July 2004

Presented to

Daniel F. Conley, Suffolk County District Attorney

Kathleen M. O’Toole, Commissioner Boston Police Department

Task Force Members

Willie Davis, Esquire, Davis, Robinson & White

John Gallagher, Superintendent, Boston Police Department

James Doyle, Esquire, Carney & Bassil

Mary Jo Harris, Esquire Legal Advisor, Boston Police Department

Michael Doolin, Esquire

Paul Joyce, Superintendent, Boston Police Department

Josh Wall, First Assistant District Attorney, Suffolk County G

ary Wells, Ph.D, Iowa State University

EXECUTIVE SUMMARY

On March 8, 2004, the Suffolk County District Attorney, Daniel F. Conley, and the Boston Police Commissioner, Kathleen M. O’Toole, announced the formation of a Task Force on Eyewitness Evidence. The Task Force was charged with reviewing the investigative process for cases in which eyewitness identification was a significant issue, and recommending any appropriate changes in the means and manner of investigation. The Task Force was born of the concern, as evidenced by a series of recently overturned convictions, that better practices in such cases would yield more reliable results and significantly reduce the potential for error.

The Task Force offers 25 separate recommendations. A number of these are, to our knowledge, without precedent in any major city police department and prosecutor’s office. Taken together, these 25 recommendations represent a dramatic leap forward in the manner in which police and prosecutors investigate and prosecute cases and would place Boston and Suffolk County in the forefront in comprehensively addressing an issue of both local and national concern.

The Task Force was co-chaired by Boston Police Superintendent John Gallagher and Suffolk County First Assistant District Attorney Josh Wall. It included three members of Boston’s defense bar: Willie Davis, Esq. of Davis, Robinson and White; Michael Doolin, Esq.; and James Doyle, Esq. of Carney and Bassil. The Task Force also included Gary Wells, Ph.D from Iowa State University, Mary Jo Harris, Legal Advisor to the Boston Police Department and Boston Police Superintendent Paul Joyce.

Professor Wells and James Doyle brought to the Task Force particular expertise in the area of eyewitness evidence. Professor Wells is the internationally recognized academic expert on eyewitness observations, human memory, and identification procedures. He has, for two decades, been the leader of efforts to advance scientific knowledge and integrate scientific procedures with actual police investigative work. James Doyle has taught, written and lectured extensively on eyewitness evidence, identification procedures, and the impact of related scientific research.

Throughout its work, the Task Force was cognizant of the high stakes involved in investigating identification cases and the serious responsibilities the Task Force owes to certain groups. Specifically, the Task Force believed its greatest responsibilities were owed to (1) the victims of crimes who deserve effective prosecution of the actual perpetrator; (2) the eyewitnesses whose honest effort to accurately identify suspects should be made under the best conditions; (3) the investigators and prosecutors who rely on their agencies to provide the most effective and updated procedures; and (4) most importantly, those innocent suspects who pay the penalty for inaccurate identifications.

After an initial review of erroneous conviction cases, the Task Force concluded that improving identification procedures is one of many things that can and should be done to reduce the risks of misidentifications, prosecution of innocent suspects and erroneous convictions. In recognition, however, of the high stakes involved in identification cases as demonstrated by the erroneous conviction cases, the Task Force decided that its initial focus on identification procedures (e.g., photo lineups) needed to expand to include other aspects of investigations and prosecutions. The Task Force recommendations reflect a comprehensive approach that encompasses not only identification procedures, but also investigative and prosecution practices, forensic scientific evidence, training, and post-conviction practices.

The Task Force makes the following recommendations:

Identification Procedures. The Task Force recommends adoption of the identification protocols described in the Department of Justice’s Eyewitness Evidence Guide. Using the DOJ protocols is an important step forward for fair and reliable identification procedures, but the Task Force concluded that there were significant steps that go beyond the DOJ Guide that need to be taken to insure the highest standards for Boston. Scientific research strongly supports sequential identification procedures and blind administration procedures as the best tools available to combat misidentifications. Historically, law enforcement in this country has not used either sequential or blind administration procedures. Research conducted in the 1980s and 1990s built academic support among psychological scientists for these new procedures. Despite the best efforts of some of its members, the DOJ national working group that produced the DOJ Guide could not agree to recommend the two biggest changes: sequential and blind administration procedures. After reviewing the scientific evidence, the Task Force concluded that sequential and blind administration procedures should, in conjunction with the DOJ protocols, be used in Boston as the most effective way to reduce misidentifications. Adopting this level of reform will set a new standard. Task Force member James Doyle, also a member of the DOJ national working group, described for the Task Force that, “this amount of reform has never before been voluntarily undertaken by a major metropolitan police department.”

