contaminate in ways that cannot easily be detected. A
victim may have been unsure about her lineup identifi cation, though
there may be no record of it, but if police told her that forensics all matched
the suspect, then her certainty would probably increase.

Similarly, when jurors consider all of the evidence, not only does the
strength of the evidence matter, but its coherence may also matter. The
prosecutors present the most coherent story they can by describing why
they believe the defendant committed the crime. They could do that par-
ticularly compellingly in the rape cases where the victim could herself

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

154 Innocence on Trial

describe what happened and then, at the end of that account, identify the
defendant as the perpetrator. However, they could also present the testi-
mony of a host of others who could all describe what had happened and
why the defendant had been connected to the crime: the police who in-
vestigated the crime scene, forensic analysts, family members of the vic-
tim, eyewitnesses, in for mants, doctors who treated the victim, and medi-
cal examiners who investigated the cause of a death or the victim’s
injuries.

A jury’s confi dence in that account of guilt may be heightened where
the defense fails to or is unable to off er a credible competing story about
innocence.20 The problem encountered by these innocent people and
their lawyers was that they typically had no way to convey their inno-
cence in a believable way. Asserting innocence was all that many exoner-
ees could do. Many simply had no compelling proof of their innocence.
That does not mean they didn’t try. More than 90% of the exonerees for
whom trial transcripts were obtained (190 of 207 trials) asserted their in-
nocence at trial. The few who did not directly assert innocence simply
argued that the State had not proved that they were guilty beyond a rea-
sonable doubt.

The trials began with the opening statements by the lawyers. Most de-
fense lawyers proclaimed their client’s innocence in their opening state-
ments. For example, Richard Alexander’s attorney opened with inno-
cence, telling the jury “this is a case of misidentifi cation.” Miguel Roman’s
lawyer closed, saying, “There’s no greater injustice than a man found
guilty of something that he didn’t do. Nothing is more horrifying, in our
system of justice.”21 Others, perhaps fi guring that the evidence was weak,
did not emphasize innocence. Recall how Earl Washington Jr.’s lawyer did
not assert innocence and simply asked for the mercy of the jury. Most ex-
onerees were also tried alone; the vast majority of criminal trials involve
single defendants.22 However, as described in Chapter 2, some exonerees
were convicted in joint trials, in some of which not all of the defendants
claimed their innocence, but rather some pleaded guilty and implicated
other innocent people in their false confessions.

Following the typically brief opening statements, the prosecution pre-
sented its case, and then, after it rested, the defense presented its case.
The defense often had a very limited case, which kept these trials short.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 155

Earl Washington Jr.’s trial of less than two days, in which the defense
case was forty minutes long, was in many respects quite representative.
Although it was a death penalty case, and only 17 of the 250 exonerees
were sentenced to death, it was of a length similar to many of the exoner-
ees trials, and it involved a similarly tepid defense case. One study has
found that typical rape trials average a little more than a day long, while
murder trials average thirty- three hours, with capital murder trials lon-
ger than that.23 The exonerees’ cases fi t this pattern: almost all of these
criminal trials were less than a week long, and the average length was fi ve
days. Only the murder trials and particularly the capital trials tended to
be longer. The defense opening and closing statements were typically
shorter that those of the prosecution, and they called fewer witnesses.

In some exonerees’ trials, the defense literally made no affi rmative
case; the defense lawyers simply rested their cases as soon as the prosecu-
tion fi nished, putting on no witnesses of their own. Carlos Lavernia and
Thomas McGowan’s cases are examples.24 More typically, the defendant
testifi ed on his own behalf and off ered an alibi, perhaps with a few wit-
nesses in support. On average, defendants presented six witnesses, with
the prosecution on average presenting fourteen witnesses, more than
twice the number presented by the defense. Some defense lawyers did
not do a vigorous job of investigating and representing their clients, al-
though, as I will discuss, it is hard to know from available rec ords what
they did or did not do and what they could have done diff erently. Some
had little time to prepare. For example, Kenneth Wyniemko’s lawyer was
appointed so late that he had only a weekend to prepare before the fi ve- day
trial.25

Innocence can be claimed in two main ways at a trial: through an alibi
defense or evidence of third- party guilt. Alibi defenses, raised by 68% of
the exonerees who had a trial that could be obtained (140 of 207 trials),
involve calling witnesses to testify that the defendant was not at the
crime scene. Claims of third- party guilt were raised by 14% of the exon-
erees who had a trial (30 of 207 trials) and involve evidence pointing to
another individual as the culprit. Those two types of defenses can be
supported by physical evidence, by testimony of witnesses, or by testi-
mony of the defendants themselves, if they choose to take the witness
stand.26 These exonerees took the stand to proclaim their innocence in

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

156 Innocence on Trial

53% of these trials (110 of 207 trials). Exonerees also had forensic evidence
that provided evidence of their innocence in 14% of these trials (30 of
207 trials). I will discuss each type of defense in the sections that follow.

Weak Alibis

Did these innocent people have compelling evidence supporting their
alibis? About two- thirds of the exonerees, or 68%, off ered an alibi de-
fense (140 of 207 trials). Alibis are often distrusted by jurors, because
criminals are all too able to make up some kind of alibi, even a seemingly
plausible one. Nor is it very easy for people to remember what they did at
a specifi c time many weeks, months, or more in the past. That is particu-
larly true for an innocent person who did nothing improper on that day
and would have no special reason to recall it. As Brandon Moon’s lawyer
argued at trial, “A guilty man knows what happened. He can explain it
away. An innocent man wasn’t there. He doesn’t know what happened.
He can’t help explain it away, because he has no earthly concept of what
occurred.”27

Most alibis are also weak. Studies suggest that very few alibis are
corroborated by physical evidence, like credit card receipts or tickets.
Most are instead corroborated by witnesses, but those witnesses are
usually family members, who would have a reason to lie to help their
loved ones, and therefore tend to be distrusted by jurors.28 We do not
usually know whether police made a real eff ort to investigate the alibis
when these innocent people fi rst became suspects. Maybe stronger evi-
dence of innocence existed at the time and it was never investigated or
was lost.

At the time of trial, most exonerees had weak alibis. Almost none had
physical evidence to support their alibis. Of the 140 exonerees who as-
serted an alibi, 121 had alibi witnesses. The others had only their own
testimony to describe what they were doing at the time of the crime.
Many exonerees did not work, so they had no rec ords of being at work
that day or coworkers who could say they were at work. Most could only
off er family members who testifi ed that the defendant was at home with
them when the crime occurred. Many of the crimes were rapes that oc-
curred at night, when there would have been few witnesses about, and

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 157

when even their family members might have been asleep and unaware of
whether they were home.

Such alibis were vulnerable to prosecution attacks and could even be
turned into a liability. The State could argue that an alibi was concocted
and that this was evidence of the defendant’s deception. For some of the
crimes, particularly the murders where the time of death could not be
established precisely, it could not be determined exactly when the crime
occurred. Without knowing the time that the crime took place, one would
have to establish an alibi over a long and vaguely defi ned time period.

A few exonerees did have remarkably strong alibis and yet were still
convicted. Timothy Durham had perhaps the strongest. Durham was far
from Tulsa, Oklahoma, at the time the rape he was convicted of occurred.
He was at the Pan American skeet shooting competition in Dallas, Texas,
the day of the crime. He had eleven alibi witnesses testify on his behalf. It
was a large shooting match, “one of the very top shoots in the country,” as
one witness explained, and the various witnesses described how they ran
into him throughout the two- day competition along with his parents.29
Then the defense introduced credit card receipts for the gas they bought
during the drive, the dinner the Durhams ate together, and the clothes
Durham bought while in Dallas.30 The prosecution argued that these
witnesses either colluded or were mistaken. And the prosecution also had
something that appeared to have been much more powerful— not just the
victim’s identifi cation of him as the culprit, but DNA test results pointing
to Durham. In fact, as described in Chapter 4, that DNA result was false
and caused by a lab error.

Some exonerees only had their own testimony to support their alibis.
Anthony Michael Green, for example, testifi ed that the night of the
crime he was running races with friends on his street and was later with
his girlfriend— but the defense called no witnesses to support those as-
sertions.31 Rickey Johnson said he did not leave home the night of the
crime because of pain in an abscessed tooth and that the next day he had
gone to the dentist. Yet the dentist’s wife, who was also his offi ce man-
ager, said that rec ords did not show that he had been to the dentist in the
days before or after the crime.32 James Richardson initially said he was
driving around for two hours in the early morning hours because he wanted
to get a special type of Dr Pepper bottle that he preferred. Yet he was the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

158 Innocence on Trial

fi rst person at the crime scene next door to his house (where he rescued the
young daughter of the victim from the fi re set by the culprit).33

A few had no alibi at all, because they conceded they were present at
the crime scene. Hector Gonzalez was at the crime scene where a man
was murdered; he said he was in a large crowd outside a nightclub that
observed the murder but had no role in it. At sentencing, Antonio Beaver
told the judge that he really wished that his alibi witness had shown up
at trial, saying “it would’ve helped . . . he’s the guy that, you know . . .
used to take me back and forth to work.”34 Arvin McGee complained
that his attorney did not call any of his alibi witnesses, telling the judge
that as a result, “I was a sitting duck.”35

Pointing to Another’s Guilt

Making a case that one is innocent is also not easy if one cannot point to
another person who did commit the crime. Fewer exonerees, only 14%
(30 of 207 cases), argued that some other person had actually committed
the crime. Several other exonerees attempted to introduce such evidence,
but judges barred them from doing so. In most states, a judge may exclude
evidence of another person’s guilt if it is not adequately supported. The
U.S. Supreme Court in Holmes v. South Carolina established a due pro-
cess right to present such a defense, but the Court ruled that judges may
still exclude such evidence if they decide it is too weak to be admissible.36

Claims of third- party guilt have not received much judicial or schol-
arly attention, maybe because few succeed in presenting them.37 After
all, the defense does not usually have substantial investigative resources
and may have no knowledge of other leads. Most surprising, a few exon-
erees had, at the time of trial, identifi ed the real perpetrator, the person
who DNA testing later showed was the culprit. Such cases suggest that
police investigating the crime may have ignored sound evidence that
could have lead them to the real culprit. Frederick Daye’s lawyer called
to the stand the mother of David Pringle, the other man charged with the
crime. She testifi ed to her son’s acquaintance with a man who matched
the eyewitness’s descriptions of the second attacker.38 Pringle’s cousin
testifi ed that shortly after the attack, Pringle’s acquaintance acted suspi-
ciously; he off ered to sell her two rings that were similar to the rings

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 159

stolen from the victim.39 None of that testimony provided concrete evi-
dence of Pringle’s guilt. However, postconviction DNA testing con-
fi rmed that Pringle was in fact the culprit. In Byron Halsey’s case his
lawyer insinuated that a neighbor was the culprit, which postconviction
DNA testing later proved to be the case.40

Robert Clark was charged with raping a woman and stealing her car,
a maroon Oldsmobile Cutlass. He was caught driving the victim’s car,
but eventually told police he got the car from his friend Tony Arnold.
Clark’s lawyer argued that Arnold, who matched the victim’s initial de-
scription of the attacker far better, was the actual rapist. Indeed, Arnold
was brought into the courtroom, where a witness identifi ed him as the
man she had seen driving a maroon car just like the victim’s, which she
thought was unusual, since he was unemployed and had no car. Postcon-
viction DNA testing proved that Arnold was in fact the culprit, but in the
meantime, Clark had spent twenty- four years in prison, and Arnold had
continued to commit a string of rapes in Georgia.41

Ronald Williamson’s lawyer argued that prosecution witness Glen
Gore “was the last known person to see” the victim alive, noting,
“Wouldn’t you think he’d be a prime suspect?”42 Recall from Chapter 5
the macabre account that Gore had given of how the victim reached out
and asked him to dance with her, to “save” her from Williamson’s requests
for dances. Postconviction DNA testing confi rmed Gore’s guilt. Other
perpetrators were all too willing to shift the blame to an innocent person.
In Bruce Nelson’s case, the actual perpetrator, Terrence Moore, had con-
fessed but implicated Nelson as lead participant in a rape and murder. In
Arthur Mumphrey’s case, DNA tests confi rmed the guilt of a codefendant
who had confessed and testifi ed against him in exchange for a reduced
sentence.

In these troubling cases, police had the real culprit in their sights, but
focused on the wrong man. The police fi les from the investigation are
typically never disclosed in full and we do not know what leads police
may have neglected to follow. In some cases, the defendant, though now
known to be innocent, may have been the lead suspect for good reason,
say because the eyewitness identifi ed him. We do not know in how many
other cases police were so focused on innocent people that they ignored
leads that could have pointed them to the true culprit. Clearly, police

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

160 Innocence on Trial

should be open to investigating alternative suspects and scenarios,
rather than rushing to a judgment that might cause a wrongful convic-
tion while the criminal goes free.

Testifying to Their Innocence

Not only did these innocent people try to develop alibis and evidence of
third- party guilt, but they could also rely on another source of evidence of
innocence: themselves. At more than half of these exonerees’ trials or 53%
(110 of 207 trials), the exonerees took the stand at trial to claim their inno-
cence, and perhaps also develop an alibi defense. They faced a delicate
task. They desired to convey their innocence, but also compassion for the
victim. Many were not eloquent or accustomed to public speaking. They
may have hoped to earn some sympathy from the jury, but the jury had just
heard about the serious crimes they were accused of committing.

Most asserted their innocence briefl y and without fl ourish or exces-
sive emotion. Dennis Brown denied committing the rape at trial, saying,
“I wouldn’t do nothing like that,” and “No, sir. I was raised better than
that.” He commented that as for the victim, “I feel sorry for her, what
somebody else did, because I didn’t do it.”43 Similarly, Charles Chatman
was asked, “Did you rape that woman that night in her house like she
says?” and he answered, “No, sir, I didn’t and that’s the honest truth.”44
Thomas Doswell testifi ed: “With God as my witness, I did not rape this
woman.”45 Clarence Elkins testifi ed succinctly, “Yes, I am innocent.”46
Dennis Fritz gave a longer statement protesting his innocence. He was
asked, “Dennis, did you kill [the victim]?” He answered:

No, I did not. I did not kill [the victim], and I don’t know anything

about the death of [the victim] whatsoever. I’ve been locked up over

in that county jail for 11 months on circumstantial evidence, and I’m

the kind of person that I’ve never taken a life. I’ve never thought

about taking a life. I’ve never wanted to rape a woman. I’ve never

had any thoughts of this kind of activity in my life.47

Exonerees not only protested their innocence, but many of those who
had falsely confessed also testifi ed and recanted their confessions. Sev-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 161

enteen of the thirty exonerees who falsely confessed and were convicted
at a trial did so. For example, Bruce Godschalk explained that the offi –
cers “started questioning me down at the station and I was quite upset,
and I was denying it most of the time, and I thought they would let me
go,” once he confessed.48 Ronald Jones, whose story we heard in Chap-
ter 1, testifi ed that he confessed to the detective “[b]ecause it seemed like
the only way he was going to stop beating me.”

At sentencing hearings, which were transcribed in only some of these
trials, at least fi fty- seven exonerees reaffi rmed their innocence, often in
particularly moving terms, begging for the mercy of the court. Ronald
Jones said, “I just wanted to say that I am sorry for the grief and the pain
that [the victim’s] family had to go through, and I am sorry for the grief
and pain my family had to go through, and that’s it.”49

Most kept their testimony brief. Anthony Capozzi stated, “Judge I
never did these crimes and I’m innocent.”50 Larry Fuller testifi ed at
greater length, stating: “I can only consider myself as being a victim of
circumstances, something that I must accept, but I cannot change . . .
I would like to receive, because I come here believing in the word of
justice, justice with eyes, not justice that would be blind.”51 Kirk
Bloodsworth, the fi rst person exonerated from death row, gave a formal
statement:

I, Kirk Noble Bloodsworth, did not commit this crime . . . I have

never in my history have ever had a violent act, especially on a child . . .

I think this has been a travesty of misjustice all the way around, not to

put the Honor on the spot or anybody else in this Court. They have to

do their job. Somebody told them what to do, and then it comes down

the line. All’s I’m saying is Your Honor, I did not commit this crime

on July the 25th 1984. If I had have, it would have been stated from the

start.52

As an innocent person facing the death penalty, he not only sought
mercy but had to constrain his own anger over his predicament. Others
also managed to address the judge with a graciousness that in retrospect
seems remarkable. Douglas Echols explained, “I didn’t rape that lady.
I didn’t kidnap that lady. I didn’t rob that lady. I didn’t do none of the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

162 Innocence on Trial

accused, you know, but I’m not angry for what ever you did, you know.
God bless you.”53

Several also appealed to God to save them. Calvin Ollins told the
judge: “I want to say that I was found guilty for something that I didn’t
do. And I always tell myself that I will be out one day and that God will
do something for me.”54

In contrast, after the verdict was read, James O’Donnell had another
understandable reaction— he exploded with anger and directed exple-
tives at the judge and jury. He then composed himself and said: “I am re-
ally sorry for my outburst. I tried to be as civil as possible. I would never
do a crime like this. And my life is over now as I know it, my wife and
kids’ life. I don’t understand how the jury did this to me. It’s really not
right, what they did. I was home in bed. I was sleeping. I would never hit
a woman. I have a wife. I never hit my kids, ever. I never forced a woman
to do anything in my whole life. That’s the God’s honest truth . . . It’s
just— I’m very sorry for my outburst. Don’t take my life away, please.”55

In the Central Park Jogger case, Korey Wise was removed from the
courtroom after hearing the prosecutor’s opening statement, when he
cried: “No. No. No. Can’t take this. O, Lord. Jesus. No . . . It’s wrong.
It’s wrong. No. No.”56

Stephan Cowans told the police offi cer who identifi ed him as a man
who shot him, that “with all due respect, I’m very disappointed in what
happened to you and what ever, but you know as well as I know I’m not
the man who did this to you.” Cowans added, “I also have a family. I have
people who love me, also. For you to do this to me— as a Boston police
offi cer, you know, I just wish you the best . . . But I think this goes
around, and goes back to other innocent people, with all due respect.”57

At the trial of Michael Evans and Paul Terry, the judge noted during
sentencing that “both defendants walked into this courtroom with a
swagger” as evidence that neither exhibited any signs of remorse. Their
attorney tried to respond, “I think that is clearly indicative of innocence,
that you don’t feel remorse for something you didn’t do.”58

Taking the stand is risky for those who have prior convictions. Most
of those who did not testify likely chose not to because they had prior
convictions that could be used to discredit them. For example, Leonard
McSherry did not testify in a case in which he was convicted of raping a

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 163

seven- year- old girl. This was no surprise. He had a prior conviction for
kidnapping and raping a six- year- old girl. Law professor John Blume has
found that “many demonstrably innocent defendants did not testify at
trial because, had they done so, they would have been impeached with
their prior convictions.”59 Those who did testify ran that risk, and the
juries heard about their prior convictions. Lacking any other evidence of
innocence, they may have testifi ed and taken that risk because otherwise
they had no case at all.

Defense Forensics and Experts

Beyond witness evidence, forensic and expert evidence also sometimes
provided these innocent people with evidence of their innocence. Foren-
sic evidence of innocence was prominent in 14% of these exonerees’ tri-
als (30 of 207 trials). Some exonerees tried to obtain DNA testing at the
time of their trial and could not, but in other exoneree trials, the foren-
sics excluded them, and jurors discounted or ignored that forensic evi-
dence of innocence. As discussed in Chapter 4, fi fteen exonerees were
excluded by DNA testing at the time of trial but were still convicted.
Fifteen more exonerees were excluded by all of the forensics in their
cases, not DNA, but other types of forensic evidence, such as serology,
hair, fi ber, or fi ngerprint evidence at the time of trial.60

Most exonerees had no experts of any kind. Most simply presented al-
ibi witnesses or the defendant’s own testimony. Only sixty- two exonerees
retained experts, not all of whom were permitted to testify at trial. Most
were forensic analysts, and several testifi ed regarding eyewitness memory
or examinations of defendants regarding their competency to stand trial
or ability to voluntarily confess.

In contrast, the State in almost every trial presented experts, includ-
ing forensic analysts and treating physicians or medical examiners. The
State had experts in at least 187 trials, often two or three or more experts.
In only seven trials was it clear the State did not retain experts.61 Why
such a striking imbalance? The State can easily obtain experts. Teams
of experts, such as medical examiners and forensic scientists, work full
time for law enforcement or the state. On the other hand, indigent crimi-
nal defendants, except for the very rare defendant who is wealthy, face

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

164 Innocence on Trial

great diffi culties securing funding for experts, and most do not have ex-
perts at trial. Rulings by the U.S. Supreme Court do not clearly entitle
defendants to experts even in many circumstances when they badly need
them to present a meaningful defense.62

The defense cases were defi cient in several diff erent ways, ranging
from the number of witnesses, to their expertise, to the limited nature of
their testimony, usually simply repeating an alibi that was itself fairly
weak. A fi rst reason for the weakness of the defense was that defendants
cannot easily investigate the crime themselves. While the police had been
investigating the crime since it was reported to them, these defendants
had no reason to think about the day of the crime, much less seek assis-
tance from a lawyer, until they became suspects or were arrested. The
prosecution had substantial investigative resources, including police and
crime lab analysts. In contrast, although criminal defendants are entitled
to lawyers at trial, they often do not receive experts and have no investiga-
tive team.

Second, it is hard to prove a negative, that the defendant did not com-
mit a crime. Although the prosecution could present a compelling story
of how a serious crime occurred, and then seemingly powerful evidence
that this defendant was the one that committed the crime, the defendant
faced an uphill battle trying to prove that he did not commit the crime,
but that some other unknown person did. Absent any evidence about
who that other person was, the defendant could present an alibi. That
was diffi cult to do, where most crimes occurred at night, and where it
would be hard to remember what happened so long before. The defen-
dant’s own word that he did not commit the crime may not be particu-
larly compelling, since a guilty person can claim to be innocent.

Third, I discuss in the next sections a fi nal reason why these defendants
did not have stronger cases: the lawyers. Although a talented attorney
might have made a more compelling case for innocence, these defendants
did not always receive particularly good defense lawyers. In addition, mis-
conduct by prosecutors sometimes made the job of the defense lawyers
more diffi cult. In some of these cases, evidence of innocence was con-
cealed from the defense.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 165

In eff ec tive Assistance of Counsel

How well did these exonerees’ defense lawyers perform? Were they in
part responsible for causing wrongful convictions? It is not easy to an-
swer those questions, but there is certainly no shortage of evidence of
subpar work by some of these defense attorneys. To take an egregious
example, Jimmy Ray Bromgard was represented by a lawyer who he re-
called was nicknamed “Jail house John Adams,” a contract attorney retained
by Yellowstone County, Montana. The lawyer had been “adjudicated in-
eff ec tive on prior occasions by federal court” and was “often found at a
bar playing cards when he was supposed to be in court representing his
clients.”63 Consistent with his apparent reputation, that lawyer “met with
him once before trial, hired no investigators or scientifi c experts, fi led no
motions to suppress evidence, made no opening statement, failed to pre-
pare Mr. Bromgard for his testimony and, after indicating he would ap-
peal, did not.”64 Adams even got Bromgard’s name wrong while address-
ing prospective jurors. Adams said, “You go by the name of Ray Bromgard,
don’t you?” Bromgard, who goes by Jimmy, said, “No.”65

The case against Bromgard hinged on an eyewitness identifi cation and
forensic evidence. Bromgard could have had a strong defense. The victim
admitted she was “not too sure” that he was the one that raped her, and
the forensic evidence was false and unscientifi c, as an FBI scientist found
in a subsequent examination. But Bromgard’s lawyer never even fi led a
pretrial motion to challenge the victim’s identifi cation, nor did he con-
duct any pretrial investigation of the forensic evidence or of anything else.
Bromgard was convicted and sentenced to forty years in prison, and he
spent fi fteen years in prison until DNA testing proved his innocence. For
years after his conviction, Bromgard tried to raise the issue of his attor-
ney’s shoddy repre sen ta tion. The Montana courts dismissed his claims
of in eff ec tive assistance of counsel.66 How could that be?

In its watershed ruling of Gideon v. Wainwright, the Supreme Court
ruled that a criminal defendant has a right to counsel in a felony trial.
But having a right to a lawyer does not mean that one has a right to a
good lawyer. Many states and localities have long provided inadequate
indigent defense funding, with predictably per sis tent poor assistance of
trial counsel as a result.67 Yellowstone County, Montana, like many

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

166 Innocence on Trial

counties in the United States, had no public defender’s offi ce. Instead, it
paid four contract attorneys, including John Adams, a fl at monthly rate
regardless how much eff ort they put into their clients’ cases, and with no
oversight of their work. Bromgard unsuccessfully sued the county for
maintaining a grossly inadequate system for repre sen ta tion of indigent
defendants, but settled claims against the state for $3.2 million, the larg-
est civil rights claim it had ever paid. In 2005, Montana fi nally created a
state public defender’s offi ce.68

Previous reports have suggested that the systemic lack of funding for
indigent defense in the United States leads to shoddy repre sen ta tion and
miscarriages of justice. Poor lawyering appears to have played a crucial
role in these exonerees’ cases, but studying the degree to which substan-
dard lawyering mattered raises many challenges. For one, take a case like
Bromgard’s, where the exoneree’s lawyer did little investigation or prepa-
ration, and may have performed badly at trial. Precisely due to his lawyer’s
in eff ec tive ness, one cannot tell from the trial rec ords what powerful argu-
ments and evidence an eff ective lawyer would have presented to the jury.
All one can say is that the trial lawyer did not put on a strong case. In addi-
tion, as I will discuss in Chapter 7, it is very diffi cult to win a postconvic-
tion claim like Bromgard’s, alleging that the trial lawyer was in eff ec tive. It
must be shown that the lawyer’s per for mance was so grossly defi cient that
it was unreasonable and could have altered the outcome at trial.

One exoneree asserted his right to adequate counsel at trial. Robert
Stinson stood up in court and asked the judge for a new lawyer. He
explained:

I would like to move the Court, with all due respect, for new coun-

selor for the following reasons.

One . . . my counselor has been misrepresenting in my case due

to the fact that there are quite a lot of facts, important facts, that he

has not brought out on cross- examination.

I feel he is ill counseling me to the degree that he only took my

case two weeks ago. There was not enough time for him to have pre-

pared for a trial in my best defense.

Three, we have a personality confl ict and cannot come to any

agreement on anything.

Your Honor, I’m facing life for something I did not commit.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 167

The judge denied Stinson’s request, and sure enough, he was sentenced
to life in prison for a crime he did not commit.69

Most exonerees were indigent and could not aff ord to hire a lawyer. Of
those for whom information could be obtained, seventy- eight had court-
appointed attorneys, seventy- one had public defenders, and fi fty- three
retained counsel.70 Those who retained their own lawyer often did not
have eff ective counsel either, nor could they hire expensive lawyers. Earl
Washington Jr.’s counsel was an example. Similarly, Nicholas Yarris re-
tained a lawyer, for just $1,500, to represent him in his capital murder
trial.

I have described a series of areas in which these exonerees’ lawyers
did not present a strong defense case, including the failure to present
strong alibis. The forensic evidence was another glaring area in which
defense lawyers were in eff ec tive. As discussed in Chapter 4, in about
half of the ninety- three cases in which prosecution forensic analysts pro-
vided unscientifi c or invalid forensic testimony, defense lawyers failed to
ask any questions about the areas in which an analyst misled the jury.
Almost all exonerees who confessed had their trial attorneys move to
suppress the confession at trial, but very few obtained experts to show
how the confession could have been coerced or produced by suggestion.
After all, state courts routinely deny funding for such experts.

Only one DNA exoneree, Donald Wayne Good, chose to represent
himself at his third and fi nal trial (his fi rst ended in a hung jury and his
second conviction was reversed). Even an inadequate lawyer might have
been a better choice. Good repeatedly interrupted to ask for the assis-
tance of his legal advisor to help him question witnesses, and in response
the judge would order him gagged and cuff ed.71

Prosecutorial and Police Misconduct

We have looked at the role played by the defense lawyers, but how about
their adversaries, the prosecutors— did they engage in conduct that con-
tributed to these wrongful convictions? The U.S. Supreme Court has
described how a prosecutor is obliged to do justice, and “while he may
strike hard blows, he is not at liberty to strike foul ones” or use “improper

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

168 Innocence on Trial

methods calculated to produce a wrongful conviction.”72 A number of
these exonerees were victims of prosecutorial misconduct that crippled
the defense’s ability to put on a meaningful case. However, it is also a
real challenge to describe, much less accurately study, what might con-
stitute prosecutorial misconduct. As with defense lawyering, the crimi-
nal procedure standards are extremely forgiving. The most common
prosecutorial misconduct claims that exonerees raised after they were
convicted were claims that prosecutors concealed evidence from the de-
fense and that prosecutors misled the jury during their opening or clos-
ing arguments.73

By its nature, misconduct involving concealed evidence may remain
hidden. We typically do not know what prosecutors had in their fi les,
much less what they failed to show to the defense. We may rarely learn
about violations of the rule established by the Supreme Court in Brady
v. Mary land, which requires that prosecutors provide to the defense evi-
dence that is “material and exculpatory”; in other words, evidence that
tends to show innocence or help the defense case, for example, by allow-
ing the defense to undercut the credibility of a witness. The defense may
not fi nd out about that evidence if prosecutors learn about evidence from
the police but do not provide it to the defense. Scholars and journalists
have gathered information on reported prosecutorial misconduct over
the years, and they have documented violations leading to troubling re-
versals in death penalty and hom i cide cases, for example. However, an-
other reason why prosecutorial misconduct often remains hidden is that
it is rarely investigated. It is almost unheard of for prosecutors to be dis-
ciplined or sanctioned for misconduct.74

Police misconduct may also play a role that is largely hidden. Indeed,
it may be hard to tell whether it was police or prosecutors that engaged
in misconduct. Prosecutors cannot fulfi ll their duty to provide the de-
fense with evidence of innocence if police never give that evidence to the
prosecutors. However, police misconduct is also very hard to defi ne or
to study. Just as we usually do not know what prosecutors had in their
fi les, we rarely fi nd out what police had in their fi les. Indeed, exonerees
claimed in some cases that police destroyed evidence helpful to their
cases or that they engaged in other misconduct, such as illegal searches
and seizures, all without any success.75 As discussed in Chapter 4, law

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 169

enforcement typically employed the forensic analysts that gave invalid or
unreliable testimony, and in some cases, when the evidence in the case
was later retested, it came to light that analysts had concealed evidence.

Although more may remain hidden to this day, we do now know that
in many of these cases the defense was never told about signifi cant evi-
dence. As I will describe in Chapter 7, twenty- nine exonerees uncovered
evidence posttrial or postconviction and argued that their cases should
be reversed because prosecutors had violated Brady by failing to dis-
close material exculpatory evidence that would have helped them to put
on an adequate defense. Four of them had their Brady claims granted by
a judge even before they obtained DNA testing; the others’ claims were
denied.

But years later, a host of additional possible Brady violations came to
light. As I will discuss in Chapter 8, many of the exonerees, after they
were vindicated through DNA evidence, fi led federal civil rights cases
asking for fi nancial compensation for their ordeals and demanding that
the police or prosecutorial practices that led to their wrongful convic-
tions be changed. This large group of exonerees blamed police miscon-
duct for their wrongful convictions; such cases invariably allege police
misconduct, since prosecutors typically enjoy civil immunity for their
acts. Such cases can be hard to win, but they can shed considerable light
on police and prosecutorial misconduct. In a federal civil rights case,
unlike a criminal case, the parties are entitled to gain access to evidence,
including not only documents but people as well. Prior to trial, the law-
yers can ask witnesses questions under oath (a “deposition”) and pose
written questions to the other side that must be answered under oath
(“interrogatories”). The lawyers of the few exonerees who fi led civil
rights cases often made document requests for the entire police fi les and
the State’s case fi les and obtained and examined them. At least seven-
teen exonerees brought claims alleging Brady violations, although many
of those cases settled without a determination of the merits of the claims.

Lawsuits were not the only way in which new evidence of Brady viola-
tions came to light. In thirty- one additional cases, lawyers, journalists,
scientists, or state investigators uncovered new information. For exam-
ple, in four cases it later came to light that police had hypnotized eyewit-
nesses who misidentifi ed exonerees. In at least twenty- two cases, it

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

170 Innocence on Trial

emerged that police failed to disclose forensic analysis helpful to the
defense. In still other cases, it later emerged that in for mants who had
denied receiving any kind of deal had in fact obtained a deal. In still other
cases, prosecutors or police had concealed evidence supporting the de-
fendant’s alibi or evidence of third- party guilt.76 For example, Clarence
Elkins discovered only years after his trial that police had notes from an
interview with a man who lived right next to the murder site, who made
self- incriminating remarks— and whom DNA testing later identifi ed as
the victim’s killer.77

After Thomas Doswell’s exoneration, an eyewitness provided his civil
rights lawyers with an affi davit explaining that her identifi cation was the
product of more nefarious coercion. Doswell had argued at trial that the
photo array procedure was suggestive, particularly since there was an
“R” written just on his photo. Years later, the witness explained that she
actually could not identify the assailant. She had seen him for only a few
seconds. She could not identify Doswell in the photo array, even with the
“R.” The detective told her he was angry, frightening her, and telling her
not to tell anyone that she could not identify anyone. She explained that
the reason she picked out Doswell at trial was because when the detective
drove her to the court house, he insisted that she identify Doswell. She
was “afraid of him” and succumbed to his pressure.78

We may never know whether more evidence of innocence or of offi cial
misconduct was concealed. These examples that have come to light,
however, certainly challenge the idea of a trial as a fair contest between
the State and the defense.

Prosecution Closings

After the defense completed its case, prosecutors delivered closing state-
ments, as did the defense, which summarized the evidence and the argu-
ments for each side. Prosecutors almost always countered any claim of
innocence quite directly in their closings arguments. Some responded
forcefully and with outrage. They used the claim of innocence against
the defendant, to argue that the defendant was not only guilty, but a liar.
In Thomas McGowan’s case, the prosecutor argued that there had “Not
been one inconsistent statement. She is telling the truth. We do not have

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 171

the wrong man.”79 In Anthony Green’s case, the prosecutor argued, in
response to his alibi, that “The defendant is a fraud, a phony and a liar.
It’s real simple, it’s as clear as could be. The defendant is a rapist and ag-
gravated robber.”80 In Chad Heins’s case, the prosecutor made fun of
the defendant, arguing, “Move over, Chubby Checkers because the twist
has been reinvented in this courtroom by that defendant right there.
And he’s taking every single thing he could that proves his guilt and he’s
tried to twist it and turn it and make it into what he wants it to be.”81 The
prosecutor in Arthur Lee Whitfi eld’s case made a slipup, and told the
jury, “We don’t want any innocent man (sic) to go free,” which he pre-
sumably did not mean to say (so the helpful court reporter added the
“sic” to the transcript), and then he added, “but Arthur Lee Whitfi eld is
not innocent.”82

Exonerees argued during their appeals that prosecutors crossed the
line during their closing arguments, by making statements designed to
infl ame the passions of the jury. Three exonerees succeeded in bringing
claims that prosecutors rendered the trial unfair during closing argu-
ments, For example, in Curtis McCarty’s case, the prosecutor had ar-
gued, “I wonder if [appellant] was grinning and laughing that night
when he murdered [the victim].” When defense counsel objected to the
prosecutor stating his personal view that “I want justice, ladies and gen-
tleman, and justice is . . . that [appellant] be convicted of murder one,”
he further argued, “He killed that girl. He needs to pay for it.” The ap-
peals court did fi nd that to be prosecutorial misconduct, noting that “It
is improper for a prosecutor to express his personal opinion of the guilt
of an accused.”83 However, such claims are very hard to win. Lawyers
are given a wide berth to make their arguments. One must show that the
arguments were so egregious that they infected the entire trial with un-
fairness. For example, in Ronald Taylor’s case the defense objected to
several statements. The prosecutor had asked the jury, “Do you want this
defendant, the bottom line, to go home with you today and go down that
elevator with you? Is that what you want? Because that is what a not-
guilty verdict will do.” The prosecutor had added, “Do you want him out
there cruising in his silver ’76 Oldsmobile in your neighborhood to-
night?” The appeals court approved such statements as a “proper plea for
law enforcement.”84

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

172 Innocence on Trial

Judges and Juries

We will never know the details of what went on during deliberations of
the jurors, after the prosecution and the defense presented their cases
and made their closing arguments, and after the judge gave the jury in-
structions on the law. The jury room is often referred to as a “black box”:
the jurors go in, they decide how to resolve the case, and they come back
out to the courtroom and give their verdict without reasons for it. But
jury rooms are black boxes that occasionally have small windows.

Despite the secrecy of the jury room, we sometimes do have some in-
formation about what jurors focused on if they asked the judge a ques-
tion during deliberations, or more dramatically, if they were deadlocked
or could not reach a verdict. We know that some jurors were troubled by
weak evidence, because they could not convict these innocent people.
Nine exonerees were tried more than once because their fi rst trial ended
with a hung jury that was dismissed because it could not reach a ver-
dict.85 Two of those exonerees had three trials because of hung juries
and mistrials. In addition, fi fteen exonerees were tried more than once
for a diff erent reason— their convictions were reversed on appeal or post-
conviction. A total of twenty- three exonerees had multiple trials on the
same off ense (one exoneree had both a hung jury and a reversal on ap-
peal), with eigh teen exonerees tried twice and fi ve more exonerees tried
three times.86 As Chapter 7 will describe, all of these innocent defen-
dants were reconvicted until DNA testing fi nally exonerated them.

Questions from jurors were not always transcribed, and sometimes
that portion of the transcript was missing, but in thirty- nine of these tri-
als, the jurors did ask questions. Rather than ask for clarifi cation about
the law or the judge’s instructions, they typically asked about the evi-
dence in the case. They did so in thirty- two cases, whereas in just thir-
teen cases they asked questions about the law or the jury instructions.
The jurors usually asked to see exhibits or transcripts of testimony, and
often focused on important issues. For example, in Roy Brown’s case, in
which the bite mark testimony was central, the jury asked to review the
testimony of the bite mark experts and to see the photographs of the bite
marks.87 In Anthony Capozzi’s trial, the jurors asked to have the victim’s
initial description of the attacker reread; that description did not partic-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 173

ularly resemble Capozzi.88 In Chad Heins’s case, the jurors asked to see
a chart explaining the DNA evidence; the judge denied the request be-
cause the chart was not admitted into evidence, but the jurors were right
to be confused, since the analysts had failed to correctly explain the sta-
tistics in that case.89 In Jerry Watkins’s trial, the jurors asked to read
again the testimony of the jail house in for mant.90

In a few cases, jurors spoke about the case after the trial. A few jurors
later spoke to the press about their deliberations. In Rolando Cruz’s case,
for example, one juror later told the press that “half of the jurors had their
minds made up before the trial even started” and that the jury foreman
announced that because “[the defendants are] here, they must have done
something,” and that the deliberations were “a mere formality, so we
might as well get on with it.”91 Arvin McGee’s third trial is a striking ex-
ample of what we can discover when jurors speak out after the trial. Mc-
Gee’s fi rst trial had ended in a mistrial, and the second trial ended with a
hung jury. After all, the victim had initially identifi ed another photo, and
McGee claimed he was physically incapable of committing the crime,
having had hernia surgery days before it occurred.92 Now, at his third
trial, in Tulsa, Oklahoma, in 1988, the jurors reported diffi culty reaching
a verdict after several hours in the jury room. The judge instructed them
to continue to deliberate “in a spirit of fairness and candor.” This time,
the jury returned a guilty verdict.

However, one juror came forward shortly after the trial and long before
the DNA testing to say she did not support the guilty verdict. She had re-
mained silent because there was a “lot of pressure” from other jurors. Two
other jurors had voted to acquit but changed their minds. For hours, the
other jurors told her that the defendant had their names and might rape
their family members, and that this rapist could not go free on the streets.
They took the “not guilty” jury forms away from her. The hold- out juror
fi nally told them “what ever you want to do, you go ahead and do it.” She
remained silent without signing the guilt verdict form, but without object-
ing. The defense moved for a new trial based on these revelations. The
judge denied the motion.93 McGee was not exonerated until more than
fourteen years later, in 2002.

In Eric Sarsfi eld’s case, as in McGee’s and several other exonerees’
cases, the jury was initially deadlocked. After “several votes” and a day

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

174 Innocence on Trial

of “active deliberation” the jurors came to the judge in frustration, say-
ing, “We are all at an impasse.” The judge gave them antideadlock in-
structions, reminding them of the reasonable doubt standard that the
State must satisfy and encouraging them to try to arrive at a verdict. The
jury deliberated again for several hours and returned a guilty verdict.94
We don’t know how they eventually reached a guilty verdict, but we do
know that at least some of the jurors initially resisted.

The selection of the jurors may have aff ected the outcomes in some of
these cases. Practices for selecting a jury vary quite a bit. The size of the
panel from which the jurors are picked varies, as does the number of
challenges that each side can make to dismiss jurors. Courts have diff er-
ent practices for what questions can be asked of prospective jurors dur-
ing voir dire.

In the trial of Kenneth Adams, Willie Rainge, and Dennis Williams,
all three of whom are black, prosecutors had struck all black jurors dur-
ing jury selection. One prospective juror announced to the judge, “It’s
obvious the state’s attorneys want an all- white jury,” and “They don’t
want me here,” resulting in applause for the prospective juror by the
courtroom audience. When the judge restored order, fi ve more prospec-
tive jurors stated that they also felt the prosecutor was targeting black
jurors— and each person who complained was then removed from the
jury panel. The fi nal jury had eleven whites and one black woman; a
second jury that decided whether Williams and Rainge should be sen-
tenced to death was all white.95 In Calvin Johnson’s case, the all- white
jury took forty- fi ve minutes to deliberate and convict him of one of the
two rapes he was charged with. Tellingly, a year later he was acquitted of
the second rape before a racially mixed jury, even though that jury heard
about his conviction of the fi rst rape. We know that several black exoner-
ees had all- white jury trials, including Marvin Anderson, Calvin John-
son, and Anthony Hicks. However, for most of the cases, I could not
obtain transcripts of the jury selection pro cess, and even where I could,
there was typically no way to discern the jury’s composition from those
rec ords. I also found that twenty- one exonerees challenged the jury se-
lection in their case on appeal or postconviction. As I describe in Chap-
ter 7, none had any success in doing so.

A fi nal insight into decision making can be gleaned from the few cases
in which a judge, not a jury, determined guilt or innocence and from the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 175

cases where judges made comments during sentencing. Twelve exoner-
ees had no jury, but instead a “bench” trial before a judge. Those trials
are fascinating because, unlike jurors, judges publicly explain the rea-
sons why they convict. For example, in Chapter 2 I discussed the judge
in Nathaniel Hatchett’s case, who explained why he decided to convict
based on Hatchett’s confession, but despite DNA tests that ruled out
Hatchett as the rapist. In Willie Davidson’s case, the judge stated that he
found Davidson guilty based on the victim’s identifi cation and the foren-
sic evidence. However, he appeared confused by the forensic evidence,
understandably so, because the analyst had presented invalid testimony.
He said that the serology evidence “had the type of a non- secretor. The
defendant is a non- secretor. That by itself isn’t totally conclusive. Forty-
two percent are of that.” Actually, 42% was not a correct percentage, since
due to the problem of masking described in Chapter 4, any male could have
been the source of the semen.96 In Richard Johnson’s case, the judge com-
mented that the victim, who was initially uncertain when she fi rst identifi ed
Johnson, “was not mistaken, no matter what the defendant says.”97 Other
judges did not provide lengthy explanations of their reasons; for example,
in Bruce Dallas Goodman’s case, the judge briefl y stated that he was “per-
suaded” by each of the prosecution witnesses, and therefore, “fi nds the de-
fendant guilty as charged.”98

We have another window into the decision- making pro cess because
judges, and not juries, often decided what a convicted person’s sentence
would be. When sentencing, judges may comment on the case and give
reasons for the length of the sentence. Perry Mitchell was convicted after
the jury deliberated for just an hour, and afterwards the judge told the
defendant that he would impose the maximum sentence of thirty years,
stating: “I am convinced until we do have a neutering law serious crimes
of violence will continue to happen. I know what neutering does to ani-
mals.” He added that rapists like Mitchell should suff er “surgical ampu-
tation of the male penis for the simple reason it has been my experience
in approaching the year of 72 that the rapist is a repeater.”99

We know, then, very little about the pro cess that juries (and judges)
went through in determining guilt. We do not have a recording of what
the jurors said to each other in the jury room. We do, however, have
something important: we know what evidence was presented to the jury
in the trial, because for most of the cases, we have the trial transcripts. We

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

176 Innocence on Trial

know what stories these innocent people’s defense lawyers told the juries,
and having described how the defense typically presented a short and
fairly weak case we can begin to understand why these juries and judges
believed that so many innocent people were guilty.

The Tilted Playing Field

All of these trials revolved around the question of innocence, and these
innocent people did not present a strong defense, in contrast to the pros-
ecution, which had access to substantial investigative resources, and
presented confession, eyewitness, forensic, and in for mant testimony that
often seemed powerful at trial. I have described how, fi rst, these people
had little ability to locate evidence of their innocence. Second, it is dif-
fi cult to prove that one did not commit a crime, particularly without any
evidence of who actually did commit the crime. One must somehow
account for all of one’s actions on a day in the past that was not eventful
at the time. Third, the defense lawyers in these cases often did not ef-
fectively challenge the prosecution evidence or try hard to develop a
more compelling alibi. Moreover, evidence of innocence was sometimes
concealed from the defense. If indigent defense were better funded, if
courts insisted on higher- quality repre sen ta tion for the poor, if lawyers
for the poor had better access to investigators and experts, if police and
prosecutors provided broader disclosures of evidence and faced sanc-
tions for misconduct, then perhaps the playing fi eld would be more even.100
As I will discuss in Chapter 9, few such changes have been made in re-
sponse to these exonerations or otherwise. These problems are deeply
ingrained and they implicate the entire structure of our criminal justice
system.

The closing statements that concluded Dennis Fritz’s trial encapsu-
lated the dynamic in so many of these exonerees’ cases. Dennis Fritz
faced the death penalty but could not aff ord to hire an attorney. He later
recalled, “[l]ike a lamb going before the slaughter, I was forced to settle
for a court- appointed civil lawyer who specialized in bankruptcy and
personal injury.”101 The defense case largely consisted of Fritz’s own tes-
timony. In contrast, the prosecutor had presented more than twenty wit-
nesses, including a series of forensic experts.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Innocence on Trial � 177

The defense closings at the sentencing stage were three paragraphs
long, with his lawyer stating only, “I’ll be very brief. I’ve never stood
before 12 people before and asked for a gentleman’s life . . . I would ask
you to search your soul and come forward with a verdict that you can live
with.”

In contrast, the prosecutor emphasized, “I’m going to ask you to come
back in this court and say, Dennis Leon Fritz, you deserve to die for what
you did.”

Fritz himself interjected during the prosecutor’s closing argument:
“I didn’t kill [the victim].”102

After deliberating for just two hours, the jury announced its sentence:
life in prison. The vote of a single juror saved him from the death pen-
alty. Fritz later wrote about how he contemplated the torment of facing a
lifetime of imprisonment, asking, “What had I done so wrong in my life
that I should deserve this kind of punishment?” He told himself, “How-
ever long it takes to prove my innocence, I will not give up.”103 Twelve
years later, postconviction DNA testing exonerated him.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

178

In 1992, Kennedy Brewer’s girlfriend’s three- year- old daughter was
abducted from their Macon, Mississippi, home, and was raped and mur-

dered. Police immediately suspected Brewer and arrested him even before
the victim’s body was found. He had been home babysitting that eve ning,
and there was no evidence of a forced entry, although the window where
the victim had been sleeping was open. In 1995 Brewer was convicted of
murder and sexual battery and sentenced to death. DNA testing was not
used at the time of trial, because the semen sample from the victim’s body
was deemed insuffi cient for testing. The chief evidence of Brewer’s guilt
came from marks on the victim’s body. The medical examiner concluded
that those marks were human bite marks. Forensic dentist Michael West
concluded that nineteen bite marks on the victim’s body came from Brew-
er’s teeth. West said he was certain, and before the jury he added his calling
card fl ourish, “indeed and without a doubt.” At the time, West had already
been removed or had resigned pending expulsion from two associations of
forensic dentists due to his unreliable methods and conclusions.1

Years later, analysts examined the evidence in Brewer’s case and de-
termined that the marks were not human bites at all. They were bites
from insects. This was logical. The victim’s body had been in a creek for
two days, in hot May weather in Mississippi.2

c h a p t e r 7

Judging Innocence

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 179

In 1998 Brewer’s appeal was dismissed by the Mississippi Supreme
Court.3 He asked for a new trial and had brought a battery of claims, in-
cluding: claims of prosecutorial misconduct at trial; objections to the
jury instructions; a challenge to the adequacy of West’s credentials; and
a challenge to the judge’s refusal to let the jury see a videotape showing
West manipulating the victim’s skin when matching the bite marks.
Brewer also claimed that he was innocent. He argued there was not
enough evidence of guilt to legally support the jury’s verdict. The court
emphasized when it rejected all of those claims that there was suffi cient
evidence of Brewer’s guilt; “the child’s body was covered with bite
marks, which were confi rmed by the State’s expert as being infl icted by
Brewer.”4 The U.S. Supreme Court declined to review the case in
1999.5

After four years, his appeal was now complete. Brewer then fi led a
state postconviction petition. States typically permit a second round of
challenges to a conviction, often to give the convict a chance to bring
claims that were not or could not be raised during the appeal. Brewer
now argued that his lawyer had been unconstitutionally denied access to
DNA testing at the time of trial in 1995. He disputed the State’s conten-
tion that the sample was insuffi cient for DNA testing.6 Because it was not
clear from the record why DNA testing was not done at the time of trial,
and due to a procedural default, the Mississippi Supreme Court dis-
missed his petition in 2000. After all of the state proceedings are over,
convicts can also fi le a habeas corpus petition with a federal judge to
challenge their conviction, although these petitions are very rarely suc-
cessful. However, Brewer did not pursue federal habeas corpus.

Instead, Brewer kept trying to get DNA tests, which the prosecutors
were not willing to conduct. In 2000, the Mississippi Supreme Court is-
sued an order sending the case back to the trial judge for DNA testing. In
2001, the results excluded Brewer and produced a single unknown male
DNA profi le. The Mississippi Supreme Court again heard the case, and
in light of these DNA results, sent the case to the trial court for an eviden-
tiary hearing.7 This time, the trial judge vacated Brewer’s conviction,
meaning that the conviction was nullifi ed and Brewer was granted a new
trial. He was not yet exonerated, however, because the prosecutors could
still decide to try him again.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

180 Judging Innocence

The prosecutors announced their intent to try Brewer again, but did
nothing. He remained in limbo, languishing in jail for fi ve years. Only in
2007, after the Mississippi attorney general intervened, additional DNA
testing was conducted, which confi rmed the guilt of one of the original
suspects, a man named Justin Albert Johnson. Johnson then confessed.
Thirteen years after his trial, in February 2008, Brewer was fi nally re-
leased based on this accumulated evidence of his innocence and he was
exonerated.

Brewer, like many others, experienced agonizing delays in his quest
for an exoneration, even after it was clear that the wrong man was con-
victed. By 2000, when DNA testing was ordered in Brewer’s case, the
technology had advanced to the point that it could provide a conclusive
answer to the question of his guilt. However, despite the availability of
this powerful tool, prosecutors opposed the DNA testing and then, once
the testing did exclude Brewer, opposed his release, and as a result he
spent eight more years in prison. Like most exonerees, Brewer obtained
no relief from judges prior to getting DNA testing.

Kennedy Brewer served fourteen years in prison between his trial and
his exoneration. That was about the average prison stay of these exoner-
ees. Compare Brewer’s experience to a far less typical exoneration case,
that of Curtis McCarty. Curtis McCarty, in contrast, served far more time
than the average exoneree—twenty- one years. Although he suff ered in
prison for many more years, in another respect the judicial system per-
formed better in McCarty’s case. In Brewer’s case, judges never granted
any relief until after the DNA testing. McCarty was one of twenty- one
exonerees who actually had success challenging his conviction even be-
fore obtaining the DNA testing that ultimately proved his innocence.
McCarty’s conviction was reversed twice based on challenges to forensic
science testimony by the Oklahoma City Police Department crime labo-
ratory analyst Joyce Gilchrist, discussed in Chapter 4.

First, McCarty had his 1986 conviction and death sentence reversed
in 1988 because of hair testimony that was invalid. Gilchrist had testifi ed
that she could tell, based on hairs found at the crime scene, that Mc-
Carty “was in fact there.” The court reversed the conviction, noting “[w]
e fi nd it inconceivable why Ms. Gilchrist would give such an improper
opinion, which she admitted she was not qualifi ed to give.”8

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 181

The victory was short- lived. Prosecutors decided to try McCarty
again, and he was convicted again in 1989. When he appealed again,
Gilchrist’s testimony at his second trial was not found to be grounds for
a reversal (this time she wisely did not claim that “he was in fact there”).9

However, a judge vacated his conviction a second time in 2005. Over
fi fteen years after his second conviction, McCarty’s lawyers had discov-
ered new misconduct by Gilchrist. She had altered and fabricated labo-
ratory reports. She had excluded McCarty as the perpetrator, so she
then changed her reports to claim a match, and when asked to produce
the hairs examined, said that she had destroyed them. The court em-
phasized that “Ms. Gilchrist’s actions alone warrant a new trial.”10 A fi rst
round of DNA testing in 2002 had excluded McCarty. Now, as a third
trial date approached, the Innocence Project secured still more DNA
tests that excluded him. McCarty was fi nally exonerated in 2007, after
serving twenty- one years in prison.11

Long delays between the conviction and the ultimate exoneration
were typical in the 250 exonerees’ cases. When most of these people
were initially convicted, DNA testing was not yet commonly available.
Although early forms of DNA testing were available by 1989, exonera-
tions occurred very slowly in the years that followed. It took an av-
erage of fi fteen years from the time that these innocent people were
convicted to their exoneration. Figure 7.1 shows the number of exoner-
ees convicted and exonerated each year. The convictions are heavily
concentrated in the 1980s, with a few in the 1970s and a few convictions
occurring in the 1990s but trailing off after 2000. Although the fi rst ex-
oneration occurred in 1989, exonerations did not begin to climb in
earnest until the mid- 1990s. The numbers then dramatically increased
starting in the late 1990s, almost a de cade after the fi rst exoneration.

Criminal appeals and postconviction review can take a long time. It
took these exonerees an average of six and a half years after their convic-
tion to exhaust all of their appeals and habeas proceedings. Some cases,
like McCarty’s, went on for de cades. In this chapter, I explore the
lengthy appeals and postconviction pro cess that accounted for the lon-
gest delays in the years that passed until their eventual exonerations.
Like Brewer, most exonerees had no luck claiming their innocence or
challenging the trial evidence until they fi nally obtained DNA tests.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

182 Judging Innocence

Why was it so hard for these innocent people to challenge their fl awed
convictions?

Challenging Trial Evidence

Mistakes occur in many trials. Sometimes a judge admits a piece of evi-
dence that, under the rules of evidence, should have been excluded, or
misstates the legal standard the jury is to apply. Sometimes a prosecutor
or defense lawyer misspeaks in a closing statement, misquoting a witness
or referring to evidence that the judge asked the lawyers not to mention
in front of the jury. One of the most diffi cult tasks of a judge is deciding
which mistakes matter and which mistakes do not. Chief Justice Roger
Traynor, of the Supreme Court of California, poetically described the
plight of the postconviction judge confronted by thousands and thou-
sands of claims of trial errors: “Errors are the insects in the world of law,
traveling through it in swarms, often unnoticed in their endless pro-
cession. Many are plainly harmless; some appear ominously harmful.
Some, for all the benign appearance of their spindly traces, mark the way
for a plague of followers that deplete trials of fairness.”12 The job of the
judge is to pinpoint those harmful errors and remedy injustice, but to
screen out inconsequential errors that did not cause an unfair trial.

1977 1980 1983 1986 1989 1992 1995 1998 2001 2004 2007

Year of conviction Year of exoneration

Figure 7.1. Number of exonerees convicted and exonerated per year

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 183

The cases brought by DNA exonerees, who we now know to have
been innocent, provide a unique opportunity to assess how well judges
sort out harmful errors from the harmless. The previous chapters have
described the serious errors common in these exonerees’ trials, from
fl awed eyewitness identifi cations and contaminated confessions to exag-
gerated forensic testimony. Now we can examine how well the criminal
justice system handled the cases of these innocent people after their tri-
als and convictions.

When I reviewed the rec ords in these exonerees’ cases I asked myself
why the judges hearing these cases on appeal or habeas review did not
correct these errors, long before DNA testing entered the picture. The
question assumes that these people could somehow show a judge that
they were innocent even without getting DNA testing. But at a more fun-
damental level, the question assumes that after a conviction, higher
courts will review the trial record and look for mistakes, to make sure
that a miscarriage of justice did not occur.

That second assumption is not a very good one. Judge Jerome Frank
and Barbara Frank, in their 1957 book about wrongful convictions, called
the notion that the court on appeal will correct the mistaken conviction of
the innocent the “Upper Court Myth.” They pointed out that the appel-
late court “knows no more than the jury and the trial judge” and has a
limited role. It is “obliged to accept the jury’s verdict” and must typically
accept the testimony of the witnesses as true rather than reconsider the
case based on a cold record.13 In the de cades after they wrote, a criminal
procedure revolution has changed the face of appellate and postconvic-
tion litigation, creating a host of new avenues to challenge a conviction,
but still, very few cases are ever reversed on appeal and postconviction
review— no more than 1% or 2%.14

When I fi rst began looking at the judicial rulings in exonerees’ cases,
I thought that perhaps the judges in these cases had taken them more
seriously than most. The exonerees earned high numbers of reversals— a
13% reversal rate— in criminal appeals and postconviction proceedings
they brought before they obtained the DNA testing that exonerated them.
This percentage was much higher than the 1%– 2% reversal rate that is
typical in postconviction cases. Even before the DNA testing, I thought,
judges saw fatal fl aws in some of these exonerees’ convictions. However,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

184 Judging Innocence

I then discovered that this 13% reversal rate was not unusual. The rever-
sal rate in these exonerees’ cases was no diff erent from the reversal rates
of other rape and murder trials.15 The grim implication is that rape and
murder trials may simply produce higher rates of reversible errors.

To fi gure out what went wrong in these cases, I needed to study them
in detail. Our criminal justice system does not collect in one place all of
the rec ords for a criminal case as it progresses. It is hard to get complete
information about the life cycle of a criminal case. I tracked down as
much information as I could about each stage in the review pro cess fol-
lowing these exonerees’ convictions, including each criminal procedure
claim they raised postconviction and each ruling a court rendered on
each of their claims. Because courts issued written decisions in about
two- thirds or 66% of the cases (165 of 250 cases), combing through this
mass of opinions does not tell us what happened in every case, but it can
allow us to make some generalizations about how courts judged inno-
cence. I studied only the exonerees who had judges issue written deci-
sions explaining their rulings during their appeals and post- conviction
proceedings.16 After all, in cases where they chose not to write an opin-
ion, we do not know the reasons why judges ruled the way they did, and
we often cannot even tell what claims they ruled on.

A troubling story emerged when I looked closely at the claims that
exonerees made and the rulings they received during their appeals and
postconviction proceedings. Many exonerees did not challenge the cen-
tral types of evidence supporting their wrongful convictions— or if they
did, judges did not think it was worth mentioning in their written opin-
ions. Of the exonerees who did receive written decisions during appeals
and postconviction, one- third (55 of 165 cases) challenged none of the
central facts that supported their convictions.

When these innocent people did challenge the evidence that had led
to their convictions, they were very rarely successful. A look at the exon-
erees with written decisions who were convicted based on par tic u lar
types of evidence sheds light on this fact. For example, how many con-
victed based on eyewitness misidentifi cations challenged those identifi –
cations postconviction? Of the 124 exonerees who were convicted on the
basis of an eyewitness identifi cation and obtained a judge’s written deci-
sion, only 56% challenged the eyewitness identifi cation (70 of 124 cases).

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 185

Only 7% were successful (5 of 70 cases). Similarly, only 32% of those
who had forensic evidence at trial challenged the forensic evidence (36 of
112 cases) and 17% succeeded (6 of 36 cases). Only 36% challenged in for-
mant testimony (16 of 45 cases) and 25% succeeded (4 of 16 cases). The
largest proportion, 59%, challenged false confessions (13 of 22 cases), but
only 8% had any success ( just 1 of 13 cases).

We now know that this trial evidence was fl awed, and that many of
those errors were in fact the kind that Chief Justice Traynor called “omi-
nously harmful.” However, the judges reviewing these exonerees’ claims
during appeals and postconviction often called those errors “harmless”
and even referred to the perceived strength of the prosecution case and
the likely guilt of the petitioners. These innocent people had no better
success raising claims asking for a new trial on the basis of evidence of
their innocence; all such innocence claims were rejected.

How did these innocent people go about challenging what had hap-
pened at their trials? Why did so many of them fail to bring factual chal-
lenges? And when they did appeal or seek habeas relief, why did judges
fail to see that they were innocent? In the next sections, I will start to ex-
plore these issues, beginning with how exonerees tried to challenge the
evidence presented at their trials.

False Confessions

Recall from Chapter 2 that almost all of the exonerees who falsely con-
fessed challenged their confessions at trial, typically by claiming that the
police coerced them into making the confession statements. To my sur-
prise, after their trial, several of the people who falsely confessed no lon-
ger challenged their confessions, even when they were otherwise appeal-
ing their convictions. Of the twenty- two innocent people who were
convicted based on false confessions and had written decisions in their
cases, only seven raised Fifth Amendment claims that their confessions
were involuntary, and three more alleged their confessions were obtained
in violation of Miranda. None of these claims was successful.17

The only successful claim was a kind of claim that can only be brought
on appeal or postconviction— that the trial lawyer provided in eff ec tive
assistance by not suffi ciently attacking the validity of the confession at

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

186 Judging Innocence

trial. Three of the exonerees made such claims relating to their confession,
and one of those three, Ronald Williamson, obtained a reversal. But in
that case, the lawyer did much more than simply fail to challenge the co-
erced confession, which was admitted despite Williamson’s manifest men-
tal illness; he also failed to challenge invalid forensic testimony, among
other failures. Indeed, before trial he had begged the judge to let him
withdraw from the case. “I can’t represent him Judge; I just can’t do it,” his
lawyer had insisted. “I’m too damned old for it, Judge. I don’t want any-
thing to do with him, not under any circumstances.”18

When exonerees did bring challenges to their confessions directly,
courts emphasized the seeming reliability of the confessions. In many of
the cases, courts noted the reliability of these confessions when denying
relief, often by highlighting the nonpublic and corroborated facts that
played such an important role at trial. The courts often also cited the
“overwhelming” nature of the evidence against the defendants, describ-
ing in detail the nonpublic and “fully corroborative” facts they each sup-
posedly volunteered.19 For example, the Illinois Supreme Court stated
that Alejandro Hernandez “did not present an argument which convinces
us that he learned the details of the crime contained in his ‘vision’ from law
enforcement offi cers.”20

In nine more cases, the exonerees allegedly made self- inculpatory
statements to police but not a full confession to any of the crimes. Four of
those exonerees also brought coerced confession claims regarding their
statements to police.21 None of these alleged voluntary statements, as re-
ported by police or witnesses, were successfully challenged on appeal or
postconviction, likely because a claim of coercion would be diffi cult to
make for a statement that was putatively volunteered.

