Convicting the Innocent:
T e x a s J u s tic e D e r a i l e d
Stories of Injustice and the Reforms
That Can Prevent Them

THE JUSTICE PROJECT

EDUCATIO
FUND

“Any wrongful conviction is a tragedy
because it leaves the guilty unpunished
and condemns the innocent to prison,
or death.”
— W allace B. J efferson , C hief J ustice
of the T exas S upreme C ourt

THE JUSTICE PROJECT

The Justice Project (TJP) is a non-profit, non-partisan
organization dedicated to improving the fairness
and accuracy of the criminal justice system. TJP’s
Campaign for Criminal Justice Reform seeks to
reaffirm America’s core commitment to fairness and
accuracy by designing and implementing national
and state-based campaigns to advance reforms that
address significant flaws in the American criminal
justice system.
The Justice Project has developed a national
program of initiatives designed to address and
affect the policies and procedures that perpetuate
errors and contribute to the conviction and
incarceration of innocent people. As such, TJP
advocates for 1) improvements in eyewitness
identification procedures; 2) electronic recording
of custodial interrogations; 3) higher standards
for admitting informant or accomplice testimony
at trial; 4) expanded discovery in criminal cases;
5) improvements in forensic testing procedures;
6) greater access to post-conviction DNA testing;
7) proper standards for the appointment and
performance of counsel in capital cases; and
8) improving prosecutorial accountability.

The Justice Project Staff
John F. Terzano, President
Joyce A. McGee, Executive Director
Robert L. Schiffer, Executive Vice President
Kirk Noble Bloodsworth, Program Officer
Edwin Colfax, Director of State Campaigns
Rosa Maldonado, Controller
Leah Lavin, Development Coordinator
Alanna Holt, Policy Coordinator
Lauren Brice, Policy Associate
Acknowledgements: The Justice Project would like
to thank Jennifer Willyard and Michelle Strikowsky for
their contributions to this report.
For information on ordering additional copies of this
report, contact Edwin Colfax at (512) 391-2320.
For more information on The Justice Project, contact
(202) 638-5855 or info@thejusticeproject.org.
1025 Vermont Avenue, NW, Third Floor,
Washington, DC 20005
(202) 638-5855 • Fax (202) 638-6056
www.thejusticeproject.org
510 S. Congress Avenue, Suite 304
Austin, TX 78704
(512) 391-2320 • Fax (512) 391-2330
www.thejusticeproject.org/state/texas
Cover: upper right, Greg Kendall-Ball / Innocence Project of Texas;
Upper left, lower left and lower right, Dallas Morning News

©2009 The Justice Project — All Rights Reserved.

Convicting the Innocent:
T e x a s J u s tic e D e r a i l e d

Clay Graham

About The Justice Project

EDUCATION
FUND

Table of Contents
Factors Leading to Wrongful Convictions......................................2
Wrongful Convictions Threaten Public Safety...............................2
The Limits of DNA Evidence..........................................................4
Texas DNA Exoneration Facts.........................................................5
Jurors Deserve the Most Reliable Evidence....................................6
Reform: Improving the Reliability of Evidence.............................6
Eyewitness Identification.................................................................6
Electronic Recording of Custodial Interrogations.........................8
Jailhouse Informant Testimony.......................................................9
Forensic Oversight..........................................................................10
Expanded Discovery.......................................................................11
Stories of Injustice: The Texas DNA Exonerated.........................12
The Texas DNA Exonerated Chart................................................28

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

S

ince 1994, Texas has exonerated
thirty-nine innocent people who served
over 500 years in prison for crimes they
did not commit. This report contains brief
overviews of these thirty-nine cases, all of which
have been exposed by DNA evidence, and analyzes the systemic problems that have resulted in the
wrongful convictions of the innocent. By identifying
the causes of wrongful convictions and implementing practical reforms, Texas can
the fairness, accuracy,
These thirty-nine DNA increase
and reliability of its criminal
cases expose a criminal justice system.
Because DNA evidence is
justice system that is
only available in a fraction of
wrought with problems cases, the wrongful convictions
described in this report are only
that lead to wrongful
the beginning. There are many
convictions.
other wrongful convictions that
have been cleared without the
benefit of DNA. While non-DNA exonerations are
more difficult to prove, they are similar to DNA cases
in that they trace back to the same flawed procedures
in need of reform. As such, these thirty-nine DNA
cases expose a criminal justice system that is wrought
with problems that lead to wrongful convictions.
Although several of the exoneration cases involve instances of intentional misconduct, inadvertent error is by far more common. It would only
compound these injustices, however, to assume that
these mistakes were inevitable. Texas cannot ignore
its broken criminal justice system. This report addresses the common causes that lead to wrongful
convictions, as echoed in each of these cases, and
presents practical reforms to prevent such errors. It
is critical for Texas to take action. When Texas gets
it wrong and convicts an innocent person, the true
perpetrator remains free to commit more crimes.

as microscopic hair comparison or serology inclusion), testimony from informants or accomplices
with incentives to lie, false confessions and guilty
pleas, suppression of exculpatory evidence, ineffective assistance of counsel, and investigative and prosecutorial tunnel vision. The last two factors, while
undoubtedly at work, are difficult to pinpoint or
quantify. As such, we have not attempted to measure
specific instances of those problems in the remainder of the report, but briefly address them here.
There is no question that the most fundamental
and important protection against wrongful conviction is access to a qualified defense attorney. With
appropriate investigative and expert resources, defenders can meaningfully test the evidence against
their clients and argue an effective line of defense.
In practice, defenders frequently go without these
much-needed resources and may often lack training,
skills, and support—all factors that put innocent defendants at risk.
Tunnel vision refers to a normal psychological
tendency to seek information that fits a theory or belief and causes one to discount or ignore informaEyewitness
tion that does not fit within
misidentification is
that theory or belief. While
investigators and prosecuby far the leading
tors must eventually comfactor in wrongful
mit to a theory of who is
responsible for a crime, too
convictions in Texas.
often this commitment to a
theory is premature in the investigative process, and
important leads or information are either rejected or
simply ignored.

Wrongful Convictions
Threaten Public Safety

C

Factors Leading to
Wrongful Convictions

onsidering the devastating consequences of
even one of these injustices, it is clear that all
reasonable steps to prevent wrongful convictions
must be taken. The injustice endured by an innocent person whose most basic liberty is denied
cannot be overstated. The nightmare of an unjust
imprisonment ruins lives and destroys families. For
the innocent, prison is a terrifying ordeal few can
even imagine.
Beyond the personal cost to those wrongfully convicted and the millions of dollars spent re­

T

he thirty-nine DNA cases analyzed in this report
clearly indicate that eyewitness misidentification is by far the leading factor in wrongful convictions in Texas. The majority of these misidentifications occurred in either photo or live lineups. Other
factors include: false forensic testimony, reliance on
unreliable or limited forensic methodologies (such


w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

investigating cases and paying compensation, there a more plausible alternative suspect, but the judge
is another price we all pay—true perpetrators go un- rejected this line of defense. In May 2008, DNA testinvestigated and unpunished, putting public safety ing proved that Jerry Wayne Johnson had indeed
at risk. In each of these innocence cases, a criminal raped the victim of the crime for which Cole was
investigation into a serious violent crime was shut imprisoned.
down prematurely when authorities prosecuted the
Johnson, however, continued to commit horwrong person.
rible crimes. On July 4, 1985, Johnson abducted a
In most cases, it took many years to bring these man and woman after a party and raped the woman
wrongful convictions to light. Most of the crimes in a cotton field. While out on bond awaiting trial
remain unsolved because the leads dried up long for that rape, Johnson raped a fifteen-year old girl
ago. Fortunately, in some of these cases, the same at knifepoint. Additionally, Johnson was suspected
DNA results that freed the innocent also identified in the murder of insurance saleswoman Mary Louthe guilty. It is in those cases that we catch a glimpse ise Smith. Smith was found beaten with a blunt
of the full cost of a wrongful conviction. The crimes object and strangled. Though he was held on a two
that are committed in the time between a wrongful million dollar bond in that case, the investigation
conviction and the identification of
stalled and Johnson was never
the true perpetrator are an immeasurtried for Smith’s murder. He was
The crimes that are
able cost to the community.
convicted of the other two rapes,
committed in the time however, and is now serving a
The Innocence Project’s recent
review of DNA exonerations has
life sentence.
between a wrongful
identified ninety-one actual perpetraIt is impossible to know for
conviction and the
tors from 233 exoneration cases—apsure what would have happened
identification
proximately thirty-nine percent. The
if more careful eyewitness idenInnocence Project also estimates that
of the true perpetrator tification procedures were used
forty-nine rapes and nineteen murin the Texas Tech rape investiare an immeasurable
ders were committed by actual pergation. By presenting a photo
petrators following wrongful conviccost to the community. lineup in which Cole’s picture
tions. Although Texas has identified
stood out as different from the
actual perpetrators in about thirty-five percent of fillers, police undermined their ability to get reliDNA exoneration cases, Texas has been unable to able eyewitness evidence and mistakenly came to
prosecute some of those perpetrators because the believe they had their man. Had the investigation
statute of limitations has passed.
continued, Johnson might have been stopped beOne of the most troubling examples of the fore victimizing others. In September 2008, the
threat posed to public safety from wrongful con- victim of the crime for which Cole was convicted,
viction is found in the case of Timothy Cole, who Michele Mallin, spoke out publicly about her expetragically died in prison in 1999 prior to exculpa- rience and her hope that reforms are implemented
tory DNA testing. Cole was convicted of one of a to avoid such mistakes in the future.
string of five rapes that occurred in 1985 near the
The public was also put at risk following the
Texas Tech campus in Lubbock based largely on a wrongful conviction of Thomas McGowan, who
victim’s eyewitness identification. The photo lineup was convicted of rape and burglary in 1985 followpresented to the victim was highly suggestive. Cole’s ing a mistaken eyewitness identification. After DNA
picture stood out because the police used a color test results cleared McGowan of involvement in the
Polaroid photo of Cole while the other photos were crime in 2008, police ran the DNA evidence through
black and white mug shots.
a national DNA database and found a match in
Cole was only convicted of one of the Texas Tech Kenneth Wayne Woodson. They found that while
rapes, but because of the similarities in the crimes, Woodson went uninvestigated for the McGowan
the police suspected that Cole was guilty of them all. case, he committed another rape and burglary in
Consequently, the investigation into the Texas Tech 1986. He was sentenced to thirty years in prison for
rapes ended once Cole was behind bars.
that crime and was paroled in January 2006 after
During his trial, Cole’s defense lawyer tried to serving twenty years. He was then convicted of robargue that a man named Jerry Wayne Johnson was bing a bank in Richardson, his parole was revoked,

