FOR RELEASE APRIL 13, 2006

CALIFORNIA COMMISSION ON THE FAIR ADMINISTRATION OF JUSTICE
REPORT AND RECOMMENDATIONS
REGARDING EYE WITNESS IDENTIFICATION PROCEDURES

The California Commission on the Fair Administration of Justice was
established by California State Senate Resolution No. 44 “to study and review the
administration of criminal justice in California, determine the extent to which that
process has failed in the past,” examine safeguards and improvements, and
recommend proposals to ensure that the administration of criminal justice in
California is just, fair and accurate. The Senate Resolution noted that study and
review in other states has resulted in recommendations for reforms in order to avoid
wrongful convictions and executions, and that California has not engaged in any
such review of the state’s criminal justice system.
The Commission began by reviewing the studies and reviews of wrongful
convictions conducted in other states, and identifying the causal factors that most
frequently recur in cases where the wrongfully convicted have been exonerated. The
Commission has assumed the accuracy of these studies without any independent
efforts to verify them. The most frequently identified causal factors include
misidentification by eyewitnesses, false confessions, perjured testimony,
mishandling of forensic evidence, withholding exculpatory evidence, and the
incompetence of defense lawyers. The Commission plans detailed inquiries into
each of these causes of wrongful convictions before it issues its final report by
December 31, 2007.
Meanwhile, the Commission has determined that there are reforms which can
improve criminal investigation techniques and thus further the cause of justice in
California. Our recommendation of these reforms need not await the issuance of our
final report. One such set of reforms involves procedures to improve the reliability
of eyewitness identifications.
A comprehensive compilation of all exonerations in the United States from
1989 through 2003 was recently published by a group of researchers at the

1

University of Michigan led by Professor Samuel R. Gross. 1 The researchers
confined their study to cases in which there was an official act declaring a defendant
not guilty of a crime for which he or she had previously been convicted, such as a
pardon based upon evidence of innocence, or a dismissal after new evidence of
innocence emerged, such as DNA testing. They identified 340 such cases, 27 of
which occurred in the State of California.
Of the 340 cases, sixty percent had been convicted of murder, and 36% had been
convicted of rape or sexual assault. They note two possible explanations for the
high prevalence of murder cases: false convictions are more likely to be discovered
in murder and death penalty cases, because of the intensive level of post-conviction
review given to these cases, or false convictions are more likely to occur in murder
and death penalty cases. There may be other explanations. We do not know
whether wrongful convictions are much more common than realized throughout the
system. What we do know is that as these cases come to light we must address their
causes.
One explanation for the high prevalence of rape and sexual assault cases
among exonerations is recent improvements in DNA technology that can now be
used not only to identify a perpetrator of rape at trial, but also to clear an individual
of the crime both before and after conviction. Mistaken eyewitness identification
was involved in 88% of the rape and sexual assault cases. This suggests that
unexposed mistaken identification could be present in other convictions that heavily
rely upon eyewitness identifications, such as robbery cases where DNA evidence is
not normally present. Among the 80 cases in which rape defendants were
subsequently exonerated and the race of both parties was known, 39 of the cases
involved black men who were wrongfully convicted of raping white women, and
nearly all of these cases involved mistaken eyewitness identifications. Since less
than 10% of all rapes in the United States involve white victims and black
perpetrators, the fact that a disproportionate number of the rape exonerations involve
white victims misidentifying black suspects suggests that the risk of error is greater
in cross-racial identifications. Research has consistently confirmed that cross-racial
identifications are not as reliable as within-race identifications. 2
The study by Professor Gross’ researchers identified seven California
exonerations involving mistaken eyewitness identifications during the fifteen year
period ending in 2003. In four of those cases, exoneration came via subsequent
1

Gross, Jacoby, Matheson, Montgomery & Patil, Exonerations in the United States 1989 Through 2003, 95 J. of
Crim. Law & Criminology 523 (2005).
2
Symposium, The Other Race Effect and Contemporary Criminal Justice: Eyewitness Identification and Jury
Decision Making, 7 Psychol., Pub. Pol’y & Law 3-262 (2001).

2

DNA testing. Additional claims of mistaken identifications leading to wrongful
conviction were called to the attention of the Commission, but we undertook no
independent investigation to verify these claims. The Commission is satisfied that
the risk of wrongful conviction in eyewitness identification cases exists in
California, as elsewhere in the country, and that reforms to reduce the risk of
misidentification should be immediately implemented in California.
In 1998, U.S. Attorney General Janet Reno assembled 34 professionals from
throughout the United States and Canada to form a Technical Working Group for
Eyewitness Evidence. Drawing upon the research of psychologists as well as the
practical perspectives of prosecutors, defense lawyers and police investigators, the
Working Group produced a comprehensive guide for law enforcement to increase
the accuracy and reliability of eyewitness evidence and decrease the numbers of
wrongful identifications. 3 Though the guidelines were not mandated, the
Department of Justice recommendations have been very influential in other states.
In the State of New Jersey, for example, Attorney General John J. Farmer
promulgated Guidelines for identification procedures based upon the U.S.
Department of Justice recommendations, for implementation by all law enforcement
agencies in the state. 4
Many of the recommendations contained in the Department of Justice
Guidelines are already being used in training by California law enforcement. For
example, the Peace Officers Standards and Training Basic Academy Workbook
chapter on identification procedures includes instruction to officers to obtain detailed
descriptions from witnesses, to remain neutral in all identification procedures, to
separate multiple witnesses, and to compose lineups with at least five fillers similar
in appearance to the suspect. 5
One California County has adopted a lineup protocol requiring double-blind and
sequential identification procedures. 6 The Commission learned from Deputy
District Attorney David Angel of the Santa Clara County District Attorneys Office
that under the leadership of District Attorney George Kennedy, all law enforcement
agencies in Santa Clara County agreed to the protocol without dissent, and the
protocol has been successfully implemented for nearly four years without complaint.