Thus, although 59% of the exonerees who falsely confessed and had
written decisions did challenge their confessions (13 of 22 cases), only one
of them succeeded, where the trial lawyer failed to challenge the confes-
sion but also made a series of other mistakes and omissions. In contrast,
courts often cited the seeming reliability of these detailed confessions
when denying relief on these and other claims. The contamination of
these confessions, described in Chapter 2, not only made the convictions
a foregone conclusion but also made it very diffi cult for these innocent
people to challenge those confessions on appeal or postconviction.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 187

Eyewitness Identifi cations

Although the vast majority of these exonerees, 67%, were convicted
based on eyewitness identifi cations (190 of 250 cases), some did not even
challenge those identifi cations, and few who did succeeded. Of those
exonerees who were convicted based on a mistaken eyewitness identifi –
cation, 56% of those who obtained a written decision postconviction
brought a claim challenging the identifi cation (70 of 124 cases).

As described in Chapter 3, many exonerees were identifi ed based on
suggestive eyewitness identifi cation procedures. Indeed, the most com-
mon type of claim involving eyewitness testimony was that police im-
properly indicated to the eyewitness who the suspect was. Thirty- nine
exonerees brought such a claim. None were granted. Four exonerees as-
serted their right, established by United States v. Wade, to have a lawyer
present at a postarrest lineup; none of the claims were granted.22

Some exonerees raised other less direct claims. For example, rather
than claim that an eyewitness identifi cation was unreliable, one might
indirectly assert that the attorney was in eff ec tive for failing to challenge
the eyewitness testimony. Thirty exonerees made indirect challenges to
the identifi cations, such as for in eff ec tive assistance of counsel, newly
discovered evidence of innocence, or improper jury instructions.

Only a handful, fi ve exonerees, earned reversals based on faulty eye-
witness identifi cations. McKinley Cromedy’s conviction was reversed
for failure to instruct the jury regarding the dangers of cross- racial iden-
tifi cations. Lesly Jean had his conviction reversed because it was not
disclosed that the victim had been hypnotized (and may have fi rst identi-
fi ed him while under hypnosis). The fi rst trial of Ronald Cotton was re-
versed because the judge had excluded evidence that one of the two vic-
tims raped in similar attacks on the same night had identifi ed another
person. Michael Evans was granted a new trial because the State did not
disclose that police had paid their star eyewitness $1,250. Mark Webb’s
conviction was reversed because there was evidence that the victim, who
had earlier not been able to identify him, violated the rule against dis-
cussing the case with other witnesses— she used a bathroom break dur-
ing the trial to talk to her sister, after which she was then suddenly able
to identify him.23

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

188 Judging Innocence

None of these unusual cases was a classic example of police using sug-
gestive eyewitness identifi cation procedures, such as a showup, or mak-
ing suggestive remarks. The only one of the fi ve cases that raises the
problems of suggestion and unreliability that were so common in these
eyewitness misidentifi cations, as discussed in Chapter 3, is Cotton’s
case, where the judges reversed the conviction because they felt the jury
should have heard about how an eyewitness had earlier been less certain
than she was at trial. Even that was an unusual case, since the eyewitness
had not only earlier failed to identify Cotton but she had identifi ed an-
other man.

What about the 44% who did not challenge their faulty eyewitness
identifi cations at all? Some exonerees may have defaulted on any such
claim by failing to preserve an objection at trial, but others may have sim-
ply decided that such a claim would be futile. As discussed in Chapter 3,
the Court held in Manson v. Brathwaite that even if the police engage in
suggestive procedures so potentially misleading that their conduct vio-
lates due pro cess, the identifi cation may still be admitted at trial if it was
otherwise “reliabl[e].”24 Even identifi cations resulting from highly sug-
gestive procedures may nevertheless be admitted given other indicia of
reliability, such as eyewitness certainty. Almost all of these eyewitnesses
expressed great confi dence at trial that they had identifi ed the right
person.

As a result, most exonerees had no successful basis for challenging
what we now know to be incorrect eyewitness identifi cations. For exam-
ple, in Steven Barnes’s case, his photo was the only one repeated in two
photo arrays. The court ruled that “the passage of two and one- half years
between the two arrays negates any possibility of suggestiveness” and
that the witness demonstrated an in de pen dent basis for the in- court iden-
tifi cation of defendant.”25 Similarly, David Bryson argued that the police
engaged in suggestion. In fi nding no error, the court emphasized how
certain the victim was. The court ruled that “Certainly, if T.T. had been
hesitant or tentative in her pre- trial identifi cation, the police comments
would have rendered the confrontation unnecessarily suggestive.”26 In
Anthony Green’s case, the court noted that his photo was the only one
repeated in two arrays, and in the fi rst, his was the only photo that was a
“wallet- type photo,” while in the second, his photo stood out as the only

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 189

“mug shot” with a physical description. Without discussing suggestive
remarks also made to the victim, the court concluded that because her
attention was “acute” during the incident, she “in fact made eye contact”
with the attacker, and “immediately recognized” him in the arrays, the
identifi cations were properly admitted.27 Given the reliance of judges on
eyewitness certainty, some defendants and their lawyers may have simply
decided that a postconviction challenge would be fruitless. Indeed, of the
thirty- nine people who did make these kinds of challenges to suggestive
identifi cation procedures, none prevailed.

Forensic Evidence

Although forensic evidence was the second leading type of evidence
supporting these erroneous convictions, once again few challenged that
evidence and even fewer obtained reversals. This seems surprising on its
face, where 61% of the prosecution forensic testimony at exonerees’ trials
was invalid testimony that misstated the relevant scientifi c principles,
and still more cases involved unreliable forensic techniques, vague testi-
mony, and concealed forensic evidence. One would think that clear de-
viation from scientifi c principles would be relatively easy to identify as
compared, say, to a challenge to an eyewitness whose memory has be-
come irreversibly contaminated. The poverty of the defendants may ex-
plain this. Without access to an expensive forensic expert, these defen-
dants’ lawyers may not have understood that the forensic testimony was
invalid. As a result, they may not have challenged the evidence at trial,
which they needed to do in order to preserve the issues for appeal.

Six exonerees did obtain reversals based on challenges relating to fo-
rensic evidence at trial, or 17% of the 36 exonerees with written decisions
who challenged the forensics in their cases. Three of those exonerees
challenged the forensic analysis itself. As I discussed earlier, Curtis
McCarty had his conviction reversed twice, the fi rst time because the
analyst claimed erroneously that she could tell, based on microscopic
hair comparison, that “he was in fact there” at the crime scene.”28 Ray
Krone’s fi rst conviction was reversed because the prosecution did not
disclose until the eve of trial a highly unusual and prejudicial videotape in
which the bite mark expert demonstrated the supposed match by holding

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

190 Judging Innocence

molds of Krone’s teeth to the deceased victim’s body.29 Testifying of
behalf of Steven Linscott, a forensic expert rejected any notion that
probabilities can be used in hair evidence. Yet in his closing statement to
the jury, the prosecutor asserted that the defense expert had endorsed
such probability testimony, which the appellate court in reversing the
conviction found to be a “rank, calculated distortion.”30 The three ad-
ditional exonerees who earned reversals related to the forensic evidence
challenged the inadequacy of their defense lawyers for not suffi ciently
disputing the forensic evidence, but also for other failures unrelated to
forensics.31

More typical were cases where judges denied relief despite unreliable
and invalid testimony by forensic analysts. Thus, in appeals brought by
William Rainge and Kenneth Adams, the Illinois appellate court ruled
that though “the State said the probability of fi nding similar hairs from
diff erent people was 1 in 4,500,” and that statistic had no support, the ar-
gument did not “suffi ciently mislead a jury as to require reversal.”32 Simi-
larly, the Montana Supreme Court denied Chester Bauer’s appeal despite
invalid serology testimony that, as I described in Chapter 4, included a
litany of errors, including ignoring the problem of masking and by
wrongly dividing the off ered statistic in half.33

These decisions are part of a real abdication of responsibility by judges
to ensure that sound science is presented in our courtrooms. Not only did
trial judges neglect their duty as gatekeepers to prevent experts from us-
ing unreliable methods or off ering exaggerated conclusions on the stand,
but after the convictions, in the cases where these errors were pointed
out, judges looked the other way and dismissively ruled that the testi-
mony was tolerable or probably did not make any diff erence at the trial.

In for mant Testimony

Cases with in for mant testimony exhibited the same pattern, in which
central evidence from trial typically went unchallenged. Of the people
convicted based on in for mant testimony who had written decisions on
appeal or postconviction, only 36% challenged in for mant testimony (16
of 45 cases). Only four of them succeeded. As I described in Chapter 5,
we now know that many of these in for mants testifi ed falsely when they

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 191

claimed that the exoneree had told them “inside information” about the
crime.

One person, Verneal Jimerson, did win a reversal based on a fabrica-
tion claim, not attacking evidence provided by a jail house in for mant but
instead attacking the coerced confession of his alleged coconspirator. As
described in Chapter 5, in Jimerson’s case police concealed that they
obtained the testimony of codefendant Paula Gray by off ering her in-
ducements. Gray’s testimony is now known to be false: she was a juvenile,
mentally retarded, innocent, and wrongly convicted along with the four
others whom she incriminated in what became known as the Ford Heights
Four case.

If most of these people did not directly claim that the in for mants had
fabricated the evidence against them, how did they attack the in for mant
testimony? Rather than claiming that the in for mant fabricated the story,
which would have been impossible to prove, they asserted that they were
denied the right to counsel during an interrogation by a government in-
for mant (Massiah claims), were denied eff ective counsel at trial (Strick-
land claims), had favorable evidence concealed from them (Brady claims),
had evidence improperly introduced against them (state law claims), or
that the judge failed to properly instruct the jury.34 Three of these claims
were successful.

Darryl Hunt brought a state law claim that the trial judge had violated
evidence law rules by allowing a prosecution witness to discuss a prior
statement by a cooperating witness that contradicted her testimony at
trial, but which was not a sworn statement.35 Rolando Cruz managed to
demonstrate that the jail house in for mant who testifi ed at his second trial
had obtained leniency that was not disclosed to the defense, but though
this revelation infl uenced the court, he did not assert a separate claim
related to the in for mant testimony.36 The federal court granted Ronald
Williamson’s habeas petition in part because his lawyer was in eff ec tive
not just for failing to attack the confession and the forensics, but also for
failing to impeach the jail house in for mant with public information that
she had served as an in for mant in his case and another case, and re-
ceived a suspended sentence for her cooperation.37

Jerry Watkins also managed to win a new trial, but ultimately the
judge decided not to address whether the police and prosecutors had

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

192 Judging Innocence

concealed the fact that they had off ered the in for mant in his case a deal
and had fed him specifi c details of the crime. The federal judge dis-
cussed this evidence at length. Watkins had denied at trial that he had
ever confessed in jail, but recalled that “a man he did not know had been
in the holding cell the day of his sentencing and had asked him lots of
questions about his case.” Watkins recalled that he “just kept buggin’
and buggin’ me trying to see what he could get out of me,” and that he fi –
nally “got tired of him askin’ me so I told him” nothing more than “what
charge I was in for.”38 Watkins called as a witness a fellow inmate who
heard the in for mant describe “a scheme for getting out of jail. The plan
was to fi nd an unsolved crime, have a confederate research the newspaper
coverage,” and then convince police he heard about the crime.39

After his conviction and during his postconviction proceedings, Wat-
kins submitted a “Motion for Relief from Judgment” that the jail house
in for mant signed under oath in October 1987 in his own criminal case.
The in for mant recanted, stating:

That the State of Indiana did in fact pay (with promises) for this pe-

titioner’s testimony, and did in fact show him not only the “death-

site,” but “grizzely” [sic] pictures of the murder in order to infl ame

this petitioner’s feelings towards the defendant (Jerry Watkins),

and thus secured (for the state) a statement from this petitioner that

could, and was used to obtain a conviction.

He went on to say that he was cheated and that “the Courts [of] the
(State of Indiana) and (Jerry Watkins) were cheated as well.”

The court concluded that if this motion was true, it was “explosive.”
If the in for mant had received the “detailed information” about the mur-
der from police or a prosecutor, “that information utterly destroys the
state’s case against Watkins” and it “described an intentional corruption
of the criminal justice system.” The court noted that the prosecution
had at trial brought in newspaper articles about the case to show that the
jail house in for mant had “specifi c, correct knowledge that he could not
have derived from newspaper articles.” During the closing arguments,
the prosecutor had argued: “Did he research it in the paper as has been
intimated? I don’t think so . . . it just couldn’t happen.”40 Although this

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 193

story had a happy ending, because the court reversed Watkins’s convic-
tion and granted him a new trial, and he was then exonerated, the fed-
eral judge failed to ensure that the prosecutors who allegedly fed facts to
the in for mant were appropriately sanctioned. Although the judge con-
cluded that the “prosecution’s failure to disclose such information would
amount to suborning perjury and corrupting the judicial pro cess,” he
declined to investigate further. Because the prosecutors had also con-
cealed a range of evidence that pointed to another culprit (a Brady viola-
tion), the judge was able to reverse on that ground alone without investi-
gating whether the prosecution had engaged in any unethical behavior
concerning the jail house in for mant that should have led to sanctions.
Typically, the postconviction judge declined to investigate much less
recommend that the prosecutor suff er any consequences at all for poten-
tially grave misconduct.

To return to the questions that I began this chapter with, why did so
few exonerees challenge the evidence at their trials, whether it was eye-
witness, confession, forensic, or in for mant evidence, which we now know
was so often deeply fl awed? Why did so few who did so succeed? The
rulings in these exonerees’ cases show how the appellate and postconvic-
tion pro cess is not designed to review factual errors. Judges view their
role chiefl y to correct legal errors. Judges are reluctant to second- guess
the verdict of the jury. Judges do not off er many legal remedies for factual
defi ciencies in criminal trials. Even where there has been a clear violation
of a legal rule, judges will often insist that the error was “harmless.” As
law professor Anthony Amsterdam has put it, “there is a remarkable un-
willingness to take claims of innocence at all seriously.”41 As we will see
in the sections that follow, judges often deny relief because they think the
convict is obviously guilty— even if they agree that constitutional rights
were clearly violated.

Put yourself in the shoes of the innocent convict. If you know that
judges will not take seriously a claim that you are innocent or that the
factual evidence at your trial was fl awed, then you may focus on trying to
raise other claims that might have a better chance of succeeding. You, or
your lawyer, may know that judges are more likely to grant procedural
claims. Those claims are technical, but may be far more commonly raised
at trial and on appeal. They are also easier claims to make. In contrast, to

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

194 Judging Innocence

bring a claim of innocence or to challenge the facts at trial, one may need
to reinvestigate the crime, try to obtain additional police fi les, review the
forensic evidence, track down alibi witnesses, or hire an expert to assess
the defendant’s mental capacity or confession statements. All of that work
is diffi cult and expensive. Nor will defense lawyers likely be held ac-
countable for their failure to develop a factual record either at trial or on
appeal. All of this in turn reinforces the unwillingness of judges to take
innocence- related claims seriously. Even in cases where such claims are
raised, where neither law enforcement nor defense lawyers develop cru-
cial facts, perhaps due to underfunding, courts may be placed in a diffi –
cult position, asked to judge innocence based on a sketchy record and
years after the trial took place.42

Exonerees also ran afoul of the myriad technical obstacles that courts
have erected postconviction, including rules requiring that claims be
exhausted in precise ways and that proper objections have been as-
serted at trial. While such procedural errors may be the lawyers’ fault,
the petitioner— here, the wrongfully convicted person— usually bears
the consequences. I next turn to those complex rules surrounding ap-
peals and postconviction proceedings to show how they stymied the
eff orts of these innocent people to obtain reversals in their cases.

Stages of Criminal Review

Which courts review a conviction after a criminal trial? After a convic-
tion, the convict can raise claims at three diff erent stages of review: the
direct appeal, state postconviction proceedings, and federal habeas cor-
pus. The direct appeal occurs immediately following the conviction. Ev-
ery convict is entitled under the U.S. Constitution to fi le an appeal and to
have a lawyer (often his trial lawyer) represent him on appeal. They can
raise a host of claims related to the evidence admitted at trial, together
with state law and federal constitutional claims. First, a motion for a new
trial is fi led with the trial judge, and if the judge rejects the motion, an ap-
peal is fi led with state intermediate courts and the state supreme court. If
all of those state courts deny the appeal, which they typically do summar-
ily and without writing an opinion, then the convict has an opportunity
to ask the U.S. Supreme Court to review the case.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 195

Once the appeal has gone through this pro cess, the conviction is said
to be fi nal. Any proceedings that follow are called “postconviction,” be-
cause they occur after the conviction is fi nalized, although they are also
commonly called “state habeas.” Such review is optional, and states do
not need to make it available, although all do make some type available.
It is also called “collateral,” or secondary, because it is separate from the
appeal, and often allows the convict to raise new claims that could not be
raised before or are based on new evidence outside the trial record.

Although states do not have to provide a lawyer to the postconviction
petitioner, and most convicts proceed without a lawyer, or pro se, this is
not easy to do, since the rules for fi ling postconviction petitions can be
byzantine in their complexity. There are rules about the proper use of
the trial record, a host of detailed timing rules, notices that must be fi led,
and affi rmations that must be sworn under oath. Claims must be pre-
served in the right form and they must be “exhausted,” or fi led at the
fi rst available opportunity. Any technical error can result in the rapid
dismissal of the petition. Most petitions are summarily dismissed based
on these technicalities. And those are just the procedural rules. If a peti-
tioner representing himself from prison actually manages to master the
deadlines, correct forms, and other technical requirements, he still must
act as his own lawyer to fi gure out what substantive legal claims he might
raise. But postconviction law is one of the most diffi cult areas to master:
the claims themselves are extremely diffi cult, the standards are often un-
clear, the cases are contradictory, and the law varies among jurisdictions
and it changes over time. Most pro se petitioners simply don’t stand a
chance. Further, one of the real diffi culties in studying how judges handle
these petitions is that so few rulings are in written form. Most judges sim-
ply “rule from the bench,” dismissing the petition orally, or rule by writ-
ing a very short statement denying the claims without giving any
reasons.

State postconviction review runs again through the state courts, be-
ginning with the trial court, up to the state supreme court, and if the peti-
tion is denied, the convict may again seek certiorari from the U.S. Su-
preme Court. Once this review is complete, the petitioner can turn from
the state to the federal courts. A federal habeas corpus petition may be
fi led in a U.S. district court, with possible appeals to a court of appeals,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

196 Judging Innocence

followed by a third opportunity to seek certiorari from the U.S. Supreme
Court. Federal courts will examine petitions only where all of the claims
have already been presented to the state courts, and they are very deferen-
tial to the factual and legal fi ndings of the state courts. Every step of the
way, judges become more deferential to the fi ndings of the judges that
came before, making it more and more diffi cult to raise new claims or new
evidence of innocence.

Completing all three levels of review takes time— many years— and
very few convicts pursue all of those types of review. All of the 165 exon-
erees for whom written decisions from their appeals or postconviction
proceedings could be located pursued direct appeals. By the time they
fi led their state postconviction petitions, the number still pursuing relief
through the courts had dropped below half: only 43% of them fi led such
a petition (71 of 165 cases). And only 21% took the next step (35 of 165
cases), fi ling a federal habeas petition, although this number is actually
quite high for this stage of litigation, as generally only 1% to 2% of state
inmates fi le a federal habeas petition.43 One explanation for the high
percentage of habeas fi lings among these exonerees may be that most
received long sentences, so they had the time and the incentive to pursue
even the slowest and lengthiest challenges to their convictions. For many
convicts, by the time their case has wandered its way through the courts
on both a direct appeal and a state habeas petition, they have fi nished
serving their prison sentence and they see no need to fi le a federal habeas
petition.

The U.S. Supreme Court did not hear all of these petitioners’ claims,
but it did rule on 38 petitions fi led by these innocent people. It summar-
ily denied each of these petitions without giving reasons, except Larry
Youngblood’s. In Youngblood’s case, the Court heard oral arguments
and issued a written opinion explaining why it rejected his claim that he
should receive a new trial because law enforcement failed to properly
preserve biological evidence from the crime scene. Twelve years later,
DNA technology had improved enough that the very evidence that had
been degraded through law enforcement’s negligence was testable. The
DNA tests exonerated Youngblood.44

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 197

Reversals, Retrials, and Vacated Convictions

Some of the exonerees did meet with partial success during their appeals
and postconviction pro cess, but this success was usually short- lived.
Some of the exonerees obtained a “reversal,” which I defi ne as an order
upheld on appeal that results in the grant of a new trial and a vacating of
the conviction or convictions. Among these actually innocent individu-
als, obtaining a reversal did not usually end their ordeal. Juries wrongly
convicted them multiple times. Twenty- one exonerees of the 165 with
written decisions in their cases received reversals, for a 13% reversal rate.
But fi fteen of these exonerees were retried after the reversal of the origi-
nal conviction. All of them were convicted again. Thirteen exonerees
had two trials, and two— Rolando Cruz and Alejandro Hernandez—
each endured three trials before being freed as a result of DNA testing.
Six more exonerees had no retrials because DNA testing was conducted
and exonerated them before their scheduled retrials. (As discussed in
the last chapter, nine exonerees also had retrials for a diff erent reason—
because of hung juries.)

The reversals were concentrated among the cases of exonerees who
had been sentenced to death. Of the twenty- one exonerees who received
reversals, eight were in capital cases while thirteen were in noncapital
cases. This is no surprise. As documented in the landmark study by
James Liebman, Jeff rey Fagan, and Valerie West of all capital cases from
1973 through 1995, there are extremely high reversal rates (68%) in all
capital cases, both in state and federal postconviction review.45 The re-
versal rate just in the capital cases among exonerees was quite high, 47%
(eight of seventeen cases). If one removes the capital cases from the anal-
ysis of these exonerees’ cases, the reversal rate falls from 13% to 9%.

Judges granted most of those reversals during the direct appeal. Of
the twenty- one who received reversals, sixteen received them during the
direct appeal, one during state habeas review, and four during federal
habeas review. Changes in the law did not play an important role, since
it did not signifi cantly change during this time period.

Criminal cases in general have a much lower reversal rate than the 9%
that the noncapital cases had here. Studies have shown that approximately
1% of federal postconviction petitioners receive relief, with similar fi gures

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

198 Judging Innocence

(1% to 2%) in state courts.46 I wondered, however, whether the reversal
rate is higher for rape and murder convictions. To fi gure out whether the
9% reversal rate was typical of the kinds of serious rape and murder con-
victions at issue in the exonerees’ cases, in a prior study I constructed a
group of randomly selected rape and murder cases from the same time
period and the same states, one for each of the noncapital cases with
written decisions among the fi rst 200 exonerations. The “matched com-
parison group” reversal rate was almost identical to the exonerees’ rate:
10% instead of 9%.47 Exonerees fared no better postconviction than the
randomly selected rape and murder cases.

The similarity in reversal rates could be because serious rape and mur-
der convictions share a reversal rate of about 9%. The trials for murder
and rape may simply be more error prone than other less serious and less
complex criminal trials. A second and related explanation may be that
judges accurately detected innocence in some exonerees’ cases and so re-
versed their convictions, and that in a similar percentage of the matched
comparison group appeals, judges did the same. In the matched com-
parison group, half of the reversals involved a ruling on a claim challeng-
ing the factual evidence at trial. Thus, half of the error rate had something
to do with a perception of innocence, or relatedly, weakness of the evi-
dence of guilt, and not procedural errors. As discussed next, most of the
reversals in the innocence group were based on such factual challenges.
Though we cannot know how many in the matched comparison group
are innocent, nor how often judges granting reversals accurately cor-
rected wrongful convictions, the incidence of reversals on factual claims
in appeals of serious convictions provides cause for concern regarding
the accuracy of such criminal trials. Because DNA evidence is not avail-
able in most cases, we can never know how many postconviction petition-
ers are really innocent.

Cases Where the Innocent Received Reversals

The cases where exonerees did receive reversals deserve special exami-
nation, because in those cases courts provided relief without the benefi t
of DNA evidence. Perhaps those cases can tell us something about how
well judges assess innocence in the vast majority of criminal cases in

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 199

which they do not have easy answers from a DNA test. Although, as
I discussed earlier in this chapter, few exonerees challenged the factual
evidence at their trials, most of the select group of innocent people who
received reversals had judges grant their claims relating to the facts sup-
porting the convictions.48 Of the exonerees who obtained reversals, two-
thirds involved such factual claims (fourteen of twenty- one cases). Recall
from earlier in the chapter that one reversal related to the trial lawyer’s
failure to challenge the confession, fi ve involved challenges to eyewit-
ness identifi cations, six related to forensic analysis, and four related to
in for mant testimony (some reversals related to more than one type of
evidence).49

In three more cases, although the reversals were not related to the reli-
ability of the state’s case at trial, they were related to the defense case at
trial, and they were based on the trial judge’s suppression of evidence of
third- party guilt.50 The other four reversals were for purely procedural
claims, such as faulty jury instructions and in eff ec tive ness of counsel
unrelated to failure to suppress or challenge factual evidence.51

There is more evidence that in some exonerees’ cases judges were fo-
cused on evidence of innocence, or weak evidence of guilt. Though it was
infrequent, when judges made a statement that suggested that an exon-
eree might be innocent, typically by way of describing how the state’s
case appeared quite weak, they often reversed. A court made such a state-
ment for eight of the eigh teen reversals. This was not typically an outright
fi nding of innocence, but rather a strong ac know ledg ment of the fl imsi-
ness of the evidence of guilt adduced at trial. For example, in Ronald
Williamson’s case, his so- called dream confession was admitted at trial
despite his manifest mental illness. The federal district court judge va-
cated his conviction, citing the “weakness of the case” against him, which
relied on evidence the court of appeals later called “largely circumstantial
and hardly overwhelming.”52 Likewise, in Ronald Cotton’s case, the state
court also vacated the conviction, noting that the excluded evidence
“tended to show that the same person committed all of the similar crimes
in the neighborhood in question on that night and that the person was
someone other than the defendant.”53

Supporting this fi nding, economist and law professor J. J. Prescott
and I conducted a series of regressions to break down this data using a

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

200 Judging Innocence

far more comprehensive set of variables than those discussed in this
book.54 We found that exonerees who made claims of innocence on
appeal— asserting claims of actual innocence, claims of insuffi cient evi-
dence, or claims of newly discovered evidence— were more likely to re-
ceive reversals of their convictions. This was true even though few pre-
vailed on those par tic u lar claims. In addition, exonerees were more likely
to receive reversals if they brought claims that challenged the reliability
of trial evidence, as opposed to purely procedural claims.

Judges may have perceived innocence in some exonerees’ cases, or at
least they did not think that those cases had such open- and- shut evi-
dence of guilt. Judges have anecdotally described how they care about
the strength of the evidence when considering appeals, and some studies
have also suggested that innocence matters to judges examining appeals
and postconviction petitions.55 However, just because judges sometimes
grant appeals or petitions because they are concerned about possible in-
nocence does not mean that they do this consistently, accurately, or very
often. Judges may far more often act on a bias toward confi rming the trial
outcome. These exonerees’ cases do not give us good reasons to think
that courts are good at judging innocence. In most of the cases of actu-
ally innocent litigants, judges simply confi rmed the trial outcome. Next,
I describe how claims of innocence were made and how most failed,
with judges instead expressing belief in the convict’s guilt.

Judging Guilt and Innocence

As Chief Justice Traynor described, the task of the judge postconviction
is to sort the errors that are “plainly harmless” from those that “appear
ominously harmful.” The U.S. Supreme Court has developed a test for
determining whether errors made at a criminal trial were so egregious
that a new trial should be granted, called the “harmless error test.” Un-
der this test, which was developed in a case called Chapman v. Califor-
nia, a court may refuse to grant a criminal appellant a new trial even if
mistakes of constitutional magnitude were made in her criminal trial so
long as the State can show “beyond a reasonable doubt” that the consti-
tutional error did not contribute to the guilty verdict at trial.56 The no-
tion is that it would be wasteful to order an entire second trial if the error

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 201

was so inconsequential that it did not truly play a role at the fi rst trial.
After all, the error would be corrected at the second trial, and the person
would just be convicted again.

Harmless error analysis, where the judge replays the entire trial and
decides holistically whether an error made a real diff erence, is notori-
ously fl exible. Judges may be susceptible to confi rmation bias. That is,
they may— unconsciously or not— credit the evidence supporting the
guilty verdict and discount fl aws in the state’s case. Indeed, a harmless
error ruling may also involve a judgment that the error would not have
impacted the jury, given outweighing evidence of guilt, though the Court
has expressly cautioned against employing harmless error analysis in that
improper fashion.57 Such analysis gives judges even greater leeway to ex-
cuse constitutional violations based on their perception that the defen-
dant did it.

Many of the 250 exonerees argued that the trial judges in their cases
had made errors that had contributed to their convictions. But in most of
their cases, the appellate judges ruled that any errors made at trial were
harmless; in other words, mistakes were made, but the defendant was
clearly guilty, or regardless whether any mistakes were made, they did no
harm. Although these defendants, we now know, did not do it, and some
of these mistakes may have played a role in their wrongful convictions, for
30% of exonerees with written decisions a court relied on harmless error
in refusing to reverse a conviction (49 of 165 cases). Some of those, 14% of
the exonerees with written decisions (23 of 165 cases), had a court agree
that a claim had merit, but nevertheless deny relief due to harmless error.
Some exonerees, 18% of the exonerees with written decisions (30 of 165
cases), received rulings similar to a harmless error ruling, but under a
slightly elevated standard by which the court found that they did not suf-
fer “prejudice” as a result of the trial error. A total of 38% of these exoner-
ees with written decisions (62 of 165 cases) had either a harmless error
ruling or a “no prejudice” ruling during their appeals and postconviction
(some had both types of rulings).

There is more evidence that some judges were animated by a belief
that these innocent people were guilty, and not just that the errors raised
were of an inconsequential nature. In writing their decisions, judges
sometimes directly labeled these people as guilty convicts. In 47% of

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

202 Judging Innocence

these exonerees’ cases with written decisions, courts referred to the
likely guilt of the exoneree (78 of 165 cases), typically by describing the
reliability of the prosecution’s case.

In 10% of these exonerees’ cases, courts were so sure of guilt that they
called the evidence of guilt “overwhelming” (16 of 165 cases).58 For exam-
ple, a court denied Jeff rey Deskovic’s appeal, stating, “There was over-
whelming evidence of the defendant’s guilt in the form of the defendant’s
own multiple inculpatory statements, as corroborated by such physical
evidence as the victim’s autopsy fi ndings.”59 What judges considered to
have been “overwhelming” varied. At times they did not clearly explain
what they meant. In retrospect, many of these cases no longer look so
strong at all. However, at the time of the appeal or habeas petition, many
cases may have seemed quite strong precisely due to the contamination of
the evidence that prior chapters have described.

Taken together, in 62% of these exonerees’ cases with written deci-
sions (102 of 165 cases), judges either commented on guilt, found error to
be harmless, or found no prejudice. All of this underscores how diffi cult
it is for judges to weigh a cold trial record— but that is exactly what ap-
pellate and postconviction judges are called on to do.

In addition to judging evidence of guilt, judges may also rule on evi-
dence of innocence. Judges (typically only state judges engaging in ha-
beas review) may ask whether newly discovered evidence of innocence
would have changed the outcome at trial. In limited circumstances, fed-
eral courts also examine new evidence of innocence. Only 25% of those
with written decisions (42 of 165 cases) raised innocence- related claims,
including claims entitling the exoneree to relief based on new evidence
of innocence, or Brady claims that prosecutors concealed signifi cant evi-
dence of innocence from the defense at the time of trial. Several exoner-
ees raised more than one innocence- related claim.

Judges rejected every claim of innocence in which an exoneree argued
that he should be released based on newly discovered evidence of inno-
cence. Only 10% of these exonerees raised state law claims seeking a new
trial based on newly discovered evidence of their innocence (17 of 165
cases). None of those exonerees received relief prior to obtaining DNA
testing. Typically such claims are diffi cult to win. One must show that
there is a reasonable probability that the newly discovered evidence

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 203

would have changed the outcome at trial. Many states provide that such
claims based on new evidence of innocence can only be brought within a
short time after the trial.60

When raising such claims of innocence, 4% of these exonerees
brought claims under the U.S. Constitution, citing Herrera v. Collins
and arguing that their conviction should be vacated based on their ac-
tual innocence (6 of 165 cases). Judges rejected all of those claims. This
comes as no surprise: no petitioner has ever received relief under that
claim.61 As I will explain in the next chapter, in Herrera the Supreme
Court only said for the sake of argument that a petitioner might receive
relief if he or she could provide a “truly persuasive” demonstration of
innocence.62 While this may come as a surprise, since a right to be
freed if one is innocent seems fundamental, the Supreme Court to this
day has not decided whether a claim of innocence can in fact be made
under the U.S. Constitution. Any right to relief based on innocence re-
mains so conjectural that these six actually innocent people who raised
such claims all failed.

Four exonerees out of the forty- two who brought innocence- related
claims were granted reversals, all on Brady claims. Though Brady claims
do not provide relief expressly on the ground that the petitioner is inno-
cent, they do relate closely to innocence. Brady claims require a showing
that the prosecutor failed to disclose evidence of innocence to the de-
fense, and this evidence was signifi cant or “material” enough that there is
a reasonable probability that concealing the evidence aff ected the out-
come at trial.63 Twenty- nine exonerees raised Brady claims. As I de-
scribed in Chapter 6, one cannot know how many exonerees were con-
victed based in part on prosecutorial or police misconduct. Evidence of
innocence that was in the possession of the police or prosecution can stay
concealed even after a DNA exoneration.

One might think that DNA evidence would be particularly powerful
evidence of innocence at a criminal trial, such that no prosecutor could
be allowed to conceal it. Yet in Darryl Hunt’s case, the Fourth Circuit
Court of Appeals rejected his Brady claim, though the State’s attorney
had told his lawyer at the time of trial that the biological material was
“too degraded to be tested.” The court held that DNA technology was
new at the time and Hunt’s lawyer “had equal access to the fl uid samples,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

204 Judging Innocence

and thus he was under an in de pen dent duty to pursue testing alterna-
tives.”64 The Fourth Circuit denied Hunt relief, even though initial DNA
testing had by that time excluded him. Ten years after DNA tests ex-
cluded him, and fi ve years after that Fourth Circuit decision, a cold hit in
a DNA database identifi ed the culprit, and Hunt was fi nally exonerated,
in 2005.65

Before they had DNA tests, most exonerees did not have new evidence
of their innocence to present to a judge, and if they did, it was sometimes
not particularly strong.66 One petitioner, Willie Jackson, did have com-
pelling non- DNA evidence of innocence: the in- court confession of the
true perpetrator, his brother. The district court granted his habeas peti-
tion, emphasizing that Jackson’s brother had convincingly confessed to
the crime and that the statute of limitations had not yet expired, but the
Fifth Circuit then summarily dismissed the petition. Years later, DNA
testing inculpated the brother.67

Other exonerees did not or could not bring innocence claims, but
they did stake out their innocence in their briefs. For example, Eddie Joe
Lloyd did not make a legal claim to relief on the basis of innocence, but
in no uncertain terms, he told the court on the fi rst page of his pro se
federal habeas brief that he was “ ‘ABSOLUTELY’, ‘TOTALLY’ and ‘COM-
PLETELY INNOCENT.’ ”68

Still more exonerees, lacking any means to directly claim innocence,
did assert in large numbers claims on suffi ciency of the evidence gov-
erned by the Supreme Court’s ruling in Jackson v. Virginia.69 Of the ex-
onerees with written decisions, 42% brought a Jackson claim (69 of 165
cases). That claim is not based on allegations of new evidence of inno-
cence, but it argues there was not suffi cient evidence presented at the trial
to convict beyond a reasonable doubt. Such suffi ciency claims sometimes
highlighted unreliable factual evidence at trial, thereby providing a quasi-
factual challenge, though one based on the context of the entire trial re-
cord. In bringing a Jackson claim, a petitioner must show that, viewing
the evidence in the light most favorable to the prosecution, no rational
juror could fi nd beyond a reasonable doubt that the prosecution proved
the essential elements of the crime. Perhaps due to this stringent stan-
dard, no exoneree who brought a Jackson claim received a reversal that
was upheld on appeal.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 205

It remains very diffi cult to obtain relief on a claim of innocence. This
explains why few of these actually innocent people raised such claims
and why almost none who did succeeded. Although the next chapter will
describe how states have introduced new statutes making it easier to
claim innocence postconviction, in many respects little has changed since
these exonerees brought their challenges. Indeed, in one way, these exon-
erees had an easier time raising their claims. The enactment of the Anti-
terrorism and Eff ective Death Penalty Act (AEDPA) in 1996 imposed a
litany of rules that restrict federal habeas corpus review, including strict
time limits, limits on fi ling more than one habeas petition, and require-
ments that federal judges defer to state court fi ndings, even to the extent
that judges must deny relief in some cases where they fi nd that the peti-
tioner’s constitutional rights were in fact violated.70 The AEDPA did not
signifi cantly impact these exonerees’ habeas petitions, since almost all
exonerees fi led their petitions before it took eff ect. The restricted federal
substantive review that left some of these exonerees without a meaningful
remedy was far more generous than the narrower habeas review that fed-
eral judges would provide today.

In eff ec tive Assistance of Counsel

In eff ec tive assistance of counsel is one of the most frequently raised
claims during postconviction proceedings,71 and 32% of these DNA ex-
onerees (52 of 165 cases) asserted that their trial was unfair because their
defense lawyer was inadequate. The Supreme Court ruled in Strickland
v. Washington that indigent defendants are constitutionally entitled to
minimally eff ective repre sen ta tion. This repre sen ta tion, however, need
only fall “within the wide range of reasonable professional assistance.”
To win a Strickland claim and have a guilty verdict overturned on
grounds of in eff ec tive assistance of counsel, a convict must show more.
The attorney’s in eff ec tive ness must have materially prejudiced the out-
come at trial, so that “there is a reasonable probability that, but for coun-
sel’s unprofessional errors, the result of the proceeding would have been
diff erent.”72

Commentators have derided the Strickland standard as nothing more
than a “foggy mirror” test— if a mirror held under the defense lawyer’s

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

206 Judging Innocence

nose during trial would fog up, indicating the lawyer was at least alive
and breathing, the lawyer has provided adequate assistance.73 Rulings
evaluating what “reasonable professional assistance” means vary widely.
Although higher standards have been imposed in certain death penalty
cases, notorious decisions have denied relief despite terribly shoddy
repre sen ta tion, including cases in which lawyers literally fell asleep or
presented no meaningful case at trial.74

Four of the exonerees earned reversals due to grossly in eff ec tive
repre sen ta tion of trial counsel. Ronald Williamson, the only exoneree
to assert a claim related to a false confession, had his conviction vacated
due to a claim that trial counsel failed to show that he was mentally in-
competent to stand trial and that another man confessed to the crime,
among other failures.75 The other three, Paula Gray, William Rainge,
and Dennis Williams, were all represented by the same lawyer, who
was later disbarred for unrelated reasons. Their lawyer failed to even
move to suppress central physical evidence, such as hair evidence, and
Gray argued that the joint repre sen ta tion of all three created confl icts of
interest.76

In every other case in which exonerees claimed in eff ec tive assistance,
judges ruled that they had failed to satisfy the Strickland standard. That
does not mean, however, that they received adequate assistance of counsel.
After all, judges may conclude that any missteps by counsel did not
“prejudice” the outcome, because the prosecution case was more than
suffi cient to convict. Recall from Chapter 6 how Earl Washington Jr.’s
trial lawyer had failed to meaningfully represent him at trial, including
by failing to even raise the fact that forensic evidence excluded him.
Washington’s new postconviction team of lawyers fi led state and federal
habeas petitions alleging that his trial lawyer was in eff ec tive. His trial
lawyer even submitted an affi davit admitting that he simply did not un-
derstand the scientifi c evidence and if he had, he would have litigated
the issue. A series of federal judges ultimately rejected his claim, empha-
sizing that Washington had confessed “as one who was familiar with the
minutiae” of the crime.77

Other exonerees raised seemingly strong Strickland claims that
judges rejected. For example, in at least twelve cases, defense counsel
failed to ask for DNA testing available at the time of trial that could have

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 207

exculpated the defendant— yet only four raised postconviction claims
on this subject. One would think that those claims would be particu-
larly clear. The DNA testing could have potentially proven their inno-
cence, as it later did. Yet even among those who raised the issue, only
one of the four, Anthony Hicks, received a reversal, and only after the
court knew he was innocent because the DNA testing had just been
conducted.78

Trial rec ords suggest that although other exonerees did not assert
Strickland claims postconviction, the eff ectiveness of their trial counsel
was limited. Only eleven exonerees raised Strickland claims regarding
failure of the defense lawyer to adequately challenge eyewitness identifi –
cations, and none of them had any success. Only fourteen exonerees
made Strickland claims regarding failures of defense lawyers to challenge
forensic evidence, yet invalid forensic evidence was presented in ninety-
three trials, and more often than not, the defense lawyers failed to bring
the errors to light. Only one exoneree, Ronald Williamson again, as-
serted that his defense lawyer failed to properly challenge a confession.
This may be because almost all of the defense lawyers did in fact move to
suppress the confession. Yet defense lawyers were not necessarily very
eff ective at challenging the confession; as discussed in Chapter 2, they
could have far more often introduced experts to show that the defendant
confessed due to mental disability or youth, and that the defendant was
suggestible.

Thus, just as exonerees often did not challenge the trial evidence
during their appeals and postconviction, they also did not often use
Strickland claims to bring to light the failures of their defense lawyers
to challenge the evidence at the trial. When they did, judges usually
denied relief, including by fi nding any error to be nonprejudicial or
harmless.

Prosecutorial Misconduct

Not only did exonerees raise claims challenging their trial lawyers as
inadequate, but even more raised claims that prosecutors played a role in
their conviction. The U.S. Supreme Court has described how a prose-
cutor is a public servant obligated to ensure that “justice shall be done,”

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

208 Judging Innocence

and not to simply secure a conviction at any cost.79 It is especially impor-
tant that prosecutors play by the rules, because they play a dominant
role in criminal cases. As described in Chapter 6, the prosecution pre-
sented most of the evidence and called most of the witnesses at these
trials. Prosecutors had access to all of the evidence that police gathered.
As a result, prosecutors also played a crucial part in these wrongful con-
victions. Exonerees claimed as much on appeal and postconviction, with
some success.

Indeed, of the twenty- one exonerees who obtained a reversal before
getting DNA testing, ten did so in part based on a claim of prosecutorial
misconduct. A range of claims relate to the conduct of prosecutors, and
47% of the exonerees with written decisions (77 of 165 cases) brought such
claims. Of those, forty- two involved claims of unduly prejudicial argu-
ments, which as discussed in Chapter 6, are very diffi cult to win, because
of the wide latitude courts give lawyers to make arguments, although three
exonerees did obtain reversals on those grounds. As discussed, twenty-
nine exonerees also raised Brady v. Mary land claims, and four obtained
reversals on the ground that the prosecution concealed exculpatory evi-
dence from them.

The U.S. Supreme Court has long held that it is “inconsistent with the
rudimentary demands of justice” for a prosecutor to knowingly put on
perjured testimony.80 Just as it is hard for a convict to fi nd out that prose-
cutors concealed evidence of innocence, a convict may have no way to
prove that prosecutors knew that their witness was lying. However, four
exonerees brought such claims. and one exoneree, Verneal Jimerson, ob-
tained a reversal on that ground.

Another twenty- one exonerees brought claims that prosecutors im-
properly selected or rejected jurors. As discussed in Chapter 6, none
prevailed on such claims, which are very diffi cult to win. For example,
if the prosecutor moved to strike minority jurors from the pool, a defen-
dant can claim that this was discriminatory, but if the prosecutor can
off er any benign reason for objecting to the juror, the claim will typi-
cally be denied.81 Kennedy Brewer, whose story I told at the beginning
of this chapter, challenged the way that the prosecutor struck six black
potential jurors. Brewer was black and alleged that this pattern was
discriminatory. However, the appellate court denied Brewer relief,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 209

citing to the prosecutor’s range of reasons to explain each strike; one
juror had a “penchant for soap operas,” another had prior run- ins with
the law, and another read publications that often “railed” against the
death penalty.82

A few exonerees also brought claims relating to other types of prosecu-
torial misconduct during the trial. Of those, Ray Krone, whose case was
discussed in Chapter 4, obtained a reversal of his conviction, because
prosecutors had failed to provide the defense until just days before the
trial a provocative and macabre video of the bite mark examiner holding
molds of Krone’s teeth to the victim’s body. Rolando Cruz obtained a re-
versal on the grounds that prosecutors improperly questioned their own
witness.83

Thus, not only did many of the exonerees who successfully obtained
reversals before they obtained DNA testing make claims challenging the
trial evidence, but many of those exonerees also challenged the conduct
of prosecutors at their trials. Ten exonerees obtained reversals based on
prosecutorial misconduct. Of course, the vast majority of these exoner-
ees failed in their eff orts to claim prosecutorial misconduct. Courts were
also unwilling to call out prosecutors on misconduct or ethical lapses,
and nor did courts recommend any sanction. We saw earlier in this
chapter how in Jerry Watkins’s case, the federal judge was unwilling to
even look into whether the prosecutor had fabricated evidence, though
he conceded such conduct “would amount to suborning perjury and
corrupting the judicial pro cess.” The prevalence of Brady violations that
came to light years later, allegations of prejudicial prosecution argu-
ments, and other claims of improper conduct in these wrongful convic-
tions, all suggest that we should take the problem of prosecutorial mis-
conduct far more seriously.

Dissents

Did some judges disagree with these rulings denying relief to the inno-
cent, perhaps seeing something that the other judges did not see? Dis-
sents are noteworthy, because they are not common. They show that a
judge felt strongly enough that the decision the other judges were reach-
ing was wrong, that she needed to speak her mind and explain why she

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

210 Judging Innocence

voted the other way. Judges authored dissents in 21% of these exonerees’
cases (34 of 165 cases). In twenty of those dissents, judges disagreed with
their fellow judges’ rulings denying exonerees relief. Many of the dis-
senting judges also commented on possible innocence, usually by noting
the weakness of the prosecution’s case. For example, in Larry Young-
blood’s case, Justice Harry Blackmun wrote a prescient dissent:

Because semen is a body fl uid which could have been tested by

available methods to show an immutable characteristic of the assail-

ant, there was a genuine possibility that the results of such testing

might have exonerated respondent. The only evidence implicating

respondent was the testimony of the victim.84

Similarly, in Lesly Jean’s case, a dissenter wrote that “Unlike the major-
ity, I believe the issue of defendant’s guilt is close.”85 In Bruce Good-
man’s case a dissenter asserted that “[t]here is no probative evidence at all
that the defendant was at the scene of the crime.”86

The reverse also occurred: seven exonerees received dissents from de-
cisions in their favor. Some dissenting judges commented on their belief
in the exonerees’ guilt. For example, in Rolando Cruz’s case, a dissenting
judge stated, “After two verdicts of guilty and 11 years after the murder, the
defendant now gets a third roll of the dice. The pressure on the prosecutor
to negotiate a plea . . . may be irresistible. In any event, justice is the
loser.”87 Now that postconviction DNA testing was conducted, we know
that justice did not lose.

Substantive Errors and Criminal Review

An innocent person who is convicted does not face an easy road to free-
dom. Judges do not encourage convicts to assert their factual innocence
on appeal or postconviction. Judges face profound diffi culties when
judging innocence postconviction. This is understandable. Judges can-
not easily review a “cold” trial record years after the trial took place.
Moreover, the hesitance of judges may be warranted. Even if judges hold
a hearing to revisit factual evidence, which they rarely do, they fi nd “ero-
sion of memory and dispersion of witnesses.”88 Years after a trial, it may

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Judging Innocence � 211

be too late to fi nd out whether evidence was contaminated or unreliable
early on in a criminal investigation.

How many innocent people are there whose cases never make the news
because the system works when appellate judges reverse their convictions
and there is no need for a DNA exoneration? We cannot answer that
question. However, the experiences of these innocent people do not pro-
vide good reasons to be optimistic that judges can eff ectively detect and
provide relief to innocent convicts. These exonerees often did not raise
their innocence or factual claims. As we saw, very few who did make such
claims succeeded. Yet the select few who did obtain reversals may have
done so because they did succeed in signaling their innocence to judges
who were concerned that something had gone wrong. Far more often,
judges found any errors to be harmless, such as by citing the strength of
the prosecution case or the defendant’s likely guilt. All of this suggests
that judges can sometimes identify evidence of innocence when it is put
before them, but our current system makes it very hard for such claims to
succeed.

What can be done to improve the accuracy of postconviction review?
If more resources are provided, they may be most useful during the di-
rect appeals, when less time has passed, when convicts have lawyers,
and when suffi ciency of the evidence claims can be raised. Most exoner-
ees who did receive relief did so during the direct appeal. Yet given how
long it took for evidence of innocence to surface in these cases, we should
also examine ways to enhance factual review during postconviction pro-
ceedings. Some states have found ways to make it easier to investigate
claims of innocence and remedy wrongful convictions. In the remainder
of the book, I will discuss several ways to do that, such as by passing
new postconviction statutes. New avenues to assert innocence, however,
are not enough, unless convicts have the resources to investigate facts
and uncover potential evidence of innocence.

The main focus of this book is on reforming criminal investigations to
prevent wrongful convictions in the fi rst instance, and not on the post-
conviction pro cess, which serves as only a backstop. Even an improved
postconviction system still must overcome the deep institutional reluc-
tance of judges to reverse criminal trial verdicts. Indeed, not only did
these exonerees confront a series of postconviction barriers before they

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

212 Judging Innocence

obtained DNA testing, but like Kennedy Brewer, whose case began this
chapter, they also faced the reluctance of prosecutors and judges to allow
DNA testing and to exonerate even after the DNA testing demonstrated
a person’s innocence. The next chapter turns to those obstacles at the
end of the road to exoneration.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

213

A string of vicious murders had been committed in the Fort Lau-
derdale area. Frank Lee Smith became a suspect in the rape and

murder of an eight- year- old when residents thought he resembled a com-
posite sketch. He was schizophrenic and had been twice convicted of
manslaughter when he was a teenager. At trial, the evidence against him
chiefl y consisted of testimony by an eyewitness. She testifi ed that she saw
him jumping out of the window of the victim’s house at around the time of
the murder. There was “no doubt” in her mind that Smith was the per-
son she saw. The detective also reported that he duped Smith into mak-
ing an admission of his guilt. The detective told Smith that the eyewit-
ness saw him commit the crime. Smith reportedly responded, “No way
that kid could have seen me. It was too dark.” The detective claimed that
only a guilty person would know that the eyewitness was a child and that
the crime occurred at night in a dark room. Smith denied ever making
such a remark. In 1986, the jury convicted Smith and sentenced him to
death.1

More than a de cade passed as Smith pursued appeals and habeas re-
lief, with no success. In 1998, Smith’s lawyers obtained a stay of execu-
tion and began seeking DNA testing to bolster their case for his inno-
cence. The State’s star eyewitness had since recanted her account. She

c h a p t e r 8

Exoneration

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

214 Exoneration

claimed that prosecutors had pressured her to identify Smith. She said
the prosecutor told her that Smith was “dangerous,” and “not to worry
about my testimony because the man would be locked up and electro-
cuted the following May . . . I was just feeling so pressured.”2 She said
that the man she saw close to the time of the murder was actually another
man named Eddie Lee Mosley.3

The prosecutor refused to agree to the DNA testing and successfully
opposed motions seeking DNA testing for years. Indeed, a detective
now claimed, contrary to his earlier testimony in the case, that he had
already showed the eyewitness a photo of Mosley and that she had not
identifi ed him. The prosecutor’s opposition to DNA testing could pre-
vent it from being done. At the time, years after DNA testing had become
available, very few states provided a right for an inmate to gain access to
DNA evidence postconviction. The evidence was in the custody of the
State, and after a conviction, the State could deny access to the evidence
or destroy it— even if the evidence could prove innocence.

Unable to gain access to the DNA testing he sought, Frank Lee Smith
languished in prison. Meanwhile, two Fort Lauderdale detectives along
with Smith’s lawyers and an investigator had been investigating Eddie
Lee Mosley. They eventually connected him to sixty rapes and seventeen
hom i cides in the area. They obtained DNA testing that linked Mosley to
those crimes, including eight rape- murders that Jerry Townsend, another
innocent man, had been convicted of. In Chapter 2, I mentioned that
Townsend was mentally ill and had falsely confessed to every unsolved
murder the police asked him about. With DNA testing implicating a se-
rial murderer and rapist in the area, law enforcement fi nally consented to
test the evidence from Frank Lee Smith’s case. DNA testing ultimately
confi rmed Mosley’s guilt and proved Smith’s innocence, but the tests
were not completed until December 2000.

That was too late. Ten months earlier, Smith had died of cancer on
Florida’s death row.4 In pain in the prison hospital, Frank Lee Smith pro-
tested his innocence on his deathbed. He died waiting for the DNA test-
ing that he had fought for so many years to obtain. His conviction was
vacated posthumously. Jerry Frank Townsend was also exonerated and
released, having spent twenty- two years in prison. During those years,
Mosley continued his “reign of terror,” committing additional murders.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 215

Innocence Project found er Barry Scheck called Mosley a “one man crime
wave.”5

In reaction to Smith’s case, Florida passed a statute in 2001 entitling a
petitioner to obtain DNA testing and, if the results are exculpatory, a re-
versal of his or her conviction. Like many states that have enacted new
innocence claims in recent years, however, Florida included several re-
strictions limiting access to testing, such as a requirement that a peti-
tioner satisfy a preliminary showing of innocence to receive DNA testing.
For example, the statute barred access to DNA testing to persons who
pleaded guilty. Jerry Townsend never would have been able to get DNA
testing under the statute, as he had pleaded guilty to several crimes to
avoid execution.6

Even after passage of that statute, judges in Florida continued to deny
DNA testing to people who requested it— including innocent people.
James Bain spent more time in prison than any other exoneree—thirty- fi ve
years. For eight of those years he sought DNA testing. He began request-
ing DNA testing in 2001, when Florida passed its statute. He fi led fi ve
separate handwritten requests from his prison cell, and he was denied
each time. He appealed each time, and judges called his requests untimely
or improper. Prosecutors argued that there was no “reasonable probabil-
ity” that the DNA tests would make any diff erence in his case. He needed
a lawyer. When lawyers from the Florida Innocence Project took his case,
they fi nally persuaded a judge to order DNA testing in 2009. With the
DNA results excluding him, the prosecutor admitted, “He’s just not con-
nected to this par tic u lar incident,” and agreed the conviction should be
vacated.7

The Long Road to Exoneration

It is diffi cult to imagine what it must be like to spend so much time be-
hind bars for a crime one did not commit. As I have mentioned, these
innocent people spent an average of thirteen years in prison, and it took
even longer, an average of fi fteen years from the conviction for them to
obtain exonerations.8 By the time they were exonerated, some of these
people were no longer in prison, since they had been released on parole
or on bail pending a new trial, but they continued to wait to obtain the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

216 Exoneration

exoneration. Why did it take so long to exonerate these innocent
people?

When I began looking at these exonerees’ cases, I expected to fi nd that
appeals and postconviction proceedings would take many years. After
all, by defi nition, these were people convicted before DNA testing was
available, and they would have to wait for it to be developed. I also knew
that even when DNA testing became possible, there was often initial op-
position to it. Yet until I analyzed these exonerees’ cases over time, I did
not appreciate just how substantial those delays were, even well into the
time period that DNA testing was readily available. I had no idea just how
resistant judges and prosecutors would be to the idea that the system
might have made a mistake.

Indeed, most of the delays these exonerees experienced occurred
while they tried to obtain DNA testing and then an exoneration. Con-
sider the average exoneree’s timeline from conviction to exoneration.
The average year of conviction was 1987, two years before the fi rst DNA
exonerations occurred. The lengthy appeals and postconviction pro cess
described in Chapter 7 took on average more than six years and was
complete, on average, in 1993. Yet that was less than half of the time
these innocent people waited to obtain an exoneration. The fi rst DNA
testing that excluded these innocent people was not conducted until, on
average, 2000— more than a de cade after the fi rst DNA exoneration.
Once those DNA test results provided powerful evidence of innocence,
these people still faced obstacles; on average, it took more than a year to
obtain the exoneration. The average year of exoneration was 2002, long
after DNA testing was in wide use and innocence projects had become
well established.

The story of Frank Lee Smith, and Florida’s halting reaction to his
case, helps to explain why it took so long for these people to obtain their
exonerations. If prosecutors opposed DNA tests, judges would often not
order the tests, and even after states enacted statutes in response to ex-
onerations, prosecutors and judges still let cases fall through the cracks.
Ronald Jones’s case, discussed in the opening pages of this book, pro-
vides another example. Recall that Jones was convicted after police in-
terrogated him and he falsely confessed to a rape- murder he did not
commit. After DNA testing became available, Jones asked to be tested,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 217

and the judge refused, asking, “What issue could possibly be resolved
through DNA testing?” But Jones persisted, and the Illinois Supreme
Court ordered the testing done. Incredibly, even once the DNA testing
showed that he was not the perpetrator, he waited for two more years for
prosecutors to decide not to retry him and one more year for a pardon.
Our system, despite the availability of DNA, still places a series of road-
blocks before prisoners seeking to prove their innocence. This chapter
describes the long and legally challenging path to exoneration.

Evolving DNA Technology

The fascinating story of the invention of the DNA technology that exon-
erated these 250 people explains in part why these par tic u lar wrongful
convictions came to light, and also when they came to light. One might
expect that once DNA testing was invented, innocent people would be
able to quickly take advantage of the new technology and win their free-
dom. However, we will see that DNA technology is most useful only in
certain types of cases, particularly rape cases. In addition, the technol-
ogy improved quite dramatically during the de cade after the fi rst DNA
exoneration occurred, in 1989. More powerful DNA testing techniques
took some time to be adopted by crime labs around the country, and the
steady advance of DNA technology partially explains the gradual rise in
the numbers of DNA exonerations.

The case of Earl Washington Jr., whose trial was described in Chap-
ter 6, spans the entire period of time during which DNA testing fi rst be-
came available and then advanced to include far more powerful types of
testing. When Washington was convicted in 1984, based on his seem-
ingly powerful confession, he was sentenced to death and would soon
face an execution date. Yet at the time, DNA testing was in its infancy.
Three- fourths of the exonerees were convicted before DNA testing was
available, and most were also convicted in the 1980s. In the mid- 1980s,
scientists developed the technology underlying DNA testing. Research-
ers discovered that portions of the code in all human nuclear DNA did
not perform any known ge ne tic function. This so- called junk DNA was
highly variable. Scientists developed technology to splice portions of
DNA and then replicate it exponentially using an enzyme chain reaction.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

218 Exoneration

This allowed them to determine with much more accuracy than ever
before the likelihood that biological evidence could come from a par tic-
u lar person. Unlike serology, which tended to show that a person was
included in a large percentage of the population that shared blood types,
DNA testing became powerful enough to show that one person in mil-
lions or billions would randomly match a ge ne tic profi le.

DNA testing turned out to be of par tic u lar use in rape cases. As Chap-
ter 4 discussed, when a woman seeks medical treatment after a rape,
hospitals prepare a rape kit that contains a swab from the victim’s vagina.
Usually, this swab will contain a mixture of skin cells from the victim and
perhaps also sperm cells from a rapist. But often, forensic analysts could
not tell whether the typically larger amount of material from the victim
was “masking” any information from sperm. Testing of the sample often
revealed only the victim’s type, and if nothing other than the victim’s type
was found, then the attacker could have been any male, with any blood
type.

But forensic scientist Edward Blake made an elegant discovery in the
mid- 1980s that changed criminal law forever. Blake developed a tech-
nique that solved the problem of masking, called “diff erential extraction.”
The pro cess took advantage of the fact that sperm cells have thicker cell
walls than skin cells. A mild detergent washed the mixed fl uid and broke
open the skin cells, releasing their ge ne tic material. During this fi rst
wash, sperm cells did not break open, because they have thicker cell
walls. A stronger detergent was then used to break open the sperm cells,
allowing the analyst to test just the ge ne tic profi le of the sperm. This
technique could rule out any danger that the evidence could have been
accidentally contaminated at the crime scene or in the laboratory. Since
the sperm cells were isolated, unless there was a reason why someone else
could have deposited sperm on the sample, the ge ne tic profi le from sperm
had to come from the rapist.

This biology explains why DNA testing can be such powerful evi-
dence of either guilt or innocence in stranger rape cases. This also ex-
plains why DNA has been used far less often in other kinds of far more
common criminal cases, such as assaults, robberies, drug possession
cases, or misdemeanor cases that receive no trial. In such cases it may be
harder to rule out contamination of the evidence, since other people

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 219

aside from the victim and the culprit may have been present at the scene
or had a reason to be in contact with the evidence. Nor is DNA testing as
useful in cases that do not usually make use of forensic evidence, such as
most rape cases, which involve disputes about consent between acquain-
tances and not the identity of the rapist.9

The fi rst DNA exonerations in 1989 sparked a criminal justice revolu-
tion, but it took some time for the full impact of DNA testing to be felt.
One reason was that early DNA tests were not very discerning. The fi rst-
generation DQ Alpha DNA testing examined only eight ge ne tic markers
shared by large populations.10 Typical results would show only that one
in, say, 6% of the population shared a profi le, an improvement over blood-
typing, but still not very probative.