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

and he was sent back to serve prison time for the Nevertheless, the devastating costs of wrongful conbank robbery.
victions make it clear that Texas cannot afford to asWoodson confessed to the crimes for which sume that these cases are merely the cost of doing
McGowan had been wrongly convicted when he business in the criminal justice system. Texas must
was confronted with the DNA evidence. He cannot recognize that the system is flawed and can be imbe charged for those crimes, however, as the stat- proved by implementing simple procedural reforms
ute of limitations has passed. Had the investigation that greatly reduce the risk of error.
into the 1985 rape not been prematurely shut down
While the majority of this report focuses on the
by bad eyewitness evidence, police might have been wrongful convictions uncovered through DNA testable to prevent Woodson from committing other ing in Texas, they are only the tip of the iceberg. The
horrible crimes. The photo lineup that led to the advent of DNA technology has given our criminal
misidentification was suggesjustice system a tool that can protive and the mistake could have
While a faulty eyewitness vide incontrovertible evidence of
been avoided.
guilt—and innocence—in cases
identification led the
The statute of limitations
where the presence of biological
has also passed in another trouinvestigators to mistakenly evidence is dispositive. Unfortubling case in which incontronately, biological evidence is preszero in on Taylor,
vertible DNA evidence came
ent in only a fraction of criminal
too late. Although Ronald Taylor
Carroll victimized at least cases. While DNA is an invaluable
was convicted of the 1993 rape
tool, it does not solve the problems
two other women.
of unreliable evidence that repeatof a Houston woman, DNA tests
later revealed the identity of the
edly surface when wrongful contrue perpetrator, Roosevelt Carroll. While a faulty victions are discovered. The vast majority of cases
eyewitness identification led the investigators to mis- simply do not have probative DNA evidence.
takenly zero in on Taylor, Carroll victimized at least
In addition to the thirty-nine DNA exoneration
two other women. Carroll was convicted of two oth- cases in this report, there are scores of exonerations
er rapes and was serving a fifteen-year sentence when in Texas for which there is no DNA evidence. While
Taylor was finally exonerated. The victim of the 1993 a defendant is innocent until the prosecution proves
rape will never get justice.
guilt, after a conviction occurs, the burden shifts to
Though these investigations and prosecutions the defendant to prove innocence. New evidence
were done in good faith, their accuracy was under- that merely casts doubt on the conviction is not
mined by unreliable evidence. If such tragic and nearly enough to overturn a conviction—which is
consequential errors in evidence were unavoidable, why DNA evidence, where it exists, is so successful
then one might be forced to accept them as “the cost in exonerating the innocent. Without DNA evidence,
of doing business.” However, the reality is that many inmates face an almost insurmountable challenge to
of the errors follow patterns that are predictable establish their innocence conclusively. Still, a surprisand preventable with the right kind of safeguards in ing number have been cleared against the odds.
place. Until those safeguards have been implementIn one high-profile set of non-DNA wronged, Texas faces the risk of more investigations end- ful convictions, dozens of people were framed in
ing prematurely while the true perpetrator remains a drug sting by a rogue undercover officer. In that
at large to commit additional crimes.
1999 case, forty-six residents of the small town of
Tulia, Texas were arrested for selling cocaine based
solely on the word of one undercover officer. After
it was discovered that the undercover officer had
fabricated all the evidence in the cases, a collective
lawsuit brought by the wrongly convicted was settled for six million dollars.
n a state with thousands of criminal convictions
Perhaps one of the most well-known Texas
every year, the number of documented wrongful wrongful convictions was that of Randall Dale Adconvictions represents only a tiny fraction. There ams. Adams was convicted of the murder of a Dallas
can be no doubt that the system usually gets it right. police officer and sentenced to death. His case was

The Limits of
DNA Evidence

I


w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

the subject of the documentary film The Thin Blue have emerged from Dallas County. The newspaper
Line. The true killer, David Harris, subsequently reviewed numerous cases and documented, for exconfessed on tape. After twelve years in prison, new ample, that eyewitness evidence is routinely relied
evidence of innocence came to light through Ad- upon in robberies, where it is rare to have DNA
ams’ appeals and he was eventually released.
evidence, even with little or no corroboration of the
Another non-DNA exoneration occurred in eyewitness evidence.
the case of Clarence Brandley, a janitor who was
It is no accident that the vast majority of DNA
wrongly convicted and sentenced to death in Mont- exonerations involve sexual assault cases where biogomery County, Texas for the rape and murder of logical evidence is often present with the potential
Cheryl Ferguson in 1981. Brandley’s conviction was to clearly indicate guilt. In these cases, DNA testing
overturned in 1989 after new evidence showed that reveals errors and weaknesses in many types of eviauthorities had suppressed exculpatory evidence. In dence and procedures. Our criminal justice system
addition, an investigation by the FBI and the U.S. relies on these same kinds of evidence and proceDepartment of Justice found additional misconduct dures in cases where DNA evidence is not available,
by the authorities in the case. All charges against putting additional innocent suspects at risk. DNA
Brandley were eventually dropped.
exonerations are but a window to the larger, unseen
In 2004, Ernest Willis was released after serv- problem. We know that the same evidence suffers
ing almost seventeen years on Death Row for the the same flaws in non-DNA cases. What we do not
arson murder of two women. Willis
had been living in the house where the
Texas DNA Exoneration Facts
fire occurred. Willis escaped uninjured
and investigators believed that he had
Texas has had more wrongful convictions exposed by DNA
intentionally set the fire. Years later, the
than any other state in the country.
Pecos County District Attorney hired
an independent arson expert who conCollectively, these thirty-nine men have spent more than
548 years in prison with an average of fourteen years.
cluded that the evidence used against
Willis was without any scientific basis,
Over $17 million dollars have been paid by state and local
and there was no evidence of arson.
governments in civil settlements and statutory compensation
The prosecutor eventually apologized
to those wrongfully convicted.
“for so many lost years” and said that
Willis “simply did not do the crime.”
Twelve counties in Texas have uncovered wrongful convictions
through DNA evidence.
In each of these cases—only a few
among many convictions overturned
Dallas County leads the state in the number of wrongful
in Texas without the benefit of DNA
convictions, a direct result of preserving DNA evidence while
evidence—flawed evidence was fortuother counties destroyed it.
itously exposed. In some cases, evidence
Nine people have been released from Texas’ death row based
of guilt evaporated under greater scruon evidence of their innocence.
tiny; in others, affirmative evidence of
innocence was found. Yet in all of these
85% of the cases involve mistaken eyewitness identification.
cases it was exceedingly difficult to undo
the damage of flawed evidence in the
18% of the cases involve false forensic testimony.
courtroom and to rectify the wrongful
28% of the cases involve the use of unreliable or limited
conviction.
forensic methodologies (e.g., microscopic hair comparison,
A recent review of evidence used
serology inclusion, bite mark matches, voiceprint analysis).
in non-DNA cases reinforces the importance of implementing procedures
13% of the cases involve informant or accomplice testimony
that enhance the quality of evidence
from witnesses with incentives to lie.
relied upon by the system. In October
13% of the cases involve false confessions or guilty pleas.
2008, the Dallas Morning News published an investigative series on the
18% of the cases involve suppression of exculpatory
alarming number of exonerations that
evidence or other misconduct.

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

know is how many innocent individuals have been
convicted based on faulty evidence.

makes me sick to my stomach that those two men
were destroyed.”
The wrongful convictions of Ochoa and Danziger are a troubling reminder of the importance
of giving jurors all the evidence they need in order to reach justice. When custodial interrogations
are not recorded, jurors miss essential information
they need to effectively and accurately evaluate the
reliability of a confession. Had the evidence been
thoroughly documented by recording the interrogation, and jurors given the full story, they might
have avoided the injustice suffered by the two men,
not to mention the millions of dollars in settlements
the taxpayers had to pay.

Jurors Deserve the
Most Reliable Evidence

I

n the criminal justice system, jurors determine
guilt or innocence by evaluating the evidence presented to them. If that evidence is incomplete, biased,
or incorrect and the innocent are wrongly convicted,
jurors are placed in a terrible situation that benefits
no one involved. As one juror from the Richard Danziger case stated, “I think any one of us resents the
position we were put in. You know, we were made a
party to ruining [Danziger’s] life.”
The only direct evidence in the case linking
Danziger to the murder of Nancy DePriest in Austin was the confession of his
friend Christopher Ochoa—a
“I think any one of us
confession that DNA evidence
resents the position we
later revealed to be false. The
jurors were not given the full
were put in. You know,
story of how that confession
we were made a party
was elicited in a threatening
and coercive interrogation that
to ruining [Danziger’s]
lasted more than twenty hours.
life.”—a danziger juror
Without that context, the confession and testimony from
Ochoa led jurors to believe that there was no decision
other than to find Danziger guilty.
Another juror from the case said, “I think we
should hear how that testimony came about because
we had no choice with what we heard. . . . And I was
uneasy about the verdict, but if you believed what
you were hearing, and it was hard not to, you had to
find him guilty.” Another juror agreed and stated, “It
would have made a difference to me if I had known
that [Ochoa] had sat there for nineteen hours before
he was allowed to go home only if he confessed.” It
took the confession of the true killer and DNA confirmation years later to finally reach the truth.
One of the Danziger jurors said when she
learned of the true killer’s confession and the DNA
evidence, “That’s when I really learned that you
can make somebody say something or admit to
something that’s just not true.” Knowing that they
sentenced Danziger to prison based on a false confession has left these jurors with anger and guilt.
One juror stated, “I feel guilty for what I’ve done. It

Reform: Improving the
Reliability of Evidence

B

y implementing a reform agenda, Texas can
ensure that the best evidence possible reaches
courtrooms and can enhance the fairness and accuracy of its criminal justice system.

Eyewitness Identification

E

yewitness identification testimony is one of
the most widespread and powerful forms of
evidence in our criminal justice system. Much like
trace physical evidence, however, eyewitness evidence is highly susceptible to contamination if it is
not collected carefully according to scientific protocols. In Texas, eighty-five percent of all wrongful
convictions exposed by DNA testing have involved
incorrect eyewitness identifications.
The criminal justice system cannot do without eyewitness evidence. Though DNA exonerations have highlighted its inherent flaws, the good
news is that extensive research conducted over
decades has identified ways to minimize the risk
of mistakes.
Texas currently has no statutory standards regarding the conduct of photo or live lineups. Furthermore, though scientifically-grounded best
practices for conducting identification procedures
have been around for more than a decade, a recent
survey by The Justice Project of over 1000 Texas
police and sheriff’s departments yielded an even
more troubling finding: eighty-eight percent of law
enforcement agencies have no written policies or


w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

procedures for the conduct of photo or live lineups.
lineup. Extensive research has demonstrated that
The few policies that do exist are more often than
cautionary instructions reduce incorrect identifinot vague and incomplete. Only a tiny fraction of
cations with no decrease in correct picks.
the departments have implemented best practices
for eyewitness procedures recommended by the • Fair lineup composition: Effective selection of
U.S. Department of Justice, the International Assofillers when composing a lineup can help reduce
ciation of Chiefs of Police, the American Bar Assothe risk of identifying an innocent suspect. Only
ciation, and others.
one suspect should appear in each lineup and at
Texas must require law enforcement agencies to
least five fillers should be included. Rather than
adopt written policies and procedures for the conduct
selecting fillers based on their resemblance to the
of photo and live lineups
suspect, which makes the witness’s task
that implement the folmore complicated, fillers should be seEighty-eight percent of
lowing key safeguards:
lected to resemble the witness’s descripTexas law enforcement
tion of the perpetrator. Most importantly,
suspect or the suspect’s photo should
• C
 omplete documentaagencies have no written the
tion of identification
not unduly stand out and should be prepolicies or procedures
procedures: Given the
sented in a uniform format to that of the
overwhelming imporfillers. Fair composition of photo and live
for the conduct of
tance of eyewitness teslineups allows authorities to judge the rephoto or live lineups.
timony and the weight
liability of an eyewitness effectively.
afforded to it by juries,
it is essential to fully document photo or live line- • Neutral blind lineup administration: The person
ups. Thorough documentation helps a jury to
who administers the photo or live lineup to a witassess the eyewitness evidence appropriately and
ness should not know the identity of the suspect.
minimizes the effects of reinforcing feedback that
The purpose of keeping the administrator “blind”
can distort the confidence level of an eyewitness
as to which person in the lineup is the suspect is
between the time of the identification and the trial.
to prevent the administrator from unintentionally
Documentation of an eyewitness identificainfluencing the results. This is generally done inadtion procedure must include the photos used in a
vertently through verbal or
photo lineup or a photograph of a live lineup and
non-verbal behavior. WitRegardless of whether
all dialogue and witness statements made during
nesses may be very motivatthe true perpetrator
the procedure. When an identification is made, it
ed to make an identification
is essential to have documentation of the witness’s
and seek to interpret the beis in a lineup, an
degree of confidence in the identification, in the
havior of the lineup admineyewitness may feel
witness’s own words and prior to any feedback
istrator for cues about the
from authorities. It is important to fully document
suspect, even if no such cues
pressure to make an
all procedures, even those that do not result in an
exist. Finally, a double-blind
identification.
identification. Electronic recording of photo and
protocol also eliminates the
live lineups provides the most complete record of
problem of investigators
these critically important investigative procedures.
interpreting ambiguous witness comments and
other behavior through the lens of their theory of
a suspect’s guilt.
• Cautionary instructions: Regardless of whether
the true perpetrator is in a lineup, an eyewitness may feel pressure to make an identification. • A
 voidance of repeated exposure of the suspect
Prior to presenting the lineup, the eyewitnesses
to a witness: Police departments must adopt
should be instructed that the perpetrator may or
policies that address the inherent risk in repeatmay not be included in the lineup and that they
edly presenting a witness with a suspect or a
should not feel compelled to make an identificasuspect’s photograph. In some exonerations, a
tion. Cautionary instructions of this sort remove
witness did not identify a suspect in an initial
some of the pressure on the eyewitness to choose
lineup but subsequently identified him in a later
a suspect when the culprit may not be in the
lineup (in which the only common person was

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

the one wrongfully convicted). Because of the
fragile nature of witness memory, particularly
the possibility of “memory transference” between events, multiple exposures undermine the
reliability of an identification.