Prosecution Practices. The Task Force is confident that reform in identification procedures will significantly reduce misidentifications, but also recognizes that even the best procedures cannot eliminate all misidentifications. The human memory, as reliable as it can be, cannot achieve perfection. In response to that reality, the Task Force proposes an approach that, to the best of our knowledge, has never been done before. Specifically, the Task Force recommends that the District Attorney’s office adopt written policies for identification cases. Reform efforts in other jurisdictions have not included written reforms for the prosecutor’s office. Our recommendations for prosecution practices are designed to give prosecutors greater abilities to investigate cases in a manner that will expose misidentifications and prevent those cases from proceeding. With that goal in mind, the Task Force recommends (a) specific, written instructions for prosecutors to use in investigating and prosecuting identification cases, (b) obtaining an attorney for every suspect who participates in a live lineup, (c) establishing an Eyewitness Evidence Committee of senior prosecutors to review identification cases, (d) require that every identification case brought in Superior Court be approved for prosecution by either the First Assistant or the Chief of Homicide, (e) continue the policy of the District Attorney’s office not to oppose post-conviction requests for DNA testing of relevant evidence, and (f) continue the District Attorney’s DNA Committee of senior prosecutors to review and make decisions on post-conviction motions that involve requests for DNA testing.

The following list, assembled in categories, includes each specific recommendation made by the Task Force.

Identification Procedures

1. Adopt in full the recommendations on eyewitness evidence set forth by the United States Department of Justice.

2. Use sequential presentation of photographs in photo lineups.

3. Use sequential presentation of persons in live lineups.

4. Blind administration of photo lineups, which requires that the photo lineup be shown by an investigator who has no knowledge of which photograph is the suspect’s.

5. Blind administration of live lineups, which requires that the live lineup be administered by an investigator who does not know which person is the suspect.

6. Use a standard printed form which provides eyewitnesses essential instructions for viewing a photo lineup or a live lineup.

7. Require detailed documentation, by means of a separate report, of every identification procedure conducted by the Boston Police.

8. Adopt as policy that every photo lineup shall consist of 8 photographs (7 fillers and 1 suspect) and that every live lineup shall consist of 6 persons (5 fillers and 1 suspect).

9. Use a live lineup subsequent to a photographic array in certain cases where testing the witness’ ability to make an in-person identification could be of significant evidentiary value.

10. Use booking photograph compilations (“mug books”) of scores of people as a possible source of identification only when all other investigative leads have been exhausted.

Investigative Practices

11. Ensure that the Boston Police Crime Laboratory at all times be maintained with excellent equipment and resources, and a full staff of trained and experienced scientists.

12. Ensure that the forensic technology units of the Bureau of Investigative Services — Identification (fingerprint) and Photography and Ballistics — receive certification from national associations.

13. Adopt a policy requiring the electronic recording of statements made by consenting suspects in custodial settings, and for that purpose providing professional quality recording equipment throughout the police department.

Prosecution Practices

14. Adopt instructions for prosecutors for investigating and prosecuting identification cases.

15. Adopt a policy that an attorney be provided for each suspect who participates in a live lineup.

16. Establish in the District Attorney’s office an Eyewitness Evidence Committee of senior prosecutors to review investigations of identification cases and charging decisions, and to direct training efforts.

17. Require by formal policy that every identification case brought in Superior Court be approved for prosecution by either the First Assistant or the Chief of Homicide.

18. Continue the policy of the District Attorney’s office not to oppose post-conviction requests for DNA testing of relevant evidence.

19. Continue the District Attorney’s DNA Committee, an existing committee of the office’s senior prosecutors to review, monitor, and make decisions on post-conviction motions that involve requests for DNA testing.

Joint Policies of the Agencies

20. Adoption by both agencies of formal policies supporting extensive and thorough use of the Grand Jury to develop and document evidence in indictable cases.

21. Adoption by both agencies of formal policies supporting extensive cooperation and collaboration between police and prosecutors from the outset of a felony investigation.

22. Adoption by both agencies of practices of extensive and thorough supervision of investigators and prosecutors by experienced supervisors in their respective agencies.

Training and Review

23. Initial and ongoing training of police officers, detectives, and prosecutors on identification procedures, investigation of identification cases, and prosecution of identification cases, including joint training where appropriate.

24. Annual internal review, for at least the next two years, of these recommendations to evaluate the implementation and the effectiveness of new procedures and policies.

25. Continuation of this Task Force for the purposes of (a) monitoring the adoption of these recommendations and (b) suggesting any revisions or additional recommendations that would improve the investigation and prosecution of identification cases.