By the late 1980s, Earl Washington Jr. had come within nine days of his
execution. The only way to forestall his execution was to fi le a state ha-
beas petition. But the State did not provide inmates, even death row in-
mates, with lawyers to fi le state habeas petitions. Another inmate who
was a skilled jail house lawyer, fi nally fi led a class action to draw attention
to the plight of inmates, like Washington, who could not possibly be ex-
pected to fi le habeas petitions on their own. Washington, after all, could
not accurately write his name, much less draft a legal brief. This bold
move caught the attention of lawyers at a New York law fi rm. A team of
attorneys took on Washington’s case pro bono. They fi led a habeas peti-
tion arguing that his confession was faulty, forensic evidence excluded
him, and his trial lawyer was in eff ec tive. Those petitions were denied, by
state and then by federal judges.

By 1993, DNA testing was Washington’s last hope. After spending nine
years on death row, with all of his appeals and habeas proceedings spent,
Earl Washington Jr. obtained DNA testing using the DQ Alpha tech-
nique. The results did not include him and cast “substantial” doubt on
his guilt, but according to the Virginia Department of Forensic Science
analyst, did not conclusively exclude him either. It was hard to imagine
how the test results, which did not match Washington, had failed to es-
tablish his innocence. During the fi nal hours of his term, Governor
Douglas Wilder commuted Earl Washington Jr.’s death sentence, sparing
him from execution. Yet despite the DNA tests, although Washington left
death row, he remained in prison to serve a life sentence.11

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

220 Exoneration

As Washington languished in prison, scientists developed still more
remarkable ge ne tic tools. By the early 1990s, Restriction Fragment
Length Polymorphism (RFLP) testing was developed to identify frag-
ments of DNA by running an electric current through biological mate-
rial dissolved in a gel medium. This RFLP testing was far more discern-
ing than DQ Alpha testing. It could generate the types of random match
probabilities that we are now familiar with, such as fi gures that one ran-
dom person in millions or billions or even trillions could be expected to
share a profi le. However, RFLP testing required a very large quantity of
nondegraded ge ne tic material, and interpretation of the results was po-
tentially subjective.12

The greatest leap forward in DNA testing technology occurred by
the mid- 1990s, with the development of modern and much more power-
ful Short Tandem Repeat (STR) DNA testing. STR testing could be
used with a polymerase chain reaction (PCR) to amplify and test mi-
nuscule samples, even just a few cells. New capillary electrophoresis
technology permitted fast and largely computerized analysis of ge ne tic
samples.13

In the late 1990s, two other forms of testing, mitochondrial DNA
(mtDNA) testing and Y-STR DNA testing were developed, further ex-
panding the kinds of cases where DNA testing could be useful.14 Both
forms of DNA testing have been used to exonerate innocent people.
MtDNA testing can be conducted on hairs that have no cells with nu-
clear DNA attached to them, on which no other form of DNA testing
could be conducted. However, mitochondrial DNA is inherited matri-
lineally, and is shared by relatives, so the results are less conclusive than
other kinds of DNA (two cousins, related on their mothers’ side of the
family, for example, could both be included as possible sources of a
sample). Y-STR testing is conducted on the Y chromosome, which is
inherited patrilineally. Y-STR testing can be useful if the evidence is
degraded and other forms of DNA testing cannot be conducted. Since
male relatives share the same profi le, two cousins related on their fa-
thers’ side would share the same Y-STR profi le. Y-STR testing was used
to show that Sally Hemings’s male descendants shared a ge ne tic profi le
with male descendants of Thomas Jeff erson.15 Thus, over more than a
de cade, DNA testing evolved into a powerful set of techniques that can
test minute particles of evidence from a crime scene.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 221

One fi nal technological development came with the rise of modern
STR DNA testing— the creation of state and then national databases, or
the Combined DNA Index System (CODIS), which now contains mil-
lions of ge ne tic profi les.16 Any unknown DNA profi le detected in a case
can now be entered into the national database system. The databases
contain over three million profi les and their size continues to expand,
particularly since all fi fty states and the federal government enacted laws
permitting collection of DNA from those convicted of serious felonies,17
and now even arrestees and detainees can have their profi les entered into
the database.18

On Monday mornings at precisely 9:00 a.m., the CODIS database
conducts an automatic search to compare all of the millions of profi les
against those from unsolved crimes.19 As new profi les from convicts and
arrestees and others are entered into the system week by week, new data-
base matches, or “cold hits,” sometimes solve those cases. Those com-
puterized database searches also help to solve the cases that these inno-
cent people were convicted of. In 45% of the DNA exonerations to date
(112 of 250 cases), postconviction DNA testing identifi ed the perpetrator,
usually through a cold hit in the CODIS database.20

These newer technologies fi nally freed Earl Washington Jr. in 2001.
Investigative journalists discovered that a second DNA test excluding
Washington had never been disclosed to his lawyers back in 1993 when
he was denied full clemency. The Virginia laboratory was now under
pressure to test the evidence again using more sophisticated STR testing.
They did so, and again “botched” the test results, claiming that they did
not exonerate Washington. A “sharply critical” in de pen dent audit later
concluded that these results were false and were due to po liti cal “pres-
sures” not to exonerate Washington.21

Another miracle intervened. Washington’s steadfast team of lawyers
discovered that a second set of slides taken from the victim had been
saved from the hospital where she had died. Those slides had never been
tested. An outside laboratory tested them using STR DNA testing. The
results did not just conclusively exclude Washington. As the Virginia lab
then confi rmed, the profi le matched in a cold hit another man who was
serving a life sentence for an unrelated rape. That man eventually pleaded
guilty to the crime. Earl Washington Jr. was fi nally exonerated and freed
from prison.22

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

222 Exoneration

The combination of the development of STR DNA testing and DNA
databases led to an acceleration in the numbers of DNA exonerations,
from a trickle in the early 1990s, to ten to twenty a year by the late 1990s.
Only thirteen people had been exonerated by the end of 1993, when
more- advanced STR DNA testing was fi rst available. Only thirty- nine
individuals had been exonerated by the end of 1997. In 2007, Jerry Miller
became the 200th person exonerated by DNA testing. Just three years
later, in February 2010, Freddie Peacock became the 250th person exon-
erated by DNA, and the last person included in the group I examine in
this book.

Do advances in the use of DNA portend the end of DNA exonerations?
You might expect DNA exonerations to drop off once DNA testing be-
came common before trial. For now, the number of DNA exonerations
per year continues to increase, and fully one- quarter of all DNA exonera-
tions still occur for cases with recent convictions from the 1990s and even
the new century. Why is it that there are trials, taking place when DNA
testing is routine, where innocent people are still wrongly convicted? In
some cases, prosecutors or defense attorneys failed to request that DNA
testing be conducted. In others, DNA technology has advanced since the
trial to make new types of testing possible. In still more cases, the State
concealed the existence of biological evidence, or worse, made testing er-
rors. Although DNA testing is now routinely done during criminal inves-
tigations, technology is only as perfect as the people using it, and as long
as we have human error we will continue to have DNA exonerations in
the future.23

The Supreme Court on the Sidelines

The improvement of DNA technology does not explain why Frank Lee
Smith could not get DNA testing when it was available, or why so many
other exonerees initially had requests for DNA testing denied. Why
couldn’t they simply ask a judge to provide them with DNA testing, over
the objections of prosecutors and police, and then be exonerated once
the tests proved they were innocent? The answer is that legal change can
be much slower than technological change. During the same time period
that DNA technology advanced and these 250 exonerees began to obtain

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 223

DNA tests, the Supreme Court repeatedly avoided addressing whether
there should be a claim of innocence— a right under the Constitution to
obtain a new trial on the grounds that one is innocent.

In a provocative 1970 article, Judge Henry Friendly asked why inno-
cence is “irrelevant” to federal habeas corpus review, rather than being
the main preoccupation of judges.24 A host of constitutional rights can be
asserted in federal courts, such as the Brady and Strickland claims that so
many exonerees brought. The one claim that no convict can easily bring
is a claim that he is innocent and should be freed for that reason alone. In
the de cades since Judge Friendly posed the question, evidence that con-
victs may in fact be actually innocent has mounted. The Supreme Court
had every opportunity to recognize a constitutional claim of actual inno-
cence but repeatedly refused to do so. Three Supreme Court decisions,
Herrera v. Collins, House v. Bell, and Osborne v. District Attorney’s Offi ce,
illustrate the Court’s indiff erent approach.

In Herrera, the Court asked whether a person can avoid execution
based on evidence of innocence. In its 1993 decision, the Court assumed
for the sake of argument the existence of a constitutional right to chal-
lenge a conviction based on “truly persuasive” evidence of “actual inno-
cence.”25 The Court never said what a “truly persuasive” showing con-
sists in, holding that what ever it was, Herrera did not satisfy it. Over the
years, lower courts similarly assumed the existence of some undefi ned
right to claim innocence in a death penalty case. Several people later
exonerated by DNA testing tried to assert such claims, but with no suc-
cess. After all, the right was just hypothetical.

In the second case, the 2006 decision in House v. Bell, the Court had
another chance to actually recognize an innocence claim.26 Paul House
asked the Court to consider new evidence of innocence, including DNA
test results. DNA results excluded House as the rapist, but the results also
included the victim’s husband and so they did not necessarily point to the
rapist and murderer’s identity. The Court ruled that this evidence of in-
nocence was powerful enough that “more likely than not,” a new jury
would not convict him.27 But the Court did not order House released.
Why? Because the Court still would not recognize a right to claim inno-
cence. Instead, the Court ruled that this evidence justifi ed the district
judge’s reconsideration of House’s otherwise defaulted claim that his trial

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

224 Exoneration

attorney was inadequate. The ruling did give his lawyers time to pursue
new DNA tests. By May 2009, a new round of DNA tests on multiple
pieces of evidence, including hair and materials from the victim’s fi nger-
nails, excluded not only House but also the victim’s husband, and pointed
to an unknown culprit.28 House was then exonerated.29

The Court’s third chance to recognize an innocence claim arrived in
the 2009 case of District Attorney’s Offi ce v. Osborne.30 A convict, Wil-
liam Osborne, sought DNA testing to prove his innocence. The State
conceded that the results could “conclusively establish Osborne’s inno-
cence.”31 Osborne would pay for the testing. The State, however, was
Alaska, one of only three states at the time that had no statute ensuring
access to postconviction DNA testing. The State adamantly opposed
Osborne’s request. The Court ruled that Osborne had “a liberty interest
in demonstrating his innocence with new evidence under state law,”32 but
still denied him access to the DNA, for unconvincing technical reasons
specifi c to his use of state procedure. The Court also revisited Herrera,
and again “assume[d] without deciding,”33 that a claim of innocence
exists.

The U.S. Supreme Court has had three major opportunities to explic-
itly recognize an innocence claim and refused to do so each time. Perhaps
this will change. The Court heard the petition of Georgia death row in-
mate Troy Davis and ordered the lower federal court to review evidence
of his innocence.34 It agreed to hear a request for DNA testing by a Texas
death row inmate.35 Yet some people who could prove their innocence
may continue to fall through the cracks and fail to obtain DNA testing.
For example, in 2009, a Tennessee death row inmate who claimed his in-
nocence was executed despite requests for DNA testing.36 Meanwhile,
the Court’s abdication of responsibility places the onus on the states. For
example, following the Osborne decision, Alaska enacted a broad DNA
access statute.37

The Exoneration Pro cess

Since the U.S. Supreme Court has dodged the issue, it has been state
court judges, not federal judges, who have been the ones primarily con-
fronted with these exonerees’ cases. Initially, as Frank Lee Smith’s case

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 225

shows, exonerees faced signifi cant obstacles to their claims. However, in
recent years, state law has improved in response to these exonerations.

The exoneration pro cess usually began when these innocent people
contacted a lawyer. The vast majority of exonerees, 77% (192 of 250
cases), initially sought DNA testing by contacting an innocence project
or requesting it through postconviction attorneys. Innocence projects
and postconviction attorneys do not request DNA testing for every pris-
oner who contacts them. The Innocence Project in New York City, for
example, pursues DNA testing in all cases in which relevant biological
evidence exists and could be useful. In many cases, prisoners contact an
innocence project claiming to be innocent, but no evidence can be lo-
cated to test. Some delay has also resulted from backlogs at innocence
projects, which are nonprofi t organizations with limited resources. For
those without a lawyer, overcoming initial opposition to DNA testing
may have been quite diffi cult. Some exonerees took matters into their
own hands when lawyers would not listen. In 10% of these cases (26 of
250), exonerees initially pursued DNA testing pro se, often by fi ling
their own petitions asking a judge for DNA testing.

In 1991 Marcus Lyons walked up the steps to the Chicago court house
where he had been convicted, dressed in his U.S. Navy reserve uniform
and carry ing an eight- by- six- foot cross. He proceeded to lift a hammer
and start to nail his foot to the cross. He later explained, “I needed some-
one to listen.” He had just been released after spending three years in
prison for a rape he said he did not commit. His lawyer never even fi led
his appeal. As a result, once he was released, he was registered as a sex
off ender. His court house display got him a $100 fi ne for disturbing the
peace, but his eff orts paid off — a new lawyer took his case and requested
DNA testing, which exonerated him.38

All of the 250 exonerees were fortunate that the relevant crime scene
evidence was actually preserved in their cases. Law enforcement does not
have a strong legal incentive to preserve evidence. Some exonerees strug-
gled to locate the crime scene evidence from their cases. For example,
Alan Newton was told that the evidence from the rape that he was con-
victed of had been lost somewhere in the vast New York City Police De-
partment ware house. Police said they had made three searches of the bar-
rel where the rape kit was stored. After more than ten years of requests,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

226 Exoneration

another search was ordered. The rape kit was found in the right barrel in
the very place where it was supposed to have been kept. The police had
apparently never looked. The DNA tests exonerated Newton.39

Although DNA exonerations have created a new awareness about the
importance of preserving crime scene evidence, here again, the U.S. Su-
preme Court played a troubling role by providing little incentive to han-
dle evidence carefully. Larry Youngblood claimed that police shoddily
stored biological evidence from the rape victim, degrading it and making
it impossible for him to have it tested using the serological techniques
available at the time. In 1989, the Court ruled that Youngblood could not
have a new trial because he could not show that the police had acted in
bad faith when they improperly stored the evidence. By 2000, the science
of DNA testing had advanced such that the degraded evidence could be
tested. The results exonerated Youngblood and produced a cold hit with
another individual in the DNA database.40

During their appeals and postconviction, twenty- four exonerees
claimed destruction of exculpatory evidence, and none had any success
pursuing those claims. Like Youngblood, each was later fortuitously able
to test degraded evidence or to locate other evidence that could be tested.
However, most people who seek DNA testing cannot get it because the
evidence was long since destroyed. The Court’s “bad faith” test permits
police to negligently destroy crucial evidence. As a result, additional pro-
tections were needed. Approximately half of the states have now passed
statutes requiring the preservation of crime scene evidence.41

Even if the crime scene evidence could be located and was of suffi cient
quality for DNA testing, the exonerees still needed to get permission from
law enforcement to test it. The crime scene evidence is typically kept in
the custody of law enforcement, which deserves credit for its lead role in
quite a few exonerations. In 11% of exonerees’ cases (27 of 250), law en-
forcement or prosecutors themselves initiated the DNA testing. This oc-
curred where law enforcement conducted DNA testing as part of a project
to test backlogged evidence, or as part of a program to retest cases where
a forensic scientist engaged in a pattern of misconduct, or as part of an
unrelated criminal investigation, or, in one case, as a result of an anony-
mous phone tip. In these cases, the State presented the exonerees with
the surprise news that DNA had proved their innocence.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 227

Prosecutors did typically consent to postconviction DNA testing—
eventually, but often not quickly. Prosecutors consented to DNA test-
ing in 81% of cases in which information was obtained on the subject
(170 of 210 cases) and opposed it in 19% (40 of 210 cases), with no infor-
mation available in 40 cases. If prosecutors do ask that testing be con-
ducted, they can override any opposition of police. However, prosecu-
tors sometimes consented only after years and years of requests or after
the lead prosecutor fi nally left and a new prosecutor took offi ce. Some
consented only at the eleventh hour after a judge was already planning
to order the DNA testing. In at least 49% of the cases (122 of 250 cases),
the exoneree had to obtain a judge’s order to get DNA testing. Although
the prosecutors may have eventually supported the requests, no judge’s
order would have been necessary had prosecutors agreed to promptly
conduct testing.

Judges also denied requests for DNA testing that could defi nitively re-
solve the question of innocence. Before DNA testing was developed, most
judges were not receptive to claims of innocence or requests for access to
evidence after a trial. States emphasized the “fi nality” of convictions, for
the understandable reason that except in unusual situations, as time
passed after a trial, evidence would get stale, memories would fade, and it
would be diffi cult to revisit the question of guilt or innocence. Judges and
legislatures adopted rules requiring that a convict must bring forth newly
discovered evidence of innocence within a few years. The notion of fi nal-
ity was sometimes taken to an absurd extreme, barring consideration of
new evidence discovered only a few weeks or months after a conviction,
or even in cases where evidence had been concealed at trial or neglected
due to failures of counsel. As for many exonerees, the refrain of “fi nality”
was repeated over and over for Earl Washington Jr. He had to rely on ob-
taining clemency from the governor once he obtained DNA testing in
1993 and then again in 2002, because Virginia had a rule that one could
assert new evidence of innocence only within twenty- one days of the trial.
The so- called 21- day rule was such a short window that the State had all
but barred convicts from claiming innocence. Washington’s trial was in
1984, so his DNA tests in 1993 came nine years too late to get relief in a
court. Most of these exonerees similarly had few avenues for seeking relief
from judges after their convictions.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

228 Exoneration

For at least eigh teen exonerees, judges initially denied motions for
DNA testing (sometimes multiple times), often referring to the evidence
of their guilt. Many more such decisions may be unreported. For exam-
ple, in Bruce Godschalk’s case, the judge denied DNA testing because
“appellant’s conviction rests largely on his own confession which con-
tains details of the rapes which were not available to the public.”42 Simi-
larly, in Byron Halsey’s case, the judge denied him DNA testing despite
a state law that entitled him to evidence that could reasonably shed light
on his innocence. The judge cited his confession as evidence of the “over-
whelming evidence of defendant’s participation in the crime,” and noted
that his lawyer off ered no alibi or other evidence of his innocence at
trial.43 In Patrick Waller’s case, the Texas judge cited the eyewitness iden-
tifi cation testimony, denying DNA testing because “three of the four vic-
tims unequivocally identifi ed appellant,” and also noting that the prose-
cutor said she would prosecute him again anyway if the DNA testing
excluded him. He was exonerated several years later when a new Dallas
County prosecutor agreed to conduct the testing.44

In Wilton Dedge’s case, the trial judge denied his request for a DNA
test because it was fi led after a two- year statute of limitations; Florida
had not yet passed a DNA access statute. When the appellate court af-
fi rmed without an opinion in 1998, one judge dissented, with a moving
conclusion stating:

One of my worst nightmares as a judge, is and has been, that persons

convicted and imprisoned in a ‘legal’ proceeding, are in fact inno-

cent. If there is a way to establish their true innocence on the basis of

a highly accurate objective scientifi c test . . . in good conscience it

should be permitted. This case calls out for such relief: the evidence

of Dedge’s guilt at trial was minimal . . .

Recall from Chapter 5 how he had been convicted— despite a strong
alibi— based on a dog scent identifi cation, the victim’s shaky identifi ca-
tion, and an elaborate story by a jail house in for mant. Dedge fought for
several more years, and in 2004 he was fi nally exonerated after DNA test
results proved his innocence.45

The obstacles to DNA testing have been gradually removed, not be-
cause many judges have reconsidered when postconviction discovery

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 229

should be granted, but because so many states have passed statutes pro-
viding a right to postconviction DNA testing. This change occurred
slowly. In 1999, ten years after investigators began to use DNA testing in
criminal cases, only New York and Illinois had passed this type of DNA
statute.46 Most of the statutes were enacted in the last fi ve years. Cur-
rently, forty- eight states, the District of Columbia, and the federal gov-
ernment have passed DNA access statutes. In at least seventy- one cases,
the exoneree obtained testing pursuant to a state statute providing for
postconviction DNA testing. States need to do more to provide adequate
avenues for claiming innocence, but they have also come a long way.

These statutes are far from perfect. Many require diffi cult prelimi-
nary showings to obtain DNA testing. For example, statutes limit access
to DNA testing to persons convicted of certain felonies. Several deny
testing to people who pleaded guilty or who did not claim that they did
not commit the crime at a trial. Some deny testing if it had not been re-
quested earlier with due diligence; if an innocent person had a trial law-
yer that failed to request the testing, then it is the innocent person who
suff ers the consequences. Two states limit testing to just death penalty
cases, which makes little sense because DNA testing is not useful in most
capital cases.47

Even in states with sensible statutes providing access to DNA testing to
those who could potentially prove their innocence, judges sometimes ar-
bitrarily deny valid requests for testing. For example, judges still some-
times cite the evidence of guilt at trial to say that the DNA evidence would
make no diff erence. They argue that because there was an eyewitness,
the DNA test would not help. Of course, as Chapter 7 described, scores
of judges had similarly denied appeals or habeas petitions based on the
seemingly strong evidence introduced at the trials of these 250 exonerees,
and they were all wrong.

Other judges concoct speculative scenarios, suggesting that perhaps
the victim could have been promiscuous or had sexual intercourse with
a consensual partner, which would explain any DNA exclusion. Such
hypotheses are often totally implausible. Testing can be done to exclude
their partners as possible contributors. If the test results matched a felon
in a DNA database, the results would be even more helpful. Such rulings
show a lack of understanding by judges of how powerful DNA tests can
be in certain cases. Some of these statutes may also be too vague. With

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

230 Exoneration

states still sometimes denying access to DNA testing to those who could
prove their innocence, the U.S. Supreme Court’s failure to recognize a
clear right in Osborne frustrates people who need DNA testing to prove
their innocence.

Upon obtaining DNA test results, the still- incarcerated among the
250 exonerees were fi nally released. Nevertheless, some waited for quite
some time before obtaining their release. Judges refused to set free at
least twelve exonerees despite at least preliminary DNA test results ex-
cluding them. Each was later exonerated by an executive or a higher
court.48 Perhaps the most high- profi le example is Texas Court of Crim-
inal Appeals judge Sharon Keller’s cruelly vacuous explanation for her
decision not to set Roy Criner free, even though DNA testing excluded
him. Keller essentially told the reporters that DNA testing could never
suffi ciently prove innocence because a scenario could always be con-
structed to explain away the results. She felt that the DNA evidence
would not have “made a diff erence in the verdict” in Criner’s case. The
sixteen- year- old victim could have been “promiscuous” in the days be-
fore her rape and murder. This allegation of promiscuity was not based
on the trial record or any other record, and in fact police had investi-
gated just that subject and concluded that the semen could have only
come from the culprit.49 Criner was pardoned after DNA tests of a ciga-
rette butt found at the crime scene matched the person who was the
source of the semen.50

Following the postconviction DNA testing, each of these exonerees
obtained a vacatur of their conviction, either because a judge ordered a
new trial, or a governor gave clemency in the form of a pardon. Prosecu-
tors eventually joined in the motions to vacate the convictions in 88% of
cases in which information was obtained on the subject (171 of 194 cases)
and opposed the motions in 12% (23 of 194 cases), with no information
available (or no vacatur proceedings held) in 56 cases. Again, however,
months or years sometimes passed after the DNA test before prosecutors
agreed that the person should be exonerated. Sixty- eight received a par-
don from their state executive, some because they lacked any available
judicial forum for relief. Only 3% of exonerees (8 of 250 cases) received
DNA testing and a vacatur through federal habeas corpus— another in-
dication of the small role played by the federal courts. The other exoner-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 231

ees received a vacatur from state judges, typically on the basis of newly
discovered evidence of innocence.

The rulings announcing the exoneration are usually unpublished,
but the hearings themselves can be dramatic. Exonerees still incarcer-
ated may be walked out of prison to meet their lawyers and loved ones.
The press often extensively covers these emotionally charged DNA ex-
onerations. And exonerees may receive something else from the judge:
an apology. For example, in James Waller’s case, the judge told him, “On
behalf of any and all public offi cials at that time, I want to apologize.”51
In Glen Dale Woodall’s case, the original trial judge told him: “I hope
that as time goes on and your bitterness fades that you can take that as a
badge of honor that you’ve made it an awful lot easier for the next man
and the next man’s family, if they are wrongly charged,” to get DNA test-
ing performed, and “that’s something that you can be proud of for the rest
of your life.”52

Actual Guilt

DNA testing not only exonerates the innocent but it helps law enforce-
ment to identify the guilty. So far, 45% of these DNA exonerations have
resulted in the inculpation of the actual perpetrator (112 of 250 cases).
The creation of vast DNA databases has made it possible to quickly
search for a cold hit with a ge ne tic profi le. In sixty- fi ve exonerees’ cases a
cold hit in a DNA database resulted in identifi cation of the actual perpe-
trator. In forty- seven cases, the actual perpetrator was identifi ed through
other methods, such as police work, or the actual perpetrator came for-
ward to confess and was subjected to DNA testing. For example, in Chris
Ochoa’s and Richard Danziger’s cases, a prisoner named Achim Josef
Marino wrote a letter confessing his guilt to then governor George W.
Bush, stating, “I do not know these men nor why they plead guilty to a
crime they never committed. I can only assume that they must have been
facing a capitol [sic] with a poor chance of acquittal, but I tell you this sir,
I did this awful crime and I was alone.”53

Particularly striking are cases where the criminal justice system ig-
nored evidence that the defendant was innocent and that, in retrospect,
pointed to the actual perpetrator. In Chapter 3, I discussed several

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

232 Exoneration

remarkable cases where lineups may have been so faulty that eyewit-
nesses failed to identify the actual perpetrator in the lineup and instead
identifi ed an innocent man. In Chapter 4, I discussed a series of exam-
ples where forensic analysts had excluded the very person who turned
out to be the guilty party. In Chapters 2, 5, and 6, I discussed cases
where in for mants, codefendants, and witnesses had cooperated or even
testifi ed, but were later shown to be the real culprits.

In a few exceptional cases, the exoneree managed to track down the
true perpetrator from prison. Roy Brown, who had maintained his inno-
cence for years, found the perpetrator from his prison cell. Recall from
Chapter 4 that Brown was convicted in part based on invalid bite mark
testimony. Years into his prison sentence, he did not give up. He used
New York State’s freedom of information laws to gain access to police
documents that were never turned over at trial. Some of those reports and
witness statements indicated that Barry Bench, the brother of the victim’s
boyfriend, was involved. Because Brown could not aff ord an attorney, he
fi led a motion himself, arguing that this proved his innocence. The judge
denied the motion. He then wrote a letter to Bench, telling him that he
suspected his guilt and was going to ask for DNA tests to prove it. Five
days after receiving Brown’s letter, Bench committed suicide by lying
down on the tracks before an oncoming Amtrak train. In 2004, the Inno-
cence Project secured DNA testing that excluded Roy Brown. The pros-
ecutor ordered that Bench’s body be exhumed and DNA tests confi rmed
his guilt.54

Some of these culprits subsequently confessed or pleaded guilty. At
least forty of these perpetrators have been convicted of crimes that they
committed while innocent persons were behind bars. They were con-
victed of approximately fi fty- six rapes and nineteen murders after inno-
cent people were convicted of their earlier crimes.55 The perpetrators
may have committed many more crimes, but were not caught or were not
successfully prosecuted. The DNA testing that eventually was done
probably prevented still more crimes. As with Eddie Mosley, had it not
been for the postconviction DNA testing, these people could have contin-
ued their crime sprees with impunity. Those cases all highlight how im-
portant it is for public safety to make sure that the right person is con-
victed. Wrongful convictions are a serious law enforcement problem.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 233

Some convicts who request postconviction DNA testing are guilty, and
the DNA tests provide powerful confi rmation of their guilt. Approxi-
mately half of the clients of the Innocence Project have their guilt con-
fi rmed by the DNA testing. This is surprising, even shocking— that so
many of those who seek DNA testing are actually innocent. In contrast,
the others sought DNA testing despite their knowledge of their actual
guilt. As Barry Scheck commented, perhaps they did “not want to admit
it, or they [ were] lying or psychopaths.”56 They may also have hoped for
an error in the DNA testing. When I fi rst discovered that about half of the
Innocence Project’s clients ended up being guilty, I thought it would be
interesting to compare their cases to the cases of people who turned out
to be innocent. It was a diffi cult comparison to make. There is no pub-
licly available list of people whose guilt was confi rmed by DNA testing.
Mainly by searching for newspaper articles, I located over seventy cases
in which postconviction DNA testing confi rmed guilt, but only very lim-
ited information could be found about those cases; for example, only half
had written decisions in their appeals, often in capital cases. The unusual
features of those who sought DNA testing despite guilt, together with the
small size of the group and scarce information about them, prevented any
meaningful comparison with cases of exonerees.57

The guilt- confi rmation group is still important. Legislatures and
judges justify denying DNA testing that could reasonably prove inno-
cence by claiming that some set of requests are “frivolous” and made by
guilty people.58 And yes, guilty convicts do request DNA testing. How-
ever, DNA tests can put to rest any question of their guilt. Further, the
cost of such testing pales in comparison to the cost of imprisoning an in-
nocent person, while the guilty person remains free to commit additional
crimes. States spend from less than a few hundred dollars to $1,700 to test
evidence from a typical rape kit; costs vary depending on the number of
items to be tested and whether an unusually small sample must be
tested.59 This cost is a pittance even compared to the costs of litigating
run- of- the- mill criminal appeals. Unlike criminal appeals or habeas mo-
tions, DNA testing is fairly quick and when appropriately used can pro-
vide clear- cut answers on the question of guilt or innocence. DNA tests
are much less expensive than the costs of incarcerating an innocent per-
son, to say nothing of the costs of letting a guilty person remain free.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

234 Exoneration

Earlier I told the story of how Larry Youngblood was exonerated, many
years after the U.S. Supreme Court denied him relief. The State of Ari-
zona spent more than $109,000 to keep him behind bars for six and a half
years, while the true perpetrator remained free. The DNA test that freed
him cost $32.60

The Unexonerated

Some convicts obtained postconviction DNA testing, the results pro-
vided strong proof of their innocence, and yet they have not been exoner-
ated. Some have applied for pardons but are still waiting to fi nd out what
will happen. Many are still in prison. Others have been released from
prison but have not yet been exonerated. In Norfolk, Virginia, four men
known as the “Norfolk Four” had confessed to a brutal rape and murder.
One after another, each was excluded by DNA testing, which matched
the ge ne tic profi le of a fi fth man who confessed to committing the crime
and says that he acted alone. Governor Tim Kaine granted each of the
four clemency and reduced their sentences to the time they had already
served in prison. However, since he did not vacate their convictions, they
were not technically exonerated.61

Life after Exoneration

In his classic 1932 book Convicting the Innocent, the fi rst study of a set
of wrongful convictions, law professor Edwin Borchard highlighted “the
plight of the innocent victim of an unjust conviction in a criminal case.”
Borchard argued, “the least that the State can do to vindicate itself and
make restitution to the innocent victim is to grant him an indemnity,
not as a matter of grace and favor but as a matter of right.”62 De cades
later, there is still no safety net for the wrongly convicted. Although the
law is improving, and more exonerees continue to obtain compensa-
tion, the typical ser vices available to convicts who are paroled are not
available to innocent convicts who are exonerated. The job training,
housing placement, counseling, treatment, or other reentry ser vices
that some states provide to released prisoners are not available to an
exoneree.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 235

Most exonerees were not highly educated or highly paid before their
false conviction. Interviews with 115 exonerees by New York Times re-
porters in 2007 found that more than half did not graduate from high
school, “only half could recall holding a job for more than a year,” and
many had substance abuse problems and prior convictions.63 After all,
wealthy, employed people with high education and clear rec ords are far
less likely to be suspected of crimes, more likely to be able to hire a skilled
lawyer, and much less likely to be wrongly convicted. In contrast, the
usual suspects, particularly people with prior convictions, are far more
likely to be searched, questioned, interrogated, and placed in lineups by
the police.64 Police know that recidivists commit most crimes, and they
operate accordingly.

Many exonerees have not obtained meaningful civil compensation for
the terrible injuries they suff ered. Only 60% have thus far received some
kind of compensation for their years of imprisonment.65 About half of
those obtained it by fi ling civil rights lawsuits against the government.
Winning this kind of lawsuit is not easy. Exonerees must be able to show
that government offi cials acted with suffi cient fault. They must overcome
immunity doctrines that often prevent a person whose civil rights were
violated from obtaining any compensation. Such cases may take many
years to litigate. At least 46% of the DNA exonerees (116 of 250) fi led civil
claims, mostly in federal courts. At least 73 who brought wrongful con-
viction lawsuits received favorable judgments or settlements. Their judg-
ments or settlements were often for many millions of dollars. For example,
Ronald Jones, whose story began this book, later received a $2.2 million
settlement from the City of Chicago.66 Arvin McGee, who was described
in Chapter 6, endured a mistrial, a hung jury, and a deadlocked jury that
fi nally convicted him at a third trial. In a civil rights case he fi led after his
exoneration, he was awarded $1 million a year for the fourteen years he
spent in prison.67 Lawsuits do more than just compensate innocent peo-
ple for the time they were incarcerated. They also provide a wake- up call
to law enforcement. Many states and municipalities have reformed their
practices after fi nding out the hard way that failing to use fair procedures
can be very expensive.

The law is changing. In part because federal courts have made civil
rights claims diffi cult and time- consuming to litigate, more states have

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

236 Exoneration

passed no- fault compensation statutes for those exonerated by DNA.
Half of the exonerees who have received some compensation obtained it
through such statutes or through special legislative bills.68 Twenty- fi ve
states, the District of Columbia, and the federal government have now
passed statutes to provide automatic compensation to the exonerated.
The statutes provide compensation far more quickly than lawsuits do,
although they range widely in the quality and quantity of the benefi ts
they provide. Most exonerees received less than $50,000 per year of
wrongful incarceration under these statutes.69 Texas now leads the way,
having changed its compensation law to provide $80,000 for each year
behind bars, lifetime annuity payments, as well as social ser vices like job
training, fi nancial advising, tuition credits, and medical treatment.70

Some exonerees have adjusted to life after prison, but the psychologi-
cal impact of a false conviction cannot be underestimated. Ronald Jones
explained, “I haven’t, still today, been able to adjust to the world. I mean
it’s like I was on another planet. Now they’ve got cell phones and com-
puters and all these other things. The closest thing I knew about a cell
phone was a telephone on the corner where you could drop a dime in the
phone. You can’t even drop a dime in the phone no more.”71

After his twenty- two years in prison, Jerry Frank Townsend walked
slowly, “shuffl ing along with his head down and his shoulders stooped,”
and was always turning around to look behind to see if someone was try-
ing to ambush him.72 The psychological harm experienced by exonerees
has been little studied. Not only did they suff er from the confi nement,
overcrowding, and monotony, as well as violence, of the harsh prison en-
vironment, but they did so knowing they were innocent. Psychologists
have studied the eff ect of incarceration upon the guilty, but coping with
imprisonment may be quite diff erent for a person who knows that he is
innocent.

Further, since so many of these exonerees were convicted of rape,
they were targets in prison. As one of Jimmy Ray Bromgard’s jailers
commented, “crimes of that sort are not well received inside,” and as a
result, “His time was hard.”73 Indeed, those who were released prior to
their exoneration had to register in sex off ender programs, and some
served additional time because they refused treatment as sex off enders.
Robert McClendon recalled that he and Joseph Fears refused to enter a

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 237

courtroom in which they would be declared sex off enders under a new
Ohio law; they were “arguing with people, saying, ‘We’re innocent!’ ”
Years later both were exonerated by DNA testing.74

The interviews with exonerees by New York Times reporters found
that “most of them have struggled to keep jobs, pay for health care, re-
build family ties and shed the psychological eff ects of years of question-
able or wrongful imprisonment.”75 About one- third of those interviewed
had found stable employment, while one- sixth had been rearrested. Ex-
onerees have had trouble obtaining jobs because they have to explain to
potential employers why they were in prison, and their stories sound in-
credible. As Neil Miller put it, “I don’t believe that anyone really under-
stands what the word ‘exonerated’ means. Every time I go to a job and
I fi ll out the application and I explain to them that I was exonerated, I
always get . . . ‘That word, what does that word mean?’ ”76 Keith Turner
says, “I keep a copy of my pardon on me. Every job, you have to explain
yourself. You have to put it on there— rape conviction— because they
check it. I always write, ‘I’ll explain at the interview.’ ”77

“All I got was a basic form letter apologizing for what happened, not
admitting they did anything wrong,” Wilton Dedge said. “But it seemed
pretty hollow, after all the names they called me in court. They weren’t
man enough to step forward and apologize to my face. That disappointed
me.”78 Others did receive that apology. For example, in Brandon Moon’s
case, the prosecutor put it simply: “I know we can’t give you back your
years . . . and for that I’m extremely sorry.”79

Others have been re- arrested since their exoneration. Many had prior
criminal rec ords, but may simply have known no other life after years of
incarceration. Steven Avery had served eigh teen years before being ex-
onerated of a rape when DNA testing excluded him and included an-
other man. However, three years after his release, he was charged and
convicted of murder. This time DNA inculpated him.80

Some died not long after being exonerated. Kenneth Waters was ex-
onerated in large part because of his sister, Betty Anne Waters. She was
a high school dropout who, after her brother’s wrongful conviction, put
herself through high school, college, and law school, determined to liti-
gate her brother’s case and prove his innocence. With the Innocence
Project’s help, she did, only after years of fi ghting for access to the DNA

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

238 Exoneration

evidence. Months after his exoneration, Kenneth Waters tripped and fell
to his death outside her house.81

Victims after Exoneration

Exonerations can also powerfully aff ect victims and the families of vic-
tims. They are confronted with the realization that an innocent man was
convicted, perhaps now becoming aware that the true perpetrator has
never been found and that the closure they thought they had achieved
was an illusion. Others learn that the perpetrator was inculpated by DNA
testing. Some remain convinced that the right person was convicted, de-
spite DNA evidence of innocence. The victim in Victor Burnette’s case
put it bluntly: “I know what I saw . . . And those tests can be wrong. I
mean, lab tests fail every day.” She added, “What is the good of bringing
this back up? . . . What am I going to say? ‘Oh, I must’ve made a mistake.
Let him go.’ I don’t care what happens to that man . . . I don’t want this
brought back up any more.”82 Similarly, the victim in Brian Piszczek’s
case commented, “I’m still 100% sure it was him,” adding, “I was there.
I know. I don’t care what those DNA tests say.”83

At least six eyewitnesses formally recanted their identifi cations and
provided the exonerees’ lawyers with statements during appeals or post-
conviction. None of the eff orts to introduce that new evidence of inno-
cence resulted in a conviction being reversed. A few of the victims came
forward after DNA testing was conducted to say that they believe they
did in fact misidentify the defendant. Jennifer Thompson, the victim in
Ronald Cotton’s case, described how police made remarks that rein-
forced her misidentifi cation of Cotton. Though she now knows he is in-
nocent, she still sees his face when she pictures the rapist. She has coau-
thored a book with Cotton and lectures with him on the need for improved
eyewitness identifi cation procedures.84

In James Curtis Giles’s case, in which the victim described a shorter,
younger man and did not describe his two prominent gold teeth, after
DNA testing the victim told the prosecutor that “She has a vivid memory
of James Curtis Giles, but she is no longer certain that the memory of
that face is emblazoned in her head from viewing the lineup and testify-
ing and seeing him at trial or from witnessing his face at the off ense.”85

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Exoneration � 239

The eyewitness in Steven Avery’s case explained in an interview that
police remarks following her identifi cation enhanced her certainty; “I
think my confi dence that I had selected the right person was boosted
when I was told that the person they arrested was the suspect that they
had in mind.”86 Few victims have spoken about these cases, so we do not
know whether such misconduct occurred in other cases.

A Shorter Road to Exoneration

Just as these exonerees faced a hard road along the way to their exonera-
tion, their path is not easy once freed. The exonerated often endured
years of litigation and other obstacles before they successfully obtained
DNA testing. Indeed, many faced barriers to obtaining relief even after
the DNA exonerated them. The DNA testing then reopened these cases,
for the victims who had to revisit painful events, but also because the
DNA tests often identifi ed the actual culprits. In response, the states have
themselves revisited many of the legal obstacles that so substantially de-
layed exoneration for these 250 innocent people. Although federal courts
still provide no clear avenue to claim innocence, almost all states have
now passed laws providing access to DNA testing. Even after exonera-
tion, some of these people had to continue to struggle in the courts for
years to obtain meaningful compensation for their injuries. That too has
changed, as states have enacted statutes to more quickly compensate the
wrongly convicted.

Nevertheless, the experiences of these exonerees illustrate how often
law enforcement, lawyers, and judges failed to take claims of innocence
seriously when they should have. We should continue to expand access
to postconviction DNA testing. We should ensure that biological mate-
rial from criminal cases is preserved, so that DNA testing can be done in
more cases. And we should ensure that judges carefully and promptly
consider claims of innocence.

Turning back to the story of Habib Abdal, whose conviction due to a
fl awed eyewitness misidentifi cation began Chapter 3, we can see how
even in a state with some sound procedures, barriers can delay justice for
the innocent. Abdal had fought for years to obtain DNA testing. He was
forty- three and not a young man when he was convicted (recall that the

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

240 Exoneration

eyewitness had initially expressed great uncertainty about the age of her
attacker). The judge denied his request for DNA testing in 1989. At the
time, New York had not yet enacted one of the fi rst statutes that entitled
convicts to DNA testing if it could reasonably prove their innocence.
Finally, in 1993, a federal judge, one of the fi rst in the United States to do
so, ordered that the DNA testing be performed.87 Even after the results
conclusively excluded him, for years prosecutors resisted dismissal of
the indictment, arguing that perhaps there was another rapist who left
his DNA and Abdal had also raped the victim without ejaculating. There
was a problem with this new speculative theory: the victim had always
been clear that one and only one man raped her. And the DNA showed
that the man was not Habib Abdal.

Once he was fi nally exonerated, Abdal recalled that in 1999, “When
they let me out, they gave me $40 and an old pair of dungarees with a
broken zipper . . . They told me, ‘Get out of here. You’re a free man.’ ”
That was not the end of his story. New York also had one of the fi rst stat-
utes, passed in 1984, designed to compensate the wrongly convicted. In
2002, he received a settlement from the State of New York. He received
two million dollars to compensate him for sixteen of the years he spent in
prison. He had initially refused to settle because he wanted his day in
court and the opportunity to tell a judge what wrongs he suff ered, but
terminally ill with cancer, he accepted the settlement. In 2005, he died.88

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

241

In 1995, Ronald Cotton’s exoneration in North Carolina received na-
tional press coverage. His DNA exoneration was followed by those of

Joseph Abbitt, Keith Brown, Dwayne Dail, Darryl Hunt, Lesly Jean,
and Leo Waters, all in North Carolina. The state’s response to this
drumbeat of DNA exonerations was dramatic: it changed the structure
of its criminal justice system. Justice I. Beverly Lake Jr., then chief justice
of the North Carolina Supreme Court, decided that preventing future
wrongful convictions would be his legacy. Before he retired in 2006, he
helped to create the fi rst entity designed to prevent wrongful convictions
in the United States, the North Carolina Actual Innocence Commission.
Chief Justice Lake called it “probably the most important thing I’ve
done.”1

Chief Justice Lake’s intrepid law clerk, Christine Mumma, helped to
design a commission that involved all of the key actors in the state, and
she later took the reins as its executive director. Though convened by
the state supreme court, it is in de pen dent of the judiciary with thirty ac-
tive participants, including law enforcement offi cers, prosecutors, de-
fense attorneys, social scientists, victim’s advocates, law professors, and
judges. The chief justice asked the commission to do two things: “re-
view our criminal justice system for potential changes” to prevent wrongful

c h a p t e r 9

Reforming the Criminal Justice System

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

242 Reforming the Criminal Justice System

convictions, and “establish a mechanism for objective review of credible
innocence claims.”2

The commission, founded in 2002, quickly charted new ground on
both fronts. At fi rst, the commission proposed standards recommending
sequential and double- blind identifi cation procedures, since eyewitness
misidentifi cations were a “leading factor” in the North Carolina exoner-
ations. That set of best practices was used to train all North Carolina law
enforcement offi cers. At the commission’s suggestion, the state legisla-
ture then enacted the Eyewitness Identifi cation Reform Act.3 The law
provided that all lineups in the state were to be conducted by an in de-
pen dent, blind administrator, and sequentially. New instructions were
to be given to all eyewitnesses before viewing a lineup, including a warn-
ing that “[t]he perpetrator might or might not be presented in the lineup,”
and that “[i]t is as important to exclude innocent persons as it is to iden-
tify the perpetrator.”4 The law also regulated the way that lineups are
created. It required that a “lineup shall be composed” to ensure that the
suspect “does not unduly stand out from the fi llers” and that the fi llers
“shall resemble, as much as practicable, the eyewitness’s description of
the perpetrator in signifi cant features, including any unique or unusual
features.”5 The law required something else that was almost never done
in the exonerees’ cases: the eyewitnesses were to be asked to express
their confi dence, in their own words, and this statement was to be docu-
mented, if practicable by using a video camera.

After the passage of this landmark eyewitness reform legislation, the
commission did not rest, but moved on to study reforms to prevent false
confessions. At the urging of commission members, North Carolina en-
acted laws in 2008 that required the recording of interrogations in hom-
i cide cases, enhanced the preservation of evidence and access to DNA
testing, and expanded post exoneration support for the wrongfully
convicted.6

The commission then turned to the set of problems raised in Chap-
ter 7— the inadequacies of existing methods for factual review of postcon-
viction innocence claims. It proposed that a new Inquiry Commission be
created to review innocence claims, similar to bodies created in the
United Kingdom and Canada. The United Kingdom, in 1997, created a
Criminal Cases Revision Commission to serve as an in de pen dent execu-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 243

tive agency with full subpoena power to review convictions. It cannot
reverse convictions, but refers cases to appellate courts, recommending
that convictions be vacated.7 Similarly, in Canada a person can request
that the minister of justice convene a Criminal Conviction Review Group
(made up of attorneys from the Department of Justice) to examine cases
and make recommendations to the minister. The minister can then order
a new trial or hearing, or refer a case to a court. The minister can order a
formal inquiry examining what caused a wrongful conviction. Several of
those inquiry reports have recommended and led to the implementation
of sweeping reforms.8

There was no such mechanism anywhere in the United States to re-
spond to a wrongful conviction until, after several years of consideration,
legislation enacted in 2006 created the proposed North Carolina Inno-
cence Inquiry Commission. When he announced the creation of the In-
quiry Commission, Governor Mike Easley announced, “As a state pros-
ecutor for more 15 years, I know that law enforcement’s greatest nightmare
is to have an innocent person in jail or on death row.”9 An eight- member
panel reviews criminal convictions. Unlike postconviction courts that
can become tangled in complex procedural bars or harmless error rules,
the commission just asks whether a person is innocent and should be
exonerated.

The Inquiry Commission panel members serve three- year terms and
include a judge, prosecutor, criminal defense lawyer, sheriff , victim’s ad-
vocate, and members of the public. Its director helps to develop rules for
deciding which cases to review and coordinates the investigations. If fi ve
panel members agree that a defendant deserves judicial review, the state
supreme court’s chief justice appoints a three- judge panel, who may over-
turn the conviction if they unanimously agree the defendant showed “clear
and convincing evidence” of factual innocence. The Inquiry Commission
has recommended judicial review in several cases and has screened hun-
dreds.10 In 2010, the Inquiry Commission recommended that based on
“clear and convincing evidence” of his innocence, Gregory Taylor be ex-
onerated, and after the three- judge panel agreed, he became the fi rst per-
son exonerated thanks to an innocence commission in the United States.11

Outside North Carolina, states have not established bodies designed
to review innocence cases. More than ten other states have created or at

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

244 Reforming the Criminal Justice System

least begun to create innocence commissions, but most lack any formal
power and few have done so much as issue a report. For example, a
Connecticut Commission on Wrongful Convictions has a purely advi-
sory role. A California commission recommended a detailed and com-
prehensive set of reforms, which the legislature passed— but which
Governor Schwarzenegger promptly vetoed, saying “current criminal
procedures provide adequate safeguards.”12 Perhaps such commissions
will become infl uential over time. So far, none approach the successes
in North Carolina.13

Criminal Procedure Reform

Two de cades into the DNA revolution, exonerations are reshaping crimi-
nal procedure. North Carolina created a new institution to respond to
wrongful convictions, but other states, including Illinois, New Jersey,
Ohio, and West Virginia, have adopted diff erent types of systemic reforms
and in very diff erent but perhaps equally eff ective ways. Across the coun-
try, prosecutors, police offi cers, defense lawyers, and judges have devel-
oped new procedures in response to exonerations.

Why have DNA exonerations had such an impact? Perhaps they em-
body a special kind of error. Serious criminal cases are reversed on appeal,
but those kinds of errors usually do not receive the front- page headlines
that DNA exonerations receive. What sets DNA exonerations apart from
typical reversals of convictions, or even from exonerations that do not in-
volve DNA, of which there have been many more?14 Perhaps the cutting-
edge science of DNA technology captivates the public. Or perhaps people
fi nd it surprising that innocence can sometimes, with the benefi t of DNA,
actually be indisputable, making people more suspicious than ever before
that our criminal justice system can make terrible mistakes.

Public fascination with DNA exonerations may also have to do with
the way that DNA technology enters a case from the outside. Wrongful
convictions were not uncovered because the criminal justice system
worked. In these cases, by defi nition, there was no exoneration until af-
ter the DNA testing. Part of the gripping drama in these exonerees’ sto-
ries is how DNA fi nally succeeded when all else had failed for so many
years. As law professor James Liebman has pointed out, DNA takes on

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 245

the quality of “divine intervention,” where if not for “the sheer accident
that a biological sample happened to be available, the miscarriage never
would have been discovered.”15 Many of these exonerees’ trials looked
unremarkable at the time they were convicted. Indeed, as we have seen,
the cases against many of them looked unusually strong, due to contami-
nation of the evidence. DNA testing shows how all cases can fail, rang-
ing from the most serious capital case, to rapes, to a handful of robberies,
including cases with fl imsy evidence but also many with seemingly strong
evidence, and in every type of jurisdiction, with all sorts of police depart-
ments, judges, jurors, and lawyers.

These DNA exonerations also came to light at a time that criminal
sentences had grown much harsher and the prison population had sky-
rocketed. Since the late 1970s, the United States has had the highest in-
carceration rate and the largest prison population in the world.16 The
grace of ge ne tic technology has revealed some prisoners as innocent, but
it has done so haphazardly, in just a few cases that happened to be ame-
nable to DNA testing. This has not happened in the vast bulk of crimi-
nal cases, which unlike rapes and murders do not usually involve bio-
logical evidence and remain impervious to the truth of DNA. We are left
wondering how many others are innocent.

Despite the costs that wrongful convictions impose on exonerees, vic-
tims, and the public, states have generally refused to ask what went wrong
in these exonerees’ cases, perhaps because they do not want to further
shake our faith in the criminal justice system. As Innocence Project
found ers Barry Scheck and Peter Neufeld have pointed out, “In the
United States there are strict and immediate investigative mea sures taken
when an airplane falls from the sky, a plane’s fuel tank explodes on the
runway, or a train derails,” but, “The American criminal justice system,
in sharp contrast, has no institutional mechanism to evaluate its equiva-
lent of a catastrophic plane crash, the conviction of an innocent person.”17
The North Carolina experience is the exception to the rule. Few exonera-
tions were followed by administrative reviews, blue- ribbon commissions,
inquiries, audits, or any offi cial eff orts to fi nd out what happened and
why.

However, as I will describe in this chapter, our criminal justice sys-
tem is fi nally starting to look forward to consider reforms. A series of

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

246 Reforming the Criminal Justice System

inadequate criminal procedure rules are implicated by these wrongful
convictions, as each chapter has developed. In Chapter 2, I discussed
the U.S. Supreme Court’s “voluntariness” test for reviewing confes-
sions, and how it fails to take into account the reliability of confessions.
In Chapter 3, I discussed the Court’s fl imsy Manson test for reviewing
eyewitness identifi cations, and how highly suggestive procedures may
be excused based on evidence of “reliability,” such as the eyewitnesses’
confi dence. Chapter 4 discussed how the Court’s Daubert ruling re-
quires that scientifi c evidence be reliable and valid, but how that require-
ment has simply been ignored by criminal courts. Chapter 5 described
how in for mant testimony is not regulated, except in the sense that in for-
mants may not be planted and deals with in for mants must be disclosed,
which was not always done in these exonerees’ cases. In Chapter 6, I de-
scribed how the Supreme Court’s Strickland test does not ensure that
adequate lawyers are provided to indigent defendants, and similarly, how
Brady v. Mary land does not ensure that defendants receive evidence of
innocence, although the full extent of concealment of exculpatory evidence
cannot be known.

Chapter 7 discussed how not only does criminal procedure fail to en-
sure that accurate evidence is presented at trial, but accuracy is even less
of a concern postconviction, in which judges rarely correct legal errors
but they almost never revisit facts. Chapter 8 discussed how, postconvic-
tion, these innocent people faced a string of legal obstacles to obtaining
DNA testing and their exonerations.

Each of those chapters showed how our law of criminal procedure
could be reformed to prevent future wrongful convictions. A change in
the U.S. Supreme Court’s interpretation of the protections aff orded to
criminal defendants under the U.S. Constitution would result in a na-
tionwide shift in the practices of law enforcement and courts in criminal
cases. But the Supreme Court has not yet reconsidered any of the crimi-
nal procedure rules that played a role in these cases, and as it does not
regularly reconsider long- standing rules, it may be quite some time be-
fore it reconsiders any of these. Instead, the legal response to wrongful
convictions is likely to be fragmented, as the fi fty individual states and
more than 17,000 law enforcement agencies experiment with their own
reforms.18 In this fi nal chapter, I show that knowledge of wrongful con-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 247

victions is changing how states and localities seek to improve accuracy
in criminal cases. Many of these changes refl ect the problems that I found
when I reviewed the exonerees’ trial transcripts. I explain which reforms
show the most promise and how much work still needs to be done.

Interrogation Reforms

As we saw in Chapter 2, false confessions are one of the most disturbing
causes of wrongful convictions. Recall Eddie Joe Lloyd, who was inter-
viewed by police while involuntarily committed in a Detroit, Michigan,
mental hospital with a preliminary diagnosis of bipolar aff ective disor-
der. It was plain that he was a vulnerable and mentally disabled individ-
ual. However, he had contacted police with information about a grue-
some rape and murder. His mental illness included “delusions that he
had a special ability to solve crimes.”19 The detective “provided me with
quite a bit of information about the case,” Mr. Lloyd later recalled. “He
said, ‘What kind of jeans was she wearing?’ I said, ‘I don’t know.’ He
said, ‘What kind do you think?’ I said, ‘Jordache.’ He said, ‘No, Gloria
Vanderbilt.’ ” Lloyd said that the detective “guided him through a sketch
of the garage” where the crime occurred, among other details. “The
emphasis was on, ‘You want to help us, right?’ he said. I said, ‘Sure, I
want to help any way I can.’ ”20 The prosecutor said that the crime scene
“was secured, no one else was allowed in . . . Aside from [the investigat-
ing offi cers], the only other person who would have known that was the
killer, Mr. Lloyd.”21

Just as in other exonerees’ cases, if Lloyd’s entire interrogations had
been recorded, it would have been apparent at the outset that Lloyd
knew nothing about how the crime occurred. Instead, Lloyd served sev-
enteen years in prison, until DNA testing exonerated him. Videotaping
is inexpensive and can shed light on what really happened in the inter-
rogation room. When Lloyd brought a civil rights lawsuit against the
Detroit Police Department seeking compensation for his wrongful con-
viction (which he reportedly settled out of court for more than four mil-
lion dollars), the police department also entered a consent decree requir-
ing that all interrogations in crimes that carry a penalty of life imprisonment
without parole should be videotaped.22

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

248 Reforming the Criminal Justice System

In the 1990s, before DNA exonerations were so numerous, the only
states to require videotaping of confessions were Alaska and Minnesota.
That has changed, in large part because the false confessions of these
exonerees vividly illustrated the costs of the failure to record entire in-
terrogations. Eleven states and the District of Columbia now require or
encourage electronic recording of at least some interrogations by statute,
and seven more state supreme courts wrote opinions either requiring or
encouraging the recording of interrogations.23 At least 500 police depart-
ments now videotape interrogations. Police in those departments have
reported positive experiences with videotaping and say that recording
does not discourage a suspect’s cooperation.24 The videotape can sim-
plify disputes over whether facts were fed or a confession coerced. As one
federal judge put it, “I don’t know why I have to sit here and sort through
the credibility of what was said in these interviews when there’s a perfect
device available to resolve that and eliminate any discussion about it.”25
In addition, a recording can “serve to protect police offi cers against false
allegations that a confession was not obtained voluntarily.”26 As the De-
troit police chief commented, “It’s a protection for the citizen that’s being
interrogated. But from a chief ’s point of view, I think the greatest benefi t
is to police because what it does is provide documentation that they
didn’t coerce.”27

While we can expect this benefi cial reform to continue to take hold, as
it should, recording entire interrogations alone is not enough. Just be-
cause an interrogation has been recorded doesn’t mean that it wasn’t co-
ercive or unreliable. Indeed, the recording may show that there was coer-
cion or contamination through the disclosure of facts. Judges should
carefully review recordings to assess the voluntariness and reliability of
interrogations. Police should adopt precautions for interrogations in-
volving vulnerable juveniles or mentally disabled suspects. A new focus
on accuracy can help to safeguard the reliability and legitimacy of
confessions.28

Eyewitness Identifi cation Reforms

Like the problem of contaminated confessions, the problem of contami-
nated eyewitness identifi cations can also be remedied. Thanks to DNA

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 249

exonerations, eyewitness identifi cation reforms are fi nally starting to
happen. Psychologist Gary Wells commented that although he and other
social scientists had conducted research on eyewitness memory for de-
cades, “before DNA they pretty much ignored the studies.”29 That has
changed, with New Jersey leading the way.

The New Jersey Supreme Court became involved in these issues when,
although it did not realize it at the time, it was confronted with the case of
an innocent man. It would respond by dramatically reshaping New Jersey
criminal procedure and acting, in eff ect, like an innocence commission.
The triggering event was exoneree McKinley Cromedy’s case. His lawyer
had argued at trial that Cromedy had been misidentifi ed, telling the jury
in the opening statement that “the evidence will show, not that she’s a liar,
but that she’s mistaken, that her identifi cation is wrong and it’s a misiden-
tifi cation.”30 The victim, a white college student, had been raped by a
black man in her apartment. A few days later, she had helped a police art-
ist draw a composite sketch of a black man with a full face and a mous-
tache, and she looked at thousands of photos of black men who had been
arrested. One of those photos was of Cromedy, and in fact the police had
him in mind as a suspect because he had been seen in the area, but she
did not identify him.

Almost eight months after the rape, she saw Cromedy crossing the street.
She thought he was her attacker, partly because of his appearance but also
because of his unusual way of walking due to a limp, “a swagger,” as she put
it.31 She called the police, who called her back fi fteen minutes later to say
that they had picked up a man matching her description. She then went to
the police station and positively identifi ed Cromedy as her attacker. The
police offi cer explained, “I’ve had a lot of experience with identifi cations
and I’m not going to lead somebody. I asked her to see if she recognized this
person.”32 Yet there was no justifi cation for conducting an inherently sug-
gestive showup in which she viewed Cromedy one- on- one from behind
one- way glass.

McKinley Cromedy’s lawyer argued that the identifi cation was im-
proper, saying that the showup was “like true or false, and to me that is
about as suggestive as a procedure you can have . . . She knows some-
body was picked up. What could be more suggestive?” The judge ruled
that the identifi cation was admissible, emphasizing, “she was very certain

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

250 Reforming the Criminal Justice System

of her identifi cation,” and noting that her composite drawing looked like
Cromedy and that “Mr. Cromedy has a very, very unique style of walking.
It’s a combination of a swagger and a roll.”33 Cromedy’s defense lawyer
asked for a special jury instruction, asking the jury to consider “whether
the cross- racial nature of the identifi cation has aff ected the accuracy of the
witness’s original perception and/or accuracy of a subsequent identifi ca-
tion.”34 The trial judge denied the request. The victim pointed to Crom-
edy in the courtroom and agreed she was “absolutely sure” he was her
attacker.

Cromedy was convicted. On appeal, though, the New Jersey Supreme
Court reversed his conviction. The court ruled in 1999 that “forty years”
of empirical studies documented a risk of heightened error when white
eyewitnesses try to identify black subjects. The court noted that some
courts, such as in California, Massachusetts, and Utah, had permitted
such instructions. The court ruled that under the facts of his case, it was
“reversible error not to have given an instruction that informed the jury
about the possible signifi cance of the cross- racial identifi cation factor, a
factor the jury can observe in many cases with its own eyes.”35 The court
reversed McKinley Cromedy’s conviction without knowing that he was
in fact innocent. After the ruling, the prosecution agreed to conduct DNA
tests. The results excluded him and he was exonerated. The victim later
commented, “I couldn’t believe that I was wrong.”36

In response to such exonerations, New Jersey began a project of re-
vamping its criminal procedure rules. The New Jersey attorney general’s
Offi ce issued guidelines to all law enforcement agencies in the state re-
quiring that detailed procedures be followed when eyewitnesses are
asked to identify a suspect.37 These guidelines were a landmark reform.
New Jersey became the fi rst state in the country to adopt double- blind
lineups. No longer would the offi cer administering the procedure know
who the suspect is. All lineups would use sequential photo arrays, where
photos are shown one at a time to prevent “comparison shopping.” Eye-
witnesses were to be instructed that the perpetrator might not appear in
the lineup. The results were to be recorded, including the witnesses’
certainty at the time of the identifi cation procedure (recall from Chapter
3 how many eyewitnesses in exonerees’ cases were absolutely certain at
trial but were not so certain at the time of the lineup).

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 251

In 2005, the New Jersey Supreme Court went further. The court con-
vened a special committee to study the problem of false confessions. The
court then issued a state rule requiring that all custodial interrogations
in hom i cide cases be electronically recorded. The next year, in 2006, the
court expanded this rule to require that police similarly record or docu-
ment all eyewitness identifi cations.38 The court noted, “Misidentifi ca-
tion is widely recognized as the single greatest cause of wrongful convic-
tions in this country.” In 2007, the court addressed jury instructions.
The court adopted a Model Jury Instruction charging all jurors not to
rely on “the confi dence level” of an eyewitness, at least not “standing
alone.” Jurors are cautioned: “Although nothing may appear more con-
vincing than a witness’s categorical identifi cation of a perpetrator, you
must critically analyze such testimony. Such identifi cations, even if made
in good faith, may be mistaken.”39

Finally, the court asked that a special master explore something more
fundamental: the U.S. Supreme Court’s Manson test for evaluating ad-
missibility of eyewitness identifi cations.40 The master held hearings, with
the participation of the New Jersey Offi ce of the Public Defender, Attor-
ney General, Association of Criminal Defense Lawyers, and the Inno-
cence Project. He recommended that the court adopt a new test for evalu-
ating eyewitness identifi cation evidence and require pretrial hearings to
evaluate all eyewitness identifi cations.41 As a result, the court may over
time completely alter the way that eyewitness identifi cations are conducted,
documented, and litigated.