While documented false confessions indicate
the need for safeguards, powerful benefits to law
enforcement have also made recording very popular with the police who do it. A survey of police departments that record interrogations conducted by
former United States Attorney Thomas P. Sullivan
• Sequential presentation option: In addition to found overwhelming support for electronic recordthe above measures, police departments should ing of interrogations. Among the benefits cited were
consider implementing sequential presentation the protections against false claims of coercion or
of lineups. Traditionally, eyewitnesses are shown misconduct, the ability to concentrate on the susa photo or live lineup in which the
pect and his demeanor rather
lineup members are presented as
than taking notes, and the useJuries will sometimes
a group. An eyewitness viewing a
fulness of recorded interrogaconvict an individual
lineup tends to make a judgment
tions for training officers. With
about which individual looks most
electronic recording, motions
based on a confession
like the perpetrator relative to the
to suppress confessions are
alone, so special care
other members of the lineup. This
reduced, and the “he said-she
natural tendency toward “comparimust be taken to ensure said” swearing matches about
son shopping” is problematic when
what took place in the interthat the suspect’s
the suspect in the lineup is not in
rogation room are essentially
fact the perpetrator. Presenting the
statements are as reliable eliminated. The result is that
photos or lineup members one at
judges and juries have the comas possible.
a time discourages the tendency to
plete story that allows them to
judge the lineup members against
effectively weigh the evidence,
each other and to make an identification through convict the guilty, and protect the innocent.
a process of elimination, in favor of a more direct
Although Texas currently requires that oral
comparison of each person to the witness’s memory. confessions be recorded to be admissible in court,
Many studies indicate that sequential presenta- there is no provision stating that the interrogation
tion reduces error, although some researchers be- preceding an oral confession must be recorded.
lieve that the superiority of sequential presenta- Further, authorities overwhelmingly rely on writtion has not been established. Texas departments ten statements signed by the suspect, which have no
may want to consider the sequential option as requirement for electronic recording at all. In any
more field-testing data is accumulated.
case, Texas currently does not require that interrogations that lead to confessions be recorded.
Texas should require the electronic recording of
full custodial interrogations in serious crimes with
the following considerations in mind:

Electronic Recording of
Custodial Interrogations

• “ Stem to stern” recording: In order to reap the
benefits that electronic recording affords police,
prosecutors, innocent suspects, and the system
as a whole, the entire custodial interrogation
must be recorded—not merely the confession.
Recording should begin at and include the delivery of the suspect’s Miranda rights and continue
uninterrupted until the end of the interview.
Implementing this requirement guarantees that
the best, most complete evidence will be available at trial. Questions as to whether Miranda
warnings were given and false claims of abuse
or coercion will be avoided, resulting in fewer

T

he idea that someone would falsely confess to
a serious crime seems counterintuitive to most
people. False confessions are a well-documented
reality, particularly among more vulnerable groups
such as juveniles and the mentally disabled, and
they have resulted in wrongful convictions in Texas
and across the country. Confessions are so powerful, in fact, that they can even overcome other exculpatory evidence. Juries will sometimes convict
an individual based on a confession alone, so special care must be taken to ensure that the suspect’s
statements are as reliable as possible.


w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

motions to suppress confessions. Recording may
also encourage guilty suspects to enter into plea
bargains rather than going to trial.
Innocent suspects will likewise be protected
from wrongful conviction by providing courts
with the information necessary to accurately assess whether a defendant’s statement is reliable
and voluntary.

fact-finding tool for the criminal justice system.
The virtue of electronic recording of custodial interrogations lies not only in its ability to help guard
against false confessions, but also in its ability to
develop the strongest evidence possible to help
convict the guilty.

Jailhouse Informant Testimony

B

• A
 udio or video: While video recording devices
are preferable, some departments have expressed
concern about the costs of implementation. Audio recording is an acceptable alternative that can
be implemented at very low cost. It should be left
to the discretion of the agency to choose the system that best fits its needs and resources.

ecause jailhouse informants are so desperate to
attain sentence reductions and other benefits,
informant testimony is widely regarded as the least
reliable testimony in the criminal justice system.
When the state offers a benefit in exchange for testimony, whether that benefit is explicit or implied,
the incentive for incarcerated
individuals to fabricate eviWhen the state offers
dence dramatically increases.
a benefit in exchange
Some informants may fabricate testimony in an effort to
for testimony, whether
curry favor with prosecutors
that benefit is explicit or
apart from any promise or implied benefit.
implied, the incentive for
The protections currently
incarcerated individuals
in place have proven inadequate to safeguard against
to fabricate evidence
unreliable testimony by witdramatically increases.
nesses with powerful incentives to lie. Remarkably, the
use of jailhouse informant testimony continues to
be largely unregulated by state legislatures or courts
despite frequent, documented cases of injustice and
instances of wanton abuse. Texas is no exception as
no statutes regulating the use of jailhouse informant
testimony exist.
About fifteen percent of all DNA exonerations
nationally included jailhouse informant testimony.
Texas must implement safeguards designed to subject this testimony to more transparency and higher
scrutiny. Specifically, Texas should require the following reforms:

• S cope of recording: Recording in all criminal cases
promises the most benefits, but at a minimum, recording interrogations conducted in connection
with felony investigations should be required. It is
especially urgent to record interrogations involving juvenile suspects and those whom authorities
have reason to believe are mentally disabled or
mentally ill.
• E
 xceptions: Recording requirements must include
reasonable exceptions so as not to place an undue
burden on law enforcement and to allow for the
admission of voluntary statements that were not
recorded for valid
reasons. For example,
False confessions are a
a suspect’s statement
well-documented reality, should be admisif officers made
particularly among more sible
a good faith effort to
vulnerable groups such
record but were unable to do so because
as juveniles and the
of equipment malmentally disabled.
function or power
outage. Additionally,
spontaneous statements made by the defendant,
or statements made during routine processing of
the defendant, may be admissible in court because
they were made outside the context of an interrogation. Statements made by a suspect who refuses
to speak if recorded might also be deemed admissible as long as the refusal itself is recorded.

• W
 ritten pretrial disclosures: Texas should require written pretrial disclosure of all inducements a jailhouse informant may have been given
or promised in exchange for testimony, including
pay, immunity from prosecution, leniency in prosecution, or other personal advantage, along with
the proffered testimony the prosecution seeks to
present. In addition, any past cooperation agreements of the informant should be disclosed along

Electronic recording of custodial interrogations has emerged as a powerful innovation and


w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

HPD lab. The establishment of the Texas Forensic
Science Commission provides an independent body
to review allegations of forensic negligence and misconduct and to recommend corrective action.
The statutory tasks of the Forensic Science Com• P
 retrial reliability hearings: Texas should follow mission are essentially backward-looking. While the
Illinois in providing for pretrial reliability hearings commission is able to make recommendations for
in cases where the prosecution
remedial action regarding comintends to use informant testiplaints of negligence and misconClearly, a lack of oversight duct, it is primarily oriented to be
mony. In a pretrial reliability
hearing, the court performs a
a reactive entity. It is essential to
of forensic labs in Texas
“gatekeeper” function where
establish a forensic oversight syshad devastating
it should be required to detertem that is more proactive in setmine that the informant’s testing quality standards in order to
consequences on the
timony is sufficiently reliable
ensure the best evidence possible.
accuracy of the criminal
to be presented to a jury.
While DNA tends to get the
attention of the media and policyjustice system—to
makers, forensic labs are engaged
• Require corroboration: Texas
date, three wrongful
should adopt corroboration
in a variety of sub-disciplines
requirements for informant
beyond DNA, and the need for
convictions have been
testimony to mitigate the inoversight and quality standards
traced to the HPD lab.
herent risks presented by witin those areas is great. The proacnesses with incentives to lie.
tive quality assurance role needed
in Texas must address all aspects of forensic science
• C
 autionary jury instructions: Texas should adopt that are relied upon in criminal trials—not just
cautionary jury instructions in all cases where the DNA evidence. The following changes would help
testimony of a jailhouse informant is used.
to implement those goals:
with other information bearing on the informant’s
credibility. Timely disclosure of this information
ensures that defendants can conduct meaningful
investigation and cross-examination.

Greater scrutiny and transparency of jailhouse
informant testimony in Texas will prevent unreliable evidence from being used in the courts and will
produce more reliable outcomes in criminal cases.

Forensic Oversight

I

naccurate forensic evidence and testimony is the
second leading cause of wrongful convictions
exposed by DNA in Texas. The Houston Police
Department’s (HPD) crime lab debacle is a striking
example of how poorly a forensics lab can operate
without proper oversight. Independent investigator
Michael Bromwich found that analysts at the HPD
crime lab repeatedly tested DNA samples incorrectly and, in some cases, made up results without actually testing evidence. Clearly, a lack of oversight of
forensic labs in Texas had devastating consequences
on the accuracy of the criminal justice system—to
date, three wrongful convictions have been traced
to the HPD lab.
Fortunately, Texas has taken important steps
forward since the disclosure of the problems in the

• T
 exas should give the Forensic Science Commission a proactive role in reviewing, setting, and enforcing quality standards: These expanded duties
should include a review of existing private accreditation program requirements and the development
of proposals for supplementing those requirements
as appropriate to best ensure objectivity and reliability. Commission staff and budget should be
augmented to accommodate these functions.
• Independence and blind testing: Texas should
develop and require all forensic science laboratories to adopt structures and policies to prevent
bias in testing and analysis, such as regulating the
amount of extraneous contextual information
an analyst receives prior to testing to reduce the
risk of confirmation bias or other observer effects. Forensic labs should also move toward becoming independent from law enforcement and
prosecutorial agencies to best insulate analysts
from the risk of “group think” that occurs from
working closely with police as part of a crimesolving “team.”
10

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Recent studies have demonstrated the risk of inad- ery to occur in advance of trial, often leaving devertent bias affecting the outcome of forensic testing. fense counsel without adequate time to review the
One 2006 study in the Journal of Forensic Identification materials and prepare. While some jurisdictions
asked experienced analysts to evaluate a series of fin- have voluntarily adopted more expansive discovery
gerprints to determine if they matched. These analysts practices, the lack of statewide standards means too
believed they were examining prints for an open, many Texans are being tried without a fair opporunsolved case, but they were in fact re-examining tunity to review the evidence the state wants to use
prints that they had previously
against them in court.
analyzed accurately. The prints
The adoption of open-file
While some jurisdictions
were given to the analysts along
discovery rules for criminal trials
have voluntarily adopted
with artificial contextual inforcreates a more level playing field
mation, such as the fact that the
ensuring that evidence can be
more expansive discovery by
suspect had confessed. In cases
meaningfully challenged and testpractices, the lack of
where analysts were given coned, and by removing much of the
textual information about the
uncertainty inherent in the discrestatewide standards
fingerprints, they were wrong
tionary disclosure decisions prosmeans too many Texans
in almost seventeen percent of
ecutors now make.
the cases. This study highlights
The record of wrongful conare being tried without
the need to ensure that extravictions has demonstrated that
a fair opportunity to
neous contextual information
exculpatory evidence can be withdoes not undermine the objecheld for years, even decades, while
review the evidence
tivity of analysts. By ensuring
an innocent person sits in prison.
the state wants to use
that labs are independent of law
Whether the state fails to disclose
enforcement and prosecutorial
evidence inadvertently or intenagainst them in court.
agencies, along with regulattionally, clear rules about what is
ing the flow of information
subject to discovery—and clear
between investigators and forensic analysts, these consequences for failure to disclose discoverable
kinds of errors can be minimized.
information—minimize the risk of these mistakes.