II. Erroneous Conviction Cases

The Task Force reviewed the misidentification cases that have resulted in erroneous (or unfair) convictions in Boston in the past 15 years. The number of erroneous convictions has been disturbing and unacceptable for anyone interested in the fair administration of justice in Boston. There has not, however, been an investigation leading to an erroneous conviction since 1997 (Stephen Cowans). The Task Force reviewed the erroneous conviction cases and compared the practices used in those cases with the current practices of the Boston Police and the Suffolk District Attorney. In reviewing the erroneous conviction cases, therefore, the Task Force had two questions: (1) were there reasons that the rate of erroneous conviction cases has declined and can those reasons, once identified, contribute to our recommendations? (2) Were there actions, events, or mistakes in the erroneous conviction cases that would suggest or compel certain reforms?

a. DNA Cases. It is important to know that the erroneous conviction cases fall into two categories: DNA and non-DNA. The DNA cases are primarily sexual assault cases (Marvin Mitchell, Anthony Powell and Neil Miller) with biological evidence left by the perpetrator. When the cases were investigated up to 15 years ago, DNA was not available and admissible in court. Once DNA became available and admissible, the preserved biological evidence underwent DNA testing and the results were used in post-conviction proceedings to exonerate Marvin Mitchell, Anthony Powell and Neil Miller.

The advances in DNA science mean that defendants in recent and future sexual assault cases will not be convicted for lack of forensic scientific testing. Current investigative practices include testing of biological evidence as part of the investigation. Since the mid-1990s, DNA testing has been a feature of the investigation rather that just the post-conviction proceedings.

The Boston Police Crime Laboratory is fully accredited and nationally recognized for its accuracy, efficiency and professionalism. Significant improvements made in the last ten years have brought the Crime Lab to this level of competence and reliability. The Crime Lab’s work with DNA and other types of physical evidence will continue to be a critical piece in the investigation of identification cases. For that reason, the Task Force recommends that the Crime Laboratory at all times be maintained with excellent equipment and resources, and a full staff of trained and experienced scientists. Nearly half of the erroneous conviction cases have not involved DNA testing. In these “non-DNA cases,” it was additional investigation, brought on by new information, that led to vacating the convictions. The post-conviction investigation that freed the wrongly-convicted defendants in non-DNA cases was done primarily, and in some cases entirely, by police and prosecutors. In reviewing the erroneous conviction cases, both DNA and non-DNA, the Task Force concluded that all the forensic sciences are critically important. In any given case, it might be the fingerprint evidence or the ballistics evidence that provides the most probative evidence implicating or clearing a suspect.

The Boston Police Crime Lab is the model of excellence for the department’s other forensic units. Supervisors in the Crime Lab set national certification as their goal and improved all aspects of the Lab’s operation in order to meet that goal. Certification would serve the same purpose for the other forensic units. The Task Force, therefore, recommends that the forensic technology units of the Bureau of Investigative Services – (a) Identification (fingerprint) and Photography, and (b) Ballistics – receive certification from national associations.

b. Lessons Since 1997.

The Task Force compared current investigative procedures with the investigations conducted in the older exoneration cases to determine if there were changes that, in combination with the availability of forensic DNA testing, have lessened significantly the pace of erroneous convictions. The Task Force identified procedures, implemented by the Boston Police and the Suffolk District Attorney in the mid to late 1990s, that significantly improved investigations and prosecutions. The Task Force also found, however, that these changes have been widely but not uniformly implemented, and the Task Force recommends the following practices – essential for fair and reliable investigations – be adopted as formal policy throughout the Police Department and the District Attorney’s office:

1. extensive cooperation and collaboration between police and prosecutors from the outset of the investigation;

2. extensive and thorough use of the Grand Jury to develop, preserve, and document evidence;

3. extensive and thorough supervision of investigators and prosecutors by experienced supervisors.

The case of Tiffany Moore’s murder in 1988 is the textbook example of what can happen when the investigation includes little communication between police and prosecutors, vague or distant supervision, and minimal Grand Jury work. Shawn Drumgold was arrested, indicted and convicted for the murder of the 12-year old girl. Drumgold’s motion for new trial led to an extensive evidentiary hearing in 2003, 4 years after the trial. At the hearing, police and prosecutors testified that they had little or no communication and collaboration during the investigation of the case and there was little or no meaningful supervision or case review (even in this high profile case) that would assist or direct the assigned investigator or prosecutor. Additionally, rather than using Grand Jury as means of presenting the important civilian witnesses (including several identification witnesses), the prosecutor presented only two witnesses (one being a police detective who presented abundant hearsay as a substitute for live witnesses).