Most other states have not adopted such reforms. For example, in Kirk
Bloodsworth’s case, the fi rst death row DNA exoneration, the Court of
Appeals of Mary land had upheld the trial court’s refusal to allow expert
testimony on the dangers of eyewitness misidentifi cations. The trial judge
excluded this testimony on the grounds that such evidence would be un-
necessary and would “confuse or mislead” the jury.42 We now know, of
course, that the jury was in fact gravely misled when it believed the eye-
witnesses in that case. Yet that decision is still the law in Mary land. Other
judges have become more open to questioning reliability of eyewitness
identifi cations, by providing juries with cautionary instructions concern-
ing eyewitness error, or admitting expert testimony explaining social sci-
ence research concerning misidentifi cations.43

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

252 Reforming the Criminal Justice System

Six states, Illinois, Mary land, North Carolina, Ohio, West Virginia
and Wisconsin, have passed statutes in response to these misidentifi ca-
tions.44 Other states, such as Georgia, Vermont, and Virginia, recom-
mended studying the problem further, while still additional jurisdictions
and departments have adopted voluntary guidelines.45 All of this marks
the beginning of a sea change, where for so long police had no written
procedures at all on identifi cations, and worse, they routinely used unre-
liable and suggestive techniques. Local police departments also increas-
ingly adopt double- blind and sequential procedures. The Dallas Police
Department, in response to more DNA exonerations than any other juris-
diction has had, adopted sequential and blind lineup procedures, as well
as limiting the use of showups and requiring documentation of eyewit-
ness procedures.46 Since so many of these DNA exonerations involve
eyewitness misidentifi cations, and because de cades of social science re-
search support use of double- blind and documented lineups, police will
likely continue to gradually adopt improved procedures. It is crucial that
they do so. Nor should our constitutional criminal procedure permit
grossly suggestive procedures so long as the identifi cation appears “reli-
able.” Judges could impose consequences if the police fail to ensure that
sound identifi cation procedures are used; at least, they could provide ju-
rors with careful instructions on the nature of eyewitness memory and
sources of error, particularly if suggestive procedures are used.

Forensic Science Reform

Flawed forensic science, like unsound eyewitness identifi cation proce-
dures, played an outsized role in these wrongful convictions, and yet re-
form has been especially slow in coming. One telling example is Glen
Woodall’s exoneration, which led to a series of court decisions reforming
the state crime laboratory, but slowly, and over more than a de cade. Wood-
all was convicted based on eyewitness identifi cations that were not very
strong, since the attacker wore a mask and neither victim saw his face, but
also based on fl awed forensic testimony by Fred Zain of the West Virginia
Crime Laboratory. Zain had claimed that 6 in 10,000 people could have
committed the crime, when in fact, due to the problem of masking, the evi-
dence could have entirely come from the victim. Any male could have been

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 253

the rapist. Zain also off ered the scientifi cally unsupportable claim that
crime scene hairs “very highly likely” came from Woodall. Five years later,
Woodall was exonerated by DNA testing.47

Woodall promptly fi led a civil rights lawsuit. An initial inquiry by the
state’s insurer and an internal audit found no other errors. However,
when Woodall’s civil suit settled for a large but undisclosed sum, a pros-
ecutor requested another audit.48 After all, Zain had conducted grossly
erroneous analysis, made up results, and off ered unscientifi c evidence on
the stand. There was no reason to think Woodall was the only victim of
this junk science. Zain was a supervisor— he had been promoted to chief
of serology at the laboratory. What did Zain do in other cases? What were
the analysts he supervised in the lab doing? How many other people were
falsely convicted based on such fl awed science?

The state’s highest court, the West Virginia Supreme Court of Ap-
peals, convened an extraordinary investigation. The court appointed the
Honorable James O. Holliday, a retired judge, to supervise an investiga-
tion of the crime lab. Judge Holliday asked scientists from the Laboratory
Accreditation Board of the American Society of Crime Laboratory Direc-
tors (ASCLD) to investigate the policies, procedures, and work of the
laboratory. The ASCLD team concluded that fraud by Zain was “the re-
sult of systematic practice rather than an occasional inadvertent error.”
The auditors found irregularities in most of the cases reviewed and rec-
ommended a more searching inquiry into more cases the lab had han-
dled.49 Judge Holliday found shocking failures. The lab had no written
quality assurance program or procedures for auditing the work of ana-
lysts. There was no routine profi ciency testing of lab technicians. The lab
had no technical review of the analysts’ work and no procedures manual,
and the testing procedures used were unscientifi c. There was inadequate
recordkeeping. There was a total breakdown in supervision and training,
which went unnoticed for years; after all, Zain was responsible for super-
vision and training.

The court ordered the state to fi le within sixty days a report outlining
the steps it would take to obtain accreditation of the laboratory. The
court ordered an inquiry into all cases in which Zain presented evidence
(more than 133). The prisoners would be notifi ed that their cases were
under review. The court would presume that Zain’s involvement tainted

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

254 Reforming the Criminal Justice System

their cases. All cases in which Zain’s evidence “would have been suffi –
cient to support the verdict” would be reversed. Nine more convictions
were reversed.50

The saga did not end there. After all, the court’s order did not address
the work by other analysts in the lab, some of whom were supervised by
Zain. All of the analysts had worked without proper supervision or quality
assurance. The court concluded in 2006 that there was evidence of larger
impropriety in the laboratory going beyond Zain’s cases. While the court
did not fi nd intentional misconduct by other serologists, it found errors
that were “frequent, recurring and multifaceted, spanning the spectrum of
examiners,” and that the errors represented “a divergence from good sci-
ence and on occasion ethical conduct” raising “a strong inference that the
problems were systemic in the Serology Division.”51 Again, that was no
surprise, but it took the court thirteen years to get there. Slowly, though,
the court’s eff orts led to reform of the crime laboratory and reinvestigation
of old cases.

Most of the few crime lab audits that have been conducted have, as in
West Virginia, uncovered systemic problems and additional errors. Au-
dits have been conducted at the FBI laboratory, and laboratories in Cali-
fornia, Illinois, Mary land, Michigan, Missouri, New York, North Caro-
lina, Ohio, Oklahoma, Texas, Virginia, and Washington, among others.
Following two DNA exonerations that both involved the fl awed use of
forensic science, the Houston Police Department crime laboratory un-
derwent a particularly extensive audit, hiring attorney Michael R. Brom-
wich to lead an in de pen dent investigation. Meanwhile, the audit re-
viewed more than 3,500 cases and conducted a comprehensive review of
crime lab operations. Auditors uncovered hundreds of cases involving
improper serology analysis beyond the two DNA exonerations that
sparked the investigation. Auditors found that 21% of serology cases ex-
amined were fl awed and 32% of DNA cases examined were fl awed, in-
cluding those of four death row inmates. They found that analysts rou-
tinely claimed to fi nd “inconclusive” results, when actually the results
conclusively ruled out the suspect.52 The audit reports led to a shutdown
and “massive restructuring” at the crime laboratory, and meanwhile the
Houston district attorney created a new unit to review convicts’ claims of
innocence.53

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 255

The systemic problems uncovered by such audits suggest that there
may be similar problems festering in other jurisdictions, including those
in which invalid and unreliable forensic analysis played a role in wrong-
ful convictions. Yet, most of the jurisdictions have never audited their
crime labs. In its report on the use of forensic science in criminal cases,
the National Academy of Sciences (NAS) proposed that there should be
a national in de pen dent scientifi c agency tasked with investigating errors
and assuring that standards are complied with. A few states have created
in de pen dent bodies to review their crime laboratories.54 However, most
have not.

While many state and local crime labs now voluntarily participate in
ASCLD/LAB (a voluntary peer accreditation and review or ga ni za tion),55
voluntary programs run by crime lab directors themselves, although a posi-
tive step, fail to address one of the most important problems in these wrong-
ful conviction cases— the lack of consistent, scientifi cally valid guidelines
for conclusions. Nor does such accreditation provide for in de pen dent and
blind auditing of crime lab work. For example, despite years of accredita-
tion of the North Carolina state laboratory, ASCLD/LAB never noticed
systemic problems that were later uncovered after a scandal. This was no
surprise. ASCLD/LAB examined actual cases only once every fi ve years,
and its reviewers would only look at the cases handpicked by the lab super-
visors.56 In response to the NAS report, the forensics community can adopt
clearer standards.57 However, unreliable techniques also require additional
scientifi c research before they can be dependably used in our courtrooms.
All labs should be subject to external oversight, including regular and blind
audits. In addition, crime labs should be in de pen dent from law enforce-
ment, so that they do not report to police. All examiners should be blind-
tested for profi ciency. The defense should have access to underlying bench
notes and laboratory reports, and to their own defense experts. In short,
our entire system of forensic science requires a complete overhaul. Judges
could also help, by following the mandates of the Supreme Court’s decision
in Daubert, which requires that only reliable and valid methods and con-
clusions are presented to a jury. Careful judicial rulings could create the
incentives for crime labs to adhere to higher standards, but so far judges
have largely abdicated their responsibility as gatekeepers of sound science
in criminal cases.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

256 Reforming the Criminal Justice System

Jail house In for mant Reforms

Unlike the other types of evidence that contributed to these wrongful con-
victions, the use of jail house in for mants has not received serious scrutiny.
Criminal procedure rules do not closely regulate reliability of in for mant
testimony. Nor do many states require reliability hearings, enhanced dis-
closure, or jury instructions regarding jail house in for mants. California re-
quires that the jury be instructed on the potential unreliability of jail house
in for mant testimony, as do several other states, most notably Oklahoma,
which also mandates “complete disclosure” of any negotiations between
prosecutors and the in for mant.58 Illinois requires that the trial judge con-
duct a hearing to evaluate the reliability of an in for mant, before the trial, as
do courts in Oklahoma and Nevada in some circumstances.59 Such eff orts
are modest improvements.

Guy Paul Morin was convicted of a rape and murder of a nine- year- old
girl in Ontario, Canada. At his trial, two jail house in for mants claimed they
overheard him cry out at night from his jail cell that he had killed the little
girl. Years later, DNA testing exonerated Morin, and it emerged that not
only did the jail house in for mants lie, but that despite testimony to the con-
trary, they had received deals from the prosecution in exchange for the tes-
timony. In response to this high- profi le exoneration, the Supreme Court of
Canada recommended that judges warn jurors about the unreliability of
jail house in for mants. Ontario’s attorney general now limits the use of jail-
house in for mants to cases approved by a committee of se nior prosecutors,
and only if “justifi ed by a compelling public interest” and “founded on an
objective assessment of reliability.”60 Such reforms have been adopted in
the United States only in Los Angeles. After a scandal in the late 1980s in-
volving a jail house in for mant who lied when testifying in dozens of cases
and then showed reporters how he had concocted false admissions by other
inmates, the Los Angeles County District Attorney’s Offi ce imposed guide-
lines to “strictly control” use of in for mants and created a database to track
the use of in for mants.61 Far more needs to be done.

Judges could assess the reliability of in for mants before permitting
them to testify. More important, since contaminated in for mant testimony
may appear uncannily reliable, all conversations with jail house in for-
mants could be electronically recorded to ensure that case information is

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 257

not disclosed and that any deals are documented. The defense should be
told of prior testimony and cooperation by the in for mant, and all infor-
mation concerning any leniency deal between the in for mant and prosecu-
tors should be formalized and disclosed up front. Prosecutors can adopt
guidelines requiring the careful use of in for mants and careful documen-
tation of their statements. Our criminal procedure should be reformed to
prevent the abuse of in for mant testimony— particularly jail house in for-
mant testimony.

Death Penalty Reforms

Revelations of innocent persons on death row have permanently altered
the death penalty debate. Regardless of whether one believes that the
death penalty is justifi ed in some circumstances, the knowledge that in-
nocent people have been sentenced to death can shake one’s confi dence
that the death penalty can be soundly imposed. Few of the DNA exon-
erations involved crimes that carried the death penalty. Seventeen of the
exonerees were sentenced to death, and several came within days of ex-
ecution. However, as we have seen, the vast majority of capital cases do
not involve biological evidence that can be DNA tested. The exonerees’
cases suggest others on death row might also be innocent. Capital con-
victions are notoriously reversal prone, and as discussed in Chapter 7,
many of these exonerees had their convictions reversed even before they
were exonerated. Six of the seventeen death penalty exonerations in-
volved jail house in for mants. Seven involved false confessions, three of
which were confessions by mentally disabled persons. Eleven were cases
in which the postconviction DNA testing not only exonerated the defen-
dant but identifi ed the actual perpetrator.62

Those few but powerful examples of DNA exonerees who faced exe-
cution, despite the fl awed evidence in their cases, have infl uenced judges
and policymakers. For example, in Baze v. Rees, Justice Stevens an-
nounced his opposition to the death penalty, citing evidence from DNA
exonerations. He wrote that “given the real risk of error in this class of
cases, the irrevocable nature of the consequences is of decisive impor-
tance to me. Whether or not any innocent defendants have actually been
executed, abundant evidence accumulated in recent years has resulted

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

258 Reforming the Criminal Justice System

in the exoneration of an unacceptable number of defendants found guilty
of capital off enses.”63 Federal district judge Jed Rakoff struck down the
federal death penalty, arguing that “We now know, in a way almost un-
thinkable even a de cade ago, that our system of criminal justice, for all
its protections, is suffi ciently fallible that innocent people are convicted
of capital crimes with some frequency.”64 But his ruling was later re-
versed by the Second Circuit Court of Appeals.65 Perhaps these seven-
teen exonerees’ cases have had an outsized impact because they are part
of a much larger group of 138 death row exonerations, most not involving
DNA testing.66 Polls also suggest that exonerations may explain lagging
public support for the death penalty.67

Reacting in part to these exonerations, states in numbers not seen in
de cades have adopted reforms and even moratoria on the death penalty.
This began in Illinois, with then governor George Ryan announcing a
moratorium and, on his last day in offi ce, commuting the sentences of all
prisoners on the Illinois death row. The Illinois Commission on Capital
Punishment, the fi rst innocence- related commission in the United States,
issued a report with eighty- fi ve detailed recommendations for reform,
some of which were adopted. The Illinois Commission then became a
model for eff orts to create a “bullet- proof ” and more accurate death pen-
alty. On the other hand, the presence of highly probative scientifi c evi-
dence, such as DNA, can also reinvigorate support for the death penalty.
After all, in some cases, as Justice Scalia has noted, it can provide a “sci-
entifi c means of establishing guilt” and a “highly eff ective way to avoid
conviction of the innocent.”68 A Mary land law followed that approach,
narrowing the death penalty and limiting its imposition to cases with
certain types of forensic or video evidence, including videotaped inter-
rogations.69 The hope was to limit death sentences to cases with more
reliable evidence. Similarly, the Massachusetts Governor’s Council on
Capital Punishment recommended a series of reforms that would, if the
death penalty were reinstated, limit it to cases where there was greater
confi dence in the accuracy of the verdict. The proposal failed and the
death penalty was not reinstated.70 In states like Florida, Texas, and Vir-
ginia that do carry out the death penalty and have large death rows, and
that have also had high- profi le death- row DNA exonerations, no such re-
forms have been adopted.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 259

Prosecution Reforms

As then U.S. attorney general (and later Supreme Court justice) Robert
Jackson famously put it, “[t]he prosecutor has more control over life,
liberty, and reputation than any other person in America.”71 Prosecutors
have far more power today, due to the rise of plea bargaining, severe sen-
tencing guidelines, and mass incarceration. With their broad discretion
and expansive power, prosecutors may be the people best situated to ef-
fectively prevent wrongful convictions.72

Dallas County has had more DNA exonerations than any other local
jurisdiction in the United States (19 of the 250 exonerees). This does not
necessarily mean that the quality of justice was worse there. Part of the
explanation is that the crime lab regularly used by Dallas County rou-
tinely saved forensic evidence from rape cases in the 1980s. A few prose-
cutors have adopted reforms and conducted case reviews to locate addi-
tional erroneous convictions. None has done so more aggressively than
Dallas County district attorney Craig Watkins. He hired a former inno-
cence project attorney to create a “Conviction Integrity” unit to look at
possible wrongful convictions. The offi ce policy is to conduct postcon-
viction DNA testing. The offi ce adopted reforms in lineup procedures
and now videotapes interrogations. The offi ce also adopted an “open-
fi le” policy providing the defense with access to prosecution fi les.73

Because the Supreme Court’s ruling in Brady v. Mary land only re-
quires prosecutors to disclose material exculpatory information, prose-
cutors need not turn over everything to the defense, or even all of the evi-
dence that might shed light on innocence, but rather only the evidence
that they decide is suffi ciently “material.” Prosecutors need not disclose
anything at all in cases resolved by a guilty plea. As a practical matter,
that means that prosecutors can encourage an innocent person to plead
guilty to a lesser charge out of fear of the harsher consequences that might
result if the case went to a jury, even if the police or prosecutor know that
there is evidence the person is innocent, without ever sharing that evi-
dence with him or her. Open- fi le policies provide enhanced discovery to
the defense, although how open they actually are depends on the policy
and practice of the prosecutor’s offi ce. In theory, open- fi le policies can
increase accuracy by improving the fl ow of information from police to

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

260 Reforming the Criminal Justice System

prosecutors as well as to defense lawyers. Indeed, if defendants are confi –
dent that there is not helpful evidence in the fi le, they may be more willing
to settle, thus saving the state the cost of a trial.

Other prosecutors’ offi ces have pushed for reforms. In Minnesota, the
Ramsey County district attorney’s offi ce hosted a statewide conference
on improving eyewitness identifi cation practices.74 The offi ces of several
state attorneys general have issued reformed eyewitness identifi cation
policies and recommendations on videotaping interrogations.75 The
Suff olk County, Massachusetts, district attorney’s offi ce convened a task
force and adopted eyewitness procedure reforms after a series of DNA
exonerations (including the Neil Miller case discussed in Chapter 3).
They credit the new eyewitness procedures with helping to convince ju-
rors that they have arrested the right person.76 Some prosecutors’ offi ces
have conducted inquiries postconviction to identify what went wrong,
such as the prosecutor’s report on Jeff rey Deskovic’s false confession.
Those offi ces are exceptional. Most prosecutors have not investigated
what went wrong after an exoneration, nor have they altered their policies
in response.77

Defense Reforms

In contrast, defense lawyers may be some of the most poorly situated to
prevent wrongful convictions, unless steps are taken to improve the qual-
ity and funding of government- provided lawyers for defendants who can-
not aff ord to hire a lawyer. In the past, criminal procedure did not focus
on the reliability of trial evidence, because judges trusted the “crucible of
the court,” in which the cross- examination of talented attorneys would
test the evidence before the jury. That trust in the trial pro cess may have
been misplaced. In all too many of these trials, the prosecution did not
have a strong adversary. Chapter 6 described how these innocent people
often received a weak defense. Even in these fairly serious felony trials,
and even where the State’s evidence was deeply fl awed, the defense often
did not expose those fl aws. They did not have funding to hire experts or
investigate leads, such as possible evidence of innocence.

Many jurisdictions grossly underpay the lawyers they provide to poor
criminal defendants, but it is not easy to demand that the government

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 261

reallocate resources for po liti cally unpop u lar criminal defendants. In
some jurisdictions, there is no public defender offi ce to take responsibil-
ity for an error. The court- appointed lawyer who took an innocent per-
son’s case may have by then retired (or sometimes even been disbarred).
A few exonerees, like Jimmy Ray Bromgard, have fi led civil rights law-
suits following their exonerations, seeking improved indigent defense
systems; few have succeeded. Some jurisdictions have made notable ef-
forts to improve training of defense lawyers, create public defender of-
fi ces, expand the defendant’s access to evidence available to the prose-
cution, and broaden access to investigators and expert witnesses. Yet
although exonerations have highlighted the problem, already extensively
documented by studies of the woeful state of indigent defense ser vices,
in many jurisdictions the problem remains unchanged.78

Federal Reforms

In contrast with states and localities, there is much less to say about what
the federal government has done in response to DNA exonerations. Con-
gress did enact an Innocence Protection Act in 2004 designed to help
prevent wrongful convictions. The act, among other things, created a
right for federal convicts to access DNA testing, rules to require preserv-
ing biological evidence in federal criminal cases, and funding for better
repre sen ta tion in state death penalty cases. Kirk Bloodsworth, the fi rst
death row DNA exoneree, spoke out repeatedly in favor of the legisla-
tion’s passage. A section of the act was named after him, creating a Kirk
Bloodsworth DNA Testing Program providing grants to states to con-
duct postconviction DNA testing. Yet Bloodsworth, along with exonerees
Marvin Anderson and Charles Chatman, returned to the U.S. Congress
in 2008 to testify at hearings that were called to investigate why the De-
partment of Justice (DOJ) had yet to distribute any of the funds to the
states. Worse, although Congress had expanded the Paul Coverdell grant
program that gives crime labs funds to improve their work, the DOJ ig-
nored a key requirement that to be eligible states must have in de pen dent
entities to investigate allegations of serious negligence or misconduct by
crime labs. For years the DOJ did not enforce that requirement (nor has
DOJ done so as of this writing).79

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

262 Reforming the Criminal Justice System

Other reform eff orts are progressing slowly. Congress appropriated
funds for the National Academy of Sciences to study the needs of the fo-
rensic science community. The NAS report, as discussed in Chapter 4,
called for an overhaul of forensic science and the creation of a federal Na-
tional Institute of Forensic Science. It remains to be seen whether Con-
gress will enact legislation to implement those recommendations.

The DOJ has a mixed record in fostering criminal justice accuracy,
shifting only slightly from one administration to the next. For example,
the DOJ opposed enactment of the Innocence Protection Act.80 The DOJ
did convene the Commission on the Future of DNA, a working group to
promote standards on preservation of biological evidence, and convened
a Working Group for Eyewitness Evidence that recommended reformed
identifi cation procedures.81 The DOJ could do more, including by study-
ing and implementing reforms. In jurisdictions plagued by systemic er-
ror, the DOJ could discourage future constitutional violations by suing
police departments.

The federal courts have also remained on the sidelines. Few exonerees
obtained relief or access to DNA testing from federal courts. Federal courts
have not reconsidered the criminal procedure rules surrounding eyewit-
ness identifi cations, confessions, forensic science, and in for mants, which
permit grossly unreliable evidence at trial. Well into the DNA age, the
U.S. Supreme Court has still not recognized a claim of innocence under
the U.S. Constitution. Perhaps over time federal judges will take heed of
lessons learned from these exonerations.

The Tip of the Iceberg Revisited

These 250 exonerations cannot tell us how many people have been
wrongly convicted in the United States. That is one of the most haunting
features of these exonerations. If known exonerations are the tip of an ice-
berg, we do not know how big the rest of the iceberg is. We cannot know.

Some hardened souls will remain untroubled by DNA exonerations.
For example, Justice Scalia has suggested that known wrongful con-
victions are an inconsequential percentage, an error rate of “.027
percent— or, to put it another way, a success rate of 99.973 percent,” if
one divides exonerations by the fi fteen million felony convictions during

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 263

the same time period.82 But should we really be so reassured by the
numbers?

If you eat at a fi ne restaurant and complain of a large bug in your soup,
you are not reassured if the waiter tells you, “Don’t worry, it will not hap-
pen again too often. There have only been a few hundred reported cases
of bugs in soup in the United States. While human error is inevitable,
with millions of bowls of soup served every year, we have an unparal-
leled sanitary soup rate.” The waiter adds, before turning away with a
fl ourish, “Because we found the bug in your soup, the system worked.”

You had better ask to talk to the manager. The restaurant did not fi nd
that bug— you did. The system did not work. What system was there?
The waiter did not tell you what the restaurant does to keep its soup
insect- free. Even if it is true that there are just a few hundred reported
cases of bugs in soup nationwide, how good are those reports? Perhaps
they rely on the restaurants’ own willingness to report bugs in their soup.
How often do less- fancy restaurants check for bugs? Perhaps health in-
spections of soup began just a few years ago. Perhaps only some kinds of
restaurants or types of soup are inspected. If nothing unusual explains
how this bug got there, perhaps it happens all the time. You also ask your-
self how many customers decide it is not worth complaining about a bug
in their soup. You then lose what remains of your appetite by wondering
how many customers do not notice bugs because they swallowed them.

Similarly, it is not reassuring to point out that millions of felony con-
victions do not result in exonerations. In fact, the argument raises even
more cause for concern, because not only are those millions of cases not
a relevant comparison group, but we have never examined those mil-
lions of felony convictions. The unremarkable criminal trials of so many
of these exonerees looked no diff erent than those of countless others.

Instead of comparing the number of known exonerations to the num-
ber of known felony convictions, we might instead ask how high the exon-
eration rate is for the types of cases that typically result in exonerations:
rapes and rape- murders.83 DNA exonerations fall within a narrow band
of cases that are a fraction of 1% of felony convictions. Most DNA exoner-
ees were convicted of a rape in the 1980s, before DNA testing was avail-
able. DNA testing is done chiefl y in rape cases, and rape convictions that
result in prison sentences make up less than 2% of felony convictions.84

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

264 Reforming the Criminal Justice System

However, the right comparison group for DNA exonerees is even
smaller. Exonerees were almost all convicted of rapes involving a stranger-
perpetrator, in which there was a real question about who committed the
crime. They were almost all convicted at a trial, unlike most convicts who
plead guilty. They received long sentences; on average they served thir-
teen years before DNA exonerated them. There is no good data on rape
convictions in the 1980s. We do know that only about a quarter of rape
prosecutions involve stranger- perpetrators, and rape convicts overwhelm-
ingly pleaded guilty and received much shorter prison sentences than
these exonerees did.85 So the right control group for the exonerees would
be stranger- rape cases from the 1980s in which the defendant took his case
to trial. The number of these cases is much, much lower than the fi fteen
million total felonies cited by Justice Scalia, most likely in the low tens of
thousands, making the number of exonerations quite troubling.

Error rates may be far higher. Recall from Chapter 1 that a federal in-
quiry in the mid- 1990s, when DNA testing before trial fi rst became com-
mon, found that 25% of suspects were cleared by DNA tests. In contrast,
most people convicted in the 1980s before DNA testing was common may
not have been able to get DNA testing. Most would have already served
their time by the time DNA testing was invented. If they did happen to
hear about DNA testing and tried to get it, they may have been refused
access, because there was no established way to get access postconvic-
tion. They may have given up hope, perhaps for good reason. In addition,
crime scene evidence was not usually preserved in the 1980s. Law en-
forcement routinely destroyed the rape kits that could have allowed
broader use of postconviction DNA testing. Some jurisdictions did pre-
serve evidence in the 1980s. If one of these jurisdictions were to conduct
a full- scale DNA audit on cold cases, it could make history, but none has
done so yet.86

Even if known exonerations do represent a high error rate among
comparable cases, far more troubling are the vast numbers of cases in-
volving other types of crimes that cannot be examined using DNA and
where we just do not know how many have been wrongly convicted. As
powerful as DNA testing is, it is limited to a small set of mostly rape
convictions, mostly from the 1980s, in which the evidence happened to
be preserved. Those rape cases both overstate and understate the prob-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 265

lem. They overstate the problem because those par tic u lar types of rape
cases are no longer as much of a risk for wrongful convictions, now that
DNA testing is routine in criminal investigations and particularly in
rape cases. They understate the problem because they do not tell us
about the vast majority of criminal cases, in which the underlying
problems remain and DNA testing cannot be used to prevent wrongful
convictions.

The problems that occurred in these cases— contaminated confessions
and eyewitness accounts, shoddy forensic analysis, and the use of inmate
informants— are just as likely to infect nonrape cases where DNA will
never be available. Tens of thousands of cases each year involve eyewit-
ness identifi cations, in which police used many of the same error- prone
procedures as in these exonerees’ cases. Just as a rape victim can pick out
an innocent person in a photo array due to suggestive remarks by the po-
lice, an eyewitness to a robbery can pick out the wrong person due to
such remarks. The vast majority of forensic work by crime labs involves
techniques similar to those used in exonerees’ trials, raising the same
questions of reliability and validity. Just as a forensic analyst can exagger-
ate the importance or meaning of blood tests to a jury in a rape case, a
ballistics expert can exaggerate the signifi cance of similarities between a
bullet and a par tic u lar make of gun in an armed robbery case.

No DNA test will correct these problems. Rather than trying to calculate
an error rate for criminal trials— an impossible task given the unavailability
of DNA testing in most cases— we need to do what we can to prevent the
dangers that these exonerees’ cases can warn us about. We cannot know
how big the iceberg is. But we can try to fi gure out what errors in the exon-
erees’ cases are also likely to be occurring in other cases, even the ones that
aren’t amenable to DNA tests.

Systemic Errors

The Supreme Court said in its Osborne ruling, denying a request for post-
conviction DNA testing, that our criminal justice system, “like any human
endeavor, cannot be perfect. DNA evidence shows that it has not been.”87
But just because the system is a “human endeavor” does not mean that we
blithely accept that things will go wrong. The question is whether human

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

266 Reforming the Criminal Justice System

failures are the result of a few incompetent people, or whether they instead
represent a larger systemic problem that can be solved. It would be very
disturbing if malicious people intended to frame these innocent defen-
dants. But in some ways it would be a relief: those bad actors would also be
an isolated problem— a few “bad apples.” When organizations make a seri-
ous error, a superfi cial explanation, that an employee was a “bad apple,”
does not tell us whether bad supervision or practices played a role. The
trial rec ords in the exonerees’ cases suggest that perhaps not all but cer-
tainly most of the police, prosecutors, forensic analysts, and others in
these cases acted in good faith and followed standard practices.

Well- intentioned people may make errors precisely because they see
themselves as members of a group. Modern psychology has shown that in
a host of ways, our beliefs and hopes and desires infl uence what we per-
ceive and how we reason and behave.88 Such eff ects are termed “cognitive
bias.” Cognitive bias can occur where people tend, for example, to see
themselves in a positive light. In criminal cases, for example, police may
see themselves as doing justice by only investigating the guilty. The result
can be “tunnel vision” or confi rmation bias. Once people form a belief,
they tend to adhere to it and look for evidence that fi ts, or confi rms, their
preconceived idea. When police form a hunch that a suspect is guilty,
they may then without realizing it discount any evidence that does not
jibe with their prior belief in the suspect’s guilt.89 We saw in Chapter 2
how detectives downplayed inconsistent details in interrogations and in-
stead emphasized that the suspect had volunteered accurate details. We
saw in Chapter 4 how forensic analysts in case after case infl ated their
statistics and conclusions that evidence supported guilt, and discounted
evidence of innocence.

A remarkable study of fi ngerprint analysis highlighted how cognitive
bias can alter conclusions of experts. The authors, Itiel Dror, David
Charlton, and Ailsa Peron, gave fi ve se nior fi ngerprint experts a pair of
prints that those same experts had declared to be a “match” in 2000.
Other experts, unknown to the fi ve participants, verifi ed that they were
indeed a match. This time, when asked if the prints matched, four out of
fi ve of these experts changed their mind and now called the prints non-
matches. What changed? This time these fi ve experts were also given a
piece of outside information. They were told that these were the prints

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 267

that had been falsely matched by the FBI to crime scene prints left by the
terrorist involved in the Madrid bombings. In that embarrassing and
high- profi le mistake, the FBI initially targeted an innocent man from
Portland, Oregon, before realizing his prints did not match. Although
fi ngerprint experts sometimes claim infallibility, most of these experts
changed their conclusions because they received outside information.
The title of the study put its point well: “Contextual information renders
experts vulnerable to making erroneous identifi cations.”90

Cognitive biases are not at all unique to police or forensics work, and
they can be counteracted. Scientists routinely use blind studies to coun-
teract bias. Experiments use placebo groups, because test subjects may
be subject to bias. Double- blind eyewitness identifi cation procedures do
precisely that; it is important that the offi cer does not know who the sus-
pect is, but also that the eyewitness knows that the offi cer does not know.
Forensic scientists do not do their work blind and with blind audits, as
some have recommended, and the result may be that, as in the fi ngerprint
study, they hear irrelevant case information that biases their conclusions.
Thus, the analyst in Frederick Daye’s case agreed at trial that before she
did any blood testing she was told about a range of facts surrounding the
victim’s ordeal.91 Sometimes the existence of outside information leads
the expert to forgo giving an opinion altogether: I noted in Chapter 2 how
in Lafonso Rollins’s case, the crime lab decided not to do DNA testing
because he had already confessed.

Good managers try to get at the “root cause” of the problem. To do
that, they ask questions about what went wrong and what could prevent
it. Those questions are typically not asked after exonerations. Take as an
example the case of Jimmy Ray Bromgard, which is representative of how
many jurisdictions respond to exonerations. He was convicted based on
invalid forensic testimony by the manager of the Montana crime lab,
whose unscientifi c testimony was discussed in Chapter 4. This was no
isolated error. Chester Bauer and Paul Kordonowy were also wrongly
convicted based on testimony by the same analyst. Those three DNA ex-
onerations have not yet led to an examination of the other cases the ana-
lyst worked on, much less statewide reforms. The State disregarded a
formal request that an audit be conducted, following a review of the
Bromgard case by top scientists. The Montana Supreme Court refused to

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

268 Reforming the Criminal Justice System

order an audit. The attorney general refused to investigate the lab either,
saying it would be an “expensive, tedious pro cess.”92

In most jurisdictions, offi cials similarly found errors too “tedious” to
be worth investigating. In some, it was apparent long before the DNA
exoneration that there was a larger problem, and yet no action was taken.
In Oklahoma, as early as 1988, a series of state courts found that forensic
analyst Joyce Gilchrist presented misleading if not false evidence. The
courts took no further action to investigate Gilchrist or the lab. In 2001,
the governor fi nally ordered review of the thousands of cases she had
worked on— only after six wrongful convictions had come to light
through DNA testing.93

It is hard to imagine an error of greater signifi cance in our criminal
justice system than a wrongful conviction. Yet the response to these er-
rors has often been inaction. What explains this? Part of the problem is
that no one is “fl ying the plane.” Criminal justice is fragmented. Local
police have separate and sometimes overlapping authority, as do prose-
cutors. Centralized judicial review occurs in the very few cases that go to
a trial, and judges typically examine possible problems in individual
cases, but not across patterns of cases.

In addition to fragmentation, our criminal justice system also suff ers
from a lack of accountability. No institution is eager to self- criticize, but
most institutions also cannot aff ord to simply ignore serious mistakes.
Criminal justice actors may be the exception. Police know that a shooting
may result in a civil rights lawsuit and immediately convene top brass to
investigate. However, police or prosecutors often do not respond to wrong-
ful convictions, since they face no consequences for doing nothing. Civil
lawsuits for wrongful convictions face often insurmountable obstacles, and
disciplinary actions against police and prosecutors who engage in miscon-
duct, much less criminal prosecutions, are almost unheard of. Since crimi-
nal justice actors are not held accountable, they have no incentive to ask
why errors happen or to fi x the underlying causes. As a result, the same
well- meaning but unreliable work and the same fl awed procedures may re-
sult in countless other fl awed convictions.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 269

Preventing Wrongful Convictions

The U.S. Supreme Court’s complacent remarks about the inevitability
of error in “any human endeavor” should trouble us. Human error may
be inevitable, but it can be minimized. As in any human endeavor, errors
in the criminal pro cess should not be cavalierly ignored, but should be
investigated and prevented. A landmark study of medical error by the
National Academy of Sciences, titled “To Err Is Human,” shows how
human error can be taken seriously, and it raises parallel problems with
eff ective solutions. Obviously doctors have very diff erent jobs than po-
lice, forensic analysts, judges, prosecutors, and defense lawyers. Just as a
wrongful conviction is a tragedy, incorrect diagnoses or medical mis-
takes can also result in grievous errors. Just as in the criminal justice
system, medical errors are frequent, and reported errors are the tip of
the iceberg because many people either do not know that their medical
caregiver made an error or do not bother or cannot aff ord to report the
error or fi le a lawsuit if they do know. And just as a lack of centralization
causes a lack of accountability in the criminal justice system, the “decen-
tralized and fragmented nature of the health care delivery system” causes
unsafe conditions.

The solutions off ered to the medical profession in the NAS report are
illustrative here. The NAS suggested shifting the focus “from blaming
individuals for past errors to a focus on preventing future errors by de-
signing safety into the system.” It called for systemwide eff orts to collect
information about errors and design systems to prevent them. The NAS
report concluded, “to err is human, but errors can be prevented.”94

A system as simple as a checklist can save lives. Infections are a lead-
ing cause of death in hospital intensive care units, and one procedure
that can lead to potentially deadly bloodstream infections is the routine
use of a catheter. Dr. Peter Provonost began a safety initiative in Michi-
gan focusing on following a list of fi ve steps when using a catheter. The
fi rst step was simple: making sure to wash one’s hands. When doctors
were trained to follow the fi ve steps, and had their compliance checked
and discussed monthly, the rate of blood infections dropped sharply,
falling to zero in many hospitals. A study of the initiative estimated that
nearly 1,800 lives had been saved over four years. A straightforward

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

270 Reforming the Criminal Justice System

system served to remind doctors of something as basic as remembering
to wash their hands, a step that may seem trivial, but focusing on that
problem saved thousands of lives in a hospital setting.95

In a criminal investigation, police must collect accurate information,
not to diagnose illness but to solve and punish crimes. Yet criminal in-
vestigations lack a quality control supervisor on the job. There is no pro-
cess for collecting and analyzing information about errors. The criminal
justice system lurches forward in assembly line fashion, pro cessing mas-
sive numbers of cases in stages, with a range of actors siphoning off cases
by exercising discretion at each stage. The pro cess begins by almost ex-
clusively relying on the memory and notes of detectives, and not on care-
ful rec ords or electronic recordings. Lawyers then rely on poorly re-
viewed forensic tests in which the underlying rec ords are not disclosed,
or on the fragile and fading memory of witnesses. The farther a case
moves down the assembly line, from police investigations, to prosecu-
tor’s charging, to defense investigation, to plea bargaining, to trial, and
to appeals and postconviction review, the harder it is to undo an error.
Once a confession is contaminated, judges and others cannot easily un-
ravel what transpired— unless the interrogation is videotaped.

The reforms that I have advocated in this book share a family resem-
blance. At their core, they each involve placing less reliance on unreliable
human memory and decisions early on in the criminal investigations.
None of these problems are new. Law professor Edwin Borchard de-
scribed eyewitness misidentifi cations and false confessions in his classic
book on wrongful convictions— written in 1932. Judges have discussed
potential for error in criminal procedure rulings dating back several de-
cades. What makes these wrongful convictions so troubling is that re-
search has exposed how we routinely rely on unreliable evidence in crim-
inal cases. Yet solutions are available, and they all involve safeguarding
the reliability of evidence.

There is the fear of the unknown— that changing criminal procedure
rules to focus on accuracy might be counterproductive, creating new bi-
ases or perverse incentives that could result in more guilty people going
free. But if it is true that the vast majority of suspects are guilty, improved
documentation may strongly benefi t the State. The assembly line pro cess
also makes mistakes by letting the guilty go free— as it did in all of these

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 271

exonerees’ cases. Prosecutors depend on the fl ow of reliable information
from police and they complain about inadequate note taking and record-
keeping. Along with police departments, prosecutors have pushed for the
adoption of checklists and documentation protocols, to improve the qual-
ity of evidence used to convict the guilty.96 A checklist reminding police
to write down crucial information like the confi dence of the eyewitness or
reminding them to videotape the confession can prevent costly mistakes,
just as the checklist for intensive care doctors reminding them to wash
their hands can prevent potentially deadly infections.

We know from de cades of social science that double- blind eyewitness
procedures reduce errors with no appreciable eff ect on true identifi ca-
tions. There is no evidence that videotaping interrogations discourages
true confessions, but videotaping helps to ward off spurious challenges to
true confessions. There is no evidence that improved quality assurance
and scientifi c standards reduce the accurate use of forensic evidence.
Other systemic reforms are untested— and they should be tested. In par-
tic u lar, since so many cases are resolved in plea bargains, we should far
more carefully scrutinize guilty pleas premised on unreliable types of
evidence.

We can also learn from successes and not just failures. In the vast major-
ity of criminal cases, DNA testing cannot be conducted or cannot answer
the question of identity. In the typical robbery, for example, no biological
material tied to the attacker may be left at the crime scene. Yet robberies
may be just as error prone. After all, they usually involve eyewitness iden-
tifi cations, and such eyewitnesses may have even less of an opportunity to
see the culprit than in a rape.97 Even in cases in which no DNA testing
can be done, errors can still be documented. If an eyewitness identifi ed a
lineup fi ller, one can be fairly sure a mistake was made. If a suspect con-
fesses, but it turns out that he was in prison when a crime occurred, police
will typically let him go. If forensic evidence clears a suspect long before a
trial, that person needs no exoneration. Successes in catching errors early
on in an investigation deserve recognition, and such near misses can teach
us important lessons about what can go wrong.

Over time, near misses will teach us more, while DNA exonerations
may fade away, although slowly. Now that DNA testing is common be-
fore a trial, many thousands of innocent people each year are cleared

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

272 Reforming the Criminal Justice System

before a trial and do not need to be exonerated. DNA exonerations still
occur regularly, but fewer are a result of recent criminal cases. Yet about
one- quarter of all DNA exonerees were convicted at a time when DNA
testing was available. Their lawyers failed to ask for DNA testing, the
State hid the fact that biological material existed, or DNA testing was
done but botched. DNA technology continues to improve, and as it im-
proves, DNA testing is increasingly available in more types of crimes,
such as robberies or other felonies. Nevertheless, a unique window into
causes of error will begin to close. That makes it all the more important
to learn what we can from this remarkable set of DNA exonerations.

Lessons from Exonerations

The jurors in the trial of Ronald Jones, whose story began this book,
were instructed by the judge at the end of the trial, before they began their
deliberations, “You have before you evidence that the Defendant made
statements relating to the off enses charged in the Indictment.” The judge
added, “It is for you to determine what weight should be given to the state-
ments. In determining the weight to be given to a statement, you should
consider all of the circumstances under which it was made.”98

The jurors that decided Jones’s fate could not fairly judge the evi-
dence. They had no record of what transpired during the interrogation.
Jones proclaimed his innocence, but was not believed. After all, the de-
tectives described a detailed confession and presented the written con-
fession that he signed. Similarly, the trial judge saw no problem with the
confession, even though Jones alleged police beat him and then spoon-
fed him the crime details. The appeals courts all denied relief. The evi-
dence of guilt seemed overwhelming, precisely because the confession
was contaminated. What makes the trials of exonerees so frightening is
that they show how the case against an innocent person may not seem
weak. The case may seem uncannily strong. Without DNA, it can be
impossible to turn back the clock on a criminal prosecution gone wrong.

A lesson from these exonerees’ criminal trials is that once central evi-
dence is contaminated at the earliest stages of a case, the damage cannot
be easily discovered or reversed. “What’s done cannot be undone,” as
Lady Macbeth put it, while sleepwalking and unsettled by her own grave

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Reforming the Criminal Justice System � 273

and irreversible wrongs. It is too late after police disclose facts during an
interrogation, leak facts to in for mants, conduct shoddy eyewitness iden-
tifi cations, or reach unscientifi c conclusions in forensic reports. It is too
late months later at a trial or years later when a habeas petition is fi led. It
is too late in the countless criminal convictions for which no DNA test-
ing can be done. Yet these are not “things without all remedy” to be put
out of mind, for even if we try to go back to sleep, some errors will return
to haunt us.99 Errors must be remedied before fi ction contaminates fact.

Kirk Bloodsworth wrote in a letter to Congress, “I am dedicated to
making sure this never happens to anyone else. I will continue my quest
to fi nd solutions to our broken and fl awed criminal justice system.”100
Should we be pessimistic or optimistic about actually fi xing the fl aws in
our criminal justice system? Wholesale forensics reform may be some-
what expensive, yet the federal government provides hundreds of mil-
lions of dollars to crime labs to eliminate their backlogs, a worthy goal,
but one that should not be the exclusive goal of federal grants.101 And
simple and inexpensive reforms have not been universally adopted. A
majority of police departments have not reformed eyewitness identifi ca-
tion procedures and do not videotape confessions. The defense lawyer in
Ronald Taylor’s case put the point well, making light of the failure to
document eyewitness identifi cations: “ ‘We don’t record it because it costs
the Houston Police Department too much money.’ Think about it. A
ninety- cent tape versus the cost of convicting an innocent man.”102

Ronald Jones also had his travails recounted on the fl oor of the United
States Senate. After his exoneration, U.S. Senator Russell Feingold de-
tailed “what happened in Illinois to Ronald Jones” and he concluded,
“The system doesn’t work. It has failed us.”103 Even so, Congress did not
pass any reform mea sures. But the Illinois legislature did enact a package
of reforms, including one that could have prevented Jones’s false convic-
tion: mandatory videotaping of interrogations in hom i cide cases. Although
our law of criminal procedure remains vulnerable to error, solutions are
available. Confession, eyewitness identifi cation, forensics, and access to
DNA reforms are all slowly gathering momentum across the country.
Wrongful convictions have altered the pop u lar consciousness, changing
the attitudes of everyone from jurors to voters, lawmakers, judges, scien-
tists, police, defense lawyers, and prosecutors.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

274 Reforming the Criminal Justice System

In this book, I have shown how the criminal justice system convicted
the innocent. The errors in these exonerees’ cases were not isolated acci-
dents. They were caused by systemic failures. Their cases display clear
patterns that can tell us how to make the system far more accurate. We
cannot give the 250 exonerees back the years they spent in prison for
crimes they did not commit. But we can dedicate ourselves to learning
from what went wrong.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Appendix

Notes

Acknowledgments

Index

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

277

Appendix

In this book, I present a wealth of information concerning the cases of
the fi rst 250 people exonerated by DNA testing. Each chapter contains a
separate study of one aspect of the cases of these innocent people. While
a complete summary of all of the information collected would be far too
long, I have made detailed summaries of all of the data available online
at http://www.law.virginia.edu/innocence. Here I present a visual over-
view of the book, with charts and graphics that display the fi ndings in
each chapter. Second, I provide a description of the methods used to
collect all of this information and then to analyze it.

Chapter 1, the introduction to the book, describes the general charac-
teristics of these 250 exonerees. As shown in Map A.1, DNA exonera-
tions have occurred throughout the United States, in thirty- three states
and the District of Columbia.

I obtained trial materials in 88% of the cases of those who had a crimi-
nal trial. I then described how, because DNA testing can most readily be
used in rape cases, most exonerees were convicted of rape, with some
convicted of murder and a handful of other crimes. Figures A.1 through
A.4 display some of these features of the 250 exonerees’ cases.

Chapters 2 through 5 discuss the prosecution evidence at trial. Figure
A.5 is an overview of the types of evidence discussed.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

1–2

3 5–

6 9–

10 19–

20 40–

Map A.1. DNA exonerations in the United States

Transcript: 88% (207)

Pleaded guilty: 6% (16)

No transcript: 12% (27)

Figure A.1. DNA exoneree trial materials

Rape: 68% (171)

Murder 9% (22):

Rape and murder 21% (52):

Other 2% (5):

Figure A.2. Crime of exoneree conviction

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Black 62% (155):

White 30% (74):

Hispanic 8% (20):

Asian 0% (1):

Figure A.3. Race of exonerees

1974–1980 7% (17):

1981–1985 33% (82):

1986–1990 40% (99):

1991–2008 21% (52):

Figure A.4. Year of conviction

Confession 16% (40):

Informant 21% (52):

Forensic evidence 74% (185):

Eyewitness 76% (190):

0 20 40 60 80 100 120 140 160 180

Figure A.5. Evidence supporting exonerees’ convictions

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

280 Appendix

No transcript

Not flawed

Flawed evidence

Eyewitnesses 88%

(141/161)

: Forensic

testimony

61% (93/153)

Jailhouse

informants

93% (26/28)

Confessions 95%

(38/40)

180

160

140

120

100

Figure A.6. Flawed trial evidence

Figure A.6 summarizes some ways that the trial evidence was found
to have been fl awed, due to use of suggestive eyewitness procedures or
an uncertain eyewitness, contamination of a confession statement or in-
culpatory statement to a jail house in for mant, or invalid conclusions by a
prosecution forensic analyst at trial.

The two most common types of evidence supporting these convic-
tions were eyewitness identifi cations and forensic evidence. Most of
the trial rec ords involved either suggestive or unreliable identifi ca-
tions, shown in Figure A.7. Similarly, most forensic testimony by pros-
ecution analysts was either invalid (61%) or vague (12%), as shown in
Figure A.8.

Figure A.9 summarizes the information discussed in Chapter 3, con-
cerning the eyewitness testimony at these exonerees’ trials. Some cases
had more than one feature.

Figure A.10 breaks down what types of forensic testimony were pre-
sented by prosecution analysts at the trials examined and, as detailed in
Chapter 4, what proportion of each type involved invalid conclusions.
Many cases had more than one type of forensic evidence.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

No record of suggestion

or unreliability 12% (20):

Unreliable or suggestive

88% (141)

Figure A.7. Unreliable and suggestive identifi cations

Nonproblematic and inculpatory: 7% (10)

Invalid testimony 61% (93/153):

Vague testimony 12% (19):

Nonprobative and nonproblematic 20% (31):

Figure A.8. Invalid and unreliable forensics

Discrepancy in description 63% (100):

Did not see face 9% (15):

Initially uncertain 21% (34):

Initial non-i.d. 40% (64):

Hypnotized 3% (5):

Suggestive remarks 28% (44):

Suggestive lineup: 34% (55)

Showup: 34% (53/161)

0 10 20 30 40 50 60 70 80 90 100

Figure A.9. Eyewitness misidentifi cations

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

282 Appendix

Chapter 6 turns to the defenses that exonerees raised at their trials.
Figures A.11 and A.12 depict the types of defenses raised and types of
defense lawyers who represented these innocent defendants.

Chapter 7 develops how in the cases with written decisions on appeal,
many exonerees did not challenge the trial evidence and very few who
did so succeeded. Figure A.13 shows how most challenges to trial evi-
dence were unsuccessful or no challenge was brought at all. However,

Defendant testified 53% (110):

Experts retained 30% (62):

Third-party guilt: 14% (30)

Alibi 68% (140):

0 50 100

Figure A.11. The defense case at trial

Voice spect. 100% (1/1):

Bite mark 71% (5/7):

Shoe print 17% (1/6):

DNA 17% (3/18):

Fingerprint 5% (1/20):

Hair % (29/75): 39

Serology 58% (67/116):

0 20 40 60 80 100

Invalid

Not invalid

Figure A.10. Types of forensic testimony

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Appendix � 283

the reversal rate was likely typical of that for similarly serious criminal
convictions.

Chapter 8 develops how exonerees faced additional delays obtaining
DNA testing and an exoneration and how postconviction testing often
identifi ed the culprit (see Figures A.14 and A.15). I also discuss the av-
erage time line to exoneration, from the conviction, to the end of the
postconviction pro cess, to the fi rst DNA exclusion, to the eventual ex-
oneration. The time line shown in Figure A.16 displays the average
years when each of those events occurred.

Court appointed 78:

Public defender 71:

Retained 53:

Unknown 48:

Figure A.12. Type of defense lawyer

120

100

Not challenged

Reversal

Unsuccessful challenges

Eyewitness

56% challenged

(70/124)

: Forensic

testimony

32% (36/112)

Informants

34% (16/45)

: Confessions

59% (13/22)

Figure A.13. Challenging evidence postconviction

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Average year

of conviction

(1987)

Average year

of first DNA test

(2000)

Average year of last

postconviction ruling

(1993)

Average year

of exoneration

(2002)

1985 1990 1995 2000 2005

Figure A.16. Average timeline from conviction to exoneration

0–5 years 9% (22):

6–10 years 22% (55):

11–15 years 22% (56):

16–20 years 27% (68):

21–35 years 20% (49):

Figure A.14. Number of years to exoneration

DNA hit 45% (112):

None 55% (138):

Figure A.15. DNA identifi ed real culprit

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Appendix � 285

T h e 2 5 0 E xon e r e e s

The project of this book began with a remarkable list— a list of 250 names
of people exonerated by postconviction DNA testing. By exonerated, I
mean that a court or executive vacated the conviction based on newly
discovered evidence of innocence, and either there was no trial and the
indictment was dismissed, or, in two cases, there was an acquittal at a
new trial. In each case, new DNA testing conducted after the conviction
excluded the defendant, and in 112 cases the DNA tests also confi rmed
the guilt of another person.

The authoritative list of DNA exonerations is maintained by the In-
nocence Project at Cardozo Law School and can be viewed on their
website, together with brief descriptions of these cases.1 I have checked
their list as others, as well as by conducting an extensive set of news
searches. The Innocence Project list has been complete and accurate,
as well as conservative in its criteria for inclusion of cases. The only
DNA- related exonerations that I uncovered that were not included on
the Innocence Project list were a handful of cases in which the Inno-
cence Project determined that the DNA testing did not play a substan-
tial role in the exoneration (the most high- profi le of which is the Paul
House case discussed in Chapter 8). I have adopted the same conserva-
tive view of what constitutes a DNA exoneration. This list does not in-
clude, for example, cases in which DNA testing excluded a person and
charges were dismissed before trial, cases in which there was an exon-
eration but it was not substantially based on DNA evidence, and cases
in which there has been no exoneration despite DNA evidence of inno-
cence. The list is not static. New DNA exonerations occur regularly.
Occasionally a new DNA exoneration comes to light, that occurred
years before but had not been reported in the news. I stopped examin-
ing new DNA exonerations when the 250th DNA exoneration occurred
in February 2010.

DNA testing is not infallible. Indeed, DNA testing errors contributed
to the convictions of three of these exonerees. As with any forensic tech-
nique, errors can arise from fl awed interpretation of results and from
mishandling of samples at a crime scene or in a laboratory. However,
DNA testing is uniquely probative of the identity of a rapist. Although it

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

286 Appendix

is possible that an erroneous DNA test could lead to the exoneration of a
guilty person, these high- profi le cases are closely scrutinized. Judges
and prosecutors are not eager to reverse a conviction and often duplica-
tive DNA tests were conducted. In addition, in 112 cases, the DNA tests
not only excluded, but also identifi ed the actual perpetrator. We can be
quite confi dent that these convicts are actually innocent.

Data C ol l e c t ion

I was able to collect such a large set of trial materials from these exoner-
ees’ cases with substantial assistance of others, to whom I am extremely
grateful. I collected several types of documents. Most time- consuming
was obtaining trial transcripts for 207 of the 234 exonerees who had
criminal trials. Substantial selections from those trial rec ords have been
posted online as a research resource. For thirteen of the sixteen who
pleaded guilty, I also located documents, such as pretrial hearings, testi-
mony in codefendant trials and confession statements. I obtained police
reports, including interrogation transcripts, for some exonerees as well.
The information collected for each of these chapters, with appendixes
describing the fi ndings concerning false confessions, eyewitness testi-
mony, forensic testimony, and in for mant testimony, can be viewed on-
line.2 They are far too detailed and lengthy to be reproduced here.

Second, I also obtained confession statements, laboratory reports, and
police reports where available from prosecutors, postconviction lawyers,
and innocence projects. Third, I collected all written judicial decisions
from exonerees’ appeals and postconviction proceedings. The methodol-
ogy for reviewing the claims raised and ruled upon in those rulings is
described in some detail in my study “Judging Innocence,” and data is
also available in a detailed appendix that has been updated online.3

Fourth, limited data was collected from news reports. Data concerning
events following the postconviction review, such as requests for DNA
testing, dates when DNA testing was obtained, whether prosecutors
consented to DNA testing, and whether exonerees obtained civil settle-
ments, was all collected chiefl y from news reports. Much of that informa-
tion was also supplemented by queries to members of the Innocence Net-
work and the Innocence Project, who were centrally involved in eff orts to

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Appendix � 287

secure postconviction DNA testing and could supplement and error- check
this data. Data concerning characteristics of exonerees, such as their race,
or the race of victims, was also supplemented by queries to the Innocence
Project and the Innocence Network.

Data A n a ly s i s

All of the more than 200 trial transcripts were scanned. Law student re-
search assistants conducted the initial coding for each of the sets of data
analyzed. I personally reviewed all of this coding. Peter Neufeld also re-
viewed the coding of the forensic science testimony, as we coauthored a
study of that data. Several scientists reviewed our methodology and an-
swered questions about par tic u lar testimony. I have made underlying
data available online, as discussed.

The eyewitness data was blind- coded in 60 cases, or half of the cases
that had then been obtained (ultimately 161 trials with eyewitness testi-
mony would be located). The eyewitness data involved asking questions
about (1) suggestion, or whether police used a suggestive identifi cation
procedure, and (2) reliability, or whether the eyewitness had admitted
earlier uncertainty or described a person who looked diff erent from the
exoneree in some clear and signifi cant way (such discrepancies in ap-
pearance, however, did not play an important role in the analysis). The
interrater reliability score was high, and I thank Jessica Kostelnik for her
help in calculating the score.4

Due to the broad scope of the data analyzed and collected, I did not
use multiple coders or obtain interrater reliability scores for all of the
other scores of questions coded. However, the high interrater reliability
score on the eyewitness data gives me confi dence as to the remaining cod-
ing, which involved questions far more discrete and not requiring signifi –
cant interpretation or subjective evaluation. For example, much of it
coded such concrete questions as what date did the trial begin, was there
an expert who testifi ed for the defense, was a confession introduced at
trial, and did the person bring a Brady claim postconviction. For all cod-
ing, more than one student looked at each set of data that was coded.
I then reviewed each set of data coded and read all of the underlying trial
transcripts and written opinions myself.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

288 Appendix

S u b s e t s S t u di e d

One complicating feature of the analysis of the 250 exonerees is that materi-
als could not be obtained for all exonerees. Further, in diff erent chapters
of this book, I look at diff erent subgroups of exonerees. The size of the rel-
evant subgroup is diff erent depending on the question being asked. In
Chapter 2, where the central question is how many exonerees falsely con-
fessed and had confession statements with detailed facts, I look at the sub-
group of forty who falsely confessed. I was able to obtain trial transcripts or
case rec ords for the entire set of forty exonerees who falsely confessed. In
contrast, Chapter 3 chiefl y examines not all 190 cases in which there was an
eyewitness identifi cation, but rather the subset of 161 cases involving eye-
witnesses in which trial rec ords could be located. Only in those cases
could the trial rec ords be reviewed to assess what identifi cation procedures
were used and whether the certainty of the eyewitness had changed before
trial. Chapter 7 examines only the 165 exonerees’ cases in which written
judicial decisions were located, since only in such cases can one know what
claims the judges ruled on and what reasons they gave for their rulings.

Sometimes I also discuss the entire group of 250 exonerees. When I
do so, I state this in the text or in the notes. Examples of information
obtained for all 250 exonerees include exoneree race, state of conviction,
crime of conviction, years until exoneration, types of evidence support-
ing their convictions, whether they obtained DNA testing upon consent
of the prosecutor, whether testing inculpated the actual perpetrator, and
whether they obtained compensation.

S e l e c t ion a n d G e n e r a l i z a b i l i t y

DNA exonerees are a unique and clearly “selected” sample. Throughout
I make clear that one cannot use their experiences to generalize, nor to
make strong conclusions about how our system treats the innocent more
generally or how many people are wrongly convicted. Indeed, what makes
these exonerations troubling is that we simply do not know to what extent
the same problems occur in the vast majority of criminal cases, in which
DNA testing cannot tell us whether there was a wrongful conviction.

These DNA exonerations largely consist in convictions in cases involv-
ing stranger- rapists, at a trial, in the 1980s, and resulting in lengthy sen-

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Appendix � 289

tences, because those are the cases in which DNA testing was not done at
trial but years later could shed light on who committed the crime. As dis-
cussed in Chapter 9, there was not adequate data from the 1980s, when
most exonerees were convicted, concerning others convicted of sexual as-
saults. For example, we do not know how many people were convicted at a
trial of nonacquaintance rape in the 1980s and served more than ten years
in prison. The only comparisons made with “matched” cases involving
nonexoneration cases were quite limited in nature. I concluded that simi-
larly serious rape and murder cases involving reported decisions in
“matched” cases involved a statistically insignifi cant diff erence in reversal
rate. I found that similar trials in three states during the 1980s included
some of the same types of invalid forensic testimony. Because DNA exon-
erations occur through such an unusual set of circumstances, they are not
susceptible to more robust comparisons. No more was done to compare
exonerees’ cases to matched reported cases, since the rec ords on those
cases were insuffi cient and did not often include adequate information.5
Where possible, however, I compared these cases to data concerning fel-
ony convictions, habeas corpus litigants, crime laboratory work, confes-
sions, and eyewitness identifi cations more generally.

Additional work was conducted to assess the robustness of fi ndings
describing how these exonerees litigated their cases. In collaboration
with Professor J. J. Prescott, this data on the case characteristics of exon-
erees was comprehensively examined using regressions, and using many
diff erent specifi cations and groups of variables. We wanted to be sure
that the prior analysis did not suff er from any omitted variable bias and
hoped that new relationships within the data might be uncovered. The
work developing those results is discussed in Chapter 7. J. J. Prescott and
I looked at what characterized cases in which exonerees obtained rever-
sals prior to obtaining DNA testing. We found evidence that courts
more often reversed the convictions of exonerees who claimed inno-
cence or who litigated substantive claims related to the trial evidence.
We have also found that claims of innocence, and substantive claims
generally, were associated with shorter durations from conviction to
exoneration.

More work should be done to examine these exonerations and no doubt
more will be done. I hope that the information and analysis in this book
will assist others in studying and learning from wrongful convictions.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

291

Notes

1. introduction

1. Ken Armstrong, “Part 6: The Far Limb of the Law,” Chicago Tribune, December 18,
2002.

2. Sharon Cohen, “Last- Minute Exonerations Fuel Death- Penalty Debate,” Los Ange-
les Times, August 15, 1999, A-1.

3. Trial Transcript, 779, 832, 862, 886, 1152– 1153, People of the State of Illinois v.
Ronald Jones, No. 85- 12043 (Ill. Cir. Ct. July 12, 1989).