Texas should implement the following changes:

Expanded Discovery

•R
 equire an open-file discovery policy to allow access to all relevant, unprivileged information in the
possession, custody, or control of the state, subject to
appropriate regulation by the courts.

D

iscovery is the process by which the prosecution discloses evidence to the defense attorney and provides a basic foundation for accuracy
and fairness. Discovery preserves the integrity of
our adversarial criminal justice system by providing the defense with a meaningful opportunity to
fully investigate and test the government’s evidence
in preparation for trial. Unfortunately in Texas, the
statutes and rules governing criminal discovery are
so minimal that they fail to guarantee the opportunity for evidence to be fully investigated and meaningfully challenged.
Unlike many other states, Texas has no statute
that mandates automatic discovery of key case documents, such as police reports and witness statements. Instead, Texas requires the defense to file
motions with the court requesting access to basic
information. The defense must also demonstrate
“good cause” in order to review such evidence.
There are no clearly defined timelines for discov-

•R
 equire automatic disclosure of key documents
such as police reports and witness statements prior
to trial.
•P
 rovide for adequate timelines to ensure access to
information to allow investigation and review.
•E
 nforce remedies in cases where discoverable
material is willfully suppressed, or when discovery
obligations are not or only partially met.
The Justice Project has published comprehensive
policy reviews on the above reforms in addition to a
policy review on post-conviction DNA testing. The
policy reviews can be downloaded from The Justice
Project’s website at www.thejusticeproject.org.
11

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Stories of Injustice: The Texas DNA Exonerated

D

espite the great injustices they have suffered,
the thirty-nine men whose stories are described
below are actually the lucky ones. Unlike most cases
in the criminal justice system, each of these cases
involve DNA evidence, which can expose the guilty
and reveal the innocent. It is fortuitous that DNA
was preserved and available for testing. Without
DNA evidence, these profiles of injustices likely
would never have been exposed, and the systemic
problems that caused these wrongful convictions
may never have been discovered. These thirty-nine
cases highlight the need for reform as it is impossible to know how many similar mistakes have been
made in cases without DNA evidence to expose
them. Texas must address these flaws and implement the reforms that can prevent them to ensure
that no innocent person suffers as these thirty-nine
men have suffered.

because she said her face was covered with a pillow
during the attack. Although the defense questioned
the strength of her eyewitness identification, the
jury found Alejandro guilty.
Because of the false testimony of forensic expert Fred Zain and the mistaken eyewitness identification, Gilbert Alejandro spent almost four years
in prison for a crime he did not commit.

Michael Blair

M

ichael Blair became the prime suspect in the
1993 murder of seven-year-old Ashley Estell
of Plano when two crime scene investigators noticed
him near the scene and thought he looked suspicious. Police interrogated Blair for over nine hours.
Throughout the interrogation, Blair maintained his
innocence and was released. An eyewitness came
forward stating that he saw Blair in the park where
the girl was abducted.
At trial, a forensic scien- Michael Blair
tist testified that strands of
spent fourteen
hair found at the crime scene
on the day of the girl’s abduc- years on death
tion appeared to match hairs row due to faulty
taken from the victim and forensic science
Blair. In addition, the forenand a mistaken
sic expert testified that hairs
found in Blair’s car closely eyewitness
resembled hairs taken from identification.
the victim, and he could not
“tell the difference, microscopically” between fibers
taken from Ashley’s body and fibers found in Blair’s
car. An FBI analyst also testified to similarities in the
chemical composition of the fibers. The jury took
only ninety minutes to convict Blair and sentence
him to death.
Post-conviction DNA testing of the hair samples in 1998, 2000, and 2002 revealed that none of
the hairs came from either the victim or Blair. In
2006, DNA testing showed that material taken from
under the victim’s fingernails did not come from
Blair. Blair’s conviction was thrown out in 2008.
Because of faulty forensic evidence and a mistaken eyewitness identification, Michael Blair spent
fourteen years on Texas’ death row for a crime he
did not commit.

Gilbert Alejandro

G

ilbert Alejandro was wrongfully convicted
of sexual assault in 1990 based largely on the
faulty testimony of forensic expert Fred Zain. At
trial, Zain testified that DNA tests established that
semen found on the victim’s clothes “could only
have originated from [Alejandro].” A reexamination of the original DNA testing years later showed
that Alejandro was actually excluded as the perpetrator. In fact, records showed that Zain’s testimony
was based on DNA testing that had
not yet been completed at the time
Gilbert Alejandro
he testified against Alejandro.
spent almost four
Inaccurate eyewitness testimoyears in prison due
ny also contributed to Alejandro’s
wrongful conviction. He initially
to false testimony
became a suspect when the victim
about DNA and a
identified him from a book of mug
mistaken eyewitness shots. After the initial identificaidentification.
tion, police then showed the victim
several photo lineups that included
Alejandro, but she failed to identify him. Regardless, police still chose to focus on Alejandro because
of the victim’s initial identification from the mug
book. The victim described her attacker as Hispanic, about six feet tall, and wearing a white cap.
She could not provide a more detailed description
12

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

A.B. Butler

the time was not advanced enough to exclude him
as the perpetrator.
Harris County prosecutors used the victim’s identification as the centerpiece of their case against Byrd.
He was convicted and sentenced to life in prison. Not
everyone was convinced of Byrd’s guilt, though. The
trial judge sent a letter to the chief of police complaining about the investigation. In addition, although
evidence was routinely destroyed in Harris County, a
deputy district clerk happened to save the trial exhibit
containing the biological evidence in Byrd’s case.
Post-conviction DNA testing provided irrefutable proof that Byrd was not the attacker. Byrd was
released and eventually granted an official pardon
from Governor George W. Bush in October 1997.
Because police and prosecutors ignored contradictions in a mistaken eyewitness’s testimony, Kevin
Byrd spent twelve years in prison for a crime he did
not commit.

n 1983, a mistaken eyewitness identification led
to the wrongful conviction of A.B. Butler for
the aggravated kidnapping of a woman in Smith
County. The victim initially identified Butler as her
attacker from a book of mug shots. She also identified him in a live lineup and again during the trial.
The defense presented three
alibi witnesses who testified
Butler was with them during
the attack. However, Butler
was convicted of aggravated
kidnapping and rape and
was sentenced to ninety-nine
years in prison.
After learning about
emerging DNA technology,
Butler repeatedly petitioned
A.B. Butler
for DNA testing of evidence
spent seventeen found at the crime scene.
In 1999, the first set of tests
years in prison
were conducted by a private
due to one
lab and the results were inmistaken
conclusive. A second set of
eyewitness
tests were conducted by a lab
in New York using a newly
identification.
developed technique that allowed scientists to better isolate male DNA and provide more sophisticated analysis. These tests proved
conclusively that Butler’s DNA did not match DNA
found on the victim. Butler was released in January
2000 and was granted an official pardon from Governor George W. Bush.
Because of one mistaken eyewitness identification, A.B. Butler spent seventeen years in prison for
a crime he did not commit.

Charles Chatman

Dallas Morning News

Clay Graham

I

I

n 1981, police placed Charles Chatman’s picture
in a photo lineup after a woman in his Dallas
County neighborhood was sexually assaulted. The
victim told police that she remembered her attacker
as a neighbor. That same day she identified Chatman
from a photo lineup and a live lineup. The victim did
not mention Chatman’s most distinguishing characteristic in her initial description of the perpetrator—
Chatman was missing his front teeth.
At trial, the state’s case consisted of little more
than the victim’s previous identifications and her
in-court identification of Chatman. The state also
presented forensic results showing the assailant’s
blood type characteristics were consistent with

Kevin Byrd

W

rongfully identified as the perpetrator of a
violent rape in Harris County in 1985, seventeen-year-old Kevin Byrd did not match the description of the attacker that the victim originally
gave to police. The victim initially described the
perpetrator as a thirty-five-year-old white male
with an unusual “honey brown” complexion. Byrd
was arrested even though he was significantly
younger than the description and more importantly, he is black. Byrd voluntarily provided blood,
saliva, and hair samples to police, but the science at
13

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

those of Chatman. These characteristics were also
consistent with forty percent of all black males.
Chatman’s defense presented alibi testimony that he
was at work, employed as a custodian, during the
attack. Chatman’s sister, who also worked at the custodial service, corroborated his alibi. Nevertheless,
Chatman was convicted of aggravated sexual assault
and sentenced to ninety-nine years in prison.
Chatman applied for post-conviction DNA
testing. After two inconclusive results, a more advanced DNA test proved that he did not commit the
rape. Chatman was released from prison on January 3, 2008, and a formal exoneration followed on
February 26, 2008. Chatman missed three chances
at parole because he refused to admit to the crime
or apologize for it.
Because of one mistaken eyewitness identification, Charles Chatman spent twenty-seven years in
prison for a crime he did not commit.

of these attempted confessions until years later. Cole
died in prison in 1999 from a severe asthma attack.
DNA testing conducted after Cole’s death revealed that Johnson had indeed raped the Texas
Tech student, not Cole. In an unprecedented hearing
in February 2009, Johnson took the stand and confessed to the crime for which Cole had been convicted. As a result, Cole was exonerated and his criminal
record was ordered expunged. Michele Mallin, the
victim who misidentified Cole, came forward and
spoke out publicly in support of eyewitness identification reform.
Because of one mistaken eyewitness identification, Timothy Cole spent thirteen years in prison
for a crime he did not commit.

Roy Criner

F

our years after the 1986 Montgomery County
sexual assault and murder of Deanna Ogg, Roy
Criner became a suspect in the case. Three men alleged that Criner made statements in which he referred to a hitchhiker with
whom he had he had sex. Roy Criner
There were numerous incon- spent ten years
sistencies among the stories
of the three men and with in prison due
to false witness
known facts in the case.
At trial, a forensic analyst testimony.
with the Texas Department
of Public Safety testified that numerous hairs collected from the crime scene and from Criner’s truck
could not link Criner to the crime. Serology tests on
semen from the crime scene did not yield probative
results either. Further, several alibi witnesses testified
that Criner was at work at the time of the crime. Despite the scant and questionable evidence tying him
to the crime, Criner was convicted of murder and
sentenced to ninety-nine years.
In 1997, DNA testing excluded Criner as the
source of the semen found on the victim, but his
ordeal did not end there. The prosecution said the
semen found on the victim was the result of consensual sex the victim had before the rape, and
they suggested Criner might have used a condom
or simply did not deposit semen during the rape.
As such, Criner was denied a new trial. Additional
DNA testing on a cigarette butt found near the victim matched that from the semen, undermining the
prosecution’s theory that the semen did not belong

Timothy Brian Cole

T

imothy Cole became the prime suspect in a series of rapes on the Texas Tech campus in 1985
after he engaged in small talk with an undercover
female officer sent to the area where the
attacks occurred. The Lubbock police
placed the undercover officer in the area
with the hope of luring out the “Texas
Tech rapist.”
The last victim in the Texas Tech rapes
was shown a photo lineup that included a
picture of Cole. The victim identified Cole
as her attacker. Cole’s picture was the only
color Polaroid photo in a group of five
black and white mug shots, and the
victim later stated that police and
Timothy Cole
prosecutors repeatedly described
was wrongfully
Cole as a “low-life hood.” At a subconvicted based on
sequent live lineup, four other Texas
Tech rape victims failed to identify
a false eyewitness
Cole as their attacker. Only the final
identification. He
victim of the most recent rape statdied in prison after
ed she had no doubt about Cole.
serving thirteen years. Cole was convicted and sentenced
to twenty-five years in prison.
Years later, Jerry Wayne Johnson, the defense
team’s prime suspect, made two attempts to confess to
the crime that placed Cole behind bars, but they went
ignored. Cole and his family did not know about either
14

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

to the perpetrator. Based on these findings, Criner
was released and he was granted an official pardon
by Governor George W. Bush on August 15, 2000.
Because of faulty witness testimony, Roy Criner spent ten years in prison for a crime he did not
commit.