The Task Force concluded that the inadequate practices in the 1988 investigation significantly contributed to a trial which convicted the defendant but was determined years later to have been conducted unfairly. The motion for new trial was allowed in 2003 and the District Attorney, in the interests of justice, decided not to retry the case. The Drumgold case is but one example of what the Task Force observed as routine practice in the earlier era that produced numerous erroneous convictions: the police investigated the case without communication and input from the prosecutor; the prosecutor used the Grand Jury merely to secure an indictment rather than to develop and document evidence, and both the police investigator and the prosecutor operated without adequate supervision and direction.

Beginning in the mid-1990s and emerging through the late-1990s, the Boston Police Department and the District Attorney’s office developed investigative practices that included cooperation, communication, collaboration, supervision and review, and thorough Grand Jury presentations. The Task Force believes that these improved practices, in combination with DNA forensic testing, are two significant reasons for the decline in erroneous conviction cases since 1997. To build on these improvements, the Task Force recommends that these practices be expanded and adopted as policy so that they are used consistently and uniformly in all cases by all investigators and prosecutors.

Supervision and communication lead to the same desired effect: more professionals involved in more discussion and review of cases. The opportunity for a greater exchange of ideas by investigators and prosecutors with different perspectives will produce more objective and more thorough investigations. An investigation that is objective and thorough will reduce the chances of an erroneous conviction based on misidentification.

The importance of Grand Jury work cannot be overstated. The presentation of witnesses to the Grand Jury has become an essential part of investigating, preserving, and documenting the facts of the case. Witnesses testifying under oath and on the record is an excellent method of making accurate and reliable information available to defendants, defense counsel, judges and jurors. A thorough Grand Jury presentation gives every person with an interest in the case the opportunity to review, analyze and critique the credibility and accuracy of the witnesses, the thoroughness and professionalism of the investigation, and the strengths and weaknesses of the case. Presenting the critical eyewitnesses and investigating police officers to the Grand Jury is a remarkably important guard against misidentifications and erroneous convictions.

The Task Force recommends that these three practices, effective in their use in recent years, become more consistent, uniform and permanent through policies and training. In short, it should not be an option to avoid cooperation between agencies, communicating with supervisors, or using the Grand Jury.

c. Lessons From Erroneous Conviction Cases. The Task Force examined the misidentifications made in the erroneous conviction cases with the expectation that there would be patterns that connected the identification errors. There was, however, no widespread characteristic or pattern that appeared in or explained a bulk of the cases. For example, a pattern of cross-racial misidentifications might be expected. There was, however, no such pattern as most identifications involved eyewitnesses identifying suspects of their own race. As another example, some might expect a pattern of admissions by the defendant that were not electronically recorded. Not one erroneous conviction case included an unrecorded interview containing admissions by the defendant.

The Task Force recognized, however, the significance of suspect interviews. Whenever a suspect decides to make a statement in a custodial interview, the suspect’s statement is likely to be of critical importance in the investigation. The interview should be recorded. Recording requires (a) the consent of the suspect, as required by statute, and (b) the investigator’s judgment, based on training and experience, on when to seek the defendant’s consent and begin recording. Investigators describe and common sense teaches that many suspects will not speak freely while being recorded. For those suspects, investigators must develop rapport, encourage communication and obtain information before recording can commence. The objectives are to develop as much information as possible and document that information. Both objectives must be met without sacrificing one for the other. The Task Force recommends required electronic recording of statements made by consenting suspects in custodial interviews, and for that purpose professional quality recording equipment should be provided throughout the police department. Without a pattern that explained several cases, the Task Force concluded that the issues were more complicated and that a comprehensive approach and set of recommendations for identification cases were necessary. The lack of a pattern is one of the factors that led the Task Force to expand its focus beyond a set of identification procedures to include more comprehensive recommendations.

III. The Scientific Approach to Eyewitness Evidence

In studying eyewitness evidence, Professor Wells uses an analogy that assisted the Task Force in its analysis and recommendations. Professor Wells likens eyewitness evidence to physical trace evidence. Physical trace evidence, such as fingerprints, fibers or blood, can help determine the facts of a crime and the identity of the perpetrator. The observations of an eyewitness are items of trace evidence contained in the witness’ memory. Like physical evidence, memory trace evidence can be contaminated, lost, destroyed or otherwise made to produce inaccurate results. Like physical trace evidence, the manner in which memory trace evidence is collected can have important consequences for the accuracy of the results.