4. Ibid., 912.
5. Ibid., 1212, 1223 (July 17, 1989).
6. Edward Blake, who conducted the testing, found a “large quantity of spermatozoa”

on the vaginal swab, suggesting that the sperm came from the person who murdered
the victim shortly thereafter. See Forensic Science Associates, Illinois v. Ronald
Jones Report (June 11, 1997), 8.

7. Ken Armstrong, “The Trials of Dick Cunningham: A Death Row Lawyer’s Search for
Mercy and Redemption, Chapter 9: Sunrise,” Chicago Tribune, December 23, 2002, 1.

8. Ibid.
9. Jones Trial Transcript, 1092– 1094 ( July 13, 1989).
10. Ibid., 1142, 1151, 1160; ibid., 64 ( July 14, 1989).
11. See ibid., 1089, 1124 ( July 13, 1989).
12. Ibid., 1213 ( July 17, 1989).
13. Ibid., 951– 954 ( July 12, 1989).
14. Those fi ve additional exonerees were convicted of other crimes, such as robbery or

attempted murder.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

292 Notes to Pages 5–7

15. Most striking, 75% of innocent rape convicts were black or Hispanic, while one
study indicates that only approximately 30% of all rape convicts are minorities. See
Sean Rosenmerkel, Matthew Durose, and Donald Farole Jr., U.S. Department of
Justice, Bureau of Justice Statistics, Felony Sentences in State Courts, 2006 (2009),
table 3.2. Chapter 3 discusses several possible explanations for this racial disparity.

16. Herrera v. Collins, 506 U.S. 390, 420 (1993) (O’Connor, J., concurring); see also
ibid., 398– 399 (citing to “constitutional provisions [that] have the eff ect of ensuring
against the risk of convicting an innocent person”).

17. Edwin M. Borchard, Convicting the Innocent: Errors of Criminal Justice (New
Haven, CT: Yale University Press, 1932), vi.

18. Hugo Adam Bedau and Michael L. Radelet, “Miscarriages of Justice in Potentially
Capital Cases,” 40 Stan. L. Rev. 21, 87 (1987).

19. The highest numbers were in Texas (40), followed by Illinois (29), New York (25),
Virginia (11), Florida (11), Oklahoma (10), Pennsylvania (10), California (9), Massa-
chusetts (9), and Ohio (8). The states with the most exonerations may not have the
most error- prone criminal courts. New York and Illinois have the oldest innocence
projects and enacted the fi rst statutes to provide access to postconviction DNA test-
ing. Texas and Virginia, on the other hand, happened to preserve biological evidence
in some 1980s cases. This suggests that many factors explain why exonerations occur
in some states more than in others.

20. See Brandon L. Garrett, “Judging Innocence,” 108 Colum. L. Rev. 55, 56– 58 (2008).
The story of the founding of the Innocence Project is told in Actual Innocence, a
book that compellingly describes the stories of the fi rst DNA exonerees and pro-
poses a series of increasingly adopted reforms. See Barry Scheck, Peter Neufeld,
and Jim Dwyer, Actual Innocence (New York: Signet, 2001). This book takes up
Barry Scheck, Peter Neufeld and Jim Dwyer’s invitation to scholars and others to
take on “the job of fi guring out what went wrong . . .” Ibid., xxiii. Before becoming
a law professor, I had the privilege to work for Peter Neufeld and Barry Scheck at the
law fi rm then called Cochran, Neufeld & Scheck, LLP, from 2002 to 2004. I repre-
sented several DNA exonerees pursuing civil wrongful conviction actions after their
exoneration.

21. Harry Kalven Jr. and Hans Zeisel, The American Jury (Chicago: University of Chi-
cago Press, 1966), 32.

22. Several exonerees’ trials could not be obtained because their cases were sealed when
they were exonerated or had already been sealed to protect the identity of the victim.
For others, the rec ords have been lost. In many jurisdictions, the court house does
not keep copies of trial transcripts, and the court reporter who transcribed the pro-
ceedings retains the rec ords. Some of those reporters retired, passed away, could not
be located, or no longer had the rec ords. For some exonerees, jury instructions or
closing statements were not transcribed, and for most, jury selection and voir dire
were not transcribed or obtained. As the Appendix describes, I have made available
online much of the data described in this book. See http:// www .law .virginia .edu/
innocence .

23. Herrera v. Collins, 506 U.S. 390, 416 (1993).

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Notes to Pages 10–16 � 293

24. United States v. Chronic, 466 U.S. 648, 656 (1984).
25. In these 250 DNA exonerees’ cases, at least 27%, or 68, received a pardon. For 84%,

or 210, exonerees, a court vacated the conviction. There is an overlap because for
some the pardon was a step that followed an exoneration by a court but was neces-
sary in order to obtain compensation. In all but eight of the cases involving a court-
ordered vacatur, a state court vacated the conviction. In the eight cases it was a fed-
eral court. In two cases, those of Gerald Davis and of Robert Miller, the prosecutor
retried the exoneree even after the postconviction DNA testing excluded them and
the court vacated the conviction. In each of those cases, the exoneree was acquitted
at the retrial.

26. See Glenna Whitley, “Chains of Evidence,” Dallas News- Observer, August 1, 2007
(describing how Dallas County’s private crime lab, the Southwestern Institute of
Forensic Sciences, was required to preserve evidence to maintain its accreditation).

27. Lauren Kern, “Innocence Lost? Despite Its Increasing Importance, DNA Evidence
Routinely Gets Destroyed Here,” Houston Press, November 30, 2000.

28. Naftali Bendavid, “Ashcroft Pledges $30 Million to Whittle Down DNA- Test Back-
log,” Chicago Tribune, August 1, 2001.

29. See Edward Connors et al., U.S. Department of Justice, Convicted by Juries, Exon-
erated by Science: Case Studies in the Use of DNA Evidence to Establish Innocence
After Trial (1996), xxviii– xxix, 20 (reporting the results of a 1995 DNA laboratory
survey, fi nding a 25% exclusion rate among the more than 10,000 cases tested by
the FBI, omitting inconclusive cases, and a higher 30% exclusion rate among the
more than 10,000 cases tested by state, local, and private laboratories); see also
William S. Sessions, “DNA Evidence and the Death Penalty,” Jurist, May 30, 2007
(noting that those statistics have remained “roughly the same” over time). Chapter
9 further discusses legal debates and scholarship concerning error rates and wrong-
ful convictions.

2 . contaminated confessions

1. Trial Transcript, 1155, 1207– 1208, 1534, State of New York v. Jeff rey Deskovic, No.
192- 90 (N.Y. Sup. Ct. Dec. 4, 1990).

2. Deskovic Trial Transcript, 1143, 1151– 1152.
3. Report on the Conviction of Jeff rey Deskovic, 2, 13 (2007), available at www

.westchesterda .net/ Jeff rey %20Deskovic %20Comm %20Rpt .
4. Deskovic Trial Transcript, 1183.
5. Ibid., 693.
6. Deskovic Report, 2.
7. Deskovic Trial Transcript, 1034.
8. Ibid., 1186.
9. Ibid., 1167, 1185, 1429, 1512– 1513.
10. Deskovic Report, 3.
11. Ibid., 19.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

294 Notes to Pages 16–19

12. Deskovic Trial Transcript, 1492.
13. Ibid., 1507– 1508.
14. Ibid., 1513, 1537.
15. Ibid., 1521.
16. Deskovic Report, 3– 4.
17. People v. Deskovic, 607 N.Y.S.2d 696, 697 (N.Y. App. Div. 1994).
18. Fernanda Santos, “Inmate Enters Guilty Plea in ’89 Killing,” New York Times,

March 15, 2007; Tony Aiello, “Deskovic Vindicated; ‘Real’ Teen Killer Confesses,”
WCBSTV.com, November 15, 2006.

19. Fernanda Santos, “DNA Evidence Frees a Man Imprisoned for Half of His Life,”
New York Times, September 21, 2006, A-1.

20. See Saul Kassin and Lawrence Wrightsman, The Psychology of Evidence and Trial
Procedure (Beverly Hills, CA: Sage Publications, 1985), 78; Fred E. Inbau et al.,
Criminal Interrogation and Confessions, 4th ed. (Gaithersburg, MD: Aspen Publish-
ers, 2001), 412.

21. Jonathan Bandler, “Deskovic Files Federal Lawsuit over His 15- year Wrongful Im-
prisonment,” Journal News (White Plains, NY), September 18, 2007.

22. John Henry Wigmore, A Treatise on the Anglo- American System of Evidence in Tri-
als at Common Law, 2nd ed., vol. 2 (Boston: Little, Brown, 1923), § 835, 867.

23. Saul Kassin et al., “Police Interviewing and Interrogation: A Self- Report Survey of
Police Practices and Beliefs,” 31 Law & Hum. Behav. 381 (2007).

24. See Kassin and Wrightsman, Psychology of Evidence, 67– 94; Richard A. Leo and
Richard J. Ofshe, “The Consequences of False Confessions: Deprivations of Liberty
and Miscarriages of Justice in the Age of Psychological Interrogation,” 88 J. Crim.,
L. & Criminology 429 (1998); Corey J. Ayling, “Corroborating Confessions: An Em-
pirical Analysis of Legal Safeguards against False Confessions,” 1984 Wis. L. Rev.
1121, 1186– 1187 (1984). For experimental work concerning false confessions, see, e.g.,
Saul Kassin and Katharine Kiechel, “The Social Psychology of False Confessions:
Compliance, Internalization, and Confabulation,” Psychological Science (1996); Saul
Kassin and Holly Sukel, “Coerced Confessions and the Jury: An Experimental Test
of the ‘Harmless Error’ Rule,” 21 Law & Hum. Behav. 27 (1997); Saul Kassin, “On
the Psychology of Confessions: Does Innocence Put Innocents at Risk?” 60 Am.
Psychol. 215, 223 (2005).

25. See Major Joshua E. Kastenberg, “A Three- Dimensional Model for the Use of Ex-
pert Psychiatric and Psychological Evidence in False Confession Defenses Before
the Trier of Fact,” 26 Seattle U. L. Rev. 783 (2003).

26. See Richard A. Leo, Police Interrogation and American Justice (Cambridge, MA:
Harvard University Press, 2008), 166 (calling on scholars to examine “the postad-
mission portion of police interrogation” and noting that “it has received far less at-
tention from scholars, lawyers, and the media”).

27. Thirty of the forty exonerees were convicted at a trial. The ten who pleaded guilty
were: Marcellius Bradford, Keith Brown, James Dean, Anthony Gray, William Kelly,
Chris Ochoa, Debra Shelden, Ada JoAnn Taylor, David Vasquez, and Thomas Win-
slow. For all of the thirty convicted at a trial, trial materials were obtained. Of those

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Notes to Page 19 � 295

who pleaded guilty, fi ve had trial materials because they testifi ed in codefendants’
trials or were tried for additional crimes that they did not commit. Bradford, Dean,
Ochoa, Taylor, and Shelden testifi ed at trials against others they had implicated. Al-
though Brown, Gray, Kelly, Vasquez, and Winslow pleaded guilty and did not testify
in a trial, I obtained court fi les and documents describing their confession state-
ments (including a video of part of the interrogation in Winslow’s case). I note that
Jerry Frank Townsend pleaded guilty to some off enses but was tried for two of the
crimes he confessed to.

28. The forty cases all involved an exoneree interrogated in a custodial setting who re-
portedly delivered self- incriminating statements and admissions of guilt to police,
though not necessarily to all of the charged acts. These forty cases do not include
nine exonerees who reportedly made self- incriminating remarks volunteered to po-
lice outside of custody and that were not full admissions to having committed any of
the charged acts. The characteristics of the forty cases are summarized in an appen-
dix accompanying my article “The Substance of False Confessions,” which is avail-
able online at a University of Virginia School of Law Library research collection
webpage together with relevant portions of these forty exonerees’ interrogation rec-
ords and trial transcripts. Written confession statements were obtained for twenty-
eight exonerees. See www .law .virginia .edu/ html/ librarysite/ garrett _falseconfess .
htm. My prior article did not include the cases of two exonerees, Keith Brown and
Anthony Gray. I have since located a copy of Keith Brown’s quite detailed confes-
sion statement (although a second statement the he wrote himself was markedly in-
consistent and was not detailed). Gray “had given a detailed confession.” Attorney
Grievance Comm’n v. Kent, 653 A.2d 909, 917 (Md. 1995). Although the court fi le
did not include the audio or written confession statements, it referenced how he
confessed to having stood watch when the victim was murdered. News reports as
well as conversations with Gray’s lawyer and former prosecutors involved in the
case confi rmed that his confession statements were detailed but also contained cer-
tain inconsistencies with crime scene evidence. See Todd Richissin, “Trying to
Right an Injustice; Murder: A Defense Attorney and Calvert County State’s Attor-
ney Say a Man Has Been Wrongly Imprisoned for the Past Seven Years,” Baltimore
Sun, February 6, 1999, 1A.

29. I fi rst explored these false confessions in a law review article. See Brandon L. Gar-
rett, “The Substance of False Confessions,” 62 Stan. L. Rev. 1051 (2010). I stopped
examining new exonerations after February 2010, when the 250th DNA exoneration
occurred. I note that three subsequent exonerations all apparently involved false
and contaminated confessions. Exoneree Ted Bradford had provided a false confes-
sion reported to have included “details that would only be known to the rapist.”
Mark Morey, “Jurors Find Bradford Innocent of Rape,” Yakima Herald- Republic
(Yakima, WA), February 11, 2010. Exoneree Anthony Caravella initially confessed
with details inconsistent with the crime, but over a series of interrogations was re-
ported to provide accurate details, including information “suggested to him by lead-
ing questions.” Paula McMahon, “DNA Result Just One Troubling Aspect in Con-
victed Man’s Case,” South Florida Sun Sentinel, September 4, 2009. Exoneree Frank

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

296 Notes to Pages 19–22

Sterling’s false confession was also reported to have included inside information
about the crime. Rachel Barnhart, “Innocence Project: Frank Sterling Cleared of
Manville Murder, Christie Confesses,” WHAM.com, March 28, 2010.

30. Arizona v. Fulminante, 499 U.S. 279, 296 (1991) (quoting Bruton v. United States,
391 U.S. 123, 139– 140 (1968)); see also Mark Costanzo, Netta Shaked- Schroer, and
Katharine Vinson, “Juror Beliefs about Police Interrogations, False Confessions,
and Expert Testimony,” 7 J. Empirical Legal Stud. 231 (2010).

31. Gross et. al., “Exonerations in the United States 1989 Through 2003,” 95 J. Crim.
L. & Criminology 523, 544 (2005).

32. Leo, Police Interrogation, ch. 5.
33. Trial Transcript, 1292, State of Oklahoma v. Robert Lee Miller, Jr., CRF- 87- 963

(Okla. D. Ct. May 15– 17, 1988).
34. See, e.g., Leo, Police Interrogation, 243 (reviewing literature, and concluding,

“Since the late 1980’s six studies alone have documented approximately 250
interrogation- inducted false confessions”).

35. Barry C. Feld, “Police Interrogation of Juveniles: An Empirical Study of Policy and
Practice,” 97 J. Crim. L. & Criminology 219, 315 (2006) (“Police in this study con-
cluded three- quarters of interrogations in thirty minutes or less, and none exceeded
one and one- half hours.”); Richard A. Leo, “Inside the Interrogation Room,” 86
J. Crim. L. & Criminology 266, 279– 280 (1996).

36. The twenty- one cases are those of: Marcellius Bradford, Rolando Cruz, Anthony
Gray, Paula Gray, Travis Hayes, Alejandro Hernandez, David Allen Jones, Ryan
Mathews, Antron McCray, Robert Miller, Christopher Ochoa, Calvin Ollins, Kevin
Richardson, Yusef Salaam, Raymond Santana, Jerry Frank Townsend, David Vasquez,
Douglas Warney, Earl Washington Jr., Ronald Williamson, and Korey Wise. In at
least eight of those cases that person subsequently confessed to the crime and often
also pleaded guilty. Cases involving confessions of the person inculpated by postcon-
viction DNA testing are those of Rolando Cruz, Alejandro Hernandez, Christopher
Ochoa, Antron McCray, Kevin Richardson, Yusef Salaam, Raymond Santana, and
Korey Wise.

37. In addition, fi ve of those twelve rape cases were in the Central Park Jogger case,
which was investigated as a murder, where the victim was in critical condition and
in a coma.

38. Anne Coughlin has written an important article examining the victim- blaming nar-
ratives endorsed by leading training manuals and employed to “minimize” the acts
of a suspect during interrogations. See Anne M. Coughlin, “Interrogation Stories,”
95 Va. L. Rev. 1599 (2009). Coughlin argues that “[v]ictim- blaming is incompatible
with the contemporary goals of rape law, and the police should stop feeding those
stock stories to accused rapists.” Ibid., 1660.

39. Frazier v. Cupp, 394 U.S. 731, 739 (1969); Laurie Magid, “Deceptive Police Interro-
gation Practices: How Far Is Too Far?” 99 Mich. L. Rev. 1168, 1169 (2001); Inbau et
al., Criminal Interrogation and Confessions, 486– 487.

40. Memorandum in Opposition to Defendant’s Motion to Suppress Statements, 3,
Commonwealth v. Vasquez, C-22,213 to – 22,216, C-22,763 (Va. Cir. Ct. Oct. 1984).

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Notes to Pages 23–29 � 297

41. The exonerees were James Dean, Jeff rey Deskovic, Byron Halsey, Travis Hayes,
Ronald Jones, John Kogut, Eddie Lowery, and Debra Shelden.

42. See generally, Saul M. Kassin et al., “Police- Induced Confessions: Risk Factors and
Recommendations,” 34 Law & Hum. Behav. 27– 32 (2010); Louis C. Senese, Anat-
omy of Interrogation Themes: The Reid Technique of Interviewing and Interrogation
(Chicago: John E. Reid and Associates, 2005).

43. Inbau et al., Criminal Interrogation and Confessions, 367.
44. Richard Ofshe and Richard Leo, “The Decision to Confess Falsely: Rational

Choice and Irrational Action,” 74 Denv. U. L. Rev. 979, 993 (1997).
45. Trial Transcript, 78– 79, State of Louisiana v. Dennis P. Brown, No. 128, 634 (La.

Dist. Ct. April 3, 1985).
46. Trial Transcript, State of New York v. Douglas Warney, Ind. No. 96- 0088 (N.Y.

Sup. Ct. Feb. 11, 1997).
47. Ibid., 570– 571.
48. Jim Dwyer, “Inmate to Be Freed as DNA Tests Upend Murder Confession,” New

York Times, May 16, 2006.
49. Those exonerees are Paula Gray (who inculpated Kenneth Adams, Verneal Jimerson,

Willie Rainge, and Dennis Williams), Antron McCray, Kevin Richardson, Raymond
Santana, Yusef Salaam, and Korey Wise (who each implicated others in the Central
Park case), James Dean, Ada JoAnn Taylor, Debra Shelden, and Thomas Winslow
(who variously implicated each other and Kathy Gonzalez and Joseph White in the
Beatrice Six case), Alejandro Hernandez and Rolando Cruz (who implicated each
other), Marcellius Bradford and Calvin Ollins (who confessed and implicated Larry
Ollins and Omar Saunders), Travis Hayes (who implicated Ryan Mathews), John
Kogut (who implicated John Restivo and Dennis Halstead), and Chris Ochoa (who
implicated Richard Danziger).

50. Trial Transcript, 942, State v. White, No. 9316 (Neb. Dist. Ct. Nov. 7, 1989).
51. Ibid., 953.
52. Ibid., 924, 931, 939.
53. Ibid., 959.
54. Patrick McCreless, “Justice . . . Finally,” Cullman Times (Culman, AL), October

22, 2008.
55. Trial Transcript, 862, State of Illinois v. Ronald Jones, No. 85- 12043 (Cir. Ct. Cook

County July 12, 1989).
56. Trial Transcript, 450, State of Oklahoma v. Ronald Keith Williamson, CRF 87- 90

(Okla. Dist. Ct. April 21– 28, 1988).
57. Ibid., 541– 542. However, in a possible inconsistency if Williamson’s statement was

interpreted to mean stabbing using a knife, Dr. Jordan also stated that he did not
believe the puncture wounds on her body were caused by a knife. Ibid., 551– 552.

58. Jim Trainum, “Editorial, Get It on Tape: A False Confession to Murder Convinced
a Cop That a Visual Record Can Help Ensure an Innocent Person Isn’t Convicted,”
Los Angeles Times, October 24, 2008, 6.

59. Motion to Suppress Hearing Transcript, 40, People v. Hatchett (Mich. Cir. Ct. Sept.
22, 1997).

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

60. Washington v. Murray, 4 F.3d 1285, 1292 (4th Cir. 1993).
61. Statement of Earl Ju nior Washington, June 4, 1982 at 6.
62. Trial Transcript, 527– 537, 540, 566, Commonwealth of Virginia v. Earl Ju nior

Washington (Va. Cir. Ct. Jan. 19, 1984).
63. Margaret Edds, An Expendable Man: The Near- Execution of Earl Washington, Jr.

(New York: New York University Press, 2003), 248.
64. Jerry Markon, “Wrongfully Jailed Man Wins Suit,” Washington Post, May 6, 2006,

B01 (describing that jury awarded $2.25 million fi nding that “Wilmore deliberately
falsifi ed evidence, which resulted in Washington’s conviction and death sentence”).

65. Trial Transcript, 5291– 5292, People v. McCray et al., No. 4762/89 (N.Y. Sup. Ct.
Aug. 8, 1990).

66. Trial Transcript, 22, Commonwealth of Pennsylvania v. Bruce Donald Godschalk,
No. 00934- 87 (Pa. Ct. Com. Pl. May 27, 1987).

67. Trial Transcript, 1292, State of Oklahoma v. Robert Lee Miller, Jr., CRF- 87- 963
(Okla. Dist. Ct. May 15– 17, 1988).

68. Trial Transcript, 146, State of Illinois v. Alejandro Hernandez et al., No.
84- CF- 361- 01- 12 (Ill. Cir. Ct. Feb. 20, 1985).

69. Transcript of Taped Interview of David Jones, 33, People v. Jones, No. BAO71698
(Cal. Super. Ct. July 15, 1993).

70. Maura Dolan and Evelyn Larrubia, “Telling Police What They Want to Hear, Even
if It’s False,” Los Angeles Times, October 30, 2004.

71. Trial Transcript, 2- 133 to 2- 135, Commonwealth v. Yarris, No. 690- 82 (Pa. Ct. Com.
Pl. June 29, 1982).

72. Leo, Police Interrogation, 261.
73. Trial Transcript, 96, State of Louisiana v. Dennis Brown, No. 128,634 (La. Dist. Ct.

April 12, 1985).
74. Trial Transcript, 77, 122, State of Illinois v. Ronald Jones, No. 85- 12043 (Ill. Cir. Ct.

Feb. 24, 1986).
75. Washington Trial Transcript, 622– 623.
76. See Leo and Ofshe, “The Decision to Confess Falsely,” 1119 (“the reliability of a

confession statement can usually be objectively determined by evaluating the fi t be-
tween a post- admission narrative and the crime facts.”).

77. In other cases, the law enforcement account of the interrogation does not describe
any statements made inconsistent with the crime, but absent a complete recording of
the interrogation, one cannot be confi dent what transpired.

78. Washington Trial Transcript, 596.
79. Ibid., 618.
80. See, e.g., Eric M. Freedman, “Earl Washington’s Ordeal,” 29 Hofstra L. Rev. 1091

(2001).
81. Trial Transcript, 714– 715, State v. Halsey, Nos. 63- 01- 86, 210- 02- 87 (N.J. Super. Ct.

Mar. 7, 1988).
82. See Godschalk Trial Transcript, 72.
83. Brooke Masters, “Missteps on the Road to Justice,” Washington Post, December 1,

2000, A1.

298 Notes to Pages 29–35

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

84. Maurice Possley, “Lab Didn’t Bother with DNA,” Chicago Tribune, August 25,
2006.

85. The eight are: Jeff rey Deskovic, Travis Hayes, Ryan Matthews, Antron McCray,
Kevin Richardson, Raymond Santana, Yusef Salaam, and Korey Wise. See Brandon
L. Garrett, “Claiming Innocence,” 92 Minn. L. Rev. 1629, 1660– 1661 (2008). A total
of sixteen exonerees were convicted despite DNA tests excluding them at the time
they were convicted (the other eight had not falsely confessed.) Those cases are dis-
cussed in Chapter 4.

86. See Deskovic Report.
87. Trial Transcript, 280, People v. Hatchett, 97- 1497- FC (Mich. Cir. Ct. Mar. 6, 1998).
88. According to Steven Drizin, there were no motions to suppress Cruz’s confession

because the defense theory was that the noncustodial confession never happened,
but rather that it was a dream statement made up by the police.

89. Colorado v. Connoley, 479 U.S. 157, 161 (1986); ibid., 175 (Brennan, J., dissenting).
90. 384 U.S. 436 (1966).
91. Richard A. Leo, “Inside the Interrogation Room,” 276; see also Steven D. Clymer,

“Are Police Free to Disregard Miranda?” 112 Yale L.J. 447, 502– 512 (2002); Richard
A. Leo, “Questioning the Relevance of Miranda in the Twenty- First Century,” 99
Mich. L. Rev. 1000, 1010 (2001); Louis Michael Seidman, “Brown and Miranda,” 80
Cal. L. Rev. 673, 745 (1992).

92. Trial Transcript, 20, People v. Lloyd, No. 85- 00376 (Mich. Rec. Ct. May 2, 1985).
93. Arizona v. Fulminante, 499 U.S. 279 (1991); Schneckloth v. Bustamonte, 412 U.S.

218, 226 (1973). The voluntariness standard regulates only confessions made in cus-
tody. Oregon v. Elstad, 470 U.S. 298, 311– 314 (1985). Nine additional exonerees did
not confess, but they did talk to police and reportedly volunteered information be-
fore any custodial interrogation began. Many of these statements were quite damag-
ing at trial, precisely for the same reason that the false confessions were; they were
reported to include nonpublic details about the crime. I describe those cases in Gar-
rett, “Substance of False Confessions,” 1106– 1107.

94. Brown v. Illinois, 422 U.S. 590, 604 (1975) (“[T]he burden of showing admissibility
rests, of course, on the prosecution.”)

95. See Steven A. Drizin and Richard A. Leo, “The Problem of False Confessions in the
Post- DNA World,” 82 N.C. L. Rev. 891, 919– 920 (2004).

96. Trial Transcript, 4:26– 4:27, State of New York v. John Kogut, Ind. 61029 (N.Y. Sup.
Ct. May 1986).

97. Trial Transcript, 2160, State of Florida v. Jerry Frank Townsend, No. 79- 7217 (Fla.
Cir. Ct. July 16, 1980).

98. Miller Trial Transcript, 1039.
99. Brown Trial Transcript, 165.
100. Trial Transcript, 125– 126, State of Ilinois v. Ronald Jones, No. 85- 12043 (Ill. Cir. Ct.

July 13, 1989).
101. Trial Transcript, 3572– 3574, State of Illinois v. Paula Gray, No. 78 C4865 (Ill. Cir.

Ct. Oct. 16, 1978).
102. Vasquez Memorandum in Opposition, 3.

Notes to Pages 35–39 � 299

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

103. Tina Kelley, “New Jersey Drops Charges for Man Imprisoned 19 Years,” New York
Times, July 10, 2007, B3.

104. Trial Transcript, 692, 695, State of New Jersey v. Byron Halsey, No. 0063- 01- 86
(N.J. Sup. Ct. App. Div. March 11– 18, 1988).

105. Trial Transcript, D-182, State of Illinois v. Lafonso Rollins, No. 93 CR 6342, (Ill.
Cir. Ct. March 2, 1994).

106. Godschalk Trial Transcript, 126– 127 (May 26, 1987).
107. Warney Trial Transcript, 6.
108. Godschalk Trial Transcript, 154.
109. Gray Trial Transcript, 1313.
110. Memorandum, Commonwealth of Virginia v. David Vasquez, C-22213–22216,

C22763 (Va. Cir. Ct. Jan. 25, 1985).
111. Return to People’s Petition for Writ of Habeas Corpus at 25, In re David Allen Jones,

No. BA071698 (Cal. Super. Ct. June 17, 2004).
112. Ardy Friedberg and Jason Smith, “Townsend Released; Judge Cites ‘An Enormous

Tragedy’; Attorneys Say Suspect Was Easily Led to Confess,” Sun- Sentinel, June
16, 2001, p. 1A.

113. Townsend Trial Transcript, 43, 50.
114. The State apparently also introduced into the record “transcriptions” of expert

evaluations they had conducted. Ibid., 100.
115. Ibid., 397– 400.
116. Townsend Trial Transcript, 516– 517.
117. Frank Lee Smith’s case is discussed in detail in Chapter 8.
118. Deskovic Report, 7.
119. See, e.g., Richard A. Leo et al., “Bringing Reliability Back In: False Confessions

and Legal Safeguards in the Twenty- First Century,” 2006 Wis. L. Rev. 479, 486
(2006).

120. Vasquez Memorandum in Opposition, 1– 2.
121. Dana Priest, “At Each Step, Justice Faltered for VA Man,” Washington Post, July 16,

1989, A1.

3 . eyewitness mis identif ications

1. Trial Transcript, 29– 33, State of New York v. Vincent H. Jenkins, No. 82- 1320- 001
(N.Y. Sup. Ct. May 31, 1983). Vincent Jenkins changed his name after his conviction
to Habib Abdal.

2. Ibid., 52– 59.
3. See Gary L. Wells and Deah S. Quinlivan, “Suggestive Eyewitness Identifi cation

Procedures and the Supreme Court’s Reliability Test in Light of Eyewitness Sci-
ence: 30 Years Later,” 33 Law & Hum. Behav. 14 (2009).

4. Jenkins Trial Transcript, 62 (June 1, 1983).
5. Ibid., 86.
6. Ibid., 36, 49– 52.

300 Notes to Pages 39–47

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

7. Ibid., 52.
8. Watkins v. Sowders, 449 U.S. 341, 352 (1981) (Brennan, J., dissenting).
9. Trial Transcript, 196, State of Illinois v. Jerry Miller, No. 81- C-7310 (Ill. Cir. Ct.

Sept. 29, 1982).
10. Gerry Smith, “Rape Conviction Gone, Stigma Isn’t,” Chicago Tribune, October 22,

2007, 1.
11. In two of those cases, those of William Gregory and Mark Webb, while the complete

trial transcript could not be located, portions were obtained because the eyewitness
identifi cation was described in police reports or portions of the trial were quoted in
judicial decisions that were obtained.

12. The online appendix details the coding criteria; the features of each of the cases; the
types of suggestive procedures used, if any; the types of unreliability identifi ed, if
any; and provides quotations from the relevant trial testimony to describe what the
trial rec ords show. See http:// www .law .virginia .edu/ innocence. In a handful of
cases, the complete trial transcripts were not obtained, but portions relevant to the
eyewitness identifi cations were obtained. Any additional information from postcon-
viction decisions or news reports is also noted.

13. Gary L. Wells et al., “Eyewitness Identifi cation Procedures: Recommendations for
Lineups and Photospreads,” 22 Law & Hum. Behav. 605 (1998).

14. See Wells and Quinlivan, “Suggestive Eyewitness Identifi cation Procedures,” 6
(discussing data from archival studies, showing that eyewitnesses in real cases make
misidentifi cations, but emphasizing that fi ller identifi cations cannot tell one at what
rate innocent people are misidentifi ed).

15. Alvin G. Goldstein, June E. Chance, and Gregory R. Schneller, “Frequency of Eye-
witness Identifi cation in Criminal Cases: A Survey of Prosecutors,” 27 Bull. Psycho-
nomic Soc’y 73 (1989).

16. Thirty- seven were identifi ed by 2 eyewitnesses, nineteen were identifi ed by 3 eye-
witnesses, eight were identifi ed by 4 eyewitnesses, three were identifi ed by 5 eyewit-
nesses, and one was identifi ed by 10 eyewitnesses.

17. Trial Transcript, 343, State of Florida v. Cody Edward Davis, Case No. 06- 004031CF
A02 (Fla. Cir. Ct. Aug. 16, 2006).

18. Samuel R. Gross, “Loss of Innocence: Eyewitness Identifi cation and Proof of
Guilt,” 16 J. Legal Stud. 400, 416 (1987) (describing misidentifi cations in which the
suspect was initially located based on appearance).

19. Amy Bradford Douglass, Caroline Smith, and Rebecca Fraser- Thill, “A Problem
with Double- Blind Photospread Procedures: Photospread Administrators Use One
Eyewitness’s Confi dence to Infl uence the ID of Another Witness,” 29 Law & Hum.
Behav. 543– 562 (2005); see Wells et al., “Eyewitness Identifi cation Procedures,”
625– 626.

20. Of the 190 exonerees who had eyewitness evidence in their cases, 159 involved a
rape conviction, 9 involved a murder, 17 involved a rape and a murder, and 5 in-
volved other crimes (a robbery, three carjackings, and an attempted murder).

21. Thirty- one of the exonerees not convicted of rape had eyewitnesses identify them.
Of the exonerees convicted of murder, 41% (nine of twenty- two) had eyewitnesses.

Notes to Pages 47–51 � 301

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Of those convicted of rape and murder, 33% (seventeen of fi fty- two) had eyewit-
nesses. All fi ve of the exonerees convicted of other crimes (not rape or murder)
had eyewitnesses testify. Only one exoneree convicted of a murder, Kevin Lee
Green, had a victim eyewitness identify him at trial. His pregnant wife had been
gravely assaulted, the murder victim was his unborn child, and perhaps in
part due to substantial memory loss and brain damage, his wife identifi ed him at
trial. See “Falsely Accused,” 16 Forensic Examiner 82 (September 22, 2007). The
other exonerees convicted of murder who had eyewitness testimony all had
testimony by nonvictim eyewitnesses. Of the exonerees convicted of murder and
rape, three had victim eyewitnesses, all where there was a victim of a rape apart
from the victim of the murder, and fourteen others had a nonvictim eyewitness
testify.

22. Only 23 of the entire group of 190 cases had identifi cations by acquaintances. See
Lawrence A. Greenfeld, U.S. Department of Justice, Bureau of Justice Statistics,
Sex Off enses and Off enders (1997), 4 (three out of four reported rapes involve off end-
ers with whom the victim had a prior relationship).

23. Elizabeth Hampson, Sari M. van Anders, and Lucy I. Mullin, “A Female Advantage
in the Recognition of Emotional Facial Expressions: Test of an Evolutionary Hy-
pothesis,” 27 Evolution & Hum. Behav. 27 (2006).

24. Manson v. Brathwaite, 432 U.S. 98, 119 (1977).
25. Gary Wells and Elizabeth Loftus, Eyewitness Testimony (Cambridge: Cambridge

University Press, 1984), 12, 28– 29; Gary L. Wells and Lisa E. Hasel, “Facial Com-
posite Production by Eyewitnesses,” 16 Current Directions Psychol. Sci. 6 (2007);
N. A. Brace et al., “Identifying Composites of Famous Faces: Investigating Memory,
Language and System Issues,” 12 Psychol. Crime & L. 351 (2006).

26. Trial Transcript, 34, State of Louisiana v. Allen H. Coco, No. 14891- 95 (La. Dist.
Ct. Nov. 6, 1997).

27. For example, of the fi fty- three showups, twenty- nine occurred in cases in which
there was also a photo array. Of the cases with lineups, forty- seven occurred in cases
in which there was also a photo array. It is a promising area for future research to ex-
plore the reinforcing eff ect of conducting multiple identifi cation procedures and of
diff erent types with an eyewitness. Studies have found that repeat viewings, or
“laps,” increase choosing rates and error rates, with particularly high error rates
among witnesses who choose to view a second time. See, e.g. Nancy K. Steblay et al.,
“Sequential Lineup Laps and Eyewitness Accuracy,” Law & Hum. Behav. (forth-
coming 2011).

28. Hearing Transcript, 163, State of California v. James Ochoa, No. 05NF2056 (Cal.
Sup. Ct. June 8, 2005); see also Steve McGonigle and Jennifer Emily, “A Blind Faith
in Eyewitnesses, 18 of 19 Local Cases Overturned by DNA Relied Heavily on Unre-
liable Testimony,” Dallas Morning News, October 12, 2008, 1A (“Most police agen-
cies don’t have written policies on identifi cation techniques, and police offi cers re-
ceive little formal training.”).

29. 432 U.S. 98 (1977).
30. United States v. Wade, 388 U.S. 218, 228 (1967).

302 Notes to Pages 51–53

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

31. See Brian L. Cutler, Steven D. Penrod, and Hedy Red Dexter, “Juror Sensitivity to
Eyewitness Identifi cation Evidence,” 14 Law & Hum. Behav. 190 (1990); Wells et
al., “Eyewitness Identifi cation Procedures,” 619– 620.

32. See Wells et al., “Eyewitness Identifi cation Procedures,” 635– 636.
33. Stovall v. Denno, 388 U.S. 293, 302 (1967).
34. Throughout this chapter, statistics are provided based on percentages of the 161 tri-

als obtained with eyewitness identifi cations. As noted, many trials had more than
one eyewitness testifying, and for those, any one eyewitness is counted. For exam-
ple, if there was a showup for any one of the eyewitnesses, or for multiple eyewit-
nesses, this is counted as one case involving a showup.

35. One study suggests that proper showups conducted on the scene may create diff erent
risks of error than a poorly handled lineup. Richard Gonzalez, Phoebe C. Ellsworth,
and Maceo Pembroke, “Response Biases in Lineups and Showups,” 64 J. Personal-
ity and Soc. Psychol. 525 (1993). In a poorly handled lineup, a witness may compari-
son shop for the person who looks most like the attacker. In a showup it is clear that
the choice is yes or no, was this the attacker or not, and the greater risk may be that a
guilty suspect might not be identifi ed. Regardless, “there is clear evidence that show-
ups are more likely to yield false identifi cations than are properly constructed line-
ups.” See Wells et al., “Eyewitness Identifi cation Procedures,” 630– 631.

36. Bibbins v. City of Baton Rouge, NO. CIV.A.04- 122- JJB, 489 F.Supp. 2d 562, 570
(M.D. La. May 11, 2007).

37. Trial Transcript, 12, 147– 149, Commonwealth of Virginia v. Willie Davidson (Va.
Cir. Ct. May 27, 1981).

38. Trial Transcript, 17, 21, 23, State of Florida v. Alan J. Crotzger, Case No. 81- 6604
(Fla. Cir. Ct. April 19, 1982).

39. Trial Transcript, 445, Commonwealth v. Neil Miller, No. 085602- 04 (Mass. Sup.
Ct. Dec. 14, 1990).

40. These cases were not coded as showups, since the police did not arrange for the
eyewitnesses to see those images. Nor were a few cases in which the victim saw
someone in the neighborhood and identifi ed the person as the attacker— unless the
police followed that identifi cation up with a showup.

41. Wells et al., “Eyewitness Identifi cation Procedures,” 630– 635; Elizabeth F. Loftus,
James M. Doyle, and Jennifer E. Dysart, Eyewitness Testimony: Civil and Criminal
(LexisNexis 4th ed. 2007) § 4- 9. Similarly, in a few cases the witness was shown
only three photos, a lineup with so few choices that it was coded as inherently sug-
gestive. Wells et al. describe how to best structure lineups or photospreads, a sub-
ject that can raise some more complex issues depending on the relationship between
what the witness described and what the suspect looks like. For the most part, the
preferred method is to pick fi llers who fi t the witnesses’ description of the perpetra-
tor. Wells et al., “Eyewitness Identifi cation Procedures,” 632.

42. Trial Transcript, 52, 99, Commonwealth of Virginia v. Marvin Lamont Anderson
(Va. Cir. Ct. Dec. 14, 1982).

43. Ibid., 22, 27– 28, 62, 198.
44. Ibid., 66.

Notes to Pages 54–58 � 303

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

45. Ibid., 201– 204.
46. Frank Green, “Va. Court Rejected First Appeal,” Richmond Times- Dispatch, July

16, 2007, A7; see also http:// www .innocenceproject .org/ Content/ 49 .php .
47. Trial Transcript, 72– 73, State of Illinois v. Ronnie Bullock, No. 83 C 5501 (Ill. Cir.

Ct. April 30, 1984).
48. Trial Transcript, 37, State of Missouri v. Lonnie Erby, No. 851- 2663 (Mo. Cir. Ct.

June 9, 1986).
49. Trial Transcript, 101, Commonwealth v. Thomas Doswell, No. CC 8603467 (Pa. Ct.

of Common Pleas Nov. 19, 1986).
50. Many more cases likely involved such repetition of just the exoneree in multiple

procedures, but it was often not clear from the testimony whether any of the fi llers
were repeated along with the exoneree. See Wells et al., “Eyewitness Identifi cation
Procedures,” 603; Ryann M. Haw, Jason J. Dickonson, and Christian A. Meissner,
“The Phenomenology of Carryover Eff ects Between Show- up and Line- up Identifi –
cation,” 15 Memory 117 (2007).

51. Trial Transcript, 26, 31– 33, State of Texas v. Larry Fuller, No. F81- 8431- P (Tex.
Dist. Ct. May 13, 1981).

52. Deah S. Quinlivan et al., “Do Prophylactics Prevent Infl ation? Post- Identifi cation
Feedback and the Eff ectiveness of Procedures to Protect against Confi dence-
Infl ation in Earwitnesses,” 33 Law & Hum. Behav. 111 (2009). Because there is less
research on voice identifi cations, and because they raise issues separate from eye-
witness identifi cations, I conservatively did not count as suggestive any of the cases
that involved single- voice identifi cations or other suggestive voice identifi cation
procedures.

53. Ibid.; Wells et al., “Eyewitness Identifi cation Procedures,” 615, 629– 630.
54. In all other cases, the eyewitness or the police described telling the eyewitness to

simply look at the photos or lineup, they could not recall any specifi c instructions,
or worse, the police told the eyewitness that a suspect had been located.

55. See Loftus, Doyle, and Dysart, Eyewitness Testimony, § 4- 8(b) (describing study by
Roy Malpas and Patricia Devine, and noting eigh teen other studies demonstrating
higher false identifi cation when such biased instructions were provided).

56. Trial Transcript, 60– 61, State of Illinois v. Richard Johnson, No. 91 CR 20794 (Ill.
Cir. Ct. Oct. 6, 1992).

57. Trial Transcript, 70– 71, State of Ohio v. Robert L. McClendon, No. 90CR- 05- 2586
(Ohio Ct. of Common Pleas Aug. 26, 1991).

58. Trial Transcript, 34, State of Illinois v. Alejandro Dominguez, 89 CP 1995 (Ill. Cir.
Ct. 1995).

59. Johnson Trial Transcript, 39.
60. Trial Transcript, 54, State of Texas v. Gilbert Alejandro, No. 90- 09- 8445- CR (Tex.

Dist. Ct. December 11, 1990).
61. Trial Transcript, 305, State of Indiana v. Larry Mayes, No. 1CR- 6- 181- 17 (Ind. Sup.

Ct. July 6, 1982).
62. Those exonerees were Edward Honaker, Lesly Jean, Larry Mayes, Leo Waters, and

Glen Woodall.

304 Notes to Pages 58–61

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

63. Jean v. Rice, 945 F.2d 82, 87 (4th Cir. 1991).
64. Due to evidence that hypnosis can create false or inaccurate memories, some courts

bar all testimony by hypnotized witnesses at trial. Other courts admit the evidence,
but only if experts also testify as to the dangers of such techniques. Council on Sci-
entifi c Aff airs, “Scientifi c Status of Refreshing Recollection by the Use of Hypno-
sis,” 253 J. Am. Med. Ass’n 1918, 1919 (1985); Lisa K. Rozzano, “The Use of Hyp-
nosis in Criminal Trials: The Black Letter of the Black Art,” 21 Loy. L.A. L. Rev.
635, 645 (1987).

65. Trial Transcript, 37– 38, State of Texas v. Larry Fuller, No. F81- 8431- P (Tex. Dist.
Ct. Aug. 24, 1981).

66. Trial Transcript, 63, State of Ohio v. Anthony Green, No. CR 228250 (Ohio Ct. of
Common Pleas Oct. 13, 1988).

67. Jennifer Thompson- Cannino and Ronald Cotton, Picking Cotton: Our Memoir of
Justice and Redemption (New York: St. Martin’s Press, 2009), 134.

68. Trial Transcript, 230– 232, State of Texas v. Thomas Cliff ord McGowan, Jr.,
85- 81070- MU (Tex. Dist. Ct. March 5, 1986). In addition, the photo array was itself
fl awed; most of the photos were hard- to- make- out, black- and- white photos or pho-
tos with captions indicating they were from a diff erent police department.

69. Trial Transcript, 148, State of Missouri v. Larry Johnson, No. 341- 00274 (Mo. Cir.
Ct. August 20, 1984).

70. I was very conservative in my criteria for counting a case as involving an “unreli-
able” identifi cation. I only included, as the online Appendix details, cases in which
(1) the eyewitness identifi ed another person, or (2) the eyewitness admitted earlier
uncertainty, or initially failed to identify the exoneree, or (3) an eyewitness admit-
tedly could not see the attacker’s face. I separately note, but did not code as unreli-
able, cases involving gross discrepancies between the eyewitness’ initial description
of the person who attacked them and the defendant’s appearance; cases involving a
lengthy duration from incident to fi rst identifi cation; cases involving very poor light-
ing conditions; and cases with a limited opportunity to view the defendant. Such
cases were extremely common in the data set but were not deemed “unreliable” on
that basis, because although they occurred frequently, discrepancies in descriptions
are not predictive of accuracy, and because almost all cases involved very limited op-
portunity to view under poor conditions.

71. Trial Transcript, 304, State of Wisconsin v. Steven A. Avery, No. 85 FE 118 (Wis.
Cir. Ct., Dec. 10, 1985).

72. Doswell Trial Transcript, 691, 742.
73. Trial Transcript, J-128, People v. Dean Cage, 94 29467 (Ill. Cir. Ct. January 9, 1995).
74. Bill Rankin, “Exonerations Urge Changes for Eyewitnesses,” Atlanta Journal-

Constitution, December 25, 2008, C1.
75. Trial Transcript, 298– 300, State of Texas v. Donald Wayne Good, F33- 31435 (Tex.

D. Ct. Sept. 15, 1987).
76. See Wells et al., “Eyewitness Identifi cation Procedures,” 622– 623, 635.
77. See Wells et al., “Eyewitness Identifi cation Procedures,” 619, 635– 636; Loftus,

Doyle, and Dysart, Eyewitness Testimony, § 6- 2.

Notes to Pages 61–65 � 305

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

78. Trial Transcript, 172– 173, State of Georgia v. Calvin Crawford Johnson, No.
12- 22011- 3 (Ga. Sup. Ct. Nov. 3, 1983).

79. Calvin C. Johnson Jr., Exit to Freedom (Athens: University of Georgia Press, 2003),
99– 100.

80. Calvin Johnson Trial Transcript, 174, 183– 185.
81. Trial Transcript, 206, State of Georgia v. John Jerome White, No. 314 (Ga. Sup. Ct.

May 29, 1980).
82. See Rankin, “Exonerations Urge Changes,” C1.
83. See Thompson- Cannino and Cotton, Picking Cotton, 134; Jennifer Thompson,

Editorial, “I Was Certain, but I Was Wrong,” New York Times, June 18, 2000, 15.
84. See Gary L. Wells, “Eyewitness Identifi cation: Systemic Reforms,” 2006 Wis. L.

Rev. 615, 627 (2006).
85. Trial Transcript, 295, People v. Anthony Capozzi, #85- 1379- 001 (N.Y. Sup. Ct. Jan.

27, 1987).
86. Trial Transcript, 3– 114, State of Mary land v. Kirk N. Bloodsworth, No. 84- CR- 3138

(Md. Cir. Ct. Mar. 1, 1985).
87. Terry Chalmers, City of Mount Vernon Police Department Reports, August–

October, 1986 (on fi le with author).
88. While other exonerees had witnesses who simply could not identify anyone, aside

from these four cases, the other cases all included eyewitness who testifi ed if asked
that they were certain about their identifi cations. The four cases are those of Kirk
Bloodsworth, Jimmy Ray Bromgard, William Dillon, and Jerry Miller.

89. Jerry Miller Trial Transcript, 226.
90. Trial Transcript, 43, 71, State of Montana v. Jimmy Ray Bromgard, No. DC 87- 148

(Mont. Cir. Ct. Nov. 17, 1987).
91. Ibid., 351– 352.
92. There is no evidence that congruence of description with the defendant predicts the

accuracy of an eyewitness identifi cation. See Gary L. Wells, “Verbal Descriptions of
Faces from Memory: Are They Diagnostic of Identifi cation Accuracy,” 70 J. Applied
Psychol. 619 (1985) (fi nding that congruence and accuracy of eyewitness reports
were not highly related); Melissa Pigott and John Brigham, “Relationship Between
Accuracy of Prior Description and Facial Recognition,” 70 J. Applied Psychol. 547–
548 (1985) (fi nding no such relationship and citing additional studies).

93. These data are limited to major diff erences in physical descriptions. They do not
include cases where witnesses described clothing defendants claimed never to have
worn or owned, or where witnesses could not describe the attacker much at all. For
example, a diff erence that could have been accounted for by the passage of time from
incident to arrest, such as a week’s growth of hair, were discounted. Nor were minor
descriptive diff erences, such as between hazel and green eyes, or light and light-
brown hair, counted. As described in the Appendix, approximately half of the data
was blind coded. I thank Jessica Kostelnik for her help calculating an interrater reli-
ability score. An interrater reliability analysis using the Kappa statistic was per-
formed to determine consistency among raters in coding the two central questions
concerning whether eyewitness identifi cations in a case involved any of the enumer-

306 Notes to Pages 65–69

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

ated indicia of police suggestion or unreliability. Kappa values from 0.40 to 0.59 are
considered moderate, 0.60 to 0.79 substantial, and 0.80 outstanding. Analyses re-
vealed outstanding interrater reliability for suggestion (Kappa 0.89, p 0.001, 95%
CI 0.77 – 1.01) and substantial interrater reliability for reliability (Kappa 0.78, p 0.001,
95% CI 0.60 – .96).

94. Trial Transcript, 326, State of Texas v. James Curtis Giles, Nos. F-83- 87258- UKJ,
F-77- 8236- KJ (Tex. Dist. Ct. June 6, 1983).

95. Avery Trial Transcript, 315– 316.
96. Because eyewitness uncertainty is such a clear indication of unreliability, I focus in

this chapter on cases where eyewitnesses were uncertain at the time of the identifi ca-
tion. As noted, although many of these cases involved such limited opportunity to
view the culprit, the only cases that I labeled as “unreliable” were the fi fteen most
extreme cases, in which the witnesses reported they could not see the culprit’s face
at all.

97. Greenfeld, Sex Off enses, 3, 11.
98. Dennis Brown Trial Transcript, 103.
99. Trial Transcript, 78– 79, State of Wisconsin v. Fredric Karl Saecker, No. 89- CF- 33

& 36 (Wis. Cir. Ct. Jan. 3. 1990).
100. See Sandra Guerra Thompson, “Judicial Blindness to Eyewitness Misidentifi ca-

tion,” 93 Marq. L. Rev. 7, 20– 21 (2009) (citing Gary L. Wells and Eric P. Seelau,
“Eyewitness Identifi cation: Psychological Research and Legal Policy on Lineups,” 1
Psychol. Pub. Pol’y & L. 765, 766 (1995)).

101. See Wells and Quinlivan, “Suggestive Eyewitness Identifi cation Procedures,” 5.
On disagreement concerning eff ects of certain estimatory variables in some types of
cases, see Steven E. Clark and Ryan D. Godfrey, “Eyewitness Identifi cation Evi-
dence and Innocence Risk,” 16 Psychonomic Bull. & Rev. 23 (2009).

102. Of the cases in which the eyewitnesses stated for how long they looked at the at-
tacker, 58 (of 161 cases) said that they saw the attacker for less than a minute, most
often just for seconds. Another 56 cases involved durations of less than twenty min-
utes. Another 38 cases involved durations of more than twenty minutes, 10 of which
involved durations of more than an hour. (In 9 cases, no information was available
where the eyewitness did not discuss duration at trial). While the relationships be-
tween the various factors involved are complex, and these exonerations involve dif-
ferent types of identifi cations under very diff erent circumstances, I note one possi-
ble hypothesis: that suggestion played a greater role in cases involving identifi cations
under conditions that would otherwise tend to be more accurate. Of those 36 cases
with long durations, 6 did not involve suggestion, 3 of which raise diff erent issues
because they were acquaintance cases. Again, however, many of those cases impli-
cated still diff erent factors aff ecting reliability, for example, where they were also
cross- racial identifi cations or involved identifi cation procedures fi rst conducted long
after the crime.

103. Gary L. Wells and D. M. Murray, “What Can Psychology Say about the Neil vs. Big-
gers Criteria for Judging Eyewitness Identifi cation Accuracy?” J. Applied Psychol.
68 (1983).

Notes to Pages 69–71 � 307

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

104. Trial Transcript, 129, State of Louisiana v. Willie Jackson, No. 87- 0205 (La. Dist.
Ct. Aug. 24, 1989).

105. Gary L. Wells and A. L. Bradfi eld, “ ‘Good, You Identifi ed the Suspect’: Feedback
to Eyewitnesses Distorts Their Reports of the Witnessing Experience,” 83 J.
Applied Psychol. 360 (1998).

106. Trial Transcript, 118, State of Texas v. Andrew William Gossett, No. F99- 22771- W
(Tex. Dist. Ct. Feb. 8, 2000); Brian L. Cutler, Eyewitness Testimony: Challenging
Your Opponent’s Witness (National Institute for Trial Advocacy, 2002), 19.

107. See Roger B. Handberg, “Expert Testimony on Eyewitness Identifi cation: A New
Pair of Glasses for the Jury,” 32 Am. Crim. L. Rev. 1013, 1023 (1995); Charles A. Mor-
gan III et al., “Acccuracy of Eyewitness Memory for Persons Encountered During
Exposure to Highly Intense Stress,” 27 Int’l J. L. & Psychol. 265 (2004).

108. See Wells and Quinlivan, “Suggestive Eyewitness Identifi cation Procedures.”
109. While in 66 cases the fi rst identifi cation occurred less than a week after the crime,

in 53 cases it was more than a week, in 15 cases it was more than three months, in
8 cases more than six months, and in 5 cases more than a year.

110. See, e.g., Gary L. Wells and Elizabeth Olson, “The Other- Race Eff ect in Eyewitness
Identifi cation: What Do We Do About It?” 7 Psychol. Pub. Pol’y & L. 230 (2001); Gary
L. Wells and Elizabeth F. Loftus, eds., Eyewitness Testimony: Psychological Perspec-
tives (Cambridge: Cambridge University Press, 1984), 1; Elizabeth F. Loftus, Eyewit-
ness Testimony (Cambridge, MA: Harvard University Press, 1979).

111. Of the 93 cross- racial identifi cation cases, 74 cases involved a black defendant and a
white eyewitness. Three of those cases involved male eyewitnesses; thus 71 involved
a male black defendant and a female eyewitness. Seven cases involved a Hispanic de-
fendant and a white eyewitness. Six involved a black defendant and a Hispanic wit-
ness. Three involved a white defendant and a black witness. Two involved a white
defendant and a Hispanic witness. One last case, that of Michael Blair, involved
a defendant who was part Asian and who self- identifi ed as white, but whom white
eyewitnesses regarded as a minority (they thought he was Hispanic). Of the 93 cross-
racial identifi cation cases, 86 involved victim identifi cations and 9 involved non-
victim eyewitnesses (two involved both types). It was not possible to obtain data on
the race of the victims for some of these cases. However, the denominator used here is
the full set of 190 cases with eyewitness identifi cations, because information about
victim race could also be obtained for many cases in which no trial transcript was
obtained.

112. Of the 171 exonerees convicted of rape, 42 were white and 129 were minorities. See
Sean Rosenmerkel, Matthew Durose, and Donald Farole Jr., U.S. Department of
Justice, Bureau of Justice Statistics, Felony Sentences in State Courts, 2006 (2009),
table 3.2 (fi nding that 30% of rape convicts were minorities); Matthew R. Durose and
Patrick A. Langan, U.S. Department of Justice, Bureau of Justice Statistics, Felony
Sentences in State Courts, 2002, (2004), 6, table 5 (fi nding in survey of 300 counties
that 37% of rape convicts were minorities). There may not have been any greater ra-
cial disparity in data from closer to the 1980s, when these exonerees were chiefl y
convicted, although data from that period were not consistently compiled. Green-

308 Notes to Pages 71–73

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

feld, Sex Off enses, 10 (fi nding 44% of rape arrestees and 47.8% of imprisoned rape
convicts were minorities); Brian A. Reaves and Pheny Z. Smith, U.S. Department of
Justice, Bureau of Justice Statistics, Felony Defendants in Large Urban Counties,
1992 (1995), 4 (“Whites (48%) and blacks (49%) comprised roughly equal percent-
ages of rape defendants”); but see Patrick A. Langan and Helen A. Graziadei, U.S.
Department of Justice, Bureau of Justice Statistics, Felony Sentences in State Courts,
1992 (1995), 5 (fi nding that 30% of rape convicts were black and 4% were “other”).

113. Greenfeld, Sex Off enses, 11. Seventy- seven exonerees were black and convicted of
raping a white victim, and 7 were Latino and convicted of raping a white victim. In
addition, 4 black exonerees were convicted of raping Latino victims, while 5 white
exonerees were convicted of raping black or Latino victims.

114. Rosenmerkel et al., Felony Sentences in State Courts, 2006, table 3.2 (fi nding that
minorities account for 40% of felony sentences in state courts); Heather C. West and
William J. Sabol, U.S. Department of Justice, Prison and Jail Inmates at Midyear
2008, (2008), 17, table 16 (fi nding that minorities account for almost 60% of prison
inmates and black men are 6.6% more likely to be incarcerated than white men).

115. There are several possible explanations for the racial disparity among exonerees.
These 250 exonerees include far more minorities than even already disproportion-
ately minority rape and murder convicts. The other- race eff ect may cause eyewit-
ness misidentifi cations disproportionately involving black defendants and white
victims. Another hypothesis is that black defendants may be overrepresented in the
kinds of serious rape and murder prosecutions common in this set of DNA exonera-
tions. One explanation could be victim related, that is, that cases involving black
defendants and white victims disproportionately receive lengthy sentences of the
sort that makes DNA testing feasible years later. The most comprehensive data on
race- of- victim eff ects is in the death penalty context. U.S. General Accounting
Offi ce, Death Penalty Sentencing: Research Indicates Pattern of Racial Disparities
(1990), GAO/GGD- 90- 57, 6 (reviewing twenty- three studies of the death penalty
after 1973, and concluding that “[i]n 82% of the studies, race- of- victim was found to
infl uence the likelihood of being charged with capital murder or receiving a death
sentence, i.e., those who murdered whites were found to be more likely to be sen-
tenced to death than those who murdered blacks. This fi nding was remarkably con-
sistent across data sets, states, data collection methods, and analytic techniques.”)
There is not much data available concerning rape prosecutions. Ethnographic stud-
ies of prosecutorial charging have observed race- of- victim eff ects. Lisa Frohmann,
“Convictability and Discordant Locales: Reproducing Race, Class, and Gender
Ideologies in Prosecutorial Decisionmaking,” 31 Law & Soc’y Rev. 531, 535 (1997);
Cassia Spohn and David Holleran, “Prosecuting Sexual Assault: A Comparison of
Charging Decisions in Sexual Assault Cases Involving Strangers, Acquaintances,
and Intimate Partners,” 18 Just. Q. 651, 652 (2001). Additional data, some empirical,
and chiefl y from mock jury studies, suggests that jurors may discriminate against
black defendants and also tend to convict in cases involving white victims. See Sam-
uel R. Sommers and Phoebe C. Ellsworth, “How Much Do We Really Know about
Race and Juries? A Review of Social Science Theory and Research,” 78 Chi.- Kent

Notes to Pages 73–74 � 309

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

L. Rev. 997 (2003); Sheri Lynn Johnson, “Black Innocence and the White Jury,” 83
Mich. L. Rev. 1611 (1985). Finally, black defendants may be disadvantaged in other
ways, for example by being disproportionately targeted by police, or disproportion-
ately needing indigent repre sen ta tion, and would thus be overrepresented among
wrongful convictions. See also Andrew E. Taslitz, “Wrongly Accused: Is Race a
Factor in Convicting the Innocent?” 4 Ohio St. J. Crim. L. 121 (2006).

116. Trial Transcript, 224, State of Texas v. Thomas Cliff ord McGowan, No. F85-
81070- MU (Tex. Dist. Ct. March 5, 1986).

117. Trial Transcript, 190, State of Texas v. Patrick Leondos Waller, No. F92- 40874
(Tex. Dist. Ct. Dec. 7, 1992).

118. Trial Transcript, 215, State of Missouri v. Larry Johnson, No. 341- 00274 (Mo. Cir.
Ct. Aug. 20, 1984).

119. Trial Transcript, 240, State of South Carolina v. Perry Renard Mitchell, No.
83- GS- 32- 479 (S.C. Cir. Ct. Jan. 19, 1984).

120. Trial Transcript, 40– 41, State of Wisconsin v. Anthony T. Hicks, No. 90CF1412
(Wis. Cir. Ct. Dec. 18, 1991).

121. State v. Cromedy, 158 N.J. 112 (1999).
122. See Michael Kennan, “Child Witnesses: Implications of Contemporary Suggestibil-

ity Research in a Changing Legal Landscape,” 26 Dev. Mental Health L. 100 (2007);
Richard Friedman, “The Suggestibility of Children; Scientifi c Research and Legal
Implications,” 86 Cornell L. Rev. 33 (2000); Maggie Bruck, Stephen J. Ceci and He-
lene Hembrooke, “Reliability and Credibility of Young Children’s Reports,” 52
Amer. Psych. 136, 140 (1998); see, e.g., State v. Michaels, 642 A.2d 1372 (1994).

123. Trial Transcript, 237– 239, People of the State of California v. Leonard McSherry,
No. A040264 (Cal. Sup. Ct. October 12, 1988).