3

U.S. Department of Justice, Eyewitness Evidence: A Guide for Law Enforcement, NCJ 178240 (October, 1999).
New Jersey Attorney General Guidelines for Preparing and Conducting Photo and Live Lineup Identification
Procedures (April 18, 2001).
5
Basic Course Workbook Series, Student Materials, Learning Domain 16, Search and Seizure, Version Three,
2006, California Commission on Peace Officers Standards and Training, Ch.6.
6
Police Chiefs’ Association of Santa Clara County, Line-up Protocol for Law Enforcement, Sept. 12, 2002.
4

3

Many of the Commissions established in other states to carry out a mission
similar to our Commission, examining the causes of wrongful convictions and
recommending reforms to avoid wrongful convictions in the future, have
recommended the adoption of guidelines for the conduct of lineups, show-ups and
photo spreads similar to the U.S. Department of Justice Guidelines. This includes
the Governor’s Commission on Capital Punishment established in Illinois, 7 the
North Carolina Actual Innocence Commission, 8
the Innocence Commission for Virginia, 9 and the Wisconsin Innocence Task
Force. 10 In addition, the American Bar Association adopted a Statement of Best
Practices for Promoting the Accuracy of Eyewitness Identification Procedures in
August, 2004, and urged all state and local governments to adopt detailed
guidelines for conducting lineups and photo spreads in a manner that maximizes
their likely accuracy, and to provide periodic training to implement them.
The Commission studied the reports of all of the aforementioned bodies,
and convened a public hearing in San Francisco on March 15, 2006 to hear the
testimony of acknowledged experts, 11 representatives of police, prosecutor and
criminal defense agencies, and concerned citizens regarding eyewitness evidence.
Based upon its consideration of the available research, the testimony of experts,
the experience of Santa Clara County, and the recommendations of other
Commissions, Task Forces and similar bodies, the California Commission on the
Fair Administration of Justice recommends the following guidelines and
procedures:

7

Report of the Governor’s Commission on Capital Punishment, State of Illinois, Recommendations 1-16 (April
2002). The Commission also considered Mecklenburg, Report to the Legislature of the State of Illinois: The
Illinois Pilot Program on Sequential Double-Blind Identification Procedures (March 7, 2006).
8
North Carolina Actual Innocence Commission Recommendations for Eyewitness Identification.
9
Innocence Commission for Virginia, A Vision for Justice, pp. 25-42.
10
Avery Task Force, Eyewitness Identification Procedure Recommendations.
11
Professor Gary Wells, Ph.D., of Iowa State University, Professor Ebbe Ebbesen, Ph.D., of the University of
California at San Diego, Ralph Norman Haber, Ph.D., and Lyn Haber, Ph.D., presented testimony before the
Commission at the San Francisco hearing.

4

RECOMMENDED PROCEDURES:
1. Double-blind identification procedures should be utilized whenever
practicable, so the person displaying photos in a photo spread or operating a
lineup is not aware of the identity of the actual suspect. When double-blind
administration is not practicable, other double-blind alternatives should be
considered.
2. When double-blind procedures are utilized, the use of sequential presentation
of photos and line-up participants is preferred, so the witness is only presented
with one person at a time. Photos or subjects should be presented in random
order, and witnesses should be instructed to say yes, no or unsure as to each
photo or participant. Sequential procedures should not be used where doubleblind administration is not available.
3. A single subject show-up should not be used if there is probable cause to arrest
the suspect. The suggestiveness of show-ups should be minimized by
documenting a description of the perpetrator prior to the show-up, transporting
the witness to the location of the suspect, and where there are multiple witnesses
they should be separated, and lineups or photo spreads should be used for
remaining witnesses after an identification is obtained from one witness.
4. All witnesses should be instructed that a suspect may or may not be in a photo
spread, lineup or show-up, and they should be assured that an identification or
failure to make an identification will not end the investigation.
5. Live lineup procedures and photo displays should be preserved on video tape,
or audio tape when video is not practicable. When video taping is not
practicable, a still photo should be taken of a live lineup. Police acquisition of
necessary video equipment should be supported by legislative appropriations.
6. At the conclusion of a lineup, photo presentation, or show-up, a witness who
has made an identification should describe his or her level of certainty, and that
statement should be recorded or otherwise documented, and preserved.
Witnesses should not be given feedback confirming the accuracy of their
identification until a statement describing level of certainty has been documented.
7. A minimum of six photos should be presented in a photo spread, and a
minimum of six persons should be presented in a lineup. The fillers or foils in
photo spreads and lineups should resemble the description of the suspect given at
5