Because of a suggestive photo lineup that led to
a mistaken eyewitness identification, Wiley Fountain spent sixteen years in prison for a crime he did
not commit.

Wiley Fountain

A

Larry Fuller

Dallas Morning News

fter a Dallas woman was attacked and raped in
her home in 1981, Larry Fuller’s picture was
included in a photo lineup, even though he had no
history of sex crimes. Fuller was a decorated Vietnam
veteran raising two young children. The victim failed
to conclusively identify Fuller as her attacker, and the
investigating officer issued a report recommending
that the investigation be suspended because the victim “was unsure of the suspect at this time.” Months
later, police showed the victim another photo
lineup. Fuller’s photograph was again included in
the photo lineup and was the only picture that was
included in both lineups. This
time the victim identified Fuller
as her attacker.
The victim initially stated
that she did not remember any
facial hair on her attacker. However, the photo of Fuller she
identified showed him with a
heavy and distinct beard. At trial, the prosecution relied heavily
on the eyewitness identification,
saying that the victim had “never
wavered” in her identification.
Larry Fuller spent
The prosecution also relied
on misleading forensic testimoalmost twenty years in
ny to convict Fuller. A serologist
prison as the result of a
testified that Fuller was includmistaken identification
ed within twenty percent of the
and misleading
population that shared blood
type characteristics of the atforensic testimony.
tacker. Because the victim’s own
blood type was not properly considered, the evidentiary value of that test was greatly overstated. Fuller
was convicted of aggravated rape and sentenced to
fifty years in prison.
An initial DNA test in 2000 was unable to obtain a profile, but a more advanced DNA test conducted in November of 2004 conclusively excluded
Fuller as the attacker. Fuller was released on October
31, 2006 and granted an official pardon by Governor Rick Perry on January 25, 2007.
Because of unreliable eyewitness and forensic

Wiley Fountain
spent sixteen
years in prison
as a result of
a mistaken
identification
from a photo
lineup.

W

Clay Graham

iley Fountain’s picture was included in
a photo lineup and shown
to the victim of a sexual
assault in Dallas. In that
lineup, Fountain was the
only man wearing a dark
baseball cap and a warmup suit, the clothing worn
by the attacker as described
by the victim. The victim
picked Fountain as her attacker and police closed
the case the following day. Even though semen was
recovered from the victim’s clothing, samples were
too small for serological analysis. Based entirely on
the eyewitness identification of the victim, Fountain was convicted of aggravated sexual assault and
sentenced to forty years in prison.
After spending fifteen years in prison, Fountain
was paroled and required to register as a sex offender. Unable to find a job, his sex offender registration
fees went unpaid and Fountain’s parole was revoked.
He was sent back to prison in 2001.
In 2002, post-conviction DNA testing proved
Fountain’s innocence. He was released from prison
on September 27, 2002 and was granted an official pardon by Governor Rick Perry on March 18,
2003.
15

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

evidence, Larry Fuller spent almost twenty years in
prison for a crime he did not commit.

cluded a photo of James Earl Giles, and he identified
James Earl Giles as one of the attackers. James Curtis
Giles was officially exonerated on June 20, 2007.
Because of faulty eyewitness evidence and misconduct by police and prosecutors, James Curtis
Giles spent ten years in prison for a crime he did
not commit.

James Curtis Giles
everal weeks after an attack on a Dallas couple
in their home, police received an anonymous tip
implicating two men in the crime. According to the
tip, one of the accomplices was named James Giles.
In a photo lineup presented to the
victims a few days later, police included a photo of James Curtis Giles,
a twenty-nine-year-old married construction worker from Duncanville.
The female victim identified Giles,
but her husband did not. Although
there were significant discrepancies
between Giles and the initial description the victims gave of their attacker
(Giles was significantly older and
had two prominent gold teeth), Giles
was arrested and charged with aggraJames Curtis Giles
vated sexual assault in 1983.
At trial, the female victim idenspent ten years in
tified Giles from the stand, telling
prison due to a
the jury she would “never forget his
mistaken eyewitness face.” Giles and his wife testified to
identification and
his alibi and explained that they lived
twenty-five miles from the crime
police and
scene. Giles was convicted and senprosecutorial
tenced to thirty years. After spending
misconduct.
ten years in prison, Giles was paroled
and required to register as a sex offender. He was not allowed to travel more than ten
miles from his home without permission and was
required to have a chaperon present to spend time
with his child.
An investigation revealed that the lead detective and prosecutor withheld exculpatory evidence
from Giles’ defense team during trial. The investigation pointed to a man named James Earl Giles who
had a history of violence and lived across the street
from the victims. In addition, his age matched the
initial description the victims gave of their attacker.
In 2003, post-conviction DNA testing revealed
the profiles of two perpetrators—neither of which
matched James Curtis Giles. The tests identified
one man with links to James Earl Giles. The female
victim was contacted and was no longer certain of
her identification. Authorities contacted the victim’s
husband and showed him a photo lineup that in-

Donald Wayne Good

D

onald Wayne Good was arrested in 1983 when a
police officer came to believe that he resembled
the composite sketch of a man wanted for a rape
and burglary. The officer placed Good’s picture in a
photo lineup shown to the victim and her daughter,
who was present during the crime. The poor quality of Good’s photo obscured his facial scar and
tattoo—two potentially important distinguishing
characteristics—and both women identified him as
the man who broke into
their home.
Good’s first trial
ended in a hung jury. At
his second trial, Good
was convicted and sentenced to life in prison.
On appeal, his conviction was overturned
based on prosecutorial misconduct. Good
made an argument that
the prosecutor had improperly told the jury to Donald Good
find him guilty because spent ten years
he did not show enough in prison based
emotion when the victim
took the stand. A third on a mistaken
trial resulted in another eyewitness
conviction, and he was identification.
again sentenced to life in
prison in 1987.
The evidence presented at all three trials was
minimal, including little else than the eyewitness
testimony of the victim and her daughter and serological testing that could not exclude Good—nor
thirty percent of the white male population—as
the perpetrator of the attack. In 2003, DNA testing
proved Good’s innocence.
Because of a mistaken eyewitness identification,
Donald Wayne Good spent ten years in prison for a
crime he did not commit.
16

w w w . T h e J u s t i c e P r o j e c t. o r g

Dallas Morning News

Greg Kendall-Ball / Innocence Project of Texas

S

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Andrew Gossett

gain and was sentenced to four years in prison. He
was released on parole after eighteen months. While
on parole, Henton was convicted of assault and two
felony drug charges. Because he was on parole, the
judge handed down a harsh sentence—twenty years
for aggravated assault and forty-two years for the
drug conviction.
In 2005, post-conviction DNA testing revealed
that Henton was indeed innocent of the 1984 sexual
assault. Henton’s 1984 conviction was overturned
and eventually his other sentence was thrown out
because it was based on his wrongful conviction. He
was re-sentenced to time already served and released
on October 26, 2007.
Because of a mistaken eyewitness identification,
Henton spent nearly two years in prison for a crime
he did not commit.

A

ndrew Gossett originally became a suspect in
the abduction and sexual assault of a Garland
woman after a patrol officer saw him in the vicinity
of the crime scene and thought he looked suspicious.
Gossett was arrested and his photo was placed in a
photo lineup. The victim identified Gossett from
the photo lineup so quickly that the officers were
taken aback by her haste. Yet her confidence that he
was her attacker ultimately assured police. At trial,
the testimony of the victim was significantly inconsistent with earlier statements made to police.
Gossett’s alibi placed him at his girlfriend’s
apartment the entire night, except for two brief trips
to a convenience store. A surveillance video at the
store showed a customer resembling Gossett making a purchase around the time of the assault. In
addition, Gossett passed a polygraph test and hair
samples taken from the victim’s car were dissimilar to those of Gossett. However, Gossett was convicted of aggravated sexual assault and sentenced to
fifty years in prison. Post-conviction DNA testing
proved Gossett’s innocence and he was released on
January 4, 2007.
Because of a mistaken eyewitness identification, Gossett spent over seven years in prison for a
crime he did not commit.

Entre Nax Karage
hen fourteen-year-old Nary Na was murdered
in Dallas in 1994, police quickly focused their
attention on her boyfriend, Entre Karage, despite
strong evidence indicating
an alternate suspect.
At trial, a forensic expert testified that a hair
sample recovered from the
victim’s body had characteristics consistent with
an African-American, and
two other hairs had Caucasian characteristics. Karage
is Cambodian. In addition, the expert testified
that Karage’s DNA did not
match DNA samples taken
from the victim.
Prosecutors argued jealousy as a Entre Karage
motive based on evidence that sexual spent seven years
intercourse had taken place and there in prison because
was no evidence of a rape. The state preDNA evidence
sented numerous witnesses, all of whom
provided remarkably different accounts pointing to the
of activities on the night of the murder. true perpetrator
However, they did present consistent was discounted.
testimony concerning the contentious
relationship between Karage and the victim.
Karage was convicted of murder and sentenced
to life in prison. Post-conviction DNA testing revealed that sperm found in the victim actually be-

Dallas Morning News

Eugene Henton

E

ugene Henton was just seventeen years old in
1984 when he was falsely identified as the attacker in a sexual assault. Although the attacker wore a
mask, the victim told police she was certain she could
identify him because he had spent such a long time
in her home. In addition, the victim also noted seeing
her attacker from her window while he waited at the
bus stop and could thus identify him.
Young and scared, Henton accepted a plea bar17

w w w . T h e J u s t i c e P r o j e c t. o r g

Dallas Morning News

W

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

longed to Keith Jordan, who was already in prison
for a similar offense. Jordan was convicted of aggravated sexual assault of a child and aggravated kidnapping in Dallas in 1997. He has since been charged
with the murder of Nary Na. Karage was officially
exonerated in December 2005 when Governor Rick
Perry pardoned him.
Because prosecutors neglected to pursue evidence
pointing to the true perpetrator, Entre Karage spent
seven years in prison for a crime he did not commit.

a year earlier. In addition to the highly problematic
nature of the photos being mailed to the witness,
the photo lineup itself was highly suggestive. Lindsey was one of only two men in
the photo lineup not wearing Johnnie Lindsey
a shirt, and the victim had described her attacker as shirtless. spent nearly
The victim identified Lindsey as twenty-six years
her attacker.
in prison as a
At trial, the victim’s identi- result of a highly
fication of Lindsey was the sole
piece of evidence against him. suggestive
Lindsey presented an alibi that photo lineup.
he was at work, pressing pants at
a commercial laundry business during the time of the
attack. He even had time cards that showed he was
at work when the crime occurred. However, Lindsey
was convicted and sentenced to life in prison.
Post-conviction DNA testing of the biological
material from his case excluded Lindsey as the perpetrator of the rape, and he was released on September 19, 2008.
Because of a mistaken eyewitness identification,
James Earl Lindsey spent almost twenty-six years in
prison for a crime he did not commit.

Carlos Lavernia

F

ourteen months after the 1983 sexual assault of
a jogger in Austin, Carlos Lavernia became the
primary suspect after the victim was shown three
photo lineups and stated that Lavernia was the only
one that “anywhere near resembled” her attacker.
Based on this identification, police
Carlos Lavernia came to suspect Lavernia for a string of
seven similar attacks in the area. The othspent nearly
er victims were also shown a photo lineup
sixteen years
that included Lavernia and two victims
identified him. Lavernia was placed in
in prison due
a physical lineup and was identified by
to a mistaken
two of the victims—the Austin jogger
eyewitness
victim and another victim who identiidentification.
fied him in the photo lineup. Two additional victims identified a filler standing
next to Lavernia in the physical lineup. Lavernia was
charged with the two rapes in which identifications
had been made, and the Austin jogger case went to
trial first. Lavernia was convicted of aggravated rape
and sentenced to ninety-nine years in prison. The
second case against Lavernia was dropped after this
conviction and sentence.
In 1999, Lavernia was questioned by Austin police about an unsolved murder, and the interview
convinced them that Lavernia might be innocent of
the Austin jogger rape. DNA testing revealed Lavernia’s innocence and his conviction was vacated.
Because of mistaken eyewitness testimony, Carlos Lavernia spent nearly sixteen years in prison for
a crime he did not commit.