124. Ibid., 107, 121.
125. Ibid., vol. 3, p. 184.
126. Ibid., 90.
127. Ibid., 124.
128. Leonard McSherry v. City of Long Beach, 423 F.3d 1015 (9th Cir. 2005).
129. Trial Transcript, 4– 110, 4– 58, 5– 105, Commonwealth v. Rodriguez U. Charles,

Criminal Action Nos. 035942- 45; 036191- 84 (Mass. Sup. Ct. Feb. 1, 1984).
130. Pretrial Hearing Transcript, 354– 355, State of West Virginia v. Larry David Hold-

ren, No. CR- 83- F-181 (W. Va. Cir. Ct. Dec. 12, 1983).
131. Bloodsworth v. State, 512 A.2d 1056 (Md. 1986).
132. The three were Steven Avery, Chester Bauer, and Darryl Hunt. Courts have increas-

ingly admitted expert testimony concerning eyewitness memory and the possibility
of a mistaken eyewitness identifi cation. See Brian L. Cutler and Margaret Bull Ko-
vera, Evaluating Eyewitness Identifi cations (New York: Oxford University Press,
2010), 10– 14; see, e.g., Utah v. Clopton, 223 P.3d 1103, 1107– 1118 (2009); Sturgeon v.
Quarterman, 615 F.Supp. 2d 546, 572– 573 (S.D. Tex. 2009); United States v. Smith-
ers, 212 F.3d 306, 311– 312 (6th Cir. 2000). Other courts have similarly denied defen-
dant’s access to experts on eyewitness memory or found it error to permit such an
expert to testify. See, e.g., State v. Young, 2010 WL 1286933 (La. 2010).

310 Notes to Pages 74–78

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

133. Loftus, Doyle, and Dysart, Eyewitness Testimony, § 8- 18.
134. Trial Transcript, 87, State of Texas v. Carlos Lavernia, No. 76,122 (Tex. Dist. Ct.

Jan. 21, 1985).
135. Trial Transcript, 781, State of West Virginia v. William O’Dell Harris, No. 86-

F-442 (W. Va. Cir. Ct. July 14, 1987).
136. Lavernia Trial Transcript, 156.
137. Hicks Trial Transcript, 586– 587.
138. Mitchell Trial Transcript, 318.
139. Trial Transcript, 896, People of the State of New York v. Alan Newton, Ind. No.

2054/84 (N.Y. Sup. Ct. May 1, 1985).
140. Commonwealth v. Yarris, 519 Pa. 571, 601– 602 (Pa. 1988).
141. Throughout I refer to misidentifi cations and not to mistaken identifi cations. That

term was chosen because these identifi cations of the innocent were often not the
product of accidents or mistakes, but rather the use of suggestive police procedures
that could predictably result in unreliable and erroneous identifi cations.

142. According to the Bureau of Justice Statistics, only 0.8% of felony defendants are con-
victed of rape and only 0.7% of felony defendants are convicted of murder. Thomas
H. Cohen and Brian A. Reaves, U.S. Department of Justice, Bureau of Justice Statis-
tics, Felony Defendants in Large Urban Counties, 2002 (2006), 27, table 28.

143. Bruce W. Behrman and Sherrie L. Davey, “Eyewitness Identifi cation in Actual
Criminal Cases: An Archival Analysis,” 25 Law & Hum. Behav. 475 (2001); Gonza-
lez et al., “Response Biases in Lineups and Showups,” 535– 536.

144. Wells, “Systemic Reforms,” 632– 635.
145. Wells and Quinlivan, “Suggestive Eyewitness Identifi cation Procedures.” On the

challenges of conducting fi eld experiments, see Nancy K. Steblay, “Commentary on
‘Studying Eyewitness Investigations in the Field’: A Look Forward,” 32 Law & Hum.
Behav. 11 (2008). On the methodological fl aws of a fi eld study conducted in Chicago,
see D. L. Schacter et. al., “Policy Forum: Studying Eyewitness Investigations in the
Field,” 32 Law & Hum. Behav. 3– 5 (2008).

146. Trial Transcript, 43– 48, 65, 107– 108, State of Louisiana v. Dennis P. Brown, No.
128,634 (La. Dist. Ct. Sept. 11– 12, 1985); see also Center on Wrongful Convictions of
Youth, Dennis Brown, at www .cwcy .org/ exonereesViewDetail .aspx ?id15. The offi –
cer administering the lineup, however, did not warn the victim that the attacker
might not be present in the lineup. Brown also later falsely confessed to having com-
mitted the crime.

147. Ibid., 627– 629; see also Loftus et. al., Eyewitness Testimony: Civil and Criminal
§ 4- 10 (LexisNexis 4th ed. 2007).

148. See Ohio Rev. Code Ann. § 2933.83(A)(6) (2010); N.C. Gen. Stat. § 15A- 284.52(b)(14)
(2007); Amy Klobuchar, Nancy K. Mehrkens Steblay, and Hilary Lindell Caliguri,
“Improving Eyewitness Identifi cations: Hennepin County’s Blind Sequential Lineup
Pi lot Project,” 4 Cardozo Pub. L. Pol’y. & Ethics J. 381, 405– 410 (2006); Otto H. Ma-
cLin, Laura A. Zimmerman, and Roy S. Malpass, “PC_Eyewitness and Sequential
Superiority Eff ect: Computer- Based Lineup Administration,” 3 Law & Hum. Behav.
303, 304– 305 (2005).

Notes to Pages 78–82 � 311

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

149. Researchers have also long recommended use of a sequential pre sen ta tion, when
also double- blind, although some recent research suggests that under some condi-
tions sequential procedure performs less well, in part because it is a “more conser-
vative” procedure. See Wells, “Systemic Reforms,” at 627– 628; Wells et al., “Eye-
witness Identifi cation Procedures,” 616– 617, 639– 640 (citing “rather impressive”
research in support of the sequential procedure). But see Paul Giannelli and Myrna
Raeder, ABA Criminal Justice Section, “Achieving Justice: Freeing the Innocent,
Convicting the Guilty,” 25 (2006).

150. See Wells et al., “Eyewitness Identifi cation Procedures,” 635– 636, 640– 641; see
also Amy Douglass and Nancy Steblay, “Memory Distortion in Eyewitnesses: A
Meta- Analysis of the Post- Identifi cation Feedback Eff ect,” 20 Applied Cognitive Psy-
chol. 859 (2006).

151. Abdal Trial Transcript, 209– 210.
152. However, juror understanding of factors aff ecting eyewitness accuracy may be im-

proving over time, due to media accounts of errors. See Sarah L. Desmarais and
J. Don Read, “After 30 Years, What Do We Know about What Jurors Know? A Meta-
Analytic Review of Lay Knowledge Regarding Eyewitness Factors,” Law & Hum.
Behav. (2010).

4 . flawed forensics

1. The victim later wrote a confessional book explaining her false testimony titled For-
give Me. Cathleen C. Webb and Marie Chapian, Forgive Me (Old Tappan, NJ: F.H.
Revell Co., 1985).

2. Rob Warden, “The Rape that Wasn’t, the First DNA Exoneration in Illinois,”
http:// www .law .northwestern .edu/ wrongfulconvictions/ exonerations/ ilDotson
Summary .html .

3. Ibid.
4. Clive A. Staff ord Smith and Patrick D. Goodman, “Forensic Hair Comparison

Analysis: Nineteenth Century Science or Twentieth Century Snake Oil?” 27 Colum.
Hum. Rts. L. Rev. 227, 242– 245 (1996).

5. Trial Transcript, 359, State of Illinois v. Gary E. Dotson, No. P.C. 4333 (Ill. Cir. Ct.
July 25, 1985).

6. Affi davit of Edward T. Blake, D. Crim., State of Illinois v. Gary E. Dotson, No. P.C.
4333 23 (Ill. Cir. Ct. July 29, 1985).

7. Ibid.; see also Michael Serrill and Laura Lopez, “Law: Cathy and Gary in Medial-
and,” Time, May 27, 1985.

8. Larry Green, “12- Year Legal Nightmare at End,” Los Angeles Times, August 15,
1989, 5.

9. See National Academy of Sciences Urges Reform, http:// www .innocenceproject
.org/ Content/ 1866 .php .

10. There were at least 185 exonerees who had forensic evidence in their cases. Of those,
153 trials were located and analyzed in this chapter. To break it down step by step: 4

312 Notes to Pages 82–89

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

exonerees of the 185 who had forensic evidence in their cases pleaded guilty, and
thus had no trial. In 2 more cases, those of Dana Holland and Earl Washington Jr.,
forensic analysis was conducted before trial, but there was no trial testimony regard-
ing that analysis. That leaves 179 exonerees who had forensics introduced at a trial.
However, in 4 of these cases, the forensic evidence was introduced at trial by stipula-
tion, so there was no testimony by a forensic analyst for either side. Six additional
exonerees had analysts testify only for the defense at their trial. Thus, in 169 cases,
analysts testifi ed for the prosecution. Finally, in 16 exonerees’ cases, although news
reports indicate there was forensic analysis at trial, the transcripts could not be lo-
cated. That left 153 cases in which prosecution analysts testifi ed and transcripts were
located.

11. Of the 185 exonerees’ cases involving forensic analysis, 125 were rape cases, 41 were
rape and murder cases, 16 were murder cases, and 3 were “other” crimes. Thus, 73%
(125 of 171) rape cases had forensics, 79% (41 of 52) rape- murder cases had forensics,
and 73% (16 of 22) murder cases had forensics. The proportion of cases involving
forensic evidence was fairly evenly distributed across crimes.

12. The National Academies, Committee on Identifying the Needs of the Forensic Sci-
ences Community, Strengthening Forensic Science in the United States: A Path For-
ward (2009), 22 (hereafter cited as NAS Report).

13. See Joseph L. Peterson and Penelope N. Markham, “Crime Laboratory Profi ciency
Testing Results, 1978– 1991, II: Resolving Questions of Common Origin,” 40 J.
Forensic Sci. 1009, 1010 (1995).

14. See Michael J. Saks and Jonathan J. Koehler, “The Individualization Fallacy in
Forensic Science,” 61 Vand. L. Rev. 199 (2008); NAS Report, 7– 8, 21– 22.

15. Daubert v. Merrell Dow Pharmaceuticals, Inc., 509 U.S. 579, 595 (1993); United
States v. Frazier, 387 F.3d 1244, 1263 (11th Cir. 2004).

16. 509 U.S. 579 (1993).
17. Samuel R. Gross and Jennifer L. Mnookin, “Expert Information and Expert Evi-

dence: A Preliminary Taxonomy,” 34 Seton Hall L. Rev. 141, 169 (2003).
18. NAS Report, 9– 13, 53, 106– 109; see also Peter J. Neufeld, “The (Near) Irrelevance of

Daubert to Criminal Justice: and Some Suggestions for Reform,” 95 Am. J. Pub.
Health S107 (2005).

19. After the NAS contacted me and requested data on the role that forensic science
played in wrongful convictions, I began a study of the forensic science testimony in
exonerees’ trials, which was published in the Virginia Law Review and coauthored
with Peter Neufeld. We found that invalid forensic science testimony was not just com-
mon but prevalent at these exonerees’ trials. Invalid testimony was identifi ed for 60% of
the exonerees whose trial transcripts were obtained and had prosecution forensic tes-
timony. See Brandon L. Garrett and Peter J. Neufeld, “Invalid Forensic Science Testi-
mony and Wrongful Convictions,” 95 Va. L. Rev. 1 (2009). This chapter updates those
fi ndings with new transcripts subsequently obtained. This chapter also has a broader
focus and examines not just validity of testimony at trials but also vague testimony and
reliability of underlying forensic techniques.

20. Michael J. Saks, “Judging Admissibility,” 35 J. Corp. L. 135, 145 (2009).

Notes to Pages 89–91 � 313

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

21. See Paul C. Giannelli, “Regulating Crime Laboratories: The Impact of DNA Evi-
dence,” 15 J. L. & Pol’y 59, 61– 67, 72 (2007).

22. The phenomenon of cognitive bias, including that of forensic scientists, is discussed
in Chapter 9.

23. Melendez- Diaz v. Massachusetts, 129 S.Ct. 2527, 2536 (2009).
24. Robert Bazell, “DNA Acquittals Shaking Up Forensic Science,” NBC News, Febru-

ary 12, 2008; Jane Campbell Moriarty, “ ‘Misconvictions,’ Science, and the Minis-
ters of Justice,” 86 Neb. L. Rev. 1 (2007).

25. “DNA Crime Labs: The Paul Coverdell National Forensic Sciences Improvement
Act: Hearing Before the Comm. on the Judiciary,” 107th Cong. 2– 3 (2001) (statement
of Sen. Orrin G. Hatch, Chairman, S. Comm. on the Judiciary).

26. See William C. Thompson, “Beyond Bad Apples: Analyzing the Role of
Forensic Science in Wrongful Convictions,” 37 Southwestern U. L. Rev. 101, 112– 119
(2009).

27. The states are: Arizona, California, Connecticut, Florida, Georgia, Idaho, Illinois,
Indiana, Kansas, Kentucky, Louisiana, Mary land, Massachusetts, Mississippi, Mis-
souri, Montana, Nebraska, Nevada, New Jersey, New York, Ohio, Oklahoma, Penn-
sylvania, South Carolina, Texas, Virginia, Wisconsin, and West Virginia.

28. Gilchrist v. Citty, 173 Fed.Appx. 675, 677 (10th Cir. 2006).
29. Quantifi cation techniques later permitted additional conclusions regarding mixed

stains, but these exonerees’ trials predated such techniques.
30. NAS Report, 160.
31. See Barry Scheck, Peter Neufeld, and Jim Dwyer, Actual Innocence (New York:

Signet 2001), 209– 210.
32. NAS Report, 160– 161.
33. Williamson v. Reynolds, 904 F. Supp. 1529, 1552– 1553 (E.D. Okla. 1995).
34. A report later issued by several of the most distinguished forensic analysts in the

nation concluded that while an experienced analyst might make an error when com-
paring a single hair, “it is highly unlikely that a competent hair examiner would in-
correctly associate” so many diff erent hairs. Richard E. Bisbing et al., “Peer Review
Report: Montana v. Jimmy Ray Bromgard,” 2.

35. Trial Transcript, 250, State v. Chester Bauer, No. 83- CR- 27 (Mont. Dist. Ct. July 16,
1983).

36. Trial Transcript, 385, State v. Timothy Edward Durham, No. CF- 91- 4922 (Okla.
Dist. Ct. Mar. 9, 1993).

37. Trial Transcript, 246, Commonwealth of Kentucky v. William Gregory (Ky. D.Ct.
1993).

38. Trial Transcript, 177, State v. Curtis Edward McCarty, No. CRF- 85- 02637 (Okla.
Dist. Ct. Mar. 24, 1986).

39. Trial Transcript, 152, State v. Larry L. Peterson, A-3034- 89T4 (N.J. Super. Ct. Mar.
6, 1989).

40. Max M. Houck et al., “The Science of Forensic Hair Comparisons and the Admis-
sibility of Hair Comparison Evidence: Frye and Daubert Considered,” Mod. Micros-
copy J. 5 (Mar. 2, 2004).

314 Notes to Pages 92–99

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

41. Trial Transcript, 2837, People v. Kharey Wise, No. 4762/89 (N.Y. Sup. Ct. Nov. 13,
1990).

42. NAS Report, 161.
43. The sixteen exonerees are: J. Abbitt, R. Alexander, J. Deskovic, C. Elkins, N. Hatch-

ett, T. Hayes, E. Karage, R. Krone, R. Matthews, A. McCray, J. Ochoa, K. Richard-
son, M. Roman, R. Santana, Y. Salaam, and K. Wise. The four cases where the DNA
tests appeared to show guilt were three that involved invalid testimony, the cases of
G. Alejandro, C. Heins, and J. Sutton, while the last case, that of T. Durham, in-
volved a lab error. James Ochoa had no trial and pleaded guilty despite a DNA exclu-
sion. Krone had exculpatory DNA results presented at a second trial, of which I was
not able to obtain a transcript. For those two cases, Ochoa and Krone, trial tran-
scripts were not obtained. Thus a total of eigh teen cases with DNA testimony at trial
were obtained. I also note that Richard Alexander was excluded at trial by DNA in an
additional rape that he was not charged with, but which prosecutors said was part of
a series of attacks by the same perpetrator.

44. Trial Testimony, 149, State v. Gilbert Alejandro, No. 90- 09- 8445- CR (Tex. Dist.
Ct. Dec. 11, 1990).

45. Trial Testimony, 168– 230, State v. Josiah Sutton, No. 800450 (Tex. Dist. Ct.
1999).

46. William Thompson, Review of DNA Evidence in State of Texas v. Josiah Sutton
(2003), http:// www .scientifi c .org/ archive/ Thompson %20Report .PDF .

47. See William C. Thompson et al., “How the Probability of a False Positive Aff ects
the Value of DNA Evidence,” 48 J. Forensic Sci. 47, 48 (2003).

48. Erin Murphy, “The Art in the Science of DNA: A Layperson’s Guide to the Subjec-
tivity Inherent in Forensic DNA Typing,” 58 Emory L.J. 489 (2008).

49. Adam Liptak, “The Nation: You Think DNA Evidence Is Foolproof? Try Again,”
New York Times, March 16, 2003.

50. C. Michael Bowers, “Scientifi c Issues,” in Faigman et al., Modern Scientifi c Evi-
dence: The Law and Science of Expert Testimony (St. Paul, MN: West Group, 2009–
2010), § 37:12– 37:13, 37:23, 37:34– 37:36.

51. See Flynn McRoberts and Steve Mills, “From the Start, a Faulty Science,” Chicago
Tribune, October 19, 2004.

52. NAS Report, 176.
53. Paul C. Giannelli, “Bite Mark Evidence,” GP Solo (Sept. 2007); Faigman et al., Mod-

ern Scientifi c Evidence, § 37:4– 37.6.
54. The fi fth exoneree with invalid bite mark testimony was Roy Brown, whose case is

discussed below; the odontologist found inconsistencies and yet called them “ex-
plainable.” The two exonerees with vague testimony concerning bite mark compari-
sons were James O’Donnell and Calvin Washington. In O’Donnell’s case the marks
were said to be “consistent” with his teeth. In Washington’s case, while Washington
was himself excluded (he was missing most of his teeth), his codefendant was said to
have teeth that were “consistent with” the bite marks.

55. Trial Transcript, 15, State of Arizona v. Ray Milton Krone, No. CR 92- 00212 (on fi le
with authors).

Notes to Pages 99–103 � 315

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

56. Robert Nelson, “About Face,” Phoenix New Times, April 21, 2005.
57. Maurice Possley and Steve Mills, “Guilty, Said Bite Expert. Bogus, Says DNA,”

Chicago Tribune, July 10, 2008.
58. Iain A. Pretty, “Reliability of Bitemark Evidence,” in Bitemark Evidence, ed. Robert

B. J. Dorion (New York: Marcel Dekker, 2005), 531, 543.
59. See Garrett and Neufeld, “Invalid Forensic Science Testimony,” 68.
60. ABFO Bitemark Methodology Guidelines at www .abfo .org/ pdfs/ ABFO %20Manual

%20 – %20Revised %2010 -5 -2009 .
61. Trial Transcript, 294, State of Idaho v. Charles I. Fain (Idaho D. Ct. Oct. 14, 1983).
62. See NAS Report, 148– 149.
63. Faigman et al., Modern Scientifi c Evidence, § 38.1.- 38.2.
64. Trial Transcript, 290, State of Texas v. David Shawn Pope, No. F85- 98755- NQ

(Tex. Dist. Ct. Feb. 4, 1986).
65. Trial Transcript, 3- 207, 3- 214–3- 225, Commonwealth of Massachusetts v.

Stephan Cowans, No. 97- 11231 (Mass. Sup. Ct. June 24, 1998); Ron Smith & As-
sociates, Inc., Reference: Request for Latent Print Consultation Ser vices (Mar. 8,
2004), 6.

66. The cases are those of: Gilbert Alejandro, Gene Bibbins, Roy Brown, David Bryson,
Ulysses Charles, Stephan Cowans, Rolando Cruz, William Gregory, Alejandro Her-
nandez, Dana Holland, Ray Krone, Curtis McCarty, Neil Miller, Marlon Pendleton,
Larry Peterson, George Rodriguez, Lafonso Rollins, Josiah Sutton, Ronald Taylor,
Earl Washington, Kenneth Waters, and Ronald Williamson. In fi ve cases, such as the
Peterson case discussed next, we do not know if the analyst intentionally or negli-
gently concealed the presence of abundant spermatozoa that could have been tested;
in such cases, later analysis reported, though, that they were readily detected and
should easily have been tested.

67. See Garrett and Neufeld, “Invalid Forensic Science Testimony,” 76– 77.
68. Margaret Edds, An Expendable Man: The Near- Execution of Earl Washington, Jr.

(New York: New York University Press, 2003), 246.
69. Trial Transcript, 740, 774, People v. Roy Brown, 91- 2099 (N.Y. Sup. Ct. Jan. 13– 23,

1992).
70. Trial Transcript, 83, State of Louisiana v. Gene Bibbins, No. 2- 87- 979 (La. Dist. Ct.

Mar. 25, 1987).
71. See Gregory v. City of Louisville, 444 F.3d 725, 732 (6th Cir. 2006).
72. Maurice Possley et al., “Scandal Touches Even Elite Labs: Flawed Work, Re sis-

tance to Scrutiny Seen Across U.S.,” Chicago Tribune, October 21, 2004, § 1, 1.
73. The Innocence Project, Larry Peterson, at www .innocenceproject .org/ Content/ 148

.php .
74. Center for Wrongful Convictions, Dennis Williams, www .law .northwestern .edu/

wrongfulconvictions/ exonerations/ ilWilliamsChart .
75. Trial Testimony, 636– 637, State of Oklahoma v. Ronald Keith Williamson, CRF

87- 90 (Okla. Dist. Ct. April 21– 28, 1988).
76. Williamson v. Ward, 110 F.3d 1508, 1522 (10th Cir. 1997).
77. Barry Scheck et al., Actual Innocence, 165.

316 Notes to Pages 104–110

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

78. The Innocence Project, Robert Miller, www .innocenceproject .org/ Content/ 219
.php .

79. Jason Volentine, “Stolen Lives: The Story of the Beatrice Six (Part Three),” KOLN/
KGIN, March 17, 2009; Lynn Safranek, “Old- style Testing of Blood and Fluids Proves
to Be No Match for Modern DNA Testing,” Omaha World- Herald, January 26, 2009.

80. Maurice Possley, “ ‘Always Knew I Was Innocent,’ Imprisoned in a 1992 Sexual As-
sault, Marlon Pendleton Is Told by His Lawyer That New DNA Tests Show That He
Was Not the Assailant,” Chicago Times- Tribune, November 24, 2006.

81. Trial Testimony, 1– 169, Commonwealth v. Neil Miller, No. 085602– 085604 (Mass.
Sup. Ct. Dec. 14, 1990).

82. Frontline, Burden of Innocence, http:// www .pbs .org/ wgbh/ pages/ frontline/ shows/
burden/ profi les/ miller .html .

83. Fred C. Zacharias, “Structuring the Ethics of Prosecutorial Trial Practice: Can
Prosecutors Do Justice?” 44 Vand. L. Rev. 45, 91 (1991); see also Model Rules of
Prof ’l Conduct R. 3.4(e).

84. Darden v. Wainwright, 477 U.S. 168, 181 (1986).
85. People v. Linscott, 142 Ill.2d 22, 38 (Ill. 1991).
86. Trial Transcript, 899, Commonwealth of Pennsylvania v. Drew Whitley, CC 89-

2462 (Pa. Ct. of Common Pleas July 21, 1989).
87. Ibid., 50– 57 (July 24, 1989).
88. Paul C. Giannelli and Edward L. Imwinkelried, Scientifi c Evidence, (4th ed. 2007),

§ 4.01–.05, 13.07.
89. Paul C. Giannelli, “Microscopic Hair Comparison: A Cautionary Tale,” 46 No. 3

Criminal Law Bulletin Art. 7 (2010).
90. Only 3% of law enforcement requests to crime labs involve requests for DNA analy-

sis. See Matthew R. Durose, Bureau of Justice Statistics, Census of Publicly Funded
Forensic Crime Laboratories 10 (2005), http:// bjs .ojp .usdoj .gov/ content/ pub/ pdf/
cpff cl05 .

91. Ninety- three exonerees had invalid forensic science testimony at their trials. Nine-
teen more had vague and potentially misleading testimony, using terms like “associ-
ated with,” “similar,” or “consistent.” Fourteen more had exculpatory forensic evi-
dence that was not disclosed and/or erroneous analysis such as laboratory errors in
their cases. Finally, at least one transcript that could not be located, that of Harold
Buntin, according to available reports had invalid testimony at his trial; another had
vague testimony; and one more, that of Thomas Webb, involved vague testimony
that evidence was “consistent.”

92. In a preliminary eff ort, I obtained a small random set of transcripts from similar
rape and murder trials. Each of those defendants was found guilty and, as far as we
know, they really were. I found that just in these exonerees’ cases, almost two- thirds
of those trials had invalid forensic science testimony. The trials included the same
types of errors and they even involved some of the same analysts who testifi ed in the
exonerees’ trials. Thirty trial transcripts in such “matched” cases were collected
from Missouri (10 transcripts), Texas (11), and Virginia (9). See Garrett and Neufeld,
“Invalid Forensic Science Testimony,” 28– 29.

Notes to Pages 110–114 � 317

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

93. See, e.g., Lopez v. State, 643 S.W.2d 431, 433 (Tex. App. 1982); State v. Bridges, 421
S.E.2d 806, 808 (N.C. Ct. App. 1992).

94. See Rob Warden and Locke Bowman, “In de pen dent Crime Labs Could Stop
Forensic Fraud,” Chicago Sun- Times, November 7, 2004.

95. See Michael R. Bromwich, Executive Summary, in Fifth Report of the In de pen dent
Investigator for the Houston Police Department Crime Laboratory and Property
Room 1– 2 (May 11, 2006), at http:// www .hpdlabinvestigation .org .

96. Comm. on Scientifi c Assessment of Bullet Lead Elemental Composition Compari-
son, Nat’l Research Council, Forensic Analysis: Weighing Bullet Lead Evidence
(2004), 90– 94.

97. Andrew E. Taslitz, “Convicting the Guilty, Acquitting the Innocent: The ABA
Takes a Stand,” 19 Crim. Just. 18– 19 (2005).

98. See, e.g., United States v. Bentham, 414 F. Supp. 2d 472, 473 (S.D.N.Y. 2006) (“False
positives— that is, inaccurate incriminating test results— are endemic to much of
what passes for ‘forensic science.’ ”).

99. Identifying the Needs of the Forensic Sciences Community, www8 .nationalacade
mies .org/ cp/ projectview .aspx ?key48741 .

100. See NAS Report, 7.
101. Ibid., 37, 53.
102. 129 S.Ct. 2527, 2537 (2009).
103. The Court merely defended confrontation as not “useless” and “one means” to en-

sure accuracy where forensic error or incompetence “may be” uncovered through
cross- examination. Ibid., 2536– 2537.

104. See NAS Report, 21– 22.

5 . trial by l iar

1. Trial Transcript, 88– 89, 104– 105, State of Illinois v. David A. Gray, Case No. 78-
CF- 124 (Ill. Cir. Ct. Sept. 29, 1978).

2. Ibid., 152, 241.
3. Ibid., 252.
4. Ibid., 74.
5. Ibid., 160– 212.
6. Ibid., 173.
7. Patrick E. Gauen, “Evidence in Alton Rape Case Is Vanishing,” St. Louis Post-

Dispatch, July 3, 1998, A1; Gray Trial Transcript, 13, (Sept. 29, 1978).
8. Gray Trial Transcript, 8.
9. Ibid., 36.
10. Gauen, “Evidence in Alton Rape Case.”
11. Gray Trial Transcript, 111.
12. Ibid., 95.
13. Ibid., 112.
14. Ibid., 110.

318 Notes to Pages 114–120

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

15. Ibid., 297.
16. Ibid., 132, 144.
17. Ted Rohrlich and Robert W. Stewart, “Jail house Snitches: Trading Lies for Free-

dom,” Los Angeles Times, April 16, 1989, 1.
18. Gray Trial Transcript, 131.
19. Ibid., 303.
20. Ibid., 308– 310.
21. For that reason, the Illinois Appellate Court dismissed David Gray’s Massiah

claim brought on appeal. See People v. Gray, 299 N.E.2d 206, 210 (Ill. App. 5 Dist.
1979).

22. Gray Trial Transcript, 12.
23. Ibid., 31.
24. Ibid., 25– 26.
25. Ibid., 37.
26. Ibid., 44.
27. Ibid., 294– 295.
28. Ibid., 19.
29. See Gauen, “Evidence in Alton Rape Case.”
30. Hoff a v. United States, 385 U.S. 293, 311 (1966) (citation omitted).
31. Steve Mills and Ken Armstrong, “Another Death Row Inmate Cleared,” Chicago

Tribune, January 19, 2000, N1 (“Jail house in for mants are considered among the least
reliable witnesses in the criminal justice system.”); see also James S. Liebman, “The
Overproduction of Death,” 100 Colum. L. Rev. 2030, 2088– 2089 n.149 (2000); Alex-
andra Natapoff , Snitching: Criminal In for mants and the Erosion of American Justice
(New York: New York University Press, 2009), 70– 72.

32. United States v. Bernal- Obeso, 989 F.2d 311, 335 (9th Cir. 1993).
33. Samuel Gross, “Lost Lives: Miscarriages of Justice in Capital Cases,” 61 Law &

Contemp. Probs. 125, 138 (1998).
34. See Peter Neufeld, Letter to Hon. Governor Phil Bredesen, January 25, 2004 (on fi le

with author).
35. 360 U.S. 264 (1959).
36. 377 U.S. 201, 203– 206 (1964).
37. Kansas v. Ventris, 129 S. Ct. 1841, 1844 (2009).
38. Daniel Richman, “Cooperating Defendants: The Costs and Benefi ts of Purchasing

Information from Scoundrels,” 8 Fed. Sent. Rep. 292, 294 (1996).
39. Hoff a, 385 U.S. at 311.
40. Trial Transcript, 222, State of New York v. John Kogut, Ind. #61029 (Nassau County

Ct. May 13, 1986).
41. Trial Transcript, 46– 47, State of Michigan v. Kenneth Wyniemko, No. CR- 94- 1595

(Mich. D. Ct. Aug. 11, 1994); ibid., 64– 65 (Nov. 3, 1994).
42. Trial Transcript, 2126, State of Illinois v. Willie L. Rainge, Kenneth E. Adams, and

Dennis Williams, Information No. 78- I6- 5186 (Ill. Cir. Ct. Sept. 27, 1978).
43. Trial Transcript, 881, Commonwealth of Pennsylvania v. Drew Whitley, No. CC

8902462, 8902609 (Pa. Ct. of Common Pleas July 18, 1989).

Notes to Pages 120–128 � 319

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

44. Trial Transcript, 1229, State of Florida v. Wilton Allen Dedge, Case No. 82- 135- CF- A
(Fla. Cir. Ct. Aug. 22, 1984).

45. Trial Transcript, 1696, State of Texas v. Calvin Edward Washington, No. 87- 08- C
(Tex. D. Ct. Nov. 30, 1987).

46. Ibid., 1338.
47. Ibid., 1339.
48. Trial Transcript, 791, State of New York v. Steven P. Barnes, Ind. 89- 96 (N.Y. Sup.

Ct. May 15, 1989).
49. See Natapoff , Snitching, 94; R. Michael Cassidy, “ ‘Soft Words of Hope’: Giglio,

Accomplice Witnesses, and the Problem of Implied Inducements,” 98 Northwestern
U. L. Rev. 1129 (2004).

50. People v. Cruz, 643 N.E.2d 636, 643– 644 (Ill. 1994).
51. Williamson v. Ward, 110 F.3d 1508, 1512 (10th Cir. 1997).
52. Waymong Dotson Statement, August 4, 1987, 7 (on fi le with author).
53. Ibid., 8.
54. Trial Transcript, 726, 778, State of Idaho v. Charles I. Fain, Criminal Case No.

C-5448 (Idaho D. Ct. Oct. 26, 1983).
55. Ibid., 764.
56. Ibid., 771.
57. Ibid., 772.
58. Barnes Trial Transcript, 538.
59. Trial Transcript, 165– 166, Commonwealth of Pennsylvania v. Bruce Donald God-

schalk, No. 00934- 87, (Pa. Ct. of Common Pleas May 27, 1987).
60. Trial Transcript, 575– 576, State of Oklahoma v. Ronald Keith Williamson, CRF

87- 90 (Okla. D. Ct. April 22, 1988).
61. Trial Transcript, 727, State of New York v. John Restivo and Dennis Halstead, Ind.

#61322 (Nassau County Ct. October 20, 1986).
62. Ibid., 760– 761.
63. Ibid., 730.
64. Whitley Trial Transcript, 879.
65. Fain Trial Transcript, 740.
66. Ibid., 737, 782– 783.
67. Ibid., 850.
68. Ibid., 1322– 1323.
69. Trial Transcript, 2488, State of Florida v. Chad Richard Heins, Case No.

94- 3965- CF (Fla. D. Ct. Dec. 10, 1996).
70. Washington v. State, 822 S.W.2d 110, 121 (Tex. App. Waco 1991).
71. Godschalk Trial Transcript, 70 (May 27, 1987).
72. Ibid., 165.
73. Dedge v. State, 442 So. 2d 429, 430 (Fla. App. 5 Dist. 1983).
74. Staff , “Cases Involving Preston,” Florida Today (Melbourne, FL), August 30, 2009,

A12; John A. Torres and Jeff Schweers, “Dog Handler Led to Bad Evidence,” Flor-
ida Today (Melbourne, FL), June 21, 2009.

75. Dedge Trial Transcript, 1214– 1215.

320 Notes to Pages 128–135

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

76. Ibid., 418
77. Ibid., 1213.
78. Ibid., 1215.
79. Ibid., 1216.
80. Ibid., 417– 418.
81. John A. Torres, “After 22 Years Dedge Again Tastes Freedom,” Florida Today (Mel-

bourne, FL), August 13, 2004, 1; Dedge Trial Transcript, 1205, 1225.
82. Staff , “Prior Zacke Knowledge May Have Been Hidden,” Florida Today (Melbourne,

FL), January 23, 2006, A1.
83. Trial Transcript, 2– 11, Commonwealth of Pennsylvania v. Nicholas Yarris, No. 690-

82 (Pa. Ct. of Common Pleas June 29, 1982). He also said Yarris had claimed to have
told police about the murder “and he was trying to put it on someone else in order to
help him out with his fi rst case,” ibid., 2– 15.

84. Ibid., 1– 84, 1– 88, 1– 121, (June 28, 1982).
85. Ibid., 2– 9.
86. Ibid., 9 (July 1, 1982).
87. Ibid., 2– 12.
88. Ibid., 2– 11.
89. Fain Trial Transcript, 847, 1323.
90. People v. Halstead, 580 N.Y.S.2d 413 (2d Dep’t 1992).
91. Yarris Trial Transcript, 2- 39- 40.
92. Fain Trial Transcript, 743.
93. Calvin Washington Trial Transcript, 1363– 1364.
94. Ibid., 1382.
95. Restivo Trial Transcript, 750.
96. Rainge, Adams, and Williams Trial Transcript, 2125– 2135.
97. Trial Transcript at 646, 665– 666, State of Oklahoma v. Dennis Fritz, CRF 87- 90

(Okla. Dist. Ct. April 7, 1988).
98. Jim Cuddy Jr., “DNA Test Fails to Link Wilkinsburg Man to /81 Murder,” Pitts-

burgh Post- Gazette, May 23, 1991 D12.
99. John Biemer, “Exonerated Earlier of Rape, Man Also Wins Related Case,” Chicago

Tribune, June 6, 2003, 6.
100. Ren E. Lee, “Rape Case Closed, But ‘Justice Was Not Served,’ ” Houston Chronicle,

February 21, 2007 A1.
101. The only codefendants who were not exonerated were those who testifi ed against

Chaunte Ott. The codefendants were also excluded by the postconviction DNA
testing; however, one was himself murdered years before, and the other had re-
ceived a plea deal and had served his fi ve years. See Meg Jones, “Man Released from
Prison on New DNA Evidence in Murder,” Journal Sentinel, January 8, 2009.

102. People v. Jimerson, 652 N.E.2d 278, 282– 286 (Ill. 1995).
103. David Protess and Rob Warden, A Promise of Justice, (New York: Hyperion Books,

1998), chap. 12, 16.
104. Trial Transcript, 35– 36, 43, State of Oklahoma v. Ronald K. Williamson and Den-

nis L. Fritz, CRF 87- 90 (Okla. Dist. Ct. July 20, 1987).

Notes to Pages 135–141 � 321

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

105. Trial Transcript, 63– 64, Commonwealth of Massachussetts v. Kenneth Waters, No.
82- 4115- 4116 (Mass. Sup. Ct. May 4, 1983); http:// www .innocenceproject .org/ Con-
tent/ 285 .php .

106. See Staff , “Wrongfully Convicted? DNA Testing May Set Convicted Murderer Wil-
liam Dillon Free After 26 Years,” Florida Today, October 10, 2007, A1.

107. Bob Burtman, “Hard Time,” Houston News, September 10, 1998; Jeff rey Rice,
“Hard Time,” Houston Chronicle, November 26, 2000, 6.

108. Jeff rey S. Neuschatz et al., “The Eff ects of Accomplice Witnesses and Jail house In-
for mants on Jury Decision Making,” 32 Law & Hum. Behav. 137 (2008) (fi nding that
“conviction rates were unaff ected by the explicit provision of information indicating
that the witness received an incentive to testify” and “the presence of a confession,
albeit a secondary confession, had a signifi cant infl uence on mock juror conviction
rates.”)

109. Hoff a v. United States, 385 U.S. 293, 311 (1966).
110. Whitley Trial Transcript, 886– 888.
111. Ibid., 883– 884, 889– 890.
112. Dodd v. State, 993 P.2d 778, 784 (Okla. Crim. App. 2000) (adopting procedure for

jail house in for mant testimony that ensures “complete disclosure”); see also Cal.
Penal Code § 1127a(b) (West 2004) (requiring courts to instruct jury on in- custody
in for mant testimony); United States v. Villafranca, 260 F.3d 374, 381 (5th Cir. 2001)
(“The testimony of a plea- bargaining defendant is admissible if the jury is properly
instructed.”); State v. Bledsoe, 39 P.3d 38, 44 (Kan. 2002) (noting that trial court
“gave a cautionary jury instruction regarding the testimony of an in for mant”); Alex-
andra Natapoff , “Beyond Unreliable: How Snitches Contribute to Wrongful Con-
victions,” 37 Golden Gate U. L. Rev. 107, 112– 115 (2006) (proposing model statute
requiring pretrial evaluations of in for mant testimony).

113. 725 Ill. Comp. Stat. Ann. 5/115- 21(d) (West Supp. 2007).
114. Staff , “Dillon Forgives Accuser at Hearing,” Florida Today, November 2, 2009.

6. innocence on trial

1. “Autopsy Reports Coming In,” Culpeper Star- Exponent, July 7, 1982; “Police Be-
lieve Williams’ Killer Knew Victim,” Culpeper Star- Exponent, September 9, 1982.

2. Trial Transcript, 47, Commonwealth of Virginia v. Earl Ju nior Washington (Va.
Cir. Ct. Jan. 19, 1984).

3. Va. Code Ann. § 19.2- 163.
4. Washington Trial Transcript, 644– 647.
5. Ibid., 629– 630, 718– 720.
6. Ibid., 805– 811.
7. Ibid., 132, 151, 13. The court had only asked that the expert evaluate Washington’s

“present status” and capacity to stand trial. The psychologist did not examine his
mental capacity at the time of the murder or interrogation, nor whether mental retar-
dation could have aff ected the voluntariness of his confession. Washington’s attor-

322 Notes to Pages 141–147

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

ney later stated that he believed that the court would have denied any request for
funding for an expert. Brooke Masters, “Missteps on the Road to Injustice,” Wash-
ington Post, November 30, 2000.

8. Affi davit of John W. Scott, November 3, 1989 (on fi le with author); Brandon L. Gar-
rett and Peter J. Neufeld, “Invalid Forensic Science Testimony and Wrongful Con-
victions,” 95 Va. L. Rev. 1, 76– 77 (2009).

9. See, e.g., Nancy M. Steblay et al., “The Eff ects of Pre- trial Publicity on Juror Ver-
dicts: A Meta- Analytic Review,” 23 Law & Hum. Behav. 219 (1999).

10. The sixteen exonerees who pleaded guilty are: Larry Bostic, Marcellius Bradford,
Keith Brown, James Dean, John Dixon, Kathy Gonzalez, Anthony Gray, Eugene
Henton, William Kelly, Michael Marshall, Christopher Ochoa, James Ochoa,
Debra Shelden, Ada JoAnne Taylor, David Vasquez and Thomas Winslow. Three
exonerees, Steven Phillips, Jerry Townsend, and Arthur Whitley, were convicted at
a trial, but pleaded guilty to an additional charge or charges. Thus, a total of nine-
teen exonerees pleaded guilty to a crime of which they were later exonerated.

11. Diane Jennings, “Two Men’s DNA Exonerations in ’88 Austin Murder Reveal Tri-
umph, Tragedy,” Dallas News, February 24, 2008.

12. Hearing Transcript, 5, State of New Jersey v. John Dixon, Ind. No. 2683- 5- 91 (N.J.
Sup. Ct. Nov. 12, 1991).

13. See American Bar Association Standing Committee on Legal Aid and Indigent De-
fendants, Gideon’s Broken Promise: America’s Continuing Quest for Equal Justice
(2004), 17.

14. Memorandum of Law, State of Mary land v. Anthony Gray, Case No. C-91- 409 (Feb.
20, 1992).

15. Matthew Durose and Patrick A. Langan, U.S. Department of Justice, Bureau of
Justice Statistics, Felony Sentences in State Courts, 2004 (2007), 1, table 4.1; see also
Brandon L. Garrett, “Judging Innocence,” 108 Colum. L. Rev. 55, 74 n.73 (2008).

16. 60 Minutes, “DNA Helps Free Inmate after 27 Years,” May 4, 2008, at www .cb
snews .com/ stories/ 2008/ 05/ 02/ 60minutes/ main4065454 .shtml .

17. Brandon L. Garrett, “Claiming Innocence,” 92 Minn. L. Rev. 1629, 1680– 1681
(2008).

18. Stephen P. Garvey et al., “Juror First Votes in Criminal Trials,” 1 J. Empir. Leg.
Stud. 396 (2004); Saul Kassin and Lawrence Wrightsman, The Psychology of Evi-
dence and Trial Procedure (New York: Hemi sphere Publishing, 1985), 8.

19. Paula Hannaford- Agor et al., Are Hung Juries a Problem? National Center for State
Courts (2002), 49.

20. See Dan Simon, “A Third View of the Black Box: Cognitive Coherence in Legal
Decision Making,” 71 U. Chi. L. Rev. 511– 586 (2004) (providing a review of the lit-
erature); Saul M. Kassin and Lawrence Wrightsman, The American Jury on Trial:
Psychological Perspectives, (New York: Hemi sphere Publishing, 1988), 153– 156.

21. Trial Transcript, 1739, State of Connecticut v. Miguel Roman, No. 54423 (Ct. Sup.
Ct. April 30, 1990).

22. Hannaford- Agor et al., Are Hung Juries a Problem? 35 (fi nding 92% of defendants in
sample were tried alone).

Notes to Pages 148–154 � 323

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

23. See Dale A. Sipes et al., On Trial: The Length of Civil and Criminal Trials, Na-
tional Center for State Courts (1988).

24. See Trial Transcript, 204, State of Texas v. Carlos Lavernia, No. 76,122 (Tex. Dist.
Ct. Jan. 21, 1985); and Trial Transcript, 352– 353, State of Texas v. Thomas Cliff ord
McGowan, No. F85- 81070- MU (Tex. Dist. Ct. March 5, 1986).

25. Kim North Shine, “Convicted Rapist Turns to DNA Tests,” Macomb Free Press,
November 27, 2002.

26. Additional exonerees, 8%, argued that no crime had occurred at all (17 of 207 trials).
Unsurprisingly, such arguments failed. After all, DNA testing ultimately excluded
these people, and therefore, almost all of the exonerees’ cases were ones in which a
crime did in fact occur, but the wrong man was convicted.

27. Trial Transcript, 549, State of Texas v. Brandon Moon, No. 50,015 (Tex. D. Ct.
January 14, 1987).

28. Tara M. Burke and John W. Turtle, “Alibi Evidence in Criminal Investigations and
Trials: Psychological and Legal Factors,” 1 Canadian J. Police & Security Ser vices
286 (2004); R. C. L. Lindsay et al., “Mock- Juror Evaluations of Eyewitness Testi-
mony: A Test of Metamemory Hypotheses,” 15 J. App. Soc. Psychol. 447 (1986).

29. Trial Transcript, 614– 615, State of Oklahoma v. Timothy Edward Durham, No. CF-
91- 4922 (Okla. Dist. Ct. March 9, 1993).

30. Ibid., 580.
31. Trial Transcript, 375, State of Ohio v. Anthony Green, No. CR 228250 (Ohio Ct. of

Common Pleas, Oct. 13, 1988).
32. Trial Transcript, 503, 506, State of Louisiana v. Rickey Johnson, No. 30,770 (La.

Dist Ct. Jan. 5, 1983).
33. Trial Transcript, 1044– 1045, 1070, State of West Virginia v. James Edmund Rich-

ardson, Jr., Criminal No. 89- F-5 (W. Va. Cir. Ct. July 27, 1989).
34. Trial Transcript, V3- 26- 27, State of Missouri v. Antonio Beaver, No.961- 2972 (Apr.

24, 1997).
35. Trial Transcript, 23, State of Oklahoma v. Arvin Carsell McGee, No. CF- 88- 886,

CF- 89- 1344 (Okla D. Ct. Sept. 14, 1988).
36. See 547 U.S. 319 (2006); Keith A. Findley and Michael S. Scott, “The Multiple Di-

mensions of Tunnel Vision in Criminal Cases,” 2006 Wis. L. Rev. 291, 343– 346.
37. For two of the few articles discussing standards for admission of third- party guilt

evidence prior to Holmes, see Brett C. Powell, “Perry Mason Meets the ‘Legitimate
Tendency’ Standard of Admissibility (And Doesn’t Like What He Sees),” 55
U. Miami L. Rev. 1023 (2001); Stephen Michael Everhart, “Putting a Burden of Pro-
duction on the Defendant Before Admitting Evidence That Someone Else Commit-
ted the Crime Charged: Is It Constitutional?” 76 Neb. L. Rev. 272 (1997).

38. Trial Transcript, 477– 478, State of California v. Frederick Rene Daye, CR- 67014/
DA- A75647 (Cal. Sup. Ct. May 21, 1984).

39. Ibid., 557– 558.
40. See Tony Perry, “DNA Test Frees Inmate after 10 Years,” Los Angeles Times, Sep-

tember 29, 1994, 3; Carl Rothman, “Suspect in 1985 Rape and Murder of 2 Kids Is
Dead,” Newark Star- Ledger, November 24, 2009.

324 Notes to Pages 155–159

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

41. S. A. Reid, “DNA That Freed One Links Another to Cobb Rape,” Atlanta Journal-
Constitution, February 16, 2007, D4; Christopher J. McFadden, “The Exoneration
of Robert Clark,” Fulton County Daily Report, June 16, 2006, 4; Trial Transcript,
145– 147, State of Georgia v. Robert Clark, Jr., Ind. No. 82- 0481 (Ga. Sup. Ct. May
24, 1982).

42. Trial Transcript, 867– 868, State of Oklahoma v. Ronald Keith Williamson, CRF
87- 90 (Okla. Dist. Ct. March 22, 1988).

43. Trial Transcript, 164, State of Louisiana v. Dennis P. Brown, No. 128,634 (La. Dist.
Ct. Sept. 11, 1985).

44. Trial Transcript, 80, State of Texas v. Charles Chatman, No. F81- 2101 QK (Tex.
Dist. Ct. 1981).

45. Trial Transcript, 391, Commonwealth of Pennyslvania v. Thomas Doswell, No. CC
8603467 (Pa. Ct. of Common Pleas November 19, 1986).

46. Trial Transcript, 1398, State of Ohio v. Clarence A. Elkins, No. 98- 06- 1415 (Ohio
Ct. of Common Pleas May 20, 1999).

47. Trial Transcript, 878, State of Oklahoma v. Dennis Leon Fritz, No. CRF- 87- 90
(Okla. Dist. Ct. April 6, 1988).

48. Trial Transcript, 180– 181, Commonwealth of Pennsylvania v. Bruce Donald God-
schalk, No. 00934- 87, (Pa. Ct. of Common Pleas May 27, 1987).

49. Ronald Jones Transcript, 84 (July 13, 1999); ibid., 183 (Aug. 19, 1989).
50. Trial Transcript, 94, State of New York v. Anthony Capozzi, #85- 1379- 001 (N.Y.

Sup. Ct. Jan. 27, 1987).
51. Trial Transcript, 197, State of Texas v. Larry Fuller, No. F81- 8431- P (Tex. Dist. Ct.

Aug. 24, 1981).
52. Trial Transcript, 28– 29, State of Mary land v. Kirk N. Bloodsworth, No. 84- CR-

3138 (Md. Cir. Ct. Mar. 1, 1985).
53. Trial Transcript, 513, State of Georgia v. Douglas Eugene Echols, No.086- 0565 (Ga.

Sup. Ct. March 23, 1987).
54. Sharon Cohen, “Jailed at 14, Youth Refused to Surrender Hope,” Los Angeles Times,

June 9, 2002.
55. Trial Transcript, 870– 872, State of New York v. James O’Donnell, No. 289- 97 (N.Y.

Sup. Ct. April 28, 1998).
56. Trial Transcript, 69, State of New York v. Kevin Richardson and Kharey Wise, No.

4762/89 (N.Y. Sup. Ct. Oct. 22, 1990).
57. Trial Transcript, 10, Commonwealth of Massachusetts v. Stephan Cowans, No.97-

11231 (Mass. Sup. Ct. July 7, 1998).
58. Trial Transcript, 1211, 1214, State of Illinois v. Michael Evans and Paul Terry No.

76- 1105, 76- 6504 (Ill. Cir. Ct. April 25, 1977).
59. John H. Blume, “The Dilemma of the Criminal Defendant with a Prior Record—

Lessons from the Wrongfully Convicted,” 5 J. Empirical Legal Stud. 492 (2009)
(reviewing case summaries of 119 exonerees and fi nding that of those who testifi ed,
43% had criminal rec ords, while of those who did not, 93% had criminal rec ords).

60. Sixteen exonerees were excluded by DNA testing at the time of their convictions:
Joseph Abbitt, Richard Alexander, Jeff rey Deskovic, Clarence Elkins, Nathaniel

Notes to Pages 159–163 � 325

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Hatchett, Travis Hayes, Entre Nax Karage, Ray Krone, Ryan Matthews, Antron
McCray, James Ochoa, Kevin Richardson, Miguel Roman, Raymond Santana,
Yusef Salaam, and Korey Wise. However, one of the sixteen, James Ochoa, did not
have a trial but pleaded guilty despite DNA testing that excluded him. In addition to
the fi fteen exonerees who were excluded by all of the non- DNA forensics in their
case, twenty- three more exonerees were excluded by some but not all of the foren-
sics in their case.

61. In forty- one trials the State had two experts, in thirty- two trials the State had
three experts, in sixteen trials the State had four experts, in fi ve trials the State
had fi ve experts, in fi ve trials the State had six, and in fi ve trials the State had
seven.

62. See, e.g., Hannaford- Agor et al., Are Hung Juries a Problem? 47; Darryl Brown,
“The Decline of Defense Counsel and the Rise of Accuracy in Criminal Adjudica-
tion,” 93 Cal. L. Rev. 1585, 1603 & n.62– 63 (2005). Chapter 4 discusses diffi culties
exonerees had in obtaining forensic experts and why under existing doctrine courts
often deny such experts. See also Paul C. Gianelli, “Ake v. Oklahoma: The Right to
Expert Assistance in a Post- Daubert World,” 89 Cornell L. Rev. 1305 (2004).

63. Brief of Appellants, Jimmy Ray Bromgard, On Appeal from the United States Dis-
trict Court for the District of Montana, Billings Division, Cause No. CV- 05- 32-
BLG- RFC 4 (2009).

64. Adam Liptak, “DNA Will Let a Montana Man Put Prison Behind Him, But Ques-
tions Still Linger,” New York Times, Oct. 1, 2002.

65. Trial Transcript, 3, State v. Jimmy Ray Bromgard, No. 88108 (Mont. Dist. Ct. Nov.
16, 1987).

66. State v. Bromgard, 285 Mont. 170 (1997).
67. See, e.g., The Spangenberg Group, State and County Expenditures for Indigent

Defense Ser vices in Fiscal Year 2002 (2003), 34– 37.
68. Clair Johnson, “Yellowstone County Wins Bromgard Case,” Billings Gazette, No-

vember 19, 2009; see also Offi ce of the State Public Defender, at http:// publicde
fender .mt .gov/ (“On July 1, 2006 the Offi ce of the State Public Defender assumed
responsibility for statewide Public Defender Ser vices, previously provided by cities
and counties”).

69. State v. Stinson, 134 Wis. 2d 224 (Wis. App. 1986).
70. Thus more exonerees had public defenders and retained counsel than the average

criminal defendant in the United States, perhaps due to the disproportionate num-
bers in large jurisdictions and in serious felony cases, as well as the desire of these
innocent individuals to vigorously defend themselves at a trial. See Steven K. Smith
and Carol J. DeFrances, U.S. Department of Justice, Indigent Defense (1996),
1 (“nearly 80% of local jail inmates indicated that they were assigned an attorney”).
Studies suggest that retaining private counsel or counsel from a public defender’s
offi ce, as opposed to a court- appointed lawyer, is associated with better outcomes at
trial and also on appeal. See, e.g., David T. Wasserman, A Sword for the Convicted:
Representing Indigent Defendants on Appeal (New York: Greenwood Press, 1990),
99, 152.

326 Notes to Pages 163–167

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

71. Trial Transcript, 332– 334, State of Texas v. Donald Wayne Good, F33- 31435 (Tex.
Dist. Ct. Sept. 15, 1987).

72. Berger v. United States, 295 U.S. 78, 88 (1935).
73. As I discuss in Chapter 7, seventy- seven exonerees brought claims relating to pros-

ecutorial misconduct during their appeals or postconviction.
74. Kathleen A. Ridolfi and Maurice Possley, “Preventable Error: A Report on Prosecu-

torial Misconduct in California, 1997– 2008” (2010); James S. Liebman, Jeff rey Fa-
gan, and Valerie West, A Broken System: Error Rates in Capital Cases, 1973– 1995
(2000), 4– 5; Marshall J. Hartman and Stephen L. Richards, “The Illinois Death
Penalty: What Went Wrong?” 34 J. Marshall L. Rev. 409, 422– 430 (2001); Ken Arm-
strong and Maurice Possley, “Trial & Error; How Prosecutors Sacrifi ce Justice to
Win; The Verdict: Dishonor,” Chicago Tribune, January 10, 1999, 1.

75. In their appeals or postconviction, nineteen exonerees made Fourth Amendment
claims regarding police search or seizure of evidence or lack of probable cause to
arrest, all without any success. Similarly, eight exonerees alleged that law enforce-
ment destroyed evidence that could have helped their case, without success.

76. Steve McGonigle and Robert Tharp, “DA Joins Fight to Clear Man,” Dallas News,
February 23, 2007.

77. Elkins v. Summit County, Ohio, No. 5:06- CV- 3004, 2009 WL 1150114 (N.D. Ohio
April 28, 2009).

78. Doswell v. City of Pittsburgh, Civil Action No. 07- 0761, 2009 WL 1734199, at *5
(W.D. Pa. June 16, 2009).

79. McGowan Trial Transcript, 396.
80. Trial Transcript, 455– 456, State of Ohio v. Anthony Green, No. CR 228250 (Ohio

Ct. of Common Pleas Oct. 13, 1988).
81. Trial Transcript, 2461– 2462, State of Florida v. Chad Richard Heins, Case No.

94- 3965- CF (Fla. D. C. Dec. 10, 1996).
82. Trial Transcript, 349, Commonwealth of Virginia v. Arthur Lee Whitfi eld, F 841- 82

(Va. Cir. Ct. Jan. 14, 1982).
83. McCarty v. State, 765 P.2d 1215, 1220 (Okla. Ct. of Crim. App. 1988).
84. Taylor v. State, 1997 WL 167849, at *3 (Tex. App.- Hous. 1 Dist. April 10, 1997).
85. Few criminal trials result in hung juries. One study of felony trials in four counties

found that about 6% of trials resulted in a hung jury and 4% more resulted in a mis-
trial. Of those, less than a third result in a second jury trial; most are dismissed or
result in guilty pleas. Hannaford- Agor et al., Are Hung Juries a Problem? 25, 31.

86. The nine exonerees who had hung juries are: Richard Alexander, Scott Fappiano,
Donald Wayne Good, David Gray, Eddie Lowery, Arvin McGee, Willie Nesmith,
Thomas McGowan, and Julius Ruffi n. McGee had a hung jury and then a mistrial
(he was tried three times). Julius Ruffi n had two hung juries and was convicted in
his third trial. The fi fteen exonerees who had multiple trials because their convic-
tions were vacated on appeal or postconviction are: Kirk Bloodsworth, Ronald Cot-
ton, McKinley Cromedy, Rolando Cruz, Wilton Dedge, Michael Evans, Donald
Wayne Good (he had both two hung juries and one new trial granted due to a vaca-
tur), Alejandro Hernandez, Darryl Hunt, Ray Krone, Johnnie Lindsey, Curtis

Notes to Pages 167–172 � 327

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

McCarty, Willie Rainge, Mark Webb, and Dennis Williams. All of those had two
trials, except Cruz and Hernandez, who had three trials. In addition to those
twenty- three exonerees who had multiple trials, Dana Holland, Steven Phillips, and
LaFonso Rollins had multiple trials for a diff erent reason, not due to hung juries or
new trials granted, but rather because they had separate trials for separate crimes
that they were charged with.

87. Trial Transcript, 1078– 1085, 774, People v. Roy Brown, 91- 2099 (N.Y. Sup. Ct. Jan.
13– 23, 1992).

88. Trial Transcript, 917– 918, People v. Anthony Capozzi, 4- 89- 657 (N.Y. Sup. Ct. Feb.
5, 1987).

89. Trial Transcript, 2525, State v. Chad Richard Heins, No. 94- 3965- CF (Fla. Cir. Ct.
Dec. 13, 1994); see also Brandon L. Garrett and Peter J. Neufeld, “Invalid Forensic
Science Testimony and Wrongful Convictions,” 95 Va. L. Rev. 1, 66 (2009).

90. Trial Transcript, 46, Jerry E. Watkins v. State of Indiana, 3CSCC- 87C8- CB- 764
(Ind. Sup. Ct. Aug. 1987).

91. Alan Berlow, “The Wrong Man,” Atlantic, November 1999.
92. David Harper, “McGee Case: City to Pay: Lawsuit Is Settled for $12.25 million,”

Tulsa World (Okla.), June 3, 2006.
93. Trial Transcript, 293, State of Oklahoma v. Arvin Carsell McGee, No. CF- 88- 886,

CF- 89- 1344 (Okla Dist. Ct. Sept. 14, 1988); ibid., 11– 16 (Sept. 28, 1988 Hearing).
94. Trial Transcript, 619, Commonwealth v. Eric Sarsfi eld, No. 87- 66 (Mass. Sup. Ct.

July 6, 1987).
95. Ken Armstrong and Maurice Possley, “Reversal of Fortune,” Chicago Tribune,

January 13, 1989.
96. Trial Transcript, Commonwealth of Virginia v. Willie Davidson, No. 919- 81 (Va.

Cir. Ct., May 27, 1981) (page numbers illegible).
97. Trial Transcript, E-33, State of Illinois v. Richard Johnson, No. 91- CR- 20794 (Ill.

Cir. Ct. Oct. 6, 1992).
98. Trial Transcript, 390– 391, State of Utah v. Bruce Dallas Goodman, No. 605 (Utah

Dist. Ct. Jan. 30, 1986).
99. Trial Transcript, 36, State of South Carolina v. Perry Renard Mitchell, No.

83- GS- 32- 479 (Ct. of Gen. Sess. January 23, 1984). The jury took only one hour to
convict. Ibid., 37.

100. For a discussion of related issues, including some advantages that defendants have
over prosecutions, but concluding that the playing fi eld in criminal cases is deci-
sively tilted, see Richard A. Uviller, A Tilted Playing Field: Is Criminal Justice Un-
fair? (New Haven, CT: Yale University Press, 1997).

101. Dennis Fritz, Journey Toward Justice (Santa Ana, CA: Seven Locks Press, 2006),
318.

102. Fritz Trial Transcript, 1071– 1072.
103. Fritz, Journey Toward Justice, 318.

328 Notes to Pages 172–177

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

7. judging innocence

1. See Shaila Dewan, “Despite DNA Test, Prosecutor Retries a ’92 Murder Case,”
New York Times, September 6, 2007.

2. See Flynn McRoberts and Steve Mills, “From the Start, a Faulty Science,” Chicago
Tribune, October 19, 2004.

3. Brewer v. State, 725 So. 2d 106 (Miss. 1998).
4. Ibid., 133.
5. Brewer v. Mississippi, 526 U.S. 1027 (Mem) (1999).
6. Brewer v. State, 819 So. 2d 1165, 1168 (Miss. 2000).
7. Brewer v. State, 819 So. 2d 1169 (Miss. 2002).
8. McCarty v. State, 765 P.2d 1215, 1218 (Okla. Cr. 1988).
9. McCarty v. State, 904 P.2d 110 (Okla. Crim. App. 1995) (denying relief on claims

regarding hair comparison and serology testimony at the second trial, but remand-
ing for new capital sentencing hearing).

10. McCarty v. State, 114 P.3d 1089 (Okla. Crim. App. 2005).
11. See Cheryl Camp, “Convicted Murderer Is Freed in Wake of Tainted Evidence,”

New York Times, May 22, 2007, A16.
12. Roger J. Traynor, The Riddle of Harmless Error (Columbus: Ohio State University

Press, 1969).
13. Judge Jerome Frank and Barbara Frank, Not Guilty (Garden City, NY: Doubleday,

1957).
14. See Victor E. Flango, National Center for State Courts, Habeas Corpus in State and

Federal Courts (1994), 62– 63; Nancy King, Fred L. Cheesman II, and Brian J. Os-
trom, National Center for State Courts, Final Technical Report: Habeas Litigation
in U.S. District Courts (2007), 58.

15. Later in the chapter I describe in greater detail the “matched comparison group”
that I constructed to compare the exonerees’ reversal rate to that of a similarly situ-
ated group of cases. In that matched group, there was no signifi cant diff erence in the
reversal rate (and just a single additional reversal).