the time of the initial interview of the witness unless this method would result in
an unreliable or suggestive presentation.
8. Photo spreads and lineups should be presented to only one witness at a time, or
where separate presentation is not practicable, witnesses should be separated so
they are not aware of the responses of other witnesses.
9. Training programs should be provided and required to train police in the use of
recommended procedures for photo spread, show-ups and lineups. The legislature
should provide adequate funding for any training necessitated by the
recommendations of this Commission.
10. Training programs should be provided and required for judges, prosecutors
and defense lawyers, to acquaint them with the particular risks of cross-racial
identifications, as well as unreliable identification procedures, and the use of
expert testimony to explain these risks to juries. The legislature should provide
adequate funding for any training necessitated by the recommendations of this
Commission.
11. The standardized jury instructions utilized in eye witness identification cases
to acquaint juries with factors that may contribute to unreliable identifications
should be evaluated in light of current scientific research regarding cross-racial
identifications and the relevance of the degree of certainty expressed by
witnesses in court.
12. The Commission recognizes that criminal justice procedures, including
eyewitness identification protocols, greatly benefit from ongoing research and
evaluation. Thus, the Commission recommends the continued study of the
causes of mistaken eyewitness identification and the consideration of new or
modified protocols.
In addition, the Commission recommends the enactment of legislation to require
the Attorney General of California to convene a task force in conjunction with
POST, local law enforcement agencies, prosecutors and defense attorneys, to
develop Guidelines for policies, procedures and training with respect to the
collection and handling of eyewitness evidence in criminal investigations by all
law enforcement agencies operating in the State of California. The Guidelines
should be consistent with the recommendations of this Commission, and should
be promulgated to all law enforcement agencies operating in the State of
California. The Task Force should report back to the legislature within one year
6

of the effective date of the legislation, describing the policies or procedures
adopted and the training implemented.
Respectfully submitted,
California Commission on the Fair Administration of Justice:
John K. Van de Kamp, Chair [Response to dissent]**
Jon Streeter, Vice Chair
Sheriff Lee Baca, Los Angeles County Sheriffs Dept.
Chief William Bratton, Los Angeles Police Department
Glen Craig, Sacramento
Jim Fox, San Mateo County District Attorney [Partial dissent]*
Rabbi Allen Freehling, Los Angeles
Michael Hersek, California State Public Defender
Prof. Bill Hing, University of California at Davis
Michael P. Judge, Los Angeles County Public Defender
Bill Lockyer, California Attorney General [Partial dissent]*
George Kennedy, Santa Clara County District Attorney
Michael Laurence, Habeas Corpus Resource Center
Alejandro Mayorkas, Los Angeles
Judge John Moulds, Sacramento
Prof. Cookie Ridolfi, Santa Clara University School of Law
Douglas Ring, Santa Monica
Greg Totten, Ventura County District Attorney [Partial dissent]*
Gerald F. Uelmen, Executive Director
Chris Boscia, Executive Assistant
California Commission on the Fair Administration of Justice
900 Lafayette St., Suite 608, Santa Clara, California 95050
Telephone 408-554-5002; FAX 408-554-5026
Website: http://www.ccfaj.org.

* We respectfully dissent from this Commission’s recommendations
numbers two and eleven.
The debate over the effectiveness of sequential lineups is not yet
settled. Many experts agree that this method produces fewer
accurate identifications. Even more disturbing is new research out
of Illinois which suggests that the sequential lineup procedures may
result in more false identifications. The sequential method appears
7

to be particularly problematic in cases involving children and the
elderly, cases involving cross racial identifications, cases involving
multiple perpetrators, and cases where a suspect has altered his or
her appearance. Given the uncertainty involving the sequential
lineup method, we feel it is premature to recommend these
procedures be adopted by California’s law enforcement officers.
We further object to this Commission’s recommendation calling
for changes to the standard jury instructions. The drafting of
criminal jury instructions has been delegated to the Judicial Council
of California by the Chief Justice which developed the current
instructions with input and review by all interested parties.
Instructions should be neutral, favoring neither party, and the law
requires trial courts to refuse an instruction that analyzes specific
evidence on a disputed question of fact to the benefit of one party or
another or one that informs jurors that particular evidence is in fact
true – or untrue. Thus, we do not believe altering the standard
instruction in order to deal with a special situation represents sound
public policy. We have lodged a letter with this Commission which
presents our objections in more detail.
Greg Totten, Ventura County District Attorney
Jim Fox, San Mateo County District Attorney
Bill Lockyer, California Attorney General

** The Commission has taken into account the dissenters’ objections. The
majority has concluded that the best scientific evidence on hand today
supports the double-blind sequential approach, noting it is the preferred
choice. With respect to jury instructions, the Commission simply asks for
a reevaluation of the jury instructions pertaining to eye witness
identification in light of the best scientific evidence available.
John K. Van de Kamp, Chair

8