Thomas McGowan

F

Clay Graham

ollowing a series of
flawed lineups, Thomas McGowan was charged
with rape and burglary.
The victim was first shown
a live lineup that included
McGowan, two other suspects, and three fillers. Experts have long known that
the inclusion of multiple
Thomas McGowan
suspects in the same lineup greatly undermines the
spent twenty-three
validity of the procedure.
years in prison
The victim did not identify
due to suggestive
anyone as her attacker in
identification
that initial lineup.
Police then showed the
procedures.
victim a photo lineup that
also included a photo of McGowan. The victim hesitantly identified McGowan as her attacker. She later
said the detective conducting the lineup told her that
she “had to make a positive ID. I had to say yes or no.”

Johnnie Earl Lindsey

I

n 1982, Johnnie Earl Lindsey’s picture was included in a photo lineup that police sent in the mail
to a White Rock Lake woman who had been raped
18

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Brandon Moon
Greg Kendall-Ball / Innocence Project of Texas

The photo lineup was equally as problematic as the
live lineup because four of the seven photos, including McGowan’s, were color originals. In addition,
three of the color photos were marked “Richardson
Police” and the other was marked “Garland Police.”
The remaining black and white photos included one
original and two photocopies. The lack of uniformity
in the photos, combined with the multiple presentations of McGowan in photo and live lineups, tainted
the identification and contributed to the witness’s
mistaken identification of McGowan.
McGowan was eventually convicted and sentenced to life in prison. Post-conviction DNA testing
excluded McGowan as the perpetrator and he was
released April 16, 2008. A DNA state database search
identified the true perpetrator, Kenneth Wayne
Woodson, who is currently serving a thirty-year sentence for another Dallas rape committed the year after McGowan was arrested.
Because of the suggestive identification procedures that led to a mistaken eyewitness identification, Thomas McGowan spent twenty-three years
in prison for a crime he did not commit.

F

ollowing the sexual assault
Brandon Moon spent
of an El Paso woman in the
spring of 1987, police created seventeen years in
a photo lineup that included prison due to false
Brandon Moon, a twenty-five- forensic testimony and
year-old military veteran and
the botched handling of
aspiring jet pilot. The victim
tentatively chose Moon’s picture exculpatory DNA.
and later firmly identified him
in a live lineup. Moon was the only person present
in both the photo and live lineups. Based on the
identification, police contacted two other women
who had been victims of similar attacks. Each victim identified Moon even though the crimes against
them had been committed years before these lineups were conducted.
At trial, a serologist testified that semen stains
from the crime scene definitively excluded the victim’s husband and indicated Moon could have been
the source. This testimony would later prove to be
patently false.
In 1989, DNA testing showed that Moon was
not the source of the semen taken from the crime
scene, but the results were considered inconclusive
because reference samples were not collected from
the victim’s husband or teenage son. Eventually,
testing revealed that the victim’s husband’s DNA
matched a stain on the comforter, but not the bathrobe. Additional testing proved that the son was not
the source of the bathrobe stain, leaving only the
rapist, whom previous tests had already revealed
could not be Moon. In December 2004, Moon was
released from prison.
Because of false forensic testimony and the
mishandling of exculpatory DNA, Brandon Moon
spent seventeen years in prison for a crime he did
not commit.

Billy Wayne Miller

I

n 1983, a Dallas woman waiting at a bus stop accepted a ride from a man who subsequently assaulted her. The victim remembered some of the
numbers from the license plate of the car driven by
her attacker. During the attack, the victim memorized street names and the location of a house where
a part of the assault took place. Following up on the
victim’s statements, police found a car in front of
the house with a license plate that differed from the
victim’s memory by only one digit. Police entered
the house and arrested Billy Wayne Miller. The car
was registered to his father.
The lead prosecutor in the case now believes
that the victim’s memory of specific details—down
to the brand of beer the attacker drank—was likely a
result of the police having fed her that information.
Post-conviction DNA testing excluded Miller
as the attacker. He was released in May 2006 and
officially pardoned by Governor Rick Perry in December of 2006.
Because of a mistaken eyewitness identification, Billy Wayne Miller spent over twenty-two years
in prison for a crime he did not commit.

19
w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Arthur Mumphrey

ten by the police and accepted the state’s offer of a life
sentence. In exchange, Ochoa pled guilty to murder
and testified against Danziger at trial. Danziger was
convicted of aggravated sexual assault and sentenced
to life in prison.
While Ochoa and Danziger were serving life
sentences, a Texas inmate named Achim Josef Marino confessed to DePriest’s murder. Marino professed he alone committed the crime and provided
a detailed description of the crime scene and information regarding the location of items stolen from
the Pizza Hut.
DNA testing of semen collected from the crime
scene conclusively proved the innocence of Ochoa
and Danziger. Ochoa and Danziger both settled
civil lawsuits with the City of Austin. Danziger sustained permanent brain damage as a result of a violent prison assault, rendering him unable to care for
himself without help.
Because of the threatening and intimidating
interrogations that led to Ochoa’s false confession,
Richard Danziger and Christopher Ochoa each
spent twelve years in prison for a crime they did not
commit.

D

uring a police interrogation in 1986, Steve
Thomas admitted to raping a thirteen-year-old
girl and implicated Arthur Mumphrey as his accomplice in the crime. Police made a deal with Thomas,
offering him a fifteen-year sentence in exchange for
his “truthful testimony” against Mumphrey.
At trial, Thomas testified to the same horrific details he had relayed to the police during his interrogation. In addition, another witness testified that he
saw Thomas and Mumphrey one night and Thomas
told him all about the crime while Mumphrey stood
silent, tacitly endorsing the story. The victim testified
that she did not look at the faces of her attackers and
could not identify them. Mumphrey was convicted.
Mumphrey was paroled in 2000, but he failed
to meet the strict terms of his parole and returned
to prison in 2002. Post-conviction DNA testing on
the rape kit evidence confirmed that Thomas was
indeed one of the perpetrators, but the other perpetrator was not Mumphrey.
Mumphrey was released on January 27, 2006,
and was granted an official pardon by Governor
Rick Perry on March 17, 2006.
Because of the false testimony of Steve Thomas,
Mumphrey spent nearly eighteen years in prison for
a crime he did not commit.

Steven Charles Phillips

I

n 1982, Steven Phillips became the chief suspect in
an unusual string of sexual assaults and burglaries in Dallas, Texas and
Kansas City, Missouri.
Police were confident that
the same assailant was responsible for the crimes
in both cities.
After Dallas police
turned their focus to Phillips, ten separate victims
picked his photo out of
dozens shown by police.
After serving
None of the photo lineups were preserved as evitwenty-six years in
dence, however, making
prison, DNA testing
it impossible to analyze
exonerated Steven
the accuracy and quality
Charles Phillips
of the procedures used.
Phillips was convicted of
of multiple crimes
sexual assault and burand exposed many
glary based largely on the
investigative errors.
eyewitness accounts, and

Christopher Ochoa and
Richard Danziger

C

Dallas Morning News

hristopher Ochoa was working at a Pizza Hut
when he and his friend, Richard Danziger, became suspects in the 1988 murder of Nancy DePriest,
a manager for another of the chain’s
Christopher Ochoa
restaurants in Austin. Employees of
the restaurant contacted police afand Richard
Ochoa and Danziger were seen
Danziger each spent ter
giving a toast in the victim’s honor.
twelve years in
Acting on the tip, police brought
prison as a result of
Ochoa in for questioning.
During the interrogation, Ochoa
a long, intimidating
was subjected to both physical and
interrogation that
verbal intimidation, and police offiproduced Ochoa’s
cers threatened him with the death
false confession.
penalty unless he confessed to the
crime. After two days of intense
questioning and threats, Ochoa eventually gave the
police what they wanted. He signed a confession writ20

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

he was sentenced to forty years in prison.
In 2007, post-conviction DNA testing on the
biological material from the Dallas case cleared
Phillips of the sexual assault and burglary charges.
Because police and prosecutors believed the same
person committed all the crimes, the Court of
Criminal Appeals fully exonerated Phillips on August 5, 2008.
A reinvestigation of the case revealed that Kansas City police had previously identified another
suspect, Sydney Alvin Goodyear, and notified the
Dallas police. At least one Dallas victim identified
Goodyear from a photo lineup, and a warrant had
been issued for his arrest prior to Dallas police focusing on Phillips. This exculpatory information
was never disclosed to Phillips’ defense.
Because of misconduct and mistaken eyewitness
identifications, Steven Charles Phillips spent twentysix years in prison for crimes he did not commit.

The victim was shown a physical lineup one
month after the photo lineup in which Pope was the
only man who had the blonde hair and tan skin described by the victim. The victim identified Pope as
her attacker.
In 1986, Pope was convicted of aggravated sexual assault and sentenced to forty-five years in prison. The prosecution’s case rested primarily on the
victim’s identification of Pope from the live lineup
and expert testimony about “voice print analysis”
that was said to match Pope’s voice to the taped calls
made to the victim. The victim testified at trial that
she not only recognized him as her attacker, but also
that she could unequivocally identify his voice.
In January of 1999, prosecutors received an
anonymous tip that someone else had committed
the crime. By that time, voice print analysis was no
longer considered reliable evidence, and prosecutors decided the anonymous tip warranted serious
consideration. Post-conviction DNA testing proved
Pope’s innocence and identified the true perpetrator:
a convicted rapist imprisoned in another state. Pope
was granted an official pardon by Governor Rick
Perry in February 2001.
Because of faulty forensic evidence and a mistaken eyewitness identification, David Shawn Pope
spent fifteen years in prison for a crime he did not
commit.

David Shawn Pope

Dallas Morning News

Ricardo Rachell

T

he 2002 sexual assault conviction of Ricardo
Rachell rested solely on testimony from the
eight-year-old victim and a friend
who was with the boy before the at- Ricardo Rachell
tack. Police relied so heavily on the
identifications made by the victim spent six years in
and his friend that they did not con- prison while the
duct DNA testing on either the bio- biological evidence
logical evidence collected in the case that would
or on the reference sample voluntarily
exonerate him
provided by Rachell.
Although both the victim and sat untested in a
his friend identified Rachell, it is dif- Houston crime lab.
ficult to understand how this identification held up in court because Rachell’s face was
significantly disfigured in a shotgun accident years
ago, making it difficult for him to talk. The victim’s
friend told police that the man spoke clearly. Rachell
took the stand in his own defense and demonstrated

I

n August 1985, a woman who lived in David
Shawn Pope’s apartment
complex was sexually
assaulted at knifepoint.
Police included a photo of Pope in a photo
lineup presented to the
victim. Even though the
victim did not identify
Pope as her attacker, police suspected him based
on the sound of his voice. In the days after the attack, the victim received several threatening phone
calls from her attacker. In addition to an answering
machine message, police recorded a phone call from
the attacker lasting about ten minutes.

DNA testing
proved David
Pope’s innocence
fifteen years after
he was wrongfully
convicted based
on junk science.

21
w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

that the gunshot wound had left him with a significant speech impediment. Despite these discrepancies, Rachell was convicted and sentenced to forty
years in prison.
In October of 2008, DNA testing established
that Rachell could not have committed the crime
and Harris county prosecutors endorsed the reversal of his conviction.
Because of two mistaken eyewitness identifications and the failure to disclose DNA evidence, Ricardo Rachell spent six years in prison for a crime
he did not commit.

own test, they still did not believe Robinson to be
an innocent man. The DA’s office argued that the
semen came from an unknown man with whom the
victim had consensual sex and that Robinson still
had something to do with the crime. There was no
evidence to back up this theory.
On November 7, 2000, the Texas Board of Pardons and Paroles unanimously voted to recommend Robinson’s pardon, and he was granted an
official pardon by Governor George W. Bush seven
days later. After spending a decade behind bars for
a crime he did not commit, Robinson went back to
school and received his law degree from the Thurgood Marshall School of Law at Texas Southern
University.
Because of a mistaken eyewitness identification, Anthony Robinson spent ten years in prison
for a crime he did not commit.