16. These results update a study that I previously published, titled “Judging Inno-
cence,” which examined written decisions during the appeals and postconviction
litigation of the fi rst 200 exonerees. See Brandon L. Garrett, “Judging Innocence,”
108 Colum. L. Rev. 55 (2008). As I explained in that study, by written decisions
I refer to decisions that were available on Westlaw or Lexis- Nexis or state court da-
tabases and that provided a reason for the decision, regardless whether they were
characterized by the judge as “reported” or “unreported.” Ibid., 68 & n.46. Most
postconviction rulings are summary or not published, and judges often rule on se-
lected claims, without saying what other claims were raised in a petition. Failure to
issue a written decision, however, may not be random. For example, judges may
tend to write decisions just discussing the claims they perceive to have the most
merit. For all of those reasons, which create challenges for any eff ort to study crimi-
nal appeals and postconviction review, the data reported here is from a selected
sample and necessarily incomplete.

Notes to Pages 178–184 � 329

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

17. See Brandon L. Garrett, “The Substance of False Confessions,” 62 Stan. L. Rev.
1051, 1107– 1109 (2010).

18. Trial Transcript, 2, 43, State of Oklahoma v. Ronald K. Williamson and Dennis L.
Fritz, CRF 87- 90 (Okla. Dist. Ct. July 20, 1987).

19. The exonerees are: James Dean, Jeff rey Deskovic, Bruce Godschalk, Byron Halsey,
Alejandro Hernandez, Nathaniel Hatchett, Stephen Linscott, Freddie Peacock,
Yusef Salaam, Jerry Townsend, Douglas Warney, Earl Washington Jr., and Ronald
Williamson. See Washington v. Murray, 4 F.3d 1285, 1292 (4th Cir. 1993) (stating
that Washington “had confessed to the crime not in a general manner, but as one
who was familiar with the minutiae of its execution”); Godschalk v. Montgomery
County Dist. Attorney’s Offi ce, 177 F. Supp. 2d 366, 367 (E.D. Pa. 2001) (quoting
unpublished state court decision fi nding “plaintiff ’s conviction ‘rests largely on his
own confession which contains details of the rapes which were not available to the
public’ ” (citation omitted)); Townsend v. State, 420 So. 2d 615, 617 (Fla. Dist. Ct.
App. 1982) (“Townsend confessed to all of the collateral crimes as well as those for
which he was charged, and he took the police to the scene and corroborated facts
known to the police which only the killer would know.”); People v. Hatchett, No.
211131, 2000 WL 33419396, at *1 (Mich. Ct. App. May 19, 2000) (stating that “the
prosecution presented overwhelming evidence” and that police “testifi ed that de-
fendant’s statement included information that only the perpetrator of the crimes
would know,” facts “fully corroborative” of the victim’s account); State v. Dean, 464
N.W.2d 782, 789 (Neb. 1991) (“[T]he presentence investigation contained numer-
ous statements made by the defendant to law enforcement offi cers. Those statements
were corroborated not only by the physical evidence found at the crime scene and
the scientifi c examination of that evidence, but also by interviews with other people
involved or intimately familiar with some details of the crimes against the deceased
as heretofore related.”); State v. Halsey, 748 A.2d 634, 636– 638 (N.J. Super. Ct.
App. Div. 2000) (citing “overwhelming” evidence of his guilt and describing each of
the facts he supposedly volunteered in his confession); People v. Peacock, 417
N.Y.S.2d 339 (N.Y. App. Div. 1979) (denying relief citing “strong evidence of guilt,
including defendant’s confession”); People v. Warney, 750 N.Y.S.2d 731, 732– 733
(N.Y. App. Div. 2002) (“Defendant confessed to the crime and gave accurate de-
scriptions of many details of the crime scene.”); People v. Salaam, 590 N.Y.S.2d 195,
196 (N.Y. App. Div. 1992) (“Details of this statement were corroborated overwhelm-
ingly by substantial physical evidence.”); Williamson v. State, 812 P.2d 384, 396
(Okla. Crim. App. 1991) (“[T]he Appellant made certain admissions, in addition to
a confession, which were corroborated by the extrinsic evidence.”).

20. People v. Cruz, 121 Ill.2d 321, 336 (Ill. 1988).
21. Those exonerees are Curtis McCarty, Marvin Mitchell, Bruce Nelson, and Walter

Snyder.
22. 388 U.S. 218 (1967).
23. State v. Cromedy, 727 A.2d 457 (N.J. 1999) (reversing conviction for failure to in-

struct jury concerning error and cross- racial identifi cations); Jean v. Rice, 945 F.2d
82, 87 (4th Cir. 1991) (fi nding violation of Brady where use of hypnosis was not dis-

330 Notes to Pages 185–187

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

closed to the defense and where there was “little” evidence to corroborate the wit-
ness identifi cations); State v. Cotton, 351 S.E.2d 277, 280 (N.C. 1987) (granting new
trial where evidence of eyewitnesses’ identifi cation of another person supported
third- party guilt theory); Webb v. State, 684 S.W.2d 800, 801 (Tex. App. 1985) (re-
versing conviction where victim violated rule against discussion with other wit-
nesses); People v. Evans, 399 N.E.2d 1333, 1337 (Ill. App. 1979) (noting that new trial
had previously been granted to Evans).

24. 432 U.S. 98 (1977).
25. People v. Barnes, 558 N.Y.S.2d 339 (4 Dept. 1990).
26. Bryson v. State, 711 P.2d 932 (Okla. Cir. 1985).
27. State v. Green, 585 N.E.2d 990 (Ohio App. 8 Dist. 1990).
28. McCarty v. State, 765 P.2d 1215, 1218 (Okla. Cr. 1988).
29. State v. Krone, 897 P.2d 621, 624– 625 (1995).
30. People v. Linscott, 159 Ill. App. 3d 71, 80– 81 (Ill. App. Ct. 1987).
31. Williamson v. Ward, 110 F.3d 1508, 1520 (10th Cir. 1997) (noting that in reversing the

conviction the district court also emphasized failure to adequately challenge the fo-
rensic evidence, and not reviewing the merits of that decision); People v. Williams,
444 N.E.2d 136, 138, 143 (Ill. 1982) (reversing after trial attorney was disbarred, cit-
ing counsel’s joint repre sen ta tion of three capital defendants before two juries and
also citing failures to move to suppress central evidence, including hair evidence);
People v. Rainge, 445 N.E.2d 535, 547 (Ill. App. Ct. 1983) (reversing on similar
grounds).

32. People v. Rainge, 445 N.E.2d 535, 551 (Ill. App. Ct. 1983).
33. State v. Bauer, 210 Mont. 298 (1984).
34. See Massiah v. United States, 377 U.S. 201, 203– 206 (1964).
35. State v. Hunt, 378 S.E.2d 754, 757– 759 (N.C. 1989) (describing improper introduc-

tion of unsworn statements).
36. People v. Cruz, 643 N.E.2d 636, 644 (Ill. 1993).
37. The Tenth Circuit later affi rmed the grant of habeas, but on other grounds, and

without reviewing rulings relating to the in for mant testimony. Williamson v. Ward,
110 F.3d 1508, 1512 (10th Cir. 1997); Williamson v. Reynolds, 904 F. Supp. 1529,
1550– 1551 (E.D. Okla. 1995).

38. Trial Transcript, 1737, Jerry E. Watkins v. State of Indiana, 3CSCC- 87C8- CB- 764
(Ind. Sup. Ct. Aug. 1987).

39. Watkins v. Miller, 92 F. Supp. 2d 824, 834 (S.D. Ind. 2000).
40. Watkins Trial Transcript, 2172.
41. Anthony G. Amsterdam, “Verbatim: Lady Justice’s Blindfold Has Been Shredded,”

Champion, May 2007, 51.
42. In a groundbreaking article, law professor William Stuntz developed these argu-

ments about the perverse incentives that the structure of criminal procedure rules
create. See William J. Stuntz, “The Uneasy Relationship between Criminal Proce-
dure and Criminal Justice,” 107 Yale L.J. 1, 37– 45 (1997).

43. John Scalia, U.S. Department of Justice, Bureau of Justice Statistics, Prisoner Peti-
tions Filed in U.S. District Courts, 2000, with Trends, 1980– 2000, 1– 2.

Notes to Pages 188–196 � 331

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

44. Arizona v. Youngblood, 488 U.S. 51 (1988).
45. James Liebman, Jeff rey Fagan, and Valerie West, A Broken System: Error Rates in

Capital Cases, 1973– 1995 (2000), 5.
46. See Flango, Habeas Corpus, 62– 63; King et al., Habeas Litigation, 58.
47. For a more detailed discussion of this matched comparison group, see Garrett,

“Judging Innocence,” 154– 157. The claims that received reversals in the matched
comparison group mirrored the claims on which exonerees received relief: fi ve state
law evidentiary claims, four in eff ec tive assistance of counsel claims (one accompa-
nied by a prosecutorial misconduct claim), a Jackson claim, a right to counsel claim,
and a suggestive eyewitness identifi cation claim.

48. The winning claims, namely those for which a new trial was granted and that ruling
was upheld on appeal, were as follows: state evidentiary claims (6); Brady claims (5);
in eff ec tive assistance of counsel claims (4); claims concerning jury instructions (2);
Bruton unconstitutional joinder claims (2); prosecutorial misconduct claims (2); due
pro cess and right to counsel claims (1); ex post facto challenge to a state statute (1); and
fabrication of evidence (1).

49. The fourteen cases include the fi ve reversals relating to eyewitness identifi cations: a
state law claim regarding failure to provide a jury instruction explaining the dangers
of cross- racial misidentifi cation (M. Cromedy); a state evidentiary violation relating
to an eyewitness identifi cation (M. Webb); a Brady claim related to payment of a re-
ward to an eyewitness (M. Evans); the trial court’s decision to bar evidence that an-
other victim of similar attacks identifi ed another person (R. Cotton); and a Brady
claim regarding hypnotism of the victim in order to elicit an identifi cation (L. Jean).
The eight additional reversals related to trial evidence: another related to expert
evidence on a bite mark central to the case (R. Krone); prosecutorial misconduct for
misrepresenting hair and blood evidence (S. Linscott); a Brady violation related to
hair analysts (C. McCarty); in eff ec tive assistance of counsel relating to expert issues
regarding competence, a confession, and forensic testimony (R. Williamson); im-
proper admission of an unsworn statement by a cooperating witness (D. Hunt); a
fabrication claim regarding testimony of a cooperating codefendant (V. Jimerson);
and two appeals involving in eff ec tive ness of counsel including failure to move to
suppress central physical evidence such as hair evidence (W. Rainge and D. Wil-
liams). The last case, not discussed earlier in this chapter, involved a state law evi-
dence claim related to a dog scent identifi cation in the case of Wilton Dedge. See
Dedge v. State, 442 So. 2d 429 (Fla. App. 1983). For just the noncapital cases, that
fi gure is seven of eleven reversals.

50. Two reversals were granted for Brady claims that alleged the state concealed police
reports relating to third- party guilt, in the cases of Kirk Bloodsworth and Jerry
Watkins. See Bloodsworth v. State, 512 A.2d 1056 (Md. 1986); Watkins v. Miller,
92 F. Supp. 2d 824, 834 (S.D. Ind. 2000). The third such reversal occurred after the
trial court in Rolando Cruz’s case barred evidence of a third party’s pattern of simi-
lar crimes and confessions, although the reversal was also on the basis that prosecu-
tors improperly impeached a witness. See People v. Cruz, 643 N.E.2d 636, 653– 661
(Ill. 1994).

332 Notes to Pages 196–199

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

51. Johnny Lindsey won a reversal due to a conviction on an amended criminal statute
that was not yet in eff ect at the time of his conviction, Lindsey v. State, 672 S.W.2d
892 (Tex. App. 1984). Donald Wayne Good’s conviction was reversed because the
prosecutor improperly referred to his demeanor when he testifi ed. Good v. State,
723 S.W.2d 734 (Tex. Ct. App. 1986). Paula Gray’s conviction was reversed due to
her defense lawyer’s confl icts in jointly representing codefendants. Gray v. Director,
Dept. of Corrections, State of Illinois, 721 F.2d 586, 598 (7th Cir. 1983). In addition
to the reversal of his second conviction for failure to allow him to present evidence of
third- party guilt and improper prosecutorial questioning, Rolando Cruz’s fi rst trial
was reversed for prejudicial joinder of his case with codefendants. People v. Cruz,
521 N.E.2d 18, 25 (Ill. 1988). His codefendant, exoneree Alejandro Hernandez, had
a reversal for the same reason. People v. Hernandez, 521 N.E.2d 25, 37– 38 (Ill. 1988).

52. See Williamson v. Ward, 110 F.3d 1508, 1520 (10th Cir. 1997).
53. State v. Cotton, 351 S.E.2d 277, 280 (N.C. 1987).
54. See Brandon L. Garrett and J. J. Prescott, “Determinants of Success in Post-

conviction Litigation by the Innocent,” (2011) (draft on fi le with author).
55. See, e.g., King et al., “Final Technical Report,” 89 (concluding that “it appeared that

the presence of an innocence claim operated somehow to make a grant of relief on a
diff erent claim more likely”); see also Thomas Y. Davies, “Affi rmed: A Study of Crim-
inal Appeals and Decision- Making Norms in a California Court of Appeal,” 1982 Am.
B. Found. Res. J. 543, 625– 628; David T. Wasserman, A Sword for the Convicted:
Representing Indigent Defendants on Appeal (Westport, CT: Greenwood Press, 1990),
138– 148.

56. See Chapman v. California, 386 U.S. 18, 24, 26 (1966). The Brecht v. Abramson test,
see 507 U.S. 619, 639 (1993), which requires that the State show that error did not
substantially infl uence the jury, applies during federal habeas corpus review; but
with fewer exonerees pursuing habeas petitions and only a handful pursuing them
after 1993 when Brecht was decided, that more stringent test was never cited in these
cases.

57. Properly applied, harmless error analysis should ask only whether the State can dem-
onstrate that error did not suffi ciently aff ect the outcome at trial and not, conversely,
whether evidence of guilt outweighed the impact of any error. See Sullivan v. Louisi-
ana, 508 U.S. 275, 279 (1993) (“The inquiry . . . is . . . whether the guilty verdict ac-
tually rendered in this trial was surely unattributable to the error. That must be so,
because to hypothesize a guilty verdict that was never in fact rendered— no matter
how inescapable the fi ndings to support that verdict might be— would violate the
jury- trial guarantee.”).

58. Those cases are: Dennis Brown, State v. Brown, No. L-82- 297, 1983 WL 6945, at
*14 (Ohio Ct. App. Sept. 16, 1983); Ronnie Bullock, People v. Bullock, 507 N.E.2d
44, 49 (Ill. App. Ct. 1987); Frederick Daye, People v. Daye, 223 Cal. Rptr. 569, 580
(Cal. Ct. App. 1986); Jeff rey Deskovic, People v. Deskovic, 607 N.Y.S.2d 957, 958
(N.Y. App. Div. 1994) (“There was overwhelming evidence of the defendant’s guilt
in the form of the defendant’s own multiple inculpatory statements, as corroborated
by such physical evidence as the victim’s autopsy fi ndings”); Bruce Godschalk,

Notes to Pages 199–202 � 333

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Godschalk v. Montgomery County Dist. Attorney’s Offi ce, 177 F. Supp. 2d 366, 367,
369 (E.D. Pa. 2001) (quoting criminal trial court); Hector Gonzalez, State v. Gonza-
lez, 696 N.Y.S.2d 696, 697 (N.Y. App. Div. 1999); Larry Holdren, Holdren v. Le-
gursky, 16 F.3d 57, 63 (4th Cir. 1994); D. Hunt, State v. Hunt, 457 S.E.2d 276, 293
(N.C. Ct. App. 1994); Leonard McSherry, People v. McSherry, 14 Cal. Rptr. 2d 630,
636 (Cal. Ct. App. 1992) (referring to “the unusual circumstances in this case, over-
whelmingly identifying appellant as the perpetrator”) (depublished); Alan Newton,
Newton v. Coombe, No. 95- 9437, 2001 WL 799846, at *6 (S.D.N.Y. July 13, 2001)
(noting evidence of guilt “extremely strong”); David Shawn Pope, Pope v. State, 756
S.W.2d 401, 403 (Tex. App. 1988); Anthony Robinson, Robinson v. State, No. C14-
87- 00345- CR, 1989 WL 102335, at *7, *10 (Tex. App. Sept. 7, 1989); Yusef Salaam,
People v. Salaam, 590 N.Y.S.2d 195, 196 (N.Y. App. Div. 1992).

59. People v. Deskovic, 607 N.Y.S.2d 696, 697 (N.Y. App. Div. 1994).
60. No written decision mentioned that an exoneree raised Schlup, the “innocence gate-

way” that excuses procedural defaults of constitutional claims on the basis of newly
discovered evidence. Schlup v. Delo, 513 U.S. 298, 326– 327 (1995).

61. See Nicholas Berg, “Turning a Blind Eye to Innocence: The Legacy of Herrera v.
Collins,” 42 Am. Crim. L. Rev. 121, 135– 137 (2005) (surveying more than 170 cases in
which actual innocence claims were asserted and concluding that no court has
granted relief solely on basis of such claims).

62. See Herrera v. Collins, 506 U.S. 390, 417 (1993) (assuming arguendo that persuasive
demonstration of actual innocence would render an execution unconstitutional, but
stating that if such a claim existed, the threshold would be “extraordinarily high”).

63. See, e.g., Kyles v. Whitley, 514 U.S. 419, 435 (1995); see also Scott E. Sundby,
“Fallen Superheroes and Constitutional Mirages: The Tale of Brady v. Mary land,”
33 McGeorge L. Rev. 643, 659 (2002) (describing Brady as a postconviction “due
pro cess safety check”).

64. See Hunt v. McDade, 205 F.3d 1333 (4th Cir. 2000).
65. Editorial, “Badly Botched Case Wrongful Convictions Undermine Faith in Justice

System,” Charlotte Observer (N.C.), December 29, 2003, 12A; Associated Press,
“DNA Tests Exonerated Man Who Spent 18 Years in Prison for Murder of a Woman
in 1984,” St. Louis Post- Dispatch, February 7, 2004, 21.

66. Seven exonerees proff ered third- party guilt evidence, seven presented police reports
suppressed at the time of trial, four presented recantations of key witnesses, two pre-
sented new alibi evidence, two presented new evidence undercutting in for mant testi-
mony, one presented evidence of police hypnosis of the victim, and one presented
new forensic expert evidence (some presented more than one type).

67. See Jackson v. Day, No. CIV.A.95- 1224, 1996 WL 225021 *1 (E.D.La May 2, 1996).
68. See Answer in Opposition to Respondent’s Motion to Dismiss, Eddie Joe Lloyd v.

Grayson, No. 88CV- 73351- DT (E.D. Mich) (on fi le with author).
69. See 443 U.S. 307, 324 (1979) (holding that habeas relief is available if petitioner

shows that no rational trier of fact “could have found proof of guilt beyond a reason-
able doubt” based on evidence presented at trial).

70. See, e.g., 28 U.S.C. § 2244(d) and § 2254(d)–(e).

334 Notes to Pages 202–205

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

71. See Flango, Habeas Corpus, 46– 47.
72. Strickland v. Washington, 466 U.S. 668, 689– 690, 693– 694 (1988).
73. Randall Coyne, Capital Punishment and the Judicial Pro cess (Durham, NC: Caro-

lina Academic Press, 1994), 148; see also Stephen B. Bright, “Counsel for the Poor:
The Death Sentence Not for the Worst Crime but for the Worst Lawyer,” 103 Yale
L.J. 1835, 1844 (1994); Robert R. Rigg, “The Constitution, Compensation, and
Competence: A Case Study,” 27 Am. J. Crim. Law 1, 7– 9 (1999).

74. See also Bright, “Counsel for the Poor”; Rigg, “The Constitution.” John Jeff ries and
Bill Stuntz have proposed that Strickland claims could be a vehicle for judges to more
broadly grant relief on innocence- related grounds. See John C. Jeff ries Jr. and Wil-
liam J. Stuntz, “In eff ec tive Assistance and Procedural Default in Federal Habeas
Corpus,” 57 U. Chi. L. Rev. 679, 691 (1990).

75. Williamson v. Ward, 110 F.3d 1508, 1522 (10th Cir. 1997).
76. People v. Williams, 444 N.E.2d 136, 138, 143 (Ill. 1982), U.S. ex. rel. Gray v. Dir.,

Dept of Corr., 721 F.2d 586, 597 (7th Cir. 1983).
77. See, e.g., Washington v. Murray, 4 F.3d 1285, 1288– 1292 (4th Cir. 1993).
78. Brandon L. Garrett, “Judging Innocence,” 108 Colum. L. Rev. 55, 115 (2008).
79. Berger v. United States, 295 U.S. 78, 88 (1935).
80. Mooney v. Holohan, 294 U.S. 103, 112 (1935).
81. Batson v. Kentucky, 476 U.S. 79 (1986).
82. Brewer v. State, 725 So.2d 106, 123– 124 (Miss. 1998).
83. State v. Krone, 897 P.2d 621, 624– 625 (Ariz. 1995); People v. Cruz, 643 N.E.2d 636,

653– 661 (Ill. 1994). As described, seventy- seven exonerees brought claims related to
prosecutorial misconduct of one kind or another. Forty- two exonerees brought claims
related to prosecution arguments, twenty- nine brought Brady claims, four brought
fabrication claims, and twenty- one brought claims related to jury instructions. Four
exonerees brought claims that made generalized claims of prosecutorial misconduct
that the court did not specify. Three more cases involved claims of improper question-
ing of witnesses, one case involved a discovery violation, and a fi nal case involved a
prosecution motion. Some cases involved more than one type.

84. See, e.g., Arizona v. Youngblood, 488 U.S. 51, 72 (1988) (Blackmun, J., dissenting).
85. State v. Jean, 311 S.E.2d 266, 274 (N.C. 1984) (Exum, J., dissenting).
86. State v. Goodman, 763 P.2d 786, 789– 790 (Utah 1988) (Stewart, J., dissenting).
87. See, e.g., People v. Cruz, 643 N.E.2d 636, 688 (Ill. 1994) (Heiple, J., dissenting).
88. Herrera, 506 U.S. 390, 403– 404 (1993).

8 . exoneration

1. See Frontline, “Requiem for Frank Lee Smith,” at www .pbs .org/ wgbh/ pages/ front
line/ shows/ smith/ eight/ .

2. Smith v. Dugger, 565 So. 2d 1293, 1296 (Fla. 1990); Sydney Freedberg, “DNA Clears
Inmate Too Late,” St. Petersburg Times, December 15, 2000, 1A.

3. Freedberg, “DNA Clears Inmate.”

Notes to Pages 205–214 � 335

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

4. Ibid.
5. Barry Scheck, “Did Frank Lee Smith Die in Vain?” www .pbs .org/ wgbh/ pages/

frontline/ shows/ smith/ ofra/ scheck .html .
6. See Fla. Stat. Ann. § 925.11(f)(3) (West 2010).
7. Bain v. State, 9 So. 3d 723 (Fla. Dist. Ct. App. 2009); Mitch Stacy, “Fla. Man Exon-

erated After 35 Years Behind Bars,” Associated Press, December 18, 2009.
8. They spent a total of more than 3,250 years behind bars. See Innocence Project,

“250 Exonerated: Too Many Wrongly Convicted,” 2– 3, at www .innocenceproject
.org/ news/ 250 .php .

9. See Cathy Maston and Patsy Klaus, U.S. Department of Justice, Criminal Victim-
ization in the United States, 2005, (2006), Statistical Tables, table 34(b) (fi nding that
31.4% of rape and sexual assault cases involved stranger- perpetrators).

10. See, e.g., Ron C. Michaelis et al., A Litigators Guide to DNA (Boston: Academic
Press/Elsevier, 2008), 45.

11. Gov. Wilder’s Executive Clemency Off er, Jan. 14, 1994, at www .pbs .org/ wgbh/
pages/ frontline/ shows/ case/ cases/ washingtonclem .html .

12. See John M. Butler, Forensic DNA Typing: Biology & Technology Behind STR Mark-
ers, 2d ed. (Boston: Academic Press/Elsevier, 2005), 146.

13. Ibid., 33– 35.
14. Ibid., 201– 298.
15. Dinitia Smith and Nicholas Wade, “DNA Test Finds Evidence of Jeff erson’s Child

by Slave,” New York Times, November 1, 1998.
16. See Federal Bureau of Investigation, Department of Justice, CODIS: Combined DNA

Index System 2 (2007), at www .fbi .gov/ hq/ lab/ pdf/ codisbrochure2 .
17. See DNA Fingerprint Act of 2005, 42 U.S.C. § 14132(a)(1) (2005) (permitting ar-

restee profi les to be entered); Michelle Hibbert, “DNA Databanks: Law Enforce-
ment’s Greatest Surveillance Tool?” 34 Wake Forest L. Rev. 767 (1999).

18. The three states are Louisiana, Texas, and Virginia, See La. Rev. Stat. Ann. § 15.609;
Tex. Gov’t Code Ann. § 411.1471; Va. Code Ann. § 19.2- 310.2:1. See also Julia Pres-
ton, “U.S. Set to Begin a Vast Expansion of DNA Sampling,” New York Times, Feb-
ruary 5, 2007 (“Federal Bureau of Investigation offi cials said they anticipated an in-
crease ranging from 250,000 to as many as 1 million samples a year. The laboratory
currently receives about 96,000 samples a year.”).

19. Jeff rey Rosen, “Ge ne tic Surveillance for All,” Slate, March 17, 2009.
20. Sixty- fi ve of the cases involved “cold hits” in a DNA database, while forty- seven in-

volved nondatabase DNA testing that confi rmed another’s guilt.
21. James Dao, “Lab’s Errors Force Review of 150 DNA Cases,” New York Times, May 7,

2005, A1.
22. Margaret Edds, An Expendable Man: The Near- Execution of Earl Washington, Jr.,

(New York: New York University Press, 2003), 244– 250.
23. Brandon L. Garrett, “Claiming Innocence,” 92 Minn. L. Rev. 1629, 1655– 1670

(2008).
24. Henry J. Friendly, “Is Innocence Irrelevant? Collateral Attack on Criminal Judg-

ments,” 38 U. Chi. L. Rev. 142, 159– 160 (1970).

336 Notes to Pages 214–223

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

25. 506 U.S. 390, 417 (1993).
26. House v. Bell, 547 U.S. 518, 540– 548 (2006).
27. Ibid., 571 (Robert, C.J., dissenting).
28. David G. Savage, “Murder Charges Dropped Because of DNA Evidence,” Los Ange-

les Times, May 13, 2009.
29. The Court has ruled in other cases in which the convict later proved innocence and

was exonerated through postconviction DNA testing. All were summary denials of
cert. except one— the case of Arizona v. Youngblood, discussed in the next section.

30. 129 S. Ct. 2308, 2316 (2009).
31. Osborne v. District Attorney’s Offi ce, Reply to Brief in Opposition 8.
32. 129 S.Ct. 2308, 2319.
33. Ibid.
34. In re Davis, 130 S. Ct. 1 (2009).
35. Skinner v. Switzer, No. 09- 9000, 2010 WL 545500 (May 24, 2010).
36. Tim Arango, “Death Row Foes See Newsroom Cuts as a Blow,” New York Times,

May 20, 2009.
37. See Alaska Stat. § 12.73.010– 12.73.090 (2010).
38. Gerry Smith, “Rape Conviction Gone, Stigma Isn’t,” Chicago Tribune, October 22,

2007.
39. Jim Dwyer, “New York Fails at Finding Evidence to Help the Wrongly Convicted,”

New York Times, July 6, 2006.
40. Arizona v. Youngblood, 488 U.S. 51, 57– 59 (1988); see also Brandon L. Garrett,

“Judging Innocence,” 108 Colum. L. Rev. 55, 95, 117 (2008).
41. See The Innocence Project, Preservation of Evidence Fact Sheet (2009), at http://

www .innocenceproject .org/ Content/ 253 .php .
42. Commonwealth v. Godschalk, 679 A.2d 1295 (Pa. Super. 1996).
43. State v. Halsey, 748 A.2d 634, 636 (N.J. Super. Ct. 2000).
44. Waller v. State, Nos. 05- 02- 00117- CR, 05- 02- 00118- CR, 05- 02- 00119- CR,

05- 02- 00120- CR, 2003 WL 22456324, at *2 (Tex. Crim. App. October 30, 2003); Jen-
nifer Emily, “Exonerated by DNA, Patrick Waller Is Released from Prison,” Dallas
News, July 3, 2008.

45. Dedge v. State, 732 So.2d 322, 323 (Fla. App. 5 Dist. 1998) (W. Sharp., J. dissent-
ing); Laurin Sellers, “DNA Testing Frees Man Accused in 1981 Rape,” Orlando
Sentinel, August 12, 2004.

46. See N.Y. Crim. Proc. Law § 440.30(1- a) (McKinney Supp. 1999); 725 Ill. Comp. Stat.
5/116- 3(a) (West Supp. 1998).

47. For a detailed analysis of these statutes, see Garrett, “Claiming Innocence,”
1673– 1682.

48. See, e.g., People v. McSherry, 14 Cal. Rptr. 2d 630 (Cal. Ct. App. 1992) (failing to
vacate conviction despite DNA exclusion, citing to serology results and the victim’s
identifi cation); Watkins v. State, No. 30A04- 9504- PC- 118, 1996 WL 42093 (Ind. Ct.
App. Jan. 29, 1996) (stating that DNA results exonerating Jerry Watkins only “sug-
gest the possibility” of another perpetrator and that the DNA evidence was merely
“cumulative” of inconclusive serology evidence at trial).

Notes to Pages 223–230 � 337

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

49. See Frontline, “The Case for Innocence,” http:// www .pbs .org/ wgbh/ pages/ front
line/ shows/ case/ interviews/ keller .html .

50. Ibid. at http:// www .pbs .org/ wgbh/ pages/ frontline/ shows/ case/ etc/ update .html .
51. Ralph Blumenthal, “12th Dallas Convict is Exonerated by DNA,” New York Times,

January 18, 2007.
52. Hearing Transcript, 3– 4, State of West Virginia v. Glen Dale Woodall, No. 87- F-46

(W. Va. Cir. Ct. May 4, 1992).
53. Lola Vollen and Dave Eggers, eds., Surviving Justice: America’s Wrongfully Con-

victed and Exonerated (San Francisco: McSweeney’s, 2005), 32.
54. Fernanda Santos, “With DNA from Exhumed Body, Man Finally Wins Freedom,”

New York Times, January 24, 2007.
55. Email, Elizabeth Webster, Communications Director, The Innocence Project,

11/9/2009 (on fi le with author); Strengthening Forensic Science in the United States:
Hearing Before the S. Comm. on the Judiciary, 111th Cong. (2009) (statement of Pe-
ter Neufeld, Co- Director, The Innocence Project).

56. See Barry C. Scheck, “Barry Scheck Lectures on Wrongful Convictions,” 54 Drake
L. Rev. 597, 601 (2006).

57. Fifteen of thirty- six with written decisions were capital cases. These death row in-
mates, though actually guilty, had a strong incentive to pursue every avenue for re-
view, regardless of whether their claims had merit. I discuss these “guilt confi rma-
tion” cases in more detail in an article. Garrett, “Judging Innocence,” 141.

58. For example, an opponent to the Virginia DNA statute cited “a strong possibility of
abuse.” Craig Timberg, “DNA Spurs Crime Panel to Debate Changes in Va.,”
Washington Post, December 2, 2000, B1.

59. U.S. Department of Justice, Cost Study of DNA Testing and Analysis Report, North
Carolina Offi ce of State Bud get and Management (2006), 7– 8, at http:// www .osbm
.state .nc .us/ fi les/ pdf _fi les/ 3 -1 -2006FinalDNAReport .

60. Gabrielle Fimbres, “Lab Work a Lot Cheaper Than Lockup,” Tucson Citizen
(Ariz.), August 23, 2000, 1A; see also Holly Shaff ter, “Postconviction DNA
Evidence: A 500 Pound Gorilla in State Courts,” 50 Drake L. Rev. 695, 735
(2002).

61. See Kaine’s Full Statement in Norfolk Four Case, at http:// voices .washingtonpost
.com/ v i rg in iapol it ics/ 2009/ 08/ k a ines _ f u l l _ statement _on _ nor fo .htm l
?sidST2009080602217. For an in- depth account and analysis of those confession
statements, see Tom Wells and Richard A. Leo, The Wrong Guys: Murder, False
Confessions, and the Norfolk Four (New York: W. W. Norton, 2008).

62. Edwin M. Borchard, Convicting the Innocent: Errors of Criminal Justice (New Ha-
ven, CT: Yale University Press, 1932), 275.

63. Janet Roberts and Elizabeth Stanton, “A Long Road Back after Exonerations, and
Justice Is Slow to Make Amends,” New York Times, November 25, 2007.

64. Chris William Sanchirico, “Character Evidence and the Object of Trial,” 101
Colum. L. Rev. 1227, 1271– 1272 (2001).

65. See Innocence Project, Making Up for Lost Time: What the Wrongfully Convicted
Endure and How to Provide Fair Compensation (2009).

338 Notes to Pages 230–235

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

66. Fran Spielman, “Former Death Row Inmate to Get $2.2 Million,” Chicago Sun-
Times, December 16, 2003, 22.

67. The case then settled for slightly less than that, avoiding an appeal. David Harper,
“McGee Case: City to Pay: Lawsuit Is Settled for $12.25 million,” Tulsa World
(Okla.), June 3, 2006.

68. Ibid.
69. For a review of these statutes, see Final Report of the New York State Bar Associa-

tion’s Task Force on Wrongful Convictions (2009), 138– 140; Adele Bernhard, “Jus-
tice Still Fails: A Review of Recent Eff orts to Compensate Individuals Who Have
Been Unjustly Convicted and Later Exonerated,” 52 Drake L. Rev. 703 (2004).

70. Jeff Carlton, “Texas DNA Exonerees Find Prosperity after Prison,” Associated
Press, September 4, 2009.

71. Hugo Kugiya, “Free of Death Row; Hard Road for 13 Former Inmates,” Newsday
(NY), May 19, 2002, A5.

72. Meg Laughlin and Daniel de Vise, “Former Convict Is Overwhelmed by His Free-
dom,” Miami Herald, July 7, 2001, 11A.

73. Adam Liptak, “DNA Will Let a Montana Man Put Prison Behind Him, but Ques-
tions Still Linger,” New York Times, October 1, 2002.

74. Mike Wagner and Geoff Dutton, “Columbus Man Finally Freed from Prison,” Co-
lumbus Dispatch, March 10, 2009.

75. Ibid.
76. Frontline, “The Burden of Innocence,” at www .pbs .org/ wgbh/ pages/ frontline/

shows/ burden/ etc/ script .html .
77. Michael Hall, “The Exonerated,” Texas Monthly, November 2008, 162.
78. Sharon Waxman, “For the Wrongly Convicted, New Trials Once the Cell Opens,”

New York Times, January 25, 2005.
79. Barbara Novovitch, “Free after 17 Years for a Rape That He Did Not Commit,” New

York Times, December 22, 2004, A22.
80. Todd Richmond, “Avery, County Settle for $400,000,” St. Paul Pioneer Press, Feb-

ruary 15, 2006.
81. That story is now the subject of a movie. See John Larrabee and Russ Olivo, “Reluc-

tant Hero,” Rhode Island Monthly, February 2010.
82. Mike Mather, “30- Year- Old DNA Clears Richmond Man of Rape,” at www .wtkr

.com/ news/ wtkr -mather -dna -story -mar05 ,0 ,7898871 .story .
83. James F. McCarty, “Wrongly Convicted, Now Free DNA Testing Clears Man Jailed

for 4 Years in Rape Case,” Cleveland Plain Dealer, October 7, 1994, 1B.
84. Jennifer Thompson- Cannino and Ronald Cotton, Picking Cotton: Our Memoir of

Justice and Redemption (New York: St. Martin’s Press, 2009).
85. Steve McGonigle, “Rape Victim Is for Exoneration,” Dallas Morning News, April 6,

2007.
86. Dateline NBC, “Profi le: Suspicion,” June 4, 2008.
87. Jenkins v. Scully, No. 91- CV- 298E, 1992 WL 205685, at *2 (W.D.N.Y. July 16, 1992);

Jenkins v. Scully, No. 91- CV- 298E(M), 1992 WL 205685, at *1 (W.D.N.Y. April 15,
1993).

Notes to Pages 235–240 � 339

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

88. Gene Warner, “What Price to Pay for Wrongful Convictions,” Buff alo News, April
12, 2008, A1.

9 . reforming the criminal justice system

1. See North Carolina Actual Innocence Commission, Mission Statement, Objectives,
and Procedures, at www .innocenceproject .org/ docs/ NC _Innocence _Commission
_Mission .html; Jack Betts, jackbetts.blogspot.com/2010/02/lake- innocence- process-
most- important.html.

2. Christine C. Mumma, “The North Carolina Actual Innocence Commission: Un-
known Perspectives Joined by a Common Cause,” 52 Drake L. Rev. 647 (2004). In
2007, the commission was renamed the North Carolina Criminal Justice Study
Commission.

3. N.C. Gen. Stat. § 15A- 284.50–53 (2002).
4. Ibid. at 52(b).
5. Ibid.
6. N.C. Gen. Stat. § 15A- 211 (2009); see also H.B. 33, 2009 Gen. Assem., 2009– 2010

Sess. (N.C. 2009).
7. See Criminal Cases Review Commission, at www .ccrc .gov .uk/ about .htm/ ; see also

Jerome M. Maiatico, “All Eyes on Us: A Comparative Critique of the North Carolina
Innocence Inquiry Commission,” 56 Duke L.J. 1345 (2007) (comparing North Car-
olina and U.K. commissions).

8. Canada Criminal Code, R.S.C., ch. C-46, § 696.1– 696.6 (1985).
9. “Gov. Easley Signs Innocence Inquiry Commission Bill,” U.S. State News, August

3, 2006.
10. www .innocencecommission -nc .gov/ inthenews .htm (“Since 2007, the Commission

has received over 300 applications and has accepted fi ve of those cases for
investigation.”).

11. Robbie Brown, “Judges Free Inmate on Recommendation of Special Panel,” New
York Times, February 17, 2010.

12. Radley Balko, “Schwarzenegger Vetoes Justice,” FOXNews.com, November 5, 2007.
13. See Innocence Commissions in the U.S., at www .innocenceproject .org/ Content/

415 .php; see also Cal. Comm’n on the Fair Admin. of Justice, Final Report (2008), at
www .ccfaj .org/ documents/ CCFAJFinalReport ; Innocence Comm’n for Va.,
A Vision for Justice: Report and Recommendations Regarding Wrongful Convictions
in the Commonwealth of Virginia (2005), at www .exonerate .org/ ICVA/ full _r ;
Commission on Capital Punishment, State of Illinois, Report of the Governor’s Com-
mission on Capital Punishment (2002), at www .idoc .state .il .us/ ccp/ ccp/ reports/
commission _report/ index .htm; Stanley Z. Fisher, “Convictions of Innocent Per-
sons in Massachusetts: An Overview,” 12 B.U. Pub. Int. L.J. 1 (2002); John T. Rago,
“A Fine Line between Chaos & Creation: Lessons on Innocence Reform from the
Pennsylvania Eight,” 12 Widener L. Rev. 359 (2006); North Carolina Innocence In-
quiry Commission, www .innocencecommission -nc .gov .

340 Notes to Pages 240–244

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

14. See Samuel R. Gross et al., “Exonerations in the United States 1989 Through
2003,” 95 J. Crim. L. & Criminology 523 (2005).

15. James S. Liebman, “The New Death Penalty Debate: What’s DNA Got to Do With
It?” 33 Colum. Hum. Rts. L. Rev. 527 (2002).

16. Adam Liptak, “U.S. Prison Population Dwarfs that of Other Nations,” New York
Times, April 23, 2008.

17. Barry C. Scheck and Peter J. Neufeld, “Toward the Formation of Innocence Com-
missions,” 86 Judicature 98, 98 (2002).

18. See Brian A. Reeves, Bureau of Justice Statistics, Census of State and Local Law
Enforcement Agencies, 2004 (2007), 1.

19. Jeremy W. Peters, “Wrongful Conviction Prompts Detroit Police to Videotape Cer-
tain Interrogations,” New York Times, April 11, 2006.

20. See Jodi Wilgoren, “Confession Had His Signature; DNA Did Not,” New York
Times, August 26, 2003.

21. Trial Transcript, 40– 41, Michigan v. Eddie Joe Lloyd, 85- 00376 (Mich. Rec. Ct.
May 2, 1985).

22. See Peters, “Wrongful Conviction.”
23. See D.C. Code Ann. § 5- 116.01 (2007) (requiring police to record all custodial in-

vestigations); 725 Ill. Comp. Stat. Ann. 5/103- 2.1 (2006) (requiring police to record
interrogations in all hom i cide cases); Md. Ann. Code, Crim. Proc. § 2- 401 (2008)
(requiring that law enforcement make “reasonable eff orts” to record interroga-
tions); Me. Rev. Stat. Ann. tit. 25, § 2803- B (2007) (mandating recording “inter-
views of suspects in serious crimes”); Mont. Code Ann. § 46- 4- 406–46- 4- 411 (2010)
(requiring recording of interrogations in all felony cases); N.C. Gen. Stat. § 15A- 211
(2010) (requiring complete electronic recording of custodial interrogations in hom-
i cide cases); Neb. Rev. Stat. § 29- 4501- 4508 (2009), eff ective July 18, 2008 (requir-
ing electronic recording of interrogations in several types of felony cases); N.M.
Stat. § 29- 1- 16 (Supp. 2006) (requiring police to record all custodial investiga-
tions); Ohio Rev. Code Ann. § 2933.81 (2010); Tex. Code Crim. Proc. Ann. art.
38.22, § 3 (Vernon Supp. 2007) (rendering unrecorded oral statements inadmissi-
ble); Wis. Stat. Ann. §§ 968.073, 972.115 (requiring recording of felony interroga-
tions and permitting jury instruction if interrogation not recorded); Ind. R. Evid.
617 (requiring that in order to be admissible, entire interrogations in felony crimi-
nal prosecutions must be recorded); Stephan v. State, 711 P.2d 1156, 1158 (Alaska
1985) (“[A]n unexcused failure to electronically record a custodial interrogation
conducted in a place of detention violates a suspect’s right to due pro cess . . .”);
Commonwealth v. DiGiambattista, 813 N.E.2d 516, 535 (Mass. 2004) (allowing
defense to point out failure to record interrogation and calling unrecorded admis-
sions “less reliable”); State v. Scales, 518 N.W.2d 587, 592 (Minn. 1994) (“[A]ll
questioning shall be electronically recorded where feasible and must be recorded
when questioning occurs at a place of detention.”); State v. Barnett, 147 N.H. 334
(2001); State v. Cook, 847 A.2d 530, 547 (N.J. 2004) (“[W]e will establish a com-
mittee to study and make recommendations on the use of electronic recordation of
custodial interrogations.”); In re Jerrell C.J., 699 N.W.2d 110, 123 (Wis. 2005)

Notes to Pages 244–248 � 341

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

(“[W]e exercise our supervisory power to require that all custodial interrogation
of juveniles in future cases be electronically recorded where feasible, and without
exception when questioning occurs at a place of detention.”); State v. Hajtic, 724
N.W.2d 449, 456 (Iowa 2006) (“electronic recording, particularly videotaping, of
custodial interrogations should be encouraged, and we take this opportunity to do
so.”).

24. A recent survey of 631 police investigators found that 81% believed that interroga-
tions should be recorded. See Saul Kassin et al., “Police Interviewing and Interro-
gation: A Self- Report Survey of Police Practices and Beliefs,” 31 Law & Hum. Be-
hav. 381 (2007).

25. Transcript of Motion to Suppress Hearing, 72– 73, United States v. Bland, No. 1:02-
CR- 93 (N.D. Ind. Dec. 12, 2002).

26. Stoker v. State, 692 N.E.2d 1386, 1390 (Ind. Ct. App. 1998).
27. See Peters, “Wrongful Conviction.”
28. See Brandon L. Garrett, “The Substance of False Confessions,” 62 Stan. L. Rev.

1051, 1109– 1118 (2010).
29. Darryl Fears, “Exonerations Have Changed Justice System,” Washington Post, May

3, 2007.
30. Trial Transcript, 182, State of New Jersey v. McKinley Cromedy, Ind. No. 1243- 07-

93 (N.J. Super. Ct. July 27, 1994).
31. Ibid., 104.
32. Ibid., 142.
33. Ibid., 164, 168.
34. State v. Cromedy, 727 A.2d 457 (N.J. 1999).
35. Ibid.
36. Tom Avril, “Eyewitness’ Blind Spot,” Philadelphia Inquirer, May 22, 2006.
37. Offi ce of the Attorney General, N.J. Department of Law and Public Safety, Attor-

ney General Guidelines for Preparing and Conducting Photo and Live Lineup Iden-
tifi cation Procedures (Apr. 18, 2001). The attorney general referred to “recent cases,
in which DNA evidence has been utilized to exonerate individuals convicted al-
most exclusively on the basis of eyewitness identifi cations.” Letter from New Jersey
Attorney General John J. Farmer, Jr., to All County Prosecutors et al. (Apr. 18,
2001), 1.

38. State v. Delgado, 188 N.J. 48, 62– 63 (N.J. 2006) (citing Court Rule 3:17).
39. State v. Romero, 191 N.J. 59, 76 (2007).
40. State v. Henderson, 2009 WL 510409 (N.J. Feb. 26, 2009).
41. Report of the Special Master, State of New Jersey v. Karry R. Henderson, No. A-8

(2010).
42. Bloodsworth v. State, 512 A.2d 1056, 1062 (Md. 1986).
43. John Monahan and Laurens Walker, Social Science in Law, 6th ed. (New York:

Foundation Press, 2006), 567– 605; see, e.g., Utah v. Dean Lomax Clopton, 2009
UT 84 (“Over the last two de cades, numerous other state courts have either re-
versed decisions to exclude or encouraged the inclusion of eyewitness expert
testimony”).

342 Notes to Pages 248–251

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

44. See 725 Ill. Comp. Stat. 5/107A- 5 (2003); Md. Code Ann., Pub. Safety § 3- 506
(2007); N.C. Gen. Stat. § 15A- 284.52 (2007); Ohio Rev. Code Ann. § 2933.83 (2010);
W. Va. Code Ann. § 62- 1E- 1 (2010); Wis. Stat. § 175.50 (2005).

45. See, e.g. Va. Code Ann. § 19.2- 390.00- 02 (2005).
46. Jennifer Emily, “New Dallas Policy Regulates, Limits Showups,” Dallas Morning

News, December 10, 2008; Jennifer Emily, “Dallas Police Drop Study, Plan Photo-
Lineup Changes,” Dallas Morning News, January 16, 2009; The Justice Project,
Eyewitness Identifi cation: A Policy Review (2006), http:// www .thejusticeproject .
org/ wp -content/ uploads/ polpack _eyewitnessid -fi n21 .

47. Trial Transcript, 508, 524– 529, State of West Virginia v. Glen Dale Woodall, Ind.
No. 87- P-46 (W. Va. Cir. Ct. July 2, 1987).

48. In re an Investigation of The W. Va. State Police Crime Lab., Serology Div., 438
S.E.2d 501, 502– 503 (W. Va. 1993).

49. See ASCLD/LAB, West Virginia State Police Crime Laboratory, Serology Divi-
sion, South Charleston, West Virginia, ASCLD/LAB Investigation Report (July 23,
1993) at http:// www .law .virginia .edu/ pdf/ faculty/ garrett/ innocence/ wva _cvrime _
lab .

50. In re an Investigation of The W. Va. State Police Crime Lab., Serology Div., 438
S.E.2d at 506; see Kit R. Roane and Dan Morrison, “The CSI Eff ect,” U.S. News &
World Report, April 25, 2005, 48.

51. In re Renewed Investigation of State Police Crime Laboratory, Serology Div., 633
S.E.2d 762, 766 (W. Va. Jun. 16, 2006).

52. See Michael R. Bromwich, Fifth Report of the In de pen dent Investigator for the Hous-
ton Police Department Crime Laboratory and Property Room, Executive Summary
(2006), 2; Roma Khanna and Steve McVicker, “Police Lab Tailored Tests to Theo-
ries, Report Says,” Houston Chronicle, May 12, 2006.

53. Editorial, “Crime Lab Legacy,” Houston Chronicle, August 12, 2009.
54. See, e.g., Minn. Stat. §299C.156 (2007) (establishing Forensic Laboratory Advi-

sory Board); N.Y. Exec. Law §§995- a to – b (McKinney 2003) (establishing forensic
science commission and requiring accreditation); Okla. Stat. Ann. tit. 74, §150.37
(2007) (requiring accreditation); Tex. Code Crim. Proc. Ann. art. 38.35(d) (Vernon
2005) (requiring accreditation by the Texas Department of Public Safety); Va.
Code Ann. §9.1- 1101 (2006) (creating Department of Forensic Science and over-
sight committee).

55. See ASCLD/LAB Home Page, www .ascld -lab .org/ .
56. Mandy Locke and Joseph Neff , “Inspectors Missed All SBI Faults,” News Observer

(Charlotte), August 26, 2010.
57. For example, in one positive step, the International Association for Identifi cation

(IAI) adopted new guidelines concerning fi ngerprint testimony, to more clearly ac-
knowledge the limitations of the method. See IAI Resolution 2010- 18.

58. See Dodd v. State, 993 P.2d 778, 784 (Okla. Crim. App. 2000) (adopting procedure
for jail house in for mant testimony that ensures “complete disclosure”); Cal. Penal
Code § 1127a(b) (West 2004) (requiring courts to instruct jury on in- custody in for-
mant testimony); United States v. Villafranca, 260 F.3d 374, 381 (5th Cir. 2001)

Notes to Pages 252–256 � 343

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

(“The testimony of a plea- bargaining defendant is admissible if the jury is properly
instructed.”); State v. Bledsoe, 39 P.3d 38, 44 (Kan. 2002) (noting that trial court
“gave a cautionary jury instruction regarding the testimony of an in for mant”); Alex-
andra Natapoff , “Beyond Unreliable: How Snitches Contribute to Wrongful Con-
victions,” 37 Golden Gate U. L. Rev. 107, 112– 115 (2006) (proposing model statute
requiring pretrial evaluations of in for mant testimony).

59. See 725 Ill. Comp. Stat. Ann. 5/115- 21(d) (West Supp. 2007); Dodd v. State, 993 P.2d
at 785 (suggesting in concurring opinion that reliability hearings be conducted);
Myers v. State, 133 P.3d 312, 321 (Okla. Crim. App. 2006) (describing post-Dodd
case in which reliability hearing was conducted, though not required by Dodd);
D’Agostino v. State, 823 P.2d 283 (Nev. 1992) (holding that before admitting testi-
mony of a jail house in for mant in a capital sentencing- phase hearing, the judge
should fi rst review its reliability).

60. See Province of Ontario, Ministry of Attorney General, In- Custody Informers,
Crown Policy Manual (March 21, 2005), R. v. Brooks, [2000] 1 S.C.R. 237 Can.;
Kent Roach, “Unreliable Evidence and Wrongful Convictions: The Case for Ex-
cluding Tainted Identifi cation Evidence and Jail house and Coerced Confessions,”
52 Crim. L. Q. 210 (2007); Hon. Fred Kaufman, Report of the Kaufman Commission
on Proceedings Involving Guy Paul Morin (1998).

61. Report of the 1989– 1990 Los Angeles County Grand Jury, Investigation of the In-
volvement of Jail house In for mants in the Criminal Justice System in Los Angeles
County (1989– 1990), 149; Henry Weinstein, “Use of Jail house In for mants Is Un-
even in State,” Los Angeles Times, September 21, 2006, 3; Staff , “Man Who Showed
How to Fake Confessions Is Indicted for Perjury,” Los Angeles Times, March 4,
1992, 1.

62. The eleven cases in which the DNA testing confi rmed the guilt of another are those
of Kirk Bloodsworth, Kennedy Brewer, Rolando Cruz, Alejandro Hernandez,
Verneal Jimerson, Ray Krone, Robert Miller, Frank Lee Smith, Earl Washington,
Dennis Williams, and Ron Williamson.

63. 553 U.S. 35, 85– 86 (2008) (Stevens, J., concurring).
64. United States v. Quinones, 196 F. Supp. 2d. 416, 420 (S.D.N.Y. 2002).
65. United States v. Quinones, 313 F.3d 49 (2d Cir. 2002).
66. See Death Penalty Information Center, Innocence and the Death Penalty, at www

.deathpenaltyinfo .org/ innocence -and -death -penalty (listing 138 death row exoner-
ees since 1973). For excellent discussions of the role that these DNA exonerations
have played in altering the death penalty debate, see Liebman, “The New Death
Penalty Debate,” and Colin Starger, “Death and Harmless Error: A Rhetorical Re-
sponse to Judging Innocence,” Colum. L. Rev. Sidebar 1, February 23, 2008.

67. See Bureau of Justice Statistics, U.S. Department of Justice, Sourcebook of Criminal
Justice Statistics 2003, 147, table 2.56; see also Frank R. Baumgartner, Suzanna L.
De Boef, and Amber E. Boydstun, The Decline of the Death Penalty and the Discov-
ery of Innocence (New York: Cambridge University Press, 2008).

68. Kansas v. March, 548 U.S. 163, 189 (2006) (Scalia, J. concurring).
69. Md. Code Ann., Crim. Law § 2- 202 (2010).

344 Notes to Pages 256–258

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

70. Massachusetts Governor’s Council on Capital Punishment, Final Report (2004), at
http:// www .lawlib .state .ma .us/ docs/ 5 -3 -04Governorsreportcapitalpunishment .
pdf .

71. Robert H. Jackson, “The Federal Prosecutor,” 24 J. Am. Judicature Soc’y 18, 18
(1940).

72. Josh Bowers, “Punishing the Innocent,” 156 U. Penn. L. Rev. 1117 (2008).
73. Editorial, “Craig Watkins Is the 2008 Texan of the Year,” Dallas News, December

28, 2008.
74. Alex Holmquist, “Ramsey County: Better ID Methods Sought for Witnesses New

Techniques to Be Focus of St. Paul Conference,” St. Paul Pioneer Press (MN), Octo-
ber 23, 2009, B1.

75. Peg Lautenschlager, Wisconsin Attorney General, “Eyewitness Identifi cation Best
Practices” (June 15, 2005); John J. Farmer, Jr., Attorney General of the State of New
Jersey, “Letter to All County Prosecutors: Attorney General Guidelines for Prepar-
ing and Conducting Photo and Live Lineup Identifi cation Procedures” (April 18,
2001); Offi ce of the Utah Attorney General, Best Practices Statement for Law En-
forcement: Recommendations for Recording of Custodial Interviews (October 2008).

76. Jonathan Saltzman, “Hom i cide Conviction Rate in ’09 Up Sharply,” Boston Globe,
December 28, 2009.

77. See, e.g., Report of the Task Force on Eyewitness Evidence, Suff olk County District
Attorney’s Offi ce (July 2004), at www .innocenceproject .org/ docs/ Suff olk _eyewit
ness .

78. See, e.g., American Bar Association Standing Committee on Legal Aid and Indigent
Defendants, Gideon’s Broken Promise: America’s Continuing Quest for Equal Justice
(2004); The Spangenberg Group, State and County Expenditures for Indigent De-
fense Ser vices in Fiscal Year 2002 (2003), 34– 37; Note, “Gideon’s Promise Unful-
fi lled: The Need for Litigated Reform of Indigent Defense,” 113 Harv. L. Rev. 2062,
2065 (2000); Darryl K. Brown, “Rationing Criminal Defense Entitlements: An Argu-
ment from Institutional Design,” 104 Colum. L. Rev. 801, 807– 808 and n.28, 815
(2004).

79. See Oversight of the Department of Justice’s Forensic Grant Programs: Hearing Be-
fore the S. Comm. on the Judiciary, 110th Cong. (2008) (statement of Glenn A. Fine,
Inspector General, U.S. Department of Justice); Oversight of the Justice For All Act:
Hearing Before the S. Comm. on the Judiciary, 110th Cong. (2008) (statement of Peter
Neufeld on Behalf of the Innocence Project).

80. Letter from Assistant Attorney General William Moschella to U.S. Senator Orrin
Hatch, April 28, 2004, H.R. Rep. No. 108- 711 at 133– 156 (2004).

81. See Department of Justice, Technical Working Group for Eyewitness Evidence,
Eyewitness Identifi cations: A Guide for Law Enforcement (1999), at www .ncjrs .gov/
pdffi les1/ nij/ 178240 ; National Institute of Justice, U.S. Department of Justice,
National Commission on the Future of DNA Evidence, www .ojp .usdoj .gov/ nij/ top
ics/ forensics/ evidence/ dna/ commission/ welcome .html .

82. Kansas v. Marsh, 548 U.S. 163, 194– 195 (2006) (Scalia, J., concurring). Much of the
thrust of his opinion, however, concerns a separate topic, the degree of error in our

Notes to Pages 258–263 � 345

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

system of capital punishment, which he argued has been “reduced to an insignifi –
cant minimum,” or at least that the public is comfortable with the risks of error that
exist. Ibid., 199. Justice Scalia was responding to a dissent by Justice Souter. Souter
discussed several studies of false capital convictions, stating that “Today, a new
body of fact must be accounted for in deciding what, in practical terms, the Eighth
Amendment guarantees should tolerate, for the period starting in 1989 has seen re-
peated exonerations of convicts under death sentences, in numbers never imagined
before the development of DNA tests.” Ibid., 207– 208 (Souter, J. dissenting). This
book does not squarely address such debates, as I do not examine any set of capital
convictions. However, as discussed earlier in this chapter, seventeen DNA exoner-
ees had been sentenced to death, some multiple times, before their exoneration. For
a discussion of the debate between Souter and Scalia, see Samuel R. Gross, “Souter
Passant, Scalia Rampant: Combat in the Marsh,” 105 Michigan Law Review First
Impressions 67 (2006).

83. While scholars have estimated troubling error rates based on exonerations in
death penalty cases, the cases that Justice Scalia was preoccupied with, such ef-
forts are possible because we know how many people have been sentenced to
death. See D. Michael Risinger, “Innocents Convicted: An Empirically Justifi ed
Wrongful Conviction Rate,” 97 J. Crim. L. & Criminology 761 (2007) (estimating
a 3.3% to 5% wrongful conviction rate for capital rape- murder exonerations in-
volving convictions in the 1980s); Samuel R. Gross and Barbara O’Brien (draft on
fi le with author).

84. See Sean Rosenmerkel, Matthew Durose, and Donald Farole Jr., U.S. Department
of Justice, Bureau of Justice Statistics, Felony Sentences in State Courts, 2006 (2009),
table 1.2.1 (fi nding that of 1,132,290 felony convictions in state courts, only 10,540
involved rape convictions resulting in a prison sentence). A study from the early
1990s reported slightly higher fi gures. See Patrick A. Langan and Jodi M. Brown,
U.S. Department of Justice, Bureau of Justice Statistics, Felony Sentences in State
Courts, 1994 (1992), 2 (fi nding that of 872,217 felony convictions, about 20,000 fel-
ons were convicted of rape, 71% of whom received a prison sentence).

85. The most complete eff ort to examine rape convictions, for example, used limited
data and included no data on how many convicts were both convicted at a trial and
received prison sentences of a given length, much less how many of those involved
stranger- perpetrators. See Lawrence A. Greenfi eld, U.S. Department of Justice,
Bureau of Justice Statistics, Sex Off enses and Sex Off enders (1997), v, 13– 14. The ag-
gregate data presented in that study suggest that exonerees’ convictions are atypical
of the vast majority of rape convictions. Approximately 75% of rape convictions in-
volve nonstranger acquaintance crimes, and of the approximately 21,000 rape con-
victions annually in the 1980s, 80% involved a plea bargain. In the 1980s, approxi-
mately two- thirds of rape convicts received a prison sentence, and among those, the
average jail term was eight months, a fraction of the average of thirteen years that
these innocent convicts served before being exonerated. Ibid. Another study sug-
gests just a few thousand rape convictions involved a trial. See also Patrick A. Lan-
gan and Helen A. Graziadei, U.S. Department of Justice, Bureau of Justice Statis-

346 Notes to Pages 263–264

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

tics, Felony Sentences in State Courts, 1992 (1995), 9 (fi nding that 3,952 rape convicts
were convicted at a trial). Only 16% of rape convicts in another 1992 study received
a maximum sentence of more than ten years, and many of those may not have served
that long. Brian A. Reaves and Pheny Z. Smith, U.S. Department of Justice, Bureau
of Justice Statistics, Felony Defendants in Large Urban Counties, 1992 (1995), 33.
Thus, the number of comparable cases during the relevant time period may not be
great, and factoring the degree to which some innocent individuals never seek DNA
testing, or where DNA tests cannot be conducted because the evidence was not
preserved, the error rate is still higher.

86. Instead, a few jurisdictions have conducted partial or fl awed retesting eff orts. For
example, an eff ort in Virginia has tested only selected cases and used outdated test-
ing technology. See, e.g., Frank Green, “DNA Retests Needed in up to 400 Cases in
Virginia,” Richmond Times- Dispatch, May 14, 2009. Other eff orts have similarly
involved small numbers of cases handpicked by law enforcement or crime laborato-
ries. See Jodi Wilgoren, “Prosecutors Use DNA Test to Clear Man in ’80s Rape,
New York Times, November 14, 2002.

87. District Attorney’s Offi ce for Third Judicial Dist. v. Osborne, 129 S. Ct. 2308, 2323
(2009).

88. D. Michael Risinger et al., “The Daubert/Kumho Implications of Observer Eff ects
in Forensic Science: Hidden Problems of Expectation and Suggestion,” 90 Cal. L.
Rev. 1 (2002).

89. Richard A. Posner, “An Economic Approach to the Law of Evidence,” 51 Stan. L.
Rev. 1477, 1495 (1999); Dan Simon, “The Limited Diagnosticity of Criminal Trials,”
63 Vand. L. Rev. (forthcoming 2011); Dan Simon, “A Third View of the Black Box:
Cognitive Coherence in Legal Decision Making,” 71 U. Chi. L. Rev. 511 (2004); Keith
A. Findley and Michael S. Scott, “The Multiple Dimensions of Tunnel Vision in
Criminal Cases,” 2006 Wis. L. Rev. 291.

90. Itiel E. Dror, David Charlton, and Ailsa E. Peron, “Contextual Information Ren-
ders Experts Vulnerable to Making Erroneous Identifi cations,” 156 Forensic Sci.
Int’l 74– 78 (2006).