Greg Kendall-Ball / Innocence Project of Texas

Anthony Robinson

George Rodriguez

F

ollowing her 1987 abduction and sexual assault,
a fourteen-year-old Houston girl gave police
basic descriptions of the two perpetrators. The victim said that one of her attackers called the other
“George,” though she told police that she believed
it was a fake name. Based on the victim’s description of the house where
the crime occurred and George Rodriguez
the surrounding area, po- spent seventeen
lice went to the home of
Manuel and Uvaldo Bel- years in prison
tran. George Rodriguez after compelling
became a suspect in the evidence of
case because he was an ac- his innocence
quaintance of one of the
was overlooked
Beltran brothers.
Police conducted a in favor of
photo lineup and the vic- a mistaken
tim identified Rodriguez eyewitness.
as one of the attackers.
Even though Manuel Beltran confessed to police
that he had sexually assaulted the girl and stated that
Isidro Yanez was his accomplice, police continued to
focus on Rodriguez. In a one-person show-up, the
victim again identified Rodriguez as her attacker.
The victim was also shown a collection of photos
that included both Rodriguez and Yanez, but the
victim again identified Rodriguez as her attacker.
At trial, a forensic analyst testified that a pubic

I

n 1986, Anthony Robinson,
a twenty-six-year-old college
graduate and U.S. Army veteran,
was put on trial for the rape of a
University of Houston woman.
The victim told police that her
attacker was a black man with a
moustache wearing a plaid shirt. That same day Robinson was on campus picking up a car for a friend.
Even though Robinson did not have a moustache,
police arrested him. Robinson was placed in a lineup
and identified by the victim. Although no physical
evidence linked him to the crime, the victim’s confident identification led to Robinson’s conviction and
he was sentenced to twenty-seven years in prison.
After being released on parole in 1997, Robinson took odd jobs to save enough money to hire
an attorney and pay for post-conviction DNA tests
on the case’s biological material. On September
19, 2000, DNA test results proved Robinson’s innocence. Even though the Harris County District
Attorney’s Office confirmed the results with their

Anthony Robinson
spent ten years in
prison due to a
mistaken eyewitness
identification.

22
w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

hair found on the victim’s underwear was microscopically similar to Rodriguez’s hair, and that the
semen from the rape kit could not exclude Rodriguez as the perpetrator, and it could not belong to
Yanez. Even though Rodriguez presented evidence
that he was at work during the time of the crime,
Rodriguez was convicted of aggravated sexual assault and kidnapping and sentenced to sixty years
in prison.
In 2004, DNA testing established that Rodriguez was not the source of the pubic hair found on
the victim’s clothing, and Yanez could not be excluded. Further testing also established that Yanez
was mistyped by the Houston Crime Lab during
the initial testing and could not have been excluded
as the source of the semen—directly contradicting
the forensic testimony at trial. Rodriguez’s conviction was vacated in August 2005 and all charges
were dismissed a month later.
Because of faulty forensic testimony and a mistaken eyewitness identification, George Rodriguez
spent seventeen years in prison for a crime he did
not commit.

attacker. Defense attorneys would later note that
Salazar was wearing a turquoise shirt in the second
picture the police showed the victim, the same color
shirt from her description.
Salazar voluntarily gave blood, saliva, and hair
samples to prove his innocence, but the results of
the forensic testing could not exclude him as the
source of the semen found on the victim. At trial, a
forensic analyst testified that the blood typing characteristics found could only come from two percent
of the Hispanic population which included Salazar.
Salazar was convicted of aggravated sexual assault
and sentenced to thirty years in prison.
Post-conviction DNA testing excluded Salazar as the source of the semen found on the victim. Two additional DNA tests were performed to
finally convince prosecutors and the parole board
that Salazar was indeed innocent. He was granted
an official pardon by Governor George W. Bush on
November 20, 1997.
Because of a mistaken eyewitness identification, Ben Salazar spent five years in prison for a
crime he did not commit.

Ben Salazar

Billy James Smith

B

n 1991, Ben Salazar’s
photo was chosen out of
a book of mug shots as the
perpetrator of a rape following an attack in an Austin home. The victim gave
a detailed description of
her attacker — a Hispanic
male, in his late twenties to
early thirties, between five
feet, five inches and five
Ben Salazar
seven inches tall and
spent five years feet,
weighing approximately
in prison due
160 to 180 pounds, and
wearing a turquoise shortto a mistaken
sleeved t-shirt and brown
eyewitness
work boots. Because she
identification.
was certain she could identify her assailant, police showed her two books with
mug shots of Hispanic males and she picked Salazar
as her attacker.
Salazar agreed to go to the station to have his
picture taken. The victim viewed several new lineups and again identified Salazar’s new photo as her

illy Smith became a suspect for rape and robbery when his building manager requested he
step out on the balcony of Dallas County apartment
in 1986. The building manager’s girlfriend stated that Smith was the man
who had robbed and raped her two
hours earlier and they immediately
called police.
An extensive search of Smith’s
belongings did not reveal any clothing matching the description given
by the victim, nor did it turn up any
physical evidence tying him to the
crime. Despite the lack of evidence,
the case proceeded to trial. The prosecution argued that the victim had
Billy James Smith
not had sex with anyone twenty-four
hours prior to the rape, and therefore,
spent nineteen
the presence of semen in the rape kit
years in prison
must prove that a rape occurred. The
due to a mistaken
prosecution then used the eyewitness
eyewitness
identification to solidify the case.
Although Smith and his sister
identification.
both testified that he was home dur23

w w w . T h e J u s t i c e P r o j e c t. o r g

Greg Kendall-Ball / Innocence Project of Texas

Clay Graham

I

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Ronald Gene Taylor

ing the time the crime occurred, he was convicted
of aggravated sexual assault and sentenced to life in
prison. Smith petitioned for post-conviction DNA
testing, but the prosecution argued that the semen
from the rape kit could belong to her live-in boyfriend and as a result could not be probative. The
argument was the exact opposite of the argument
made at trial, where prosecutors maintained that the
presence of semen meant a rape occurred.
Eventually, Smith was given access to testing
that revealed he did not rape the victim. Smith was
released in July 2006 and officially exonerated on
December 13, 2006.
Because of a mistaken eyewitness identification,
Billy James Smith spent nineteen years in prison for
a crime he did not commit.

onald Gene Taylor became a suspect in a 1993
rape in Houston when a neighbor told police
that he saw Taylor near the crime scene on the night
of the attack. Police placed Taylor in a video lineup
because the victim was unable to travel to the police station to view a live
lineup. After two viewings
of the lineup, the victim
identified Taylor as her
attacker.
At trial, the victim’s
identification testimony
was the primary evidence
against Taylor. When the
witness included new details that were not present
in her initial statement,
Ronald Gene
however, Taylor’s attorneys
Taylor spent
argued that police leaked
fourteen
details of the investigation
years in prison
to her and that her identification of Taylor was
due to forensic
tainted. A forensic analyst
work and a
from the Houston Police
mistaken
Department also testified
eyewitness
for the prosecution stating
that there was no semen
identification.
on the sheets found at the
crime scene and Taylor could not be eliminated as a
suspect through DNA testing. Taylor was convicted
and sentenced to sixty years in prison.
A re-examination of the sheets from the crime
scene revealed that the sheets actually did contain
biological evidence that could be used to extract a
profile. Post-conviction DNA testing excluded Taylor and revealed the identity of the true perpetrator,
Roosevelt Carroll. Carroll lived just a mile from the
victim at the time of the attack and had already been
convicted of two other rapes. He was serving a fifteenyear sentence for failing to register as a sex offender
when Taylor was finally exonerated. Unfortunately,
the statute of limitations had already passed on the
1993 rape case and Carroll could not be charged.
Taylor was released on October 9, 2007.
Because of faulty forensic work and a mistaken
eyewitness identification, Taylor spent fourteen
years in prison for a crime he did not commit.

Josiah Sutton

I

n October 1998, the victim of an abduction and
rape in Houston initially described her attacker as
five feet, seven inches tall and weighed 135 pounds.
Even though he was six
feet tall weighing 200
An eyewitness
pounds, Josiah Sutton
identification
was identified as the
and false DNA
perpetrator and was artestimony resulted rested.
The Houston Crime
in the wrongful
Lab compared DNA
conviction of
samples from Sutton to
Josiah Sutton.
the two DNA profiles
obtained from the biological evidence at the crime scene. According to the
lab, Sutton’s DNA was a match. At trial, an analyst
from the crime lab testified that the DNA match
was solid—the DNA profile was shared by only one
person in almost 700,000. Sutton was convicted and
sentenced to twenty-five years in prison.
In 2002, journalists with KHOU-TV in Houston exposed pervasive flaws with the Houston Police
Department Crime Lab, causing several cases to be
reexamined, including Sutton’s. Independent experts
concluded that the forensic testimony at Sutton’s trial was false. When the lab retested the evidence, the
DNA results excluded Sutton as the perpetrator.
Because of faulty forensic testing, false forensic
testimony, and a mistaken eyewitness identification,
Sutton spent over four years in prison for a crime he
did not commit.
24

w w w . T h e J u s t i c e P r o j e c t. o r g

Clay Graham

R

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Victor Larue Thomas

On appeal, Turner’s conviction was reversed
based on a prosecutorial error at trial. A higher
court overruled that decision and Turner’s sentence
was reinstated.
Turner was paroled after spending four years
in prison. In 2005, post-conviction DNA testing
proved Turner’s innocence and he received an official pardon by Governor Rick Perry on December
22, 2005.
Because of a mistaken eyewitness identification, Keith Turner spent four years in prison for a
crime he did not commit.

W

hen a woman was raped during the robbery
of a Waxahachie convenience store in 1985,
Victor Larue Thomas was a suspect in an unrelated
sexual assault that occurred in the area. Although
their initial suspicions of Thomas in the unrelated
case were unsubstantiated and police did not pursue that case against Thomas, he was taken in for
questioning for the convenience store rape.
Statements made by Thomas during his interrogation eventually led police to suspect him in the
convenience store rape case. The victim identified
Thomas as her attacker, and he was arrested and
charged with robbery, kidnapping, and rape. Based
almost entirely on the victim’s in-court eyewitness
identification, Thomas was convicted of all charges
and sentenced to life in prison.
Post-conviction DNA testing proved that Thomas had not committed the crime, and he was released
on June 27, 2001. The District Court Judge noted that
Thomas should never have been prosecuted. Thomas was granted an official pardon by Governor Rick
Perry in 2002.
Because of a mistaken eyewitness identification,
Victor Larue Thomas spent fifteen years in prison
for a crime he did not commit.

James Douglas Waller

W

Greg Kendall-Ball / Innocence Project of Texas

hile shopping in a convenience store in his neighborhood, James Waller was identified by a twelve-year-old boy as the
man who had sexually assaulted
him in his family’s apartment. The
boy initially described the intruder
as an African-American man who
wore a cowboy hat and a bandana
around the lower half of his face.
Waller and his family lived in the
same apartment complex and
were the only African-American
residents of the complex.
Police arrested Waller based
on the boy’s identification. The
apartment manager had also reported seeing an unknown male Questionable voice and
with a cowboy hat and bandana.
eyewitness identifications
After the boy’s identification, the
apartment manager also identified put James Waller
in prison for eleven
Waller as the person she had seen.
At trial, the state’s case rested years for a crime
almost solely on the testimony of
he did not commit.
the two eyewitnesses. The victim
testified that he knew Waller was his attacker based
on Waller’s eyes and voice. The boy admitted that
he had been unable to see his assailant’s face. Moreover, Waller was substantially taller and heavier than
the man described by the victim in his initial conversation with the police. The apartment manager
also identified Waller at trial as the man she spotted wearing the hat and bandana near the scene.
However, her testimony was inconsistent with her
original description because she had previously de-

Keith E. Turner

Dallas Morning News

Keith E. Turner
spent four years
in prison and
eight years on
parole after a rape
victim incorrectly
identified him.