91. Trial Transcript, 189– 190, State of California v. Frederick Rene Daye, D 002073
(Cal. Ct. App. May 24, 1984).

92. Maurice Possley, “Exonerated by DNA, Guilty in Offi cial’s Eyes,” Chicago Tribune,
May 28, 2007.

93. Judith Graham, “Crime Labs Contaminate Justice,” Chicago Tribune, June 21,
2001.

94. Linda T. Kohn, Janet M. Corrigan, and Molla S. Donaldson, To Err Is Human:
Building a Safer Health System (Washington, D.C.: National Academy Press, 2000),
3– 7.

95. See Peter Pronovost et al., “An Intervention to Decrease Catheter- Related Blood-
stream Infections in the ICU,” 355 New Engl. J. Med. 2725– 2732 (2006).

96. See Keith Findley, “Proceedings of the Conference on New Perspectives on Brady
and Other Disclosure Obligations: What Really Works: Report of the Working
Groups on Best Practices,” 31 Cardozo L. Rev. 1961, 1974– 1975 (2010).

Notes to Pages 264–271 � 347

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

97. Gary L. Wells and Deah S. Quinlivan, “Suggestive Eyewitness Identifi cation Pro-
cedures and the Supreme Court’s Reliability Test in Light of Eyewitness Science:
30 Years Later,” 33 Law & Hum. Behav. 1, 1 (2009) (“DNA exoneration cases can
only represent a fraction, probably a very small fraction, of the people who have
been convicted based on mistaken eyewitness identifi cation.”).

98. Trial Transcript, 76, State of Illinois v. Ronald Jones, No. 85- 12043 (Ill. Cir. Ct. July
17, 1989).

99. William Shakespeare, Macbeth, act. 3, sc. 2, 8– 12, act. 5, sc. 5, 69.
100. Letter from Kirk Bloodsworth to Congress (May 24, 2004), at www .thejusticepro

ject .org/ testimony/ kirk -bloodsworth -letter -to -congress/ .
101. Ben Protess, “The DNA Debacle,” ProPublica, May 5, 2009 (describing half- billion

dollars in grant for DNA backlog elimination allocated under the 2004 Justice For
All Act).

102. Trial Transcript, 321, State of Texas v. Ronald Gene Taylor (Tex. Dist. Ct. April 28,
1995).

103. See 145 Cong. Rec. S14533- 02.

appendix

1. The Innocence Project Home Page, at www .innocenceproject .org (providing count
of postconviction DNA exonerations; the number as of July 2010 is 255).

2. These documents are available online at a set of University of Virginia School of
Law Library research collection webpages. See Brandon L. Garrett, “Convicting
the Innocent,” at http:// www .law .virginia .edu/ innocence. Links are available at that
webpage to appendixes and resources related to each book chapter. For example, for
rec ords relating to exoneree false confessions, see www .law .virginia .edu/ html/
librarysite/ garrett _falseconfess .htm. For rec ords relating to forensic analysis, see
Brandon L. Garrett, “Exoneree Trials: Testimony by Forensic Analysts,” at www
.law .virginia .edu/ html/ librarysite/ garrett _exoneree .htm .

3. Brandon L. Garrett, Judging Innocence: An Update (2009), at www .law .virginia
.edu/ html/ librarysite/ garrett _exonereedata .htm .

4. An interrater reliability analysis using the Kappa statistic was performed to
determine consistency among raters in coding the two central questions concern-
ing whether eyewitness identifi cations in a case involved indicia of police sugges-
tion or unreliability. Kappa values from 0.40 to 0.59 are considered mod-
erate, 0.60 to 0.79 substantial, and 0.80 outstanding (Landis and Koch, 1977).
Analyses revealed outstanding interrater reliability for suggestion (Kappa 0.89,
p .0.001, 95% CI 0.77 – 1.01) and substantial interrater reliability for reliability
(Kappa 0.78, p .0.001, 95% CI 0.60 – .96). See J. R. Landis and G. G. Koch, “The
Mea sure ment of Observer Agreement for Categorical Data,” 33 Biometrics 59– 174
(1977).

5. For example, trial rec ords obtained often do not provide basic data, such as the race
of the defendant and the race of the victim, data that is readily available in exonerees’

348 Notes to Pages 271–289

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

comparatively well- publicized cases. Indeed, trial rec ords are sealed in sexual as-
sault cases in several key jurisdictions. For cases in which the only rec ords are writ-
ten judicial decisions, data is even more scant. Even data as basic as the year of con-
viction, or types of evidence presented at trial, was not typically reported in
“matched” cases identifi ed through judicial decisions.

Note to Page 289 � 349

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

351

Ac know ledg ments

This book benefi ted from invaluable comments on the manuscript by
Kerry Abrams, Samuel Gross, Richard Lempert, Greg Mitchell, John
Monahan, Maurice Possley, George Rutherglen, Colin Starger, Zahr
Stauff er, Steven Sun, and Larry Walker. Several of the chapters began as
law review articles studying aspects of DNA exonerations: “Judging In-
nocence,” 108 Columbia Law Review 55 (2008); “Claiming Innocence,”
92 Minnesota Law Review 1629 (2008); “Invalid Forensic Science Testi-
mony and Wrongful Convictions,” 95 Virginia Law Review 1 (2009) (with
Peter Neufeld); “The Substance of False Confessions,” 62 Stanford Law
Review 1051 (2010); and “DNA and Due Pro cess,” 78 Fordham Law Re-
view (2010). I benefi ted greatly from collaborating with Peter Neufeld on
the study of the forensic testimony in exonerees’ trials, and with J. J.
Prescott on a project in progress analyzing the litigation of exonerees us-
ing a series of regressions.

I thank for their comments on book chapters or previous law review
articles that led to this book: Anthony Amsterdam, Rachel Barkow,
Tony Barkow, Josh Beaton, Adele Bernhard, Edward Blake, Richard
Bonnie, Darryl Brown, Rebecca Brown, Albert Choi, Anne Coughlin,
Steven Drizin, Jeff Fagan, Jonathan Haidt, Bernard Harcourt, Toby Hey-
tens, Roger Hood, Richard Hynes, Jim Jacobs, Saul Kassin, Jody Kraus,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

352 Acknowledgments

Eric Lander, Richard Leo, Richard Lewontin, James Liebman, Paul Ma-
honey, Nina Morrison, Erin Murphy, Caleb Nelson, Peter Neufeld,
Sinead O’Doherty, Richard Ofshe, Carolyn Ramsey, D. Michael Ris-
inger, Austin Sarat, Barry Scheck, Richard Schragger, Stephen Schul-
hofer, Elizabeth Scott, Dan Simon, Nancy Steblay, Susan Sturm, William
Thompson, Gary Wells, and Rob Warden. The book benefi ted from pre-
sen ta tions at Chicago, Georgia, Harvard, New York University, University
of Virginia, Toronto, and Tulsa law schools, and conferences including
the Annual Conference of the American Society of Criminology, the An-
nual Conference of the Law and Society Association, the Innocence Net-
work Conference, the Conference on Empirical Legal Studies, and the
Fourth Meeting of the National Academy of Sciences, Committee on
Identifying the Needs of the Forensic Sciences Community. I thank stu-
dents in my Wrongful Convictions seminars and my Habeas Corpus
course for their comments over the years.

Assembling this body of rec ords from the cases of the fi rst 250 DNA
exonerees was possible only with the help of many others. I am very
grateful to the law fi rm Winston & Strawn, which had scanned much of
the Innocence Project’s fi les and provided access to those documents.
I give special thanks to Maddy DeLone, Huy Dao, Emily West, and oth-
ers at the Innocence Project for all of their help in both locating materials
and exchanging data, as well as Christine Mumma of the North Carolina
Center on Actual Innocence and the staff of the Center on Wrongful
Convictions for their help locating materials in several cases. I obtained
still more trial rec ords with the assistance of the superlative staff of the
UVA Law Library, and in par tic u lar Michelle Morris, Ben Doherty, and
Kent Olson. Perhaps the largest group of rec ords was obtained with the
help of countless court clerks, court librarians, court reporters, prosecu-
tors, defense lawyers, postconviction lawyers, and innocence projects,
who graciously searched through archives and fi les. The Dallas District
Attorney’s Offi ce was particularly helpful, and provided trial transcripts
for a number of cases that otherwise would not have been located. I give
special thanks to Dianne Johnson and Betty Snow of UVA Law School
for all of their assistance scanning these fi les.

Once these materials were collected and scanned, a team of superlative
research assistants helped to review and code all of the documents over

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Acknowledgments � 353

several years: Jeff rey Bender, Catherine Byrd, James Cass, Christine
Chang, Veronica Dragalin, Rebecca Ivey, Tifani Jones, Bradley Justus,
Jessica King, David Koenig, Shannon Lang, Josephine Liu, Rebecca
Martin, Erin Montgomery, Sinead O’Doherty, Nilakshi Pardigamage,
T. J. Parnham, Rebecca Reeb, Richard Rothblatt, Kerry Shapleigh, Nat-
alie Shonka, Jason Shyung, Elizabeth Studdard, Steven Sun, Elizabeth
Tedford, and Justin Torres. I thank Jessica Kostelnik for her help calcu-
lating an interrater reliability score concerning eyewitness coding and
Edward Kennedy for his help analyzing data and preparing charts dis-
playing several of the results. I thank Mary Wood for assistance creating
the law library research resource websites that have made some of the
materials from these exonerees’ cases available online to the public.

This book is a testament to the per sis tence and courage of these inno-
cent people, who typically struggled for many years to prove their inno-
cence. I also thank the exonerees to whom I reached out when researching
this book, who were kind enough to answer questions about their cases
and to provide trial materials or other rec ords.

I had the privilege to work for Peter Neufeld and Barry Scheck at the
law fi rm then called Cochran, Neufeld & Scheck, LLP, from 2002 to
2004. While there, I had the opportunity to represent several DNA ex-
onerees with respect to civil wrongful conviction actions fi led following
their exoneration.

I thank Elizabeth Knoll for her editorial suggestions and guidance, as
well as all of those at Harvard University Press whose work improved the
manuscript.

This book would not have been possible without the intellectual con-
tributions of my colleagues at the University of Virginia School of Law.
This research was also fi nancially supported by generous grants from the
Olin Program at UVA Law, summer research grants from UVA Law, and
a very generous grant over several years from the Open Society Institute,
Criminal Justice Fund.

My most heartfelt appreciation is reserved for my family, who sup-
ported this work over many years. My amazing children were both born
as I worked on this book, and they have brought joy to my every day. My
parents, Theodore and Bonnie Garrett, my sister, Natalie Garrett, and
my in- laws, Richard and Janice Abrams, have all encouraged and helped

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

354 Acknowledgments

me in countless ways. My greatest thanks are to my wife, Kerry Abrams,
who not only is a lovely and brilliant partner in all things, but also made
the most invaluable, insightful, and thought- provoking contributions to
this book in all of its stages, as an unparalleled reader, editor, and sound-
ing board.

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

355

Index

Abbitt, Joseph, 241
Abdal, Habib: misidentifi cation of, 9,

45– 48, 53, 61, 71– 72; voice showup, 59;
trial of, 82; DNA testing, exoneration,
239– 240

ABFO. See American Board of Forensic
Odontology

A-B- O blood- typing. See Serology
Actual guilt. See DNA testing, and culprit

identifi cation; Guilt
Adams, Kenneth, 140, 174, 190
AEDPA. See Antiterrorism and Eff ective

Death Penalty Act
Alaska, 224, 248
Alejandro, Gilbert, 61, 70, 101
Alexander, Richard, 51, 154
Alibis: weak, 10, 156– 158; in for mant

undermining of, 119– 120, 123– 124,
135– 137; diffi culty of having, 148– 149;
proving innocence with, 155, 160;
percentage use at trials, 155– 156, 282f

American Bar Association, 115, 151
American Board of Forensic Odontology

(ABFO), 104– 105
American Psychology- Law Society, 60
American Society of Crime Laboratory

Directors (ASCLD), 253, 255
Amsterdam, Anthony, 193
Anderson, Marvin, 57– 58, 67, 174, 261
Antiterrorism and Eff ective Death Penalty

Act (AEDPA), 205
Apologies to exonerees, 144, 231, 237
Appeals, 10– 11; guilty pleas, 152; pro-

cessing time, 181, 216; challenging trial

evidence, 182– 185, 193, 226, 282, 283f;
false confessions, 186; pro cess, 194– 196;
direct, 194– 196, 211; reversals, retrials,
vacated convictions, 197– 198; judging
innocence, 200– 202, 210; Strickland
claims, 205– 207; prosecutorial miscon-
duct, 208; cost, 233. See also Postconvic-
tion proceedings

ASCLD. See American Society of Crime
Laboratory Directors

Association of Criminal Defense Lawyers,
251

Avery, Steven, 63, 69, 237, 239

Bain, James, 215
Barnes, Steven, 128, 131, 188
Bauer, Chester, 97, 190, 267
Baze v. Rees, 257
Beatrice Six case, 26– 27, 110
Beaver, Antonio, 158
Bench trials, rulings, 36, 175, 195
Bibbins, Gene, 55, 109
Bite mark comparison: percentage use at

trials, 90, 282f; pro cess/use of, 102– 105;
concealing evidence, 108; jury questions
regarding, 172; fl awed analysis, 178– 179;
postconviction challenges, 189– 190

Black box, 172
Blackmun, Justice Harry, 210
Blair, Michael, 57
Blake, Edward, 88, 110, 218
Bloodsworth, Kirk, 67, 78, 161, 251, 261,

273

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

356 Index

Blood typing. See Serology
Blume, John, 163
Borchard, Edwin, 234, 270
Boston Police Department, 94, 106– 107
Bowman, Locke, 115
Bradford, Marcellius, 38
Brady v. Mary land: signifi cance of, 122,

126, 168, 259; claims, 169, 191, 202– 203,
208, 223, 287; violations, 169– 170, 193,
209

Brennan, Justice William, 48
Brewer, Kennedy, 103, 178– 180, 208, 212
Bromgard, Jimmy Ray, 68, 97, 165– 166,

236, 261, 267
Bromwich, Michael R., 115, 254
Brown, Dennis, 24, 33, 38, 70, 81– 82, 160
Brown, Keith, 241
Brown, Roy, 89, 108, 131, 172, 232
Bryson, David, 188
Bullock, Ronnie, 58
Burnette, Victor, 238
Bush, George W., 231

Cage, Dean, 63, 69
California: jail house in for mants, 143, 256;

criminal justice reforms, 244; cross-
racial identifi cations, 250; crime
laboratory audit, 254

Canada: Criminal Conviction Review
Group, 242– 243; jail house in for mants,
256; Supreme Court, 256

Capillary electrophoresis, 220
Capozzi, Anthony, 67, 69, 161, 172– 173
Center for Wrongful Convictions, 115
Central Park Jogger case, 22, 31, 38, 99,

140, 153
Certainty. See False confi dence
Certiorari, 194– 196
Chalmers, Terry, 67– 68
Chapman v. California, 200
Charles, Ulysses, 77
Chatman, Charles, 160, 261
Checklists, 269, 271
Chicago, IL: rape- murder case, 1– 5; Ford

Heights Four, 39, 110, 115, 140, 191;
Police Department, 110; Crime
Laboratory, 111; court house, 225;
exoneree settlement, 235

Child witnesses, 75– 77
Civil rights lawsuits, 166, 169, 235, 247,

261, 268
Clark, Robert, 60, 159
Clemency, 221, 227, 230, 234
Closing arguments: prosecution, 31,

79– 80, 112, 170– 171; eyewitness
identifi cations, 79– 80; forensic
evidence, 88, 112; procedure rules, 112;
defense, 155; appellate review, 190

Coco, Allen, 52
Codefendant testimony, 139– 140
CODIS. See Combined DNA Index

System
Coerced- compliant confession, 18
Coerced confessions. See Confessions,

contaminated
Coercion, physical, 3– 4, 38– 39, 161
Cognitive bias, 12, 93, 266– 267
Cold hits, 5, 204, 221, 226, 231
Collateral proceedings. See Postconviction

proceedings
Combined DNA Index System (CODIS),

221
Commission on the Future of DNA, 262
Commission on Wrongful Convictions,

244
Compensation for exonerees, 234– 236, 239
Compensation statutes, 236, 239
Composite images, 52, 57, 150
Confessions, of actual culprit, 17, 58, 66,

76– 77, 159, 204, 232, 296n36
Confessions, coerced- compliant, 18
Confessions, contaminated: physical

coercion, 3– 4, 38– 39, 161; case study,
14– 18; in exoneree cases, 19– 21; law
enforcement practices, 22– 23; corrobo-
rated, nonpublic facts, 23– 27; disclosing
facts, denial of, 28– 31; recorded false
interrogations, 32– 33; inconsistent facts,
33– 36; judicial review, 36– 40; experts,
40– 42; reforming interrogations, 42– 44,
247– 248; postconviction challenges to,
185– 186

Confessions, false: in exoneree cases, 8,
19– 21, 279f, 280f; polygraph examina-
tions, 15, 23; psychology of, 18– 19,
294n24; materials examined, 19,
294– 295nn27,28; implicating others,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Index � 357

26– 27, 297n49; judicial review, 36– 40;
postconviction challenges to, 185– 186;
North Carolina reform, 242

Confessions, voluntary: standards for, 16,
36– 40; Constitutional requirement,
19– 20, 37; written statements, 39; mental
disability, 40– 42

Confi rmation bias, 266
Congress, 115– 116, 261– 262, 273
Consistent/similar analysis, 85– 86,

99– 100, 102, 107
Constitution: right to claim innocence, 7,

203, 223, 262; voluntary confession,
19– 20, 37; Due Pro cess Clause, 53; right
to appeal, 194; protection of defendants,
246

Contamination, in confessions. See
Confessions, contaminated

Contamination, eyewitnesses. See
Eye witness identifi cations/mis-
identifi cations

Contamination, in for mants. See In for-
mants; Jail house in for mants

Convicting the Innocent (Borchard), 234
Conviction Integrity unit, 259
Convictions, wrongful. See Wrongful

convictions
Cooperating witnesses, 141– 142
Corroborated, nonpublic facts, 23– 27,

130– 134
Cotton, Ronald: eyewitness misiden-

tifi cation, 62, 67, 78, 238; conviction
reversal, 187– 188, 199; exoneration, 241

Counsel. See Defense counsel
Coverdell, Paul, 261
Cowans, Stephan, 106– 107, 162
Crime laboratories: Houston, 12, 101, 115,

254; types of, 92; analysis errors, 93,
109– 110; audits, 93, 109, 115, 253– 255,
343n54; Montana State, 97, 267– 268;
Virginia Department of Forensic
Science, 108, 219, 221; New Jersey State
Police, 109; Oklahoma State, 110, 254;
Oklahoma City, 111, 113; Illinois, 111,
115, 254; Allegheny County, 112;
Kansas City, 113; reform, 114– 117,
252– 255; West Virginia, 252; ASCLD,
253, 255; Dallas County, 259. See also
Forensic analysts

Criminal Cases Revision Commission,
242– 243

Criminal Conviction Review Group,
242– 243

Criminal investigation before trial,
149– 150

Criminal justice system, reforming. See
Reform

Criminal review: stages of, 194– 196;
substantive errors, 210– 212. See also
Postconviction proceedings

Criner, Roy, 141– 142, 230
Cromedy, McKinley, 74, 187, 249– 250
Cross- examinations: eyewitnesses, 54;

forensic evidence, 91, 113– 114, 116;
in for mant testimony, 127; crucible of the
court, 260

Cross- racial identifi cations, 9, 51, 72– 74,
187, 250, 308– 310nn111,112,115

Crotzger, Alan, 56
Cruz, Rolando: false confession, 31, 36;

jail house in for mant, 129, 191; jury
deliberations, 173; three trials, 197, 210;
reversal, 209

CSI eff ect, 100
Culprit identifi cation. See DNA testing and

culprit identifi cation

Dail, Dwayne, 241
Dallas, TX: exonerations, 12, 228, 259;

Police Department, 106, 252; District
Attorney’s Offi ce, 259

Danziger, Richard, 151, 231
Data collection/analysis, 7, 19, 285– 287,

294– 295nn27,28
Daubert v. Merrell Dow Pharmaceuticals,

Inc., 91, 255
Davidson, Willie, 56, 175
Davis, Cody, 50
Davis, Troy, 224
Daye, Frederick, 158, 267
Deals with in for mants, 127– 130
Dean, James, 26– 27
Death penalty, 5, 146, 155, 205, 224,

257– 258
Death row exonerations, 5, 78, 143, 161, 261
Dedge, Wilton, 126, 128, 134– 136, 138, 228,

237

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

358 Index

Defense cases: alibis, 10, 155– 158; experts,
false confessions, 40– 42; experts, 113,
149, 163– 164; challenges for, 148– 149,
153– 156, 176– 177; before trial, 149– 150;
few witnesses, 149, 155; guilty pleas,
150– 153, 323n10; third- party guilt, 155,
158– 160; testifying innocence, 155– 156,
160– 163; forensics, 156, 163– 164;
prosecutorial, police misconduct,
167– 170; judges, juries, 172– 176; pro se,
195, 225; open- fi le policy, 259– 260;
reform, 260– 261; defense used, case
percentages, 282, 282f. See also
Postconviction proceedings

Defense counsel, 10; false confessions, 36;
right to, 54, 127, 138, 165– 166, 187, 191,
205; forensics, 111– 117; inexperience,
146– 148; diffi culties for, 149, 260; guilty
pleas, 150– 152; meet ’em and plea ’em,
151; in eff ec tive assistance, 165– 167, 187,
190; in eff ec tive assistance, postconvic-
tion claims, 205– 207; type retained, case
percentages, 283, 283f

Degree of attention, 71– 72
Department of Justice (DOJ), 261– 262
Description discrepancies, 68– 70
Deskovic, Jeff rey: false confession, 8,

14– 20, 42, 202, 260; DNA testing, 35– 36,
100

Diff erential extraction, 218
Dillon, William, 141, 143– 144
Direct appeals, 194– 196, 211. See also

Appeals
Disclosing facts: contaminating confes-

sions, 19– 21; interrogation practices,
22– 23; corroborated, nonpublic facts,
23– 27; police denial of, 28– 31; in for mant
knowledge of facts, 124, 130– 134. See also
Confessions, contaminated; Jail house
in for mants

Dissents, 209– 210
District of Columbia, 229, 236, 248
Dixon, John, 151
DNA exonerations. See Exonerations;

Exonerees
DNA testing: cost of, 1, 233– 234; vs.

previous forensic science, 2– 3, 85, 89;
denying requests for, 3, 151, 214– 215,
222, 227– 230; pretrial suspect exclusion,

12, 264, 293n29; at time of trial, 16, 35,
100– 102, 163, 282f, 299n85, 315n43;
limited scope of, 100– 102, 263– 265, 271;
mechanics of, 101, 217– 221; elimination
testing, 110– 111; guilty pleas, 152, 229;
access statutes, 215– 216, 224, 228– 229;
technology evolution, 217– 222; Supreme
Court, 222– 224; exoneration pro cess,
224– 231; confi rming guilt, 233; criminal
procedure reform, 244– 245

DNA testing and culprit identifi cation: from
cold hit, 5, 204, 221, 226, 231; percentage
of cases, 5, 221, 231, 284 f, 285; confes-
sions, 17, 58, 66, 76– 77, 159, 204, 232,
296n36; previous forensic misidentifi ca-
tion, 110; of codefendant, 139– 140, 159;
of witness, 141; of original suspect, 180;
determining actual guilt, 231– 234

Dog scent identifi cation, 135
DOJ. See Department of Justice
Dominguez, Alejandro, 61
Doswell, Thomas, 59, 63, 160, 170
Dotson, Gary, 9, 84– 89, 94, 111, 115
Double- blind lineups, 81– 82, 252, 267, 271
DQ Alpha DNA testing, 219– 220
Drizin, Steven, 32
Dror, Itiel, 266
Due pro cess test, 53– 54
Durham, Timothy, 98, 102, 157

Echols, Douglas, 161– 162
Elimination testing, 110– 111
Elkins, Clarence, 160, 170
Erby, Lonnie, 59
Ethnicity. See Race/ethnicity
Evans, Michael, 162, 187
Evidence: preservation, 12, 225– 226, 239,

264; concealing, 35, 108– 109, 168– 170,
176; postconviction challenges to,
182– 185; destruction of, claims, 226;
type, case percentages, 277, 279f;
fl awed, case percentages, 280, 280f. See
also Forensic evidence

Exclusion, 37, 58, 69, 78, 85, 158
Exhausted claims, 194– 195
Exonerations: from death row, 5, 78, 143,

161, 261; locations of, 5, 277, 278f,
292n15; timeline of, 11, 180– 181, 182f,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Index � 359

215– 217, 283, 284 f; defi ned, 11, 285,
293n25; Supreme Court, DNA testing,
222– 224; pro cess of, 224– 231; unexoner-
ated, 234; life after, 234– 238; victims
after, 238– 239; shortening time until,
239– 240; impact of, 244– 245; limited
scope of, 262– 265; lessons from,
272– 274; case selection, generalizability,
288– 289; matched cases, 289,
348– 349n5

Exonerees: gender of, 5; overview of,
5– 6; juveniles, 5, 8, 21, 38, 191;
race, ethnicity, 5, 73, 279f, 292n15,
308– 310nn111,112,115; crimes convicted
for, percentages, 5, 277, 278f; right to
counsel, 54, 127, 138, 165– 166, 187, 191,
205; testifying innocence, 154– 156,
160– 163; alibi defenses, 155– 158;
third- party guilt defenses, 155, 158– 160;
civil rights lawsuits, 166, 169, 235, 247,
261, 268; bench trials, 175; year of
conviction, 182f, 279f; criminal review
pro cess, 194– 196; reversals, retrials,
vacated convictions, 197– 198; cases of
reversals, 198– 200; innocence claims,
202– 203; compensation, 234– 236, 239;
life after exoneration, 234– 238; death
penalty reforms, 257– 258; list of,
285– 286

Exonerees, and mental disabilities. See
Mental disabilities, and exonerees

Experts: prohibitive cost of, 40, 149,
163– 164, 189; for prosecution, 40– 41,
149, 163, 326n61; use of, contaminated
confessions, 40– 42; for defense, 40– 42,
113, 149, 163– 164, 282f; use of, forensics,
113, 163– 164. See also Forensic analysts

Eyewitness identifi cations/misidentifi ca-
tions: occurrence percentages, 9, 48, 50,
279f, 280, 280f, 281f; cross- racial, 9, 51,
72– 74, 187, 250, 308– 310nn111,112,115;
case study, 45– 48; description discrep-
ancies, 47, 68– 70, 306– 307nn92,93;
suggestive, unreliable, 48– 50; false
confi dence, 49, 63– 68; multiple
witnesses, 50– 51; by victims, 50– 52,
301– 302n21; by acquaintances, 51,
302n22; police procedures, 52– 53;
multiple procedure use, 52– 53, 302n27;

composite images, 52, 57, 150; due
pro cess test, 53– 54; reliability test,
53– 54, 62– 63, 305n70; suggestive
procedures, remarks, 54– 55, 59– 62;
opportunity to view, 70– 71, 307n102;
degree of attention, 71– 72; time passage,
72, 308n109; child witnesses, 75– 77;
judging, 77– 79; closing arguments, jury
instructions, 79– 80; reform of proce-
dures, 80– 83, 242, 248– 252; postconvic-
tion challenges to, 184– 185, 187– 189;
data analysis, 287

Eyewitness identifi cations/misidentifi ca-
tions in lineups. See Lineups

Eyewitness identifi cations/misidentifi ca-
tions and showups. See Showups

Fagan, Jeff rey, 197
Fain, Charles, 105, 130, 132, 137– 138
False confessions. See Confessions, false
False confi dence, 49, 63– 68
False testimony, in for mants. See In for-

mants; Jail house in for mants
FBI. See Federal Bureau of Investigation
Fears, Joseph, 237
Federal Bureau of Investigation (FBI): hair

comparison symposium, 99; shoe print
comparison, 105; voice comparison, 106;
invalid testimony by, 115; crime
laboratory audit, 254; Madrid bomb-
ings, fi ngerprint comparison error, 267

Federal habeas corpus. See Habeas corpus
Federal reforms, 261– 262
Feingold, Russell, 273
Fifth Amendment claims, 185
Fillers, 50, 52, 57, 64, 242
Fillers, identifi cation of, 50, 61, 64– 67, 76,

271
Finality of convictions, 227
Fingerprint comparison, 90, 106– 107, 109,

266– 267, 282f
Florida: rape- murder cases, 40, 134,

213– 215; exoneree compensation, 143;
Innocence Project, 215; DNA testing
statute, 215, 228; death penalty, 258

Ford Heights Four, 39, 110, 115, 140, 191
Forensic analysts: role of, 85– 86, 91– 93,

312– 313n10; invalid testimony, errors,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

360 Index

Forensic analysts (continued)
89– 92, 109– 111, 267– 268; serology,

comparison methods, 93– 100, 105– 107;
DNA testing, 100– 102; odontologists,
102– 105, 108, 178; concealing evidence,
108– 109; reforming practice, 114– 117,
252– 255; postconviction testimony
challenges, 189– 190

Forensic evidence: use at trials, 9, 89– 91,
279f, 280, 280f, 281f, 282f,
312– 313nn10,11; concealing, 35, 108– 109,
316n66; case study, 84– 89; exclusion/
inclusion analysis, 85; errors in analysis,
89– 91, 109– 110; gatekeeping, 91;
lawyers, judges, 91, 111– 113; hair
comparison, 95– 100; bite mark compari-
son, 102– 105; shoe print comparison,
105– 106; voice comparison, 106; fi nger-
print comparison, 106– 107; elimination
testing, 110– 111; reforms, 114– 117,
252– 255; for defense, 163– 164; postcon-
viction challenges to, 185, 189– 190. See
also DNA testing; Serology

Forensic hair microscopy. See Hair
comparison

Forensic odontology, 102– 105, 108, 178. See
Bite mark comparison

Fourth Circuit Court of Appeals, 203– 204
Frank, Barbara, 183
Frank, Judge Jerome, 183
Friendly, Judge Henry, 223
Fritz, Dennis, 110, 139, 141, 160, 176– 177
Fuller, Larry, 59, 62, 161

Georgia: rape cases, 65, 159; death row
inmate, 224; eyewitness reform, 252

Gideon v. Wainwright, 165
Gilchrist, Joyce, 92– 93, 98, 110– 111,

180– 181, 268
Giles, James Curtis, 69, 238
Godschalk, Bruce: contaminated

confession, 31, 35, 39– 40, 161; eyewit-
ness misidentifi cation, 67, 126; jail house
in for mant, 131, 134; DNA testing, 228

Gonzalez, Hector, 158
Good, Donald Wayne, 63, 167
Good Cop/Bad Cop, 22
Goodman, Bruce Dallas, 175, 210

Gore, Glen, 110, 141, 159
Gray, Anthony, 151
Gray, David, 9– 10, 118– 123, 126, 128,

133– 134, 137– 138
Gray, Paula, 39, 140, 191, 206
Green, Anthony, 62, 157, 171, 188
Gregory, William, 98, 109
Gross, Samuel, 19, 126
Guilt: actual, DNA test confi rming, 5,

231– 233; overwhelming evidence of, 11,
17, 36, 186, 202; third- party, 155,
158– 160; judging, 200– 205

Guilt, and culprit identifi cation. See DNA
testing and culprit identifi cation

Guilty pleas, 7, 26, 150– 153, 323n10

Habeas corpus: review, 10– 11; guilty pleas,
152; federal, 179, 205, 223, 230;
pro cessing time, 181; challenging trial
evidence, 182– 185, 282, 283f; criminal
review stages, 194– 196; reversals
through, 197. See also Postconviction
proceedings

Hair comparison: fl awed analysis, 85– 86,
88, 107; percentage use at trials, 90,
282f; pro cess/use of, 95– 100; concealing
evidence, 109; postconviction chal-
lenges, 189– 190

Halsey, Byron, 24, 39, 159, 228
Halstead, Dennis, 138
Hand, Judge Learned, 5– 6, 123
Harmless error analysis, 200– 202
Harris, William, 79
Hatch, Orrin, 92
Hatchett, Nathaniel, 28, 36, 175
Hayes, Travis, 20
Heins, Chad, 101, 133, 171, 173
Hemings, Sally, 220
Hernandez, Alejandro, 31, 186, 197
Herrera v. Collins, 203, 223
Hicks, Anthony, 74, 79, 174, 207
Holdren, Larry, 77– 78
Holland, Dana, 140
Holliday, Judge James O., 253
Holmes v. South Carolina, 158
House, Paul, 223– 224, 285
House v. Bell, 223– 224
Houston Crime Laboratory, 12, 101, 115, 254

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Index � 361

Hung juries, 118– 119, 172, 327– 328n86
Hunt, Darryl, 191, 203– 204, 241
Hypnosis, 55, 61, 187

Illinois: rape- murder cases, 1– 5, 84– 89,
119; Department of Law Enforcement,
85; exonerations, 115; Governor’s
Commission, 115; Crime Laboratory,
115, 254; appellate court, 190; DNA
testing statute, 229; criminal procedure
reform, 244; eyewitness identifi cation
reform, 252; jail house in for mants, 256;
Commission on Capital Punishment,
258; death penalty reform, 258;
interrogation recording, 273. See also
Chicago, IL

Illinois Supreme Court, 3, 112, 129, 186, 217
Inclusion, 85
Inconsistent facts, 33– 36
Indigent defendants: right to repre sen ta-

tion, 7– 8, 205; experts, 40, 149, 163– 164,
189; inadequate funding for, 146,
165– 167, 176; reforming ser vices for,
260– 261

In for mants, 9– 10, 123– 125, 141– 142. See
also Jail house in for mants

Innocence: right to claim, 7, 203, 223, 262;
on trial, 153– 156; testifying to, 154– 156,
160– 163, 282f; ways to claim, 155– 156;
third- party guilt, 155, 158– 160; claims
of, 200, 202– 205, 223– 224; judging,
200– 205

Innocence, and alibis. See Alibis
Innocence Commission, 11, 241– 244
Innocence Network, 6, 287
Innocence Projects: found ers, 6, 245;

Cardozo Law School, 6, 285; data from,
7, 285– 287; obtaining DNA testing
through, 181, 215– 216, 225, 232, 237, 251;
Florida, 215; New York City, 225;
confi rming guilt through, 233

Innocence Protection Act, 261– 262
Interrogations: polygraph examinations,

15, 23; confession- making phase, 20;
pro cess, techniques, 22– 23; false,
recorded, 32– 33; reform, 42– 44,
247– 248; recording, 43– 44, 242,
247– 248, 260, 271, 273, 341– 342n23

Invalid forensics. See Forensic evidence
Invalid testimony, defi ned, 87

Jackson, Justice Robert, 259
Jackson, Willie, 68, 71, 103, 204
Jackson v. Virginia, 204
Jail house in for mants: percentage use at

trials, 10, 124, 257, 279f, 280f; case
study, 118– 123; bolstering prosecution’s
case, 120, 134– 138; use of, 123– 125;
evidence contamination, 124, 133; in
murder cases, 125– 127; deals with,
127– 130; knowledge of facts, 130– 134;
testimony motives, 138– 139; reforming
use of, 142– 144, 256– 257, 344n59;
postconviction challenges to testimony
of, 185, 190– 194

Jean, Lesly, 61, 187, 210, 241
Jeff erson, Thomas, 220
Jimerson, Verneal, 140, 191, 208
Johnson, Calvin, 65– 66, 174
Johnson, Justin Albert, 180
Johnson, Larry, 62, 74
Johnson, Richard, 60– 61, 175
Johnson, Rickey, 157
Jones, David Allen, 32, 40
Jones, Ronald: case description, 1– 5;

physical coercion, 3– 4, 38– 39, 161;
contaminated confession, 7, 27, 33, 42,
153; testifying innocence, 161; DNA
testing, 216– 217; compensation, 235; life
after exoneration, 236; lessons from
case, 272– 273

Judges/judging: denying DNA testing
requests, 3, 151, 227– 229; false confes-
sions, 36– 40; eyewitness misidentifi ca-
tions, 77– 80; gatekeeping forensic
evidence, 91; fl awed forensics, 111– 117;
jail house in for mants, 142– 143, 256– 257;
convicting innocent defendants,
172– 176; bench trials, 175, 195; post- trial
evidence challenges, 182– 185; unwilling-
ness to believe innocence, 193– 194, 196,
201– 202, 210– 211; reversals, 198– 200;
harmless error analysis, 200– 202; guilt,
innocence, 200– 205, 333– 334n58;
dissents, 209– 210

“Judging Innocence” (Garrett), 286

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

362 Index

Juries: eyewitness identifi cation instruc-
tions, 79– 80, 251; hung, 118– 119, 172,
327– 328n86; jail house in for mant
instructions, 142– 143, 256; prosecution
evidence strength, 153– 154; alibi
distrust, 156; deadlock, 172– 174;
convicting innocent defendants,
172– 176; race, ethnicity of, 174, 208– 209;
selection of, 174, 208– 209

Juvenile exonerees, 5, 8, 21, 38, 191
Juvenile witnesses, 75– 77

Keller, Judge Sharon, 230
Kirk Bloodsworth DNA Testing Program,

261
Kogut, John, 38, 127
Kordonowy, Paul, 267
Krone, Ray, 103– 104, 189– 190, 209

Laboratory Accreditation Board,
American Society of Crime Laboratory
Directors, 253, 255

Lake, Chief Justice I. Beverly Jr., 241
Lander, Eric, 114
Lavernia, Carlos, 69, 78– 79, 155
Law enforcement. See Police
Lawyers. See Defense counsel; Prosecution
Liebman, James, 197, 244
Life sentence, 5, 247
Lincoln, John Otis, 58
Lineups: witness suggestion, 43, 46,

49– 50, 55, 57– 62; fi ller identifi cation,
50, 61, 64– 67, 76, 271; live, 52, 54;
procedure training, 53; right to counsel,
54, 187; stacked, 57– 59; voice, 59;
reform, 81– 82, 242, 250, 252, 259;
double- blind, 81– 82, 252, 267, 271; dog
scent, 135. See also Photo arrays

Linscott, Steven, 112, 190
Live lineup, 52, 54
Lloyd, Eddie Joe, 37, 204, 247
Loftus, Elizabeth, 81
Lyons, Marcus, 48, 225

Manson v. Brathwaite, 53, 188; Due
Pro cess test, 53– 54; reliability factors of,

63, 70– 72; fl aws, 65, 68– 69, 72, 75, 77,
81– 82, 246; reform, 82, 251

Mary land: Supreme Court, 78; Court of
Appeals, 251; eyewitness identifi cation
reform, 251– 252; crime laboratory audit,
254; death penalty reform, 258

Masking, 4– 5, 87, 94, 218
Massachusetts: Boston Police Department,

94, 106– 107; State Police, 107; cross-
racial identifi cations, 250; death penalty
reform, 258; Governor’s Council on
Capital Punishment, 258; eyewitness
identifi cation reform, 260

Massiah v. United States, 126– 127, 191
Match, 88, 99, 107
Matched comparison groups, 198, 289,

332n47, 348– 349n5
Mayes, Larry, 61
McCarty, Curtis, 98, 113, 171, 180– 181, 189
McClendon, Robert, 60, 237
McCray, Antron, 31
McGee, Arvin, 158, 173, 235
McGowan, Thomas, 62, 74, 155, 170
McMillan, Clark, 70, 126
McSherry, Leonard, 75– 77, 162– 163
Media, and pretrial reporting, 150
Medical error reform, 269– 270
Melendez- Diaz v. Massachusetts, 116
Mental disabilities, and exonerees, 5; false

confessions, 8, 21, 23, 38; experts,
confession, 40– 42, 194; guilty pleas, 152;
police precautions, 248; death penalty
exonerations, 257

Michigan, 247, 254, 269
Microscopic hair comparison. See Hair

comparison
Miller, Jerry, 48, 68, 222
Miller, Neil, 56– 57, 78, 93– 94, 111, 237
Miller, Robert, 20, 31, 38, 110
Minorities. See Race/ethnicity
Miranda v. Arizona, 14, 37
Miranda warnings, 8, 14– 15, 22, 37, 40– 42,

185
Mississippi, 178– 180
Mistrials, 113, 172– 173
Mitchell, Marvin, 94– 95
Mitchell, Perry, 74, 80, 95, 175
Mitochondrial DNA (mtDNA) testing,

220

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Index � 363

Montana, public defense in, 165– 166
Montana State Crime Laboratory, 97,

267– 268
Montana Supreme Court, 190, 267– 268
Moon, Brandon, 156, 237
Morin, Guy Paul, 256
Mosley, Eddie Lee, 42, 214– 215, 232
MtDNA testing. See Mitochondrial DNA

testing
Mumma, Christine, 241
Mumphrey, Arthur, 140, 159
Murder cases: percentages, 5, 277, 278f;

false confessions, 21; jail house in for-
mants, 125– 127; guilty pleas, 151;
reversal rates, 183– 184, 198

Murder- rape cases, percentages, 5, 21, 277,
278f

Napue v. Illinois, 126
National Academy of Sciences (NAS):

report, 90, 96, 99, 102, 262; spectro-
graph report, 106; forensic science
reform, 115– 116, 255; “To Err Is Human”
report, 269

National Institute of Forensic Science,
262

Nebraska, Beatrice Six case, 26
Nelson, Bruce, 139– 140, 159
Neufeld, Peter, 6, 102, 245, 287
Nevada, and jail house in for mants, 256
New Jersey, criminal review reforms in,

244, 249– 251
New Jersey Supreme Court, 74,

249– 251
Newton, Alan, 80, 225– 226
New York: rape- murder case, 14– 18;

Central Park Jogger case, 22, 31, 38, 99,
140, 153; rape case, 45– 48; State Police,
108; Allegheny County Crime
Laboratory, 112; DNA statute, 229, 240;
freedom of information laws, 232;
compensation statute, 240; crime
laboratory audit, 254

New York Times, exoneree interviews, 235,
237

No- fault compensation statutes, 236
Nonpublic facts, 23– 27, 130– 134
Norfolk Four, 234

North Carolina: exonerees, 241; Supreme
Court, 241; eyewitness identifi cation
reform, 252; crime laboratory audit,
254– 255

North Carolina Innocence Commission,
11, 241– 244

North Carolina Inquiry Commission,
242– 243

Ochoa, Christopher, 32, 150– 151, 231
Ochoa, James, 53
O’Connor, Justice Sandra Day, 5
O’Donnell, James, 162
Odontologists. See Forensic odontology
Ohio: sex off ender law, 237; criminal

procedure reform, 244; eyewitness
identifi cation reform, 252; crime
laboratory audit, 254

Oklahoma: State Police, 92; State
Crime Laboratory, 110, 254; murder
case, 141; jail house in for mants, 143,
256; rape case, 157, 173; fl awed
forensics, 268

Ollins, Calvin, 162
Open- fi le policy, 259– 260
Opening statements, 111, 154– 155, 168
Opportunity to view, 70– 71, 307n102
Osborne v. District Attorney’s Offi ce,

223– 224, 230, 265
Other- race eff ect, 72. See also Cross- racial

identifi cations
Overwhelming evidence of guilt, 11, 17,

186, 202

Pardon, 11, 88, 230, 234
PCR. See Polymerase chain reaction
Peacock, Freddie, 20, 222
Pendleton, Marlon, 111
Peterson, Larry, 98, 109
Photo arrays: false identifi cations from,

45– 46, 61– 62, 65– 68; suggestive, 51,
56– 59, 77– 78, 170, 188– 189; procedures
for use, 52– 54; reforms, 82, 250. See also
Lineups

Physical coercion, 3– 4, 38– 39, 161
Piszczek, Brian, 238
Plea bargaining, 151– 153, 259, 271

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

364 Index

Police: physical coercion, 3– 4, 38– 39, 161;
confession contamination, Deskovic
case, 15– 18; interrogation practices,
22– 23; confessions, fact feeding, 23– 27,
33– 35; denying disclosing facts, 28– 31;
recorded false interrogations, 32– 33;
reforming interrogations, 42– 44,
247– 248; eyewitness suggestion, 46,
48– 50, 54– 55, 59– 62; eyewitness
procedure, 52– 53, 252; forensic analysts,
91– 93; in for mant contamination, 124,
133, 141– 142, 144; jail house in for mants,
deals with, 127– 130; third- party guilt
leads, 159– 160; misconduct, 167– 170

Polygraph examinations, 15, 23
Polymerase chain reaction (PCR), 220
Pope, David Shawn, 106
Postconviction proceedings, 10– 11; length

of time for, 181; challenging trial
evidence, 182– 185, 282, 283f; eyewitness
misidentifi cation claims, 184– 185,
187– 189; false confession claims,
185– 186; fl awed forensics claims, 185,
189– 190; false in for mant testimony
claims, 185, 190– 194; stages of criminal
review, 194– 196; defi ned, 195; reversals,
retrials, vacated convictions, 197– 198;
cases of reversals, 198– 200; judging
guilt, innocence, 200– 205; in eff ec tive
counsel claims, 205– 207; prosecutorial
misconduct claims, 207– 209; substan-
tive errors, 210– 212; reform, 211– 212;
evidence destruction claims, 226. See
also Appeals; Habeas corpus

Posthumously vacated convictions, 214
Prejudice standard, 201– 202
Prescott, J. J., 199, 289
Pretrial: DNA testing, suspect exclusion,

12, 264, 293n29; identifi cations, 78, 251;
challenges, 149– 150; publicity, 150

Pro se, 195, 225
Prosecution: closing arguments, 31, 79– 80,

112, 170– 171; experts, 40– 41, 149, 163,
326n61; opening statements, 111,
154– 155, 168; fl awed forensics, 111– 117;
in for mant contamination, 124, 133;
jail house in for mant use, 125– 127,
134– 138; jail house in for mant deals,
127– 130; misconduct, 167– 170; advantage

of, 176– 177; challenging misconduct of,
207– 209; consenting to DNA testing,
226– 227; reform, 259– 260

Protess, David, 140
Provonost, Peter, 269
Psychological interrogation tactics, 23
Public defenders, 167. See also Defense

counsel

Race/ethnicity: exonerees, 5, 73, 279f,
292n15, 308– 310nn111,112,115; cross-
racial identifi cations, 9, 51, 72– 74, 187,
250, 308– 310nn111,112,115; victims, 73,
308– 310nn111,115; juries, 174, 208

Rachell, Ricardo, 69
Rainge, William, 140, 174, 190, 206
Rakoff , Judge Jed, 258
Ramsey County, MN, 260
Rape cases: percentages, 5, 263– 264, 277,

278f; false confessions, 21; eyewitness
misidentifi cation, 50– 51, 80– 81; fl awed
forensics, 89– 90; serology testing,
93– 95; jail house in for mants, 126; guilty
pleas, 151; reversal rates, 183– 184, 198

Rape kits, 84, 218, 233, 264
Rape- murder cases, percentages, 5, 21, 277,

278f
Recantations: of confessions, 3, 25,

160– 161; by witnesses, 88, 141, 213, 238;
by in for mant, 144, 192

Recording interrogations: false, 32– 33; as
fi rst step, 43– 44, 248; requiring, 242,
247– 248, 260, 273, 341– 342n23; true
confessions, 271

Reform: contaminated confessions, 42– 44;
interrogations, 42– 44, 247– 248;
eyewitness identifi cations, 80– 83, 242,
248– 252; lineups, 81– 82, 242, 250, 252,
259; crime laboratories, 114– 117,
252– 255; forensic science, 114– 117,
252– 255; jail house in for mants, use of,
142– 144, 256– 257, 344n59; postconvic-
tion proceedings, 211– 212; shorter time
for exoneration, 239– 240; North
Carolina Innocence Commission,
241– 244; criminal procedure, 244– 247;
death penalty, 257– 258; prosecution,
259– 260; defense, 260– 261; federal,

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Index � 365

261– 262; necessity for further investiga-
tion, 262– 265; systemic errors, 265– 268;
wrongful convictions, preventing,
269– 272

Reid Technique, 22– 23
Reliability test, 62– 63, 305n70. See also

Manson v. Brathwaite
Restivo, John, 131– 132, 138
Restriction Fragment Length Polymor-

phism (RFLP) testing, 220
Retained lawyer, 167. See also Defense

counsel
Retrials, 11, 197– 198
Reversals: instances of, 183– 184, 197– 200,

332– 333nn48,49,50,51; false confession,
186; eyewitness misidentifi cation, 187;
fl awed forensics, 189– 190; in for mant
testimony, 190– 194; defi ned, 197;
matched case comparison group, 198,
332n47; in eff ec tive counsel, 206;
prosecutorial misconduct, 208

RFLP testing. See Restriction Fragment
Length Polymorphism (RFLP) testing

Richardson, James, 157
Right to claim innocence, 7, 203, 223, 262
Right to confront witnesses, 8, 116
Right to counsel, 54, 127, 138, 165– 166, 187,

191, 205
Robinson, Anthony, 55
Rollins, Lafonso, 35, 39, 267
Roman, Miguel, 133, 154
Ryan, George H., 3, 258

Saecker, Fredric, 70
Salaam, Yusef, 22
Sarsfi eld, Eric, 173
Scalia, Justice Antonin, 116, 258, 262, 264
Scheck, Barry, 6, 215, 233, 245
Schwarzenegger, Arnold, 244
Scientifi c Working Group on Shoeprint

and Tire Tread Evidence, 105
Secretors, 86, 95
Serology: fl awed analysis, 2– 5, 87– 88, 90,

109– 110; masking, 4– 5, 87, 94, 218;
pro cess of, 86, 93– 95; percentage use at
trials, 90, 94, 282f

Shelden, Debra, 26– 27
Shoe print comparison, 90, 105– 106, 282f

Short Tandem Repeat (STR) DNA testing,
220– 222

Showups: defi ned, 46, 49, 52, 55; case
percentage used, 52, 55; faulty use, 53,
55– 57, 81, 249; limiting use of, 252

Similar/consistent analysis, 85– 86, 99– 100,
102, 107

Six- pack photo array, 52
Smith, Frank Lee, 11, 42, 213– 216, 224
Specifi c fact knowledge. See Corroborated,

nonpublic facts
Spectrograph, 106
State habeas. See Habeas corpus: review;

Postconviction proceedings
Stevens, Justice John Paul, 257
Stinson, Robert Lee, 103– 104, 166– 167
STR DNA testing. See Short Tandem

Repeat (STR) DNA testing
Strickland v. Washington, 191, 205– 207,

223
Suff olk County, MA, 260
Suggestive identifi cation procedures,

48– 50, 54– 55, 59– 62
Supreme Court: defendants’ rights, 7– 8,

164, 246; actual innocence claim, 7, 203,
223, 262; Miranda protections, 8, 22, 37,
40– 42, 185; voluntariness standard, 16,
37– 38, 40; due pro cess, 53– 54; expert
evidence, 91– 92, 116; in for mants, 123,
126– 127, 142; prosecutors, role of, 168,
207– 208; certiorari, 194– 196; harmless
error test, 200– 201; DNA testing,
222– 224; destruction of evidence, 226;
human error, 269

Sutton, Josiah, 101– 102

Taylor, Ada JoAnn, 26– 27
Taylor, Ronald, 171, 273
Terry, Paul, 162
Testimony: invalid, defi ned, 87; of

codefendants, 139– 140; of cooperating
witnesses, 141– 142. See also Confessions,
contaminated; Defense cases; Experts;
Jail house in for mants

Texas: DNA exonerations, 12; Houston
Crime Laboratory, 12, 101, 115, 254;
crime laboratory audits, 115, 254; death
row inmate, DNA testing, 224; denying

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

366 Index

Texas (continued)
DNA testing, 228; Court of Criminal

Appeals, 230; exoneree compensa-
tion, 236; death penalty, 258.
See also Dallas, TX

Third- party guilt, 155, 158– 160, 282f
Thompson, Jennifer, 67, 238
“To Err Is Human” (NAS), 269
Totality of the circumstances, 37
Townsend, Jerry Frank, 38, 40, 214– 215,

236
Trainum, James, 28
Traynor, Chief Justice Roger, 182, 185, 200
Trial errors, 200– 201, 210– 212
Trial evidence. See Evidence; Forensic

evidence
Trials. See Defense cases; Jail house

in for mants; Judges/judging; Juries;
Prosecution

Trials, and bench rulings, 175, 195
Trial transcripts, 7, 277, 278f, 286
Tunnel vision, 266
Turner, Keith, 237
21- day rule, Virginia, 227
250 exonerees, 5– 6, 285– 286. See also

Exonerations; Exonerees

Unexonerated, the, 234
United Kingdom, Criminal Cases Revision

Commission, 242– 243
United States v. Wade, 187
Unreliable forensics. See Forensic evidence
Unreliable identifi cations. See Eyewitness

identifi cations/misidentifi cations
U.S. Supreme Court. See Supreme Court

Vacated convictions, 197– 198, 214
Vacatur, 230– 231
Validity, 87
Vasquez, David, 22, 39, 40, 43– 44
Venue transfer, 150
Victims: misidentifi cations, 50– 52,

301– 302n21; race, ethnicity, 73,
308– 310nn111,115; after exoneration,
238– 239

Videotaping interrogations. See Recording
interrogations

Virginia: rape- murder case, 29– 30, 33,
145– 148; 21- day rule, 227; Norfolk Four,
234; eyewitness reform, 252; crime
laboratory audit, 254; death penalty,
258

Virginia Department of Forensic Science,
108, 219, 221

Voice comparison, 90, 106, 282f
Voice lineup, 59
Voluntariness standard, 16, 37– 38
Voluntary confessions. See Confessions,

voluntary

Waller, James, 231
Waller, Patrick, 74, 228
Warden, Rob, 115
Warney, Douglas, 25– 26, 39
Washington, Calvin, 128, 130, 134, 138
Washington, D.C. See District of Columbia
Washington, Earl, Jr.: faulty defense case,

10, 154– 155, 206; false confession, 29– 31,
33– 35, 42; concealed forensic evidence,
108; case description, 145– 148;
indigence, counsel, 167; DNA testing,
217, 219– 221, 227

Waters, Betty Anne, 237– 238
Waters, Kenneth, 141, 237– 238
Waters, Leo, 241
Watkins, Craig, 259
Watkins, Jerry, 173, 191– 193, 209
Weapon: described in confession, 16, 25,

27, 153; eyewitness focus, 71– 72;
in for mant testimony, 131

Webb, Mark, 187
Wells, Gary, 81, 249
West, Michael, 178– 179
West, Valerie, 197
West Virginia: criminal procedure

reforms, 244, 252– 254; crime laboratory,
252; Supreme Court of Appeals, 253

White, John Jerome, 66
White, Joseph, 26– 27, 37
Whitfi eld, Arthur Lee, 171
Whitley, Drew, 112, 128, 132, 142
Wigmore, John Henry, 18
Williams, Dennis, 110, 128, 139– 140, 174,

206
Williams, Willie Otis, 63

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Index � 367

Williamson, Ronald: false confession, 27,
199; habeas petition, 96, 191; fl awed
forensic evidence, 110; in for mant
testimony, 129, 131; actual perpetrator as
witness, 141, 159; reversal, 186; vacated
conviction, 206– 207

Wise, Korey, 162
Witnesses: right to confront, 8; juvenile,

75– 77; DNA testing, culprit identifi ca-
tion, 141; cooperating, 141– 142; few
for defense cases, 149, 155. See also
Eyewitness identifi cations/mis –
identifi cations

Woodall, Glen Dale, 231, 252– 253
Woodard, James Lee, 151– 152
Working Group for Eyewitness Evidence,

262

Written decisions, 184, 196– 197,
329n16

Wrongful convictions: reasons for, 6– 11;
unknown scope of, 11– 13, 262– 265;
cross- racial identifi cations, 72– 74;
eyewitness misidentifi cation, 80;
preventing, 269– 272. See also
Exonerations

Wyniemko, Kenneth, 126– 127, 137, 155

Yarris, Nicholas, 32, 80, 136– 138, 167
Youngblood, Larry, 196, 210, 226, 234
Y-STR DNA testing, 220

Zain, Fred, 92, 101, 252– 254

Garrett, Brandon L.. Convicting the Innocent : Where Criminal Prosecutions Go Wrong, Harvard University Press, 2011. ProQuest
Ebook Central, http://ebookcentral.proquest.com/lib/umboston/detail.action?docID=3300940.
Created from umboston on 2024-09-29 18:46:23.

C
op

yr
ig

ht
©

2
01

1.
H

ar
va

rd
U

ni
ve

rs
ity

P
re

ss
. A

ll
rig

ht
s

re
se

rv
ed

Rhetorical Analysis Peer Review Sheet

Read your peer’s essay.

2.
Annotate the essay, make comments, etc.

3.
Discuss your thoughts with your peers.

4.
Fill out the below form and give a copy to your peer.

5.
Upload a copy of the filled-out document to canvas.

Whose essay are you reviewing?

What is this essay’s thesis statement? Does it make sense? Is it specific? Interesting?

Does their introduction follow the 3 introductory moves?

Do the body paragraphs contain a mix of TEAL? (
Topic,
Evidence,
Analysis,
Link) Though not every paragraph needs all 4 elements, are there any paragraphs that feel overstuffed or undercooked?

Which parts of the essay stand out to you? What’s working well? Are there any elements you would consider incorporating into your own writing?

Where do you see the rhetorical strategies we’ve discussed? Does the author hedge? Boost? Do they ask rhetorical questions or provide directives? Where might they consider adding these in?

What suggestions do you have for ways your peer could improve their essay? Where are you confused? Where is your attention drifting?

I HAVE ATTACHED MY FRIEND DRAFT BELOW, YOU HAVE TO DO A PEER REVIEW ACCORDING TO THE INSTRUCTIONS GIVEN ABOVE.

Sarah Reis

English Composition

Professor C Lavigne

September 2024

Rhetorical Analysis

In her essay, “Four Things Social Media Can Teach You about College Writing—and One

Thing It Can’t,” Ann N. Amicucci explores how students can transfer rhetorical skills they

already practice on social media to their academic writing. She identifies four key composing

practices that can be repurposed for college essays: imitating successful formats, targeting

specific audiences, shaping one’s ethos, and making strategic literary choices. Amicucci extends

this with ideas of intertextuality and interdiscursivity. This explains how writers ‘borrow’ from

pre-existing texts and rhetorical strategies. She concludes through practical examples such as,

“riffing off” model essays, thinking about “hashtag audiences,” and choosing “filters”and how

these practices can enrich academic representation.

This essay is important to the field of composition studies because it links daily digital habits of

students with their academic writing assignments. The approach connects their informal,

everyday digital communication with formal conventions at an academic level.

Addressing students who may struggle to write at the collegiate level, Amicucci places

academic writing in social media practices in a way that makes complex rhetorical theories more

understandable. Her essay contributes to the ongoing conversations in composition methods on

how to incorporate students’ digital literacies into the classroom and offers students a set of

strategies to help transition from informal to formal writing contexts. The approach engages

more students and develops writing abilities.

Amicucci applies real-life social media examples to critically analyze rhetorical moves

while emphasizing compositional values as relevance, approachability, and skill transferability.

A good composition argument is one that links theory with practice. In discussing intertextuality

and interdiscursivity through the use of social media examples, like Hashtags in an Instagram

post, Amicucci is trying to get students to interpret rhetorical choices in meaningful ways. These

interpretations help them learn about complex concepts involved in writing.

Amicucci argues that “students can effectively transfer rhetorical resources from social

media to academic writing.”. She explains how writers “borrow” from existing texts and

rhetorical moves through the use of intertextuality and interdiscursivity. She identifies four key

strategies: imitating formats that work, targeting audiences, shaping one’s ethos, stylistic choices.

She illustrates how to do these things in social media-specific ways that then can be transferred

into academic writing to support students in using their digital literacy skills in formal writing.

Amicucci uses hedges and boosters as linguistic means of negotiating the details of

writing. Hedges, as shown in her statement, “You might think of social media as completely

separate from the writing you do for college, but the truth is, the two aren’t that different,” make

propositions more tentative and avoid strong claims. In that example, the verb “might” shows

uncertainty and invites a cautious and reflective style of composing. This attention to the

complexities in the writing process speaks to the current emphasis in composition studies on

critical thought and multiple points of view.

On the other hand, boosters make a statement sure and confident; they reinforce an

argument. In this context, her remark, “nothing you write exists in a vacuum,” is telling that all

writing is joined. By balancing the boosters and hedges, Amicucci elicits a text to be persuasive

and nuanced to enable the student to transfer their skills from informal to formal with sensitivity

to diverse views.

Reader pronouns, like “you” and “we,” establish a connection with the readers and allow

the argument to be relatable. For instance, she asks, “Have you ever had a writing assignment to

work on but find yourself scrolling through social media instead?” This gives a direct invitation

for the readers to think about their experiences. The use of such inclusive language will bring a

sense of unity among the users of social media and allow the text to sound even more

conversational.

Rhetorical devices differ according to discipline: they reflect the values and norms of

each discipline. Which are more or less common can sometimes provide a way of gauging what

types of arguments and evidence are most convincing in which contexts. Terms, for example, and

citation style illustrate how disciplinary conventions are reproduced. The ability to understand

these rhetorical conventions provides insight into the epistemological assumptions and

argumentative moves that underlie various academic disciplines.

The essay will, therefore, effectively illustrate how rhetorical techniques apply to various

forms of writing, from social media posts to academic essays. It introduces the generally

applicable concepts of intertextuality and interdiscursivity. It is explained that rhetorical moves

may transcend genres. Amicucci points out the convergence of rhetorical strategies in the use of

social media and academic writing and emphasizes the possibility of students’ skills transfer

across contexts.

The essay gives emphasis on the integration of writing skills in social media into

academics, but, however, also provides recognition regarding tensions between the casual tone of

social media and traditional academic conventions. An invitation for stylistic experimentation

and personal voice may run in conflict with expectations of formality, particularly in certain

disciplines.

Admittedly, much of the rhetorical work present in day-to-day writing transfers into the

academic work. In this regard, students can begin to consider the ways in which they already

know so much about rhetoric and learn to accommodate that knowledge within the disciplinary

norms of the class. A clear academic voice arises from a balance between one’s personal style

and the conventions of the field and allows for a more active and inclusive writing.

1. Amicucci, Ann N. “Four Things Social Media Can Teach You about College

Writing—and One Thing It Can’t.” 2018.

2. Thaiss, Chris, and Tara Zawacki. “What is Academic? What is Alternative?” , 2006.

Expert paper writers are just a few clicks away

Place an order in 3 easy steps. Takes less than 5 mins.

Calculate the price of your order

Type of paper needed:

Pages:

You will get a personal manager and a discount.

Academic level:

We'll send you the first draft for approval by at

Total price:

$0.00