I

n 1982, Keith Turner was wrongly identified as the
perpetrator of a Dallas coworker’s rape. The victim
called police when she thought she recognized Turner at work as her attacker. She also picked Turner out
of a lineup police arranged that same day. At trial, the
victim testified that she could identify Turner based
on his basic physical appearance and the sound of his
voice. Turner provided an alibi for his whereabouts
during the attack, but he was convicted of rape and
sentenced to twenty years in prison.
25

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

scribed an “unknown” man and she knew Waller as
a resident of the building.
Despite the discrepancies in the identifications,
and alibi testimony given by Waller’s girlfriend,
Waller was convicted of aggravated sexual abuse
and sentenced to thirty years in prison.
Waller was paroled in 1993 after spending
eleven years in prison and was required to register
as a child sex offender. In 2003, post-conviction
DNA tests were unable to extract a DNA profile
from the small amount of biological evidence available, but Waller received permission to have previously unavailable advanced DNA testing conducted
on liquid extracts of the semen evidence that had
been preserved. In 2006, those results conclusively
proved that Waller did not commit the rape. Waller
was officially pardoned by Governor Rick Perry in
March 2007.
Because of the mistaken eyewitness identifications in his case, James Douglas Waller spent eleven
years in prison for a crime he did not commit.

it matched with Byron Demond Bell, a man who
had been paroled in February of 2008 after serving
fifteen years of a forty-five-year sentence for burglary. When questioned, Bell admitted to the attack
and named Lemondo Simmons as his accomplice in
the crimes. Both men testified to their guilt in front
of a grand jury, but neither would be prosecuted for
the crimes because the statute of limitations had
expired. Bell’s parole officer stated that the crimes
would be taken into consideration if Bell ever violated the terms of his parole.
Because of mistaken eyewitness identifications, Patrick Waller spent more than fifteen
years in prison for a crime he did not commit.

Gregory Wallis

F

our months after police circulated a flier and
composite sketch in a local jail about a 1988
sexual assault of an Irving woman, an inmate told
police that Gregory Wallis had a tattoo similar to
the description given by
the victim. A photo of Gregory
Wallis was included in a Wallis spent
photo lineup presented to
eighteen years
the victim and she identified Wallis as her attacker. in prison due
Investigators found no to a mistaken
physical evidence linking eyewitness
Wallis to the attack.
identification
Although biological
evidence was taken from the scene, forensic DNA
analysis was not available at the time. At trial, the
prosecution relied entirely on the victim’s eyewitness
testimony, and Wallis was convicted and sentenced
to fifty years in prison.
In 2004, DNA testing could not entirely exclude Wallis as the attacker, but the test results had
cast enough doubt to motivate prosecutors to offer
a deal. Prosecutors offered Wallis his freedom if he
agreed to spend the rest of his life as a registered
sex offender. Wallis refused and remained in prison.
A few months later, a more sophisticated DNA test
proved that Wallis was not the perpetrator. Wallis
was released in March 2006.
Because of a mistaken eyewitness identification, Gregory Wallis spent eighteen years in prison
for a crime he did not commit.

Patrick Waller

Clay Graham

P

atrick Waller was
identified by four victims as one of two attackers in a 1992 robbery and
sexual assault in Dallas.
Though all four victims
identified Waller as one of
the attackers, one witness
was not certain. Records
indicate that the detective
pointed toward Waller’s
False eyewitness picture and said that two
identifications
other eyewitnesses had alsent Patrick
ready identified Waller as
Waller to prison the perpetrator. Even so,
eyewitness remained
for over fifteen the
uncertain and did not tesyears.
tify at trial.
Waller was convicted
of aggravated robbery and sentenced to life in prison.
He pled guilty to two counts of kidnapping to avoid
more jail time, but he proclaimed his innocence
throughout his appeals.
In June of 2008, DNA testing excluded Waller
in the crime. Investigators ran the DNA profile from
the crime scene through the state DNA database and
26

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Calvin Washington

James Lee Woodard

J

J

ames Lee Woodard quickly became a suspect in
the December, 1980 rape and murder of his former girlfriend, Beverly Ann Jones. Woodard was
taken into custody on New Year’s Day.
Two eyewitnesses placed Woodard with the
victim close to the time when the crime occurred,
but there was no evidence that tied Woodard to the
crime. The defense provided
two alibi witnesses who stated
that Woodard had been thirty
miles away from the scene on
the night of the murder. Woodard was convicted and sentenced to life in prison.
DNA testing excluded
Woodard as the perpetrator of
the sexual assault, but because
the other charge against him
was murder, he needed to prove James Lee Woodard
that the same person commit- spent twenty-seven
ted both crimes.
years in prison due
An investigation by Woodard’s lawyers revealed that to prosecutorial
the Dallas County prosecutors misconduct and two
withheld key evidence from the mistaken eyewitness
defense in 1981. Several days identifications.
before Woodard’s trial, authorities learned that three men had been with the victim
on the night of her death: a man told investigators
he went with Jones to a South Dallas convenience
store where she got into another car with three men
and left. He could not identify the car or the men
in question. In addition, police neglected to investigate the three men as suspects, even though one
was in prison on a charge of aggravated rape during
Woodard’s trial. This finding, along with a forensic
pathologist’s conclusion that the rape and murder
“were tied together in such a way that the rape results would conclusively show who the perpetrator
was,” helped prove Woodard’s innocence. Woodard
was released on April 29, 2008.
Because of two mistaken eyewitness identifications and prosecutorial misconduct, James Lee
Woodard spent twenty-seven years in prison for a
crime he did not commit, more time than any other
DNA exoneree in the country.
Dallas Morning News

ailhouse informants provided damaging testimony against Calvin Washington and Joe Sydney Williams for the March 1986 rape and murder
of Juanita White of Waco. One man testified that
he walked past a hotel room in the middle of the
night and overheard Williams and Washington implicating themselves. Several other witnesses testified that Washington and Williams were spotted
in the victim’s car after the murder and that they
had sold some of her property. The testimony of a
dental expert, Homer Campbell, who claimed that
the marks on the victim’s body were bite marks that
matched Williams’ teeth, helped convict both men.
Williams’ conviction was overturned on direct appeal, and he was released when the prosecutor declined to retry him.
Washington remained in prison. He was released when post-conviction DNA testing showed
that blood found on a shirt in Washington’s home
was not the victim’s, as the prosecutor had alleged.
Also, DNA found on the victim implicated another
man, Bennie Carroll, who had previously confessed
to raping Juanita White’s next door neighbor prior
to committing suicide.
Because of unreliable jailhouse informants,
Calvin Washington spent thirteen years in prison
for a crime he did not commit.

Mark Webb

M

ark Webb was wrongfully convicted of a 1985
rape based on the victim’s mistaken eyewitness identification.
Although biological evidence from the crime
scene was tested and the results were presented in
court, the tests at the time were not sophisticated
enough to exclude Webb as the attacker. The tests
also could not exclude a large portion of the rest of
the population. The defense called several alibi witnesses who testified that Webb was at work at the
time of the crime. Unfortunately, the eyewitness
identification held more weight with the jury, and
Webb was convicted of rape and sentenced to thirty
years in prison. In 2001, post-conviction DNA testing proved Webb’s innocence.
Because of a mistaken eyewitness identification, Mark Webb spent thirteen years in prison for a
crime he did not commit.
27

w w w . T h e J u s t i c e P r o j e c t. o r g

THE JUSTICE PROJECT

EDUCATION
FUND

Convicting the Innocent: Texas Justice Derailed

Michael Nawee

1994 2008

Collin

rape, murder

Butler

A.B.

1983 2000

Smith

rape, kidnapping

Byrd

Kevin

1985 1997

Harris

rape

Chatman

Charles

1981 2007

Dallas

rape

Cole*

Timothy Brian

1986 2008

Lubbock

rape

Criner

Roy

1990 2000

Montgomery

rape, murder

Danziger

Richard

1990 2001

Travis

rape, murder

Fountain

Wiley

1986 2003

Dallas

rape

Fuller

Larry

1981 2006

Dallas

rape

Giles

James Curtis

1983 2007

Dallas

rape

Good

Donald

1984 2004

Dallas

rape

Gossett

Andrew

2000 2007

Dallas

rape

Henton

Eugene

1984 2006

Dallas

rape

Karage

Entre Nax

1997 2004

Dallas

murder

Lavernia

Carlos

1985 2000

Travis

rape

Lindsey

Johnnie Earl

1983 2008

Dallas

rape

McGowan

Thomas

1985 2008

Dallas

rape, burglary

Miller

Billy Wayne

1984 2006

Dallas

rape

Moon

Brandon

1988 2005

El Paso

rape

Mumphrey

Arthur

1986 2006

Montgomery

rape

Ochoa

Christopher

1989 2001

Travis

murder

Phillips

Steven Charles

1983 2007

Dallas

rape, burglary

Pope

David Shawn

1986 2001

Dallas

rape

Rachell

Ricardo

2003 2008

Harris

child sex assault

Robinson

Anthony

1987 2000

Harris

rape

Rodriguez

George

1987 2004

Harris

rape, kidnapping

Salazar

Ben

1992 1997

Travis

rape

Smith

Billy James

1987 2006

Dallas

rape

Sutton

Josiah

1999 2004

Harris

rape

Taylor

Ronald

1995 2007

Harris

rape

Thomas

Victor Larue

1986 2002

Ellis

rape

Turner

Keith E.

1983 2005

Dallas

rape

Waller

James

1983 2007

Dallas

rape

Waller

Patrick

1992 2008

Dallas

robbery, kidnapping

Wallis

Gregory

1989 2007

Dallas

rape

Washington Calvin

1987 2001

McLennan

rape, murder

Webb

Mark

1987 2001

Tarrant

rape

Woodard

James Lee

1981 2008

Dallas

murder, rape

P
P
P
P
P
P

P
P

Approximate Years
in Prison

Blair

P
P
P
P
P
P

Suppression of
Exculpatory Evidence
or Other Misconduct

1990 1994

False Confession
or Plea

Gilbert

Jailhouse Informant
& Accomplice
Testimony

First Name

Alejandro

Unreliable or Limited
Forensic Methods

rape

Last Name

Faulty Forensic
Testimony

Crime

Uvalde

Mistaken Eyewitness
Identification

County

Year Exonerated

Year Convicted

The Texas DNA Exonerated

4

P

14
17

P
P
P
P

12
27

P
P

P

13*
10

P

12
16

P

20

P
P

P

10
10
7

P

2
7

P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P
P

16
26
23
22

P
P
P

33

17
18

P
P

12

P

26
15

P

6
10

P

P
P

P

17
5
19

P
P

4
14
15
4
11

P

16
18

P

P
P

TOTALS

P

P

P

13
13

7

*Died in prison in 1999
28
w w w . T h e J u s t i c e P r o j e c t. o r g

11

5

5

P

27

7

548

“Law’s evolution is never done, 
and for every improvement made 
there is another reform that is overdue.”
— J ustice W illiam J. B rennan , J r .

Convicting the Innocent:
T e x a s J u s tic e D e r a i l e d
Stories of Injustice and the Reforms
That Can Prevent Them

Texas has had more wrongful convictions exposed
by DNA evidence than any other state in the country.
The thirty-nine cases in this report—all of which
have been exposed by DNA evidence—highlight the
systemic problems that have resulted in the wrongful
convictions of the innocent. By identifying the causes
of wrongful convictions and implementing practical
reforms, Texas can increase the fairness, accuracy,
and reliability of its criminal justice system.

THE JUSTICE PROJECT
510 S. Congress Ave., Suite #304
Austin, TX 78704
(512) 391-2320
Fax: (512) 391-2330
www.thejusticeproject.org/state/texas

EDUCATION
FUND

1025 Vermont Ave., NW, Third Floor
Washington, DC 20005
(202) 638-5855
Fax: (202) 638-6056
www.thejusticeproject.org