The Task Force concluded that a more scientific approach to collecting and analyzing eyewitness evidence should be the guiding principle for our recommendations. Police protocols for the collection, preservation and interpretation of physical evidence are dictated largely by forensic scientists, and the practice of physical evidence collection and examination has tried to borrow as much as possible from science. The analysis of physical evidence, especially biological traces, has advanced rapidly in the past decade. In fact, it is the advancement in DNA forensic science that has made evident the weaknesses in eyewitness evidence. These recommendations for eyewitness evidence are based on a scientific model. All members of the Task Force — representing police, prosecutors, the defense bar, and psychological scientists — agree that the recommendations will improve the justice system’s ability to collect, preserve and analyze eyewitness evidence. Our goals are to (1) make it less likely that an eyewitness will make a mistaken identification, (2) give police and prosecutors investigative practices that will expose mistaken identifications, and (3) increase the reliability and strength of cases that are based on accurate identifications.

IV. Collecting and Preserving Eyewitness Evidence

The failure of the criminal justice system to adopt a scientific model for eyewitness evidence is attributable, in part, to a lack of understanding of how human memory works. Many assume that information stored in memory remains largely unchanged over time, is impervious to suggestion, and that memory failures are primarily failures to retrieve information. In fact, however, memory can be influenced by post-event information, is susceptible to suggestion, and can fail or err in many different ways. Although research established that mistaken identification rates increase under certain conditions, research also established that many of these conditions could actually be avoided by the use of more scientific procedures for photo lineups and live lineups. The Task Force makes the following recommendations for the scientific collection and preservation of eyewitness evidence:

1. Adopt the recommendations of the Eyewitness Evidence Guide developed by the Department of Justice;

2. Adopt a sequential procedure for photo lineups and live lineups; and

3. Adopt blind administration procedures for photo lineups and live lineups.

In recommending these new procedures, the Task Force is not criticizing the current identification procedures of the Boston Police Department. The current procedures have been used objectively and fairly by Boston detectives for many years. The procedures fully comport with federal and state constitutional requirements. The current procedures have produced countless reliable and accurate eyewitness identifications. As Professor Gary Wells has stated, “Boston’s previous procedures were already at or above the national level and there is no reason to think that Boston’s old procedures placed innocent suspects at a unique level of risk.” The Task Force in no way intends that these recommendations be used to undermine reliable identifications and reliable cases built on current procedures. The Task Force does intend that these recommendations be used to improve on the current constitutionally-sound practices.

DOJ Guide.

The Task Force recommends that the Boston Police Department adopt the recommendations of the Eyewitness Evidence Guide developed in 1999 by the United States Department of Justice’s Technical Working Group for Eyewitness Evidence. The DOJ procedures are based on scientific research. The DOJ Guide gives procedures for the 911 operator, the responding officer, and the investigating detective. The procedures include the best methods for getting accurate information from eyewitnesses about the incident and the suspect’s description, and also provide specific steps for identification procedures.

Many of the steps described in the DOJ Guide are currently in use by some detectives and some units of the Boston Police. The steps are not, however, in use in their entirety throughout the department. The DOJ recommendations need to be the focus of training and implementation department-wide. Superintendent John Gallagher of the Task Force has drafted a departmental Rule that incorporates the recommendations. Adoption of the Rule will implement the protocols described in the DOJ Guide. Some of the more significant recommendations are described in the remainder of this section.

(a) The manner of questioning the witness at the scene and in the follow-up interview will affect how much accurate information is obtained. Information should be gathered through open-ended questions (e.g., “What can you tell me about the car?); augmented by closed-end questions (e.g., “What color was the car?”); avoiding suggestive or leading questions (e.g., “Was the car red?”). All information obtained from the witness is documented in a written report.

(b) The DOJ Guide also describes instructions that the lineup administrator should give to each eyewitness who views a photo or live lineup. The Task Force recommends that the Boston Police use a printed set of instructions for every lineup identification procedure; and toward that end Superintendent Gallagher developed a standard printed form. The standard form will include those instructions developed from scientific research that create the optimal conditions for the eyewitness. By using a standard form, signed by the witness, the Boston Police will have procedures that ensure that every eyewitness is given the identical set of effective instructions.

(c) By adopting the DOJ recommendations, the Boston Police will also be taking the significant step of asking for and recording the witness’ own statement about his level of confidence in an identification if one is made. Because research demonstrates that a witness’ confidence can be increased over time by extraneous factors, it is important to ascertain and document the certainty level at the time of the identification, before other factors intervene.

(d) The DOJ Guide also includes detailed procedures for conducting showup identifications, which are conducted with witnesses by presenting a suspect in person shortly after the incident. The opportunity to see a suspect in person within minutes of the event can lead to important evidence, but procedures are necessary for fairness and reliability given that there are no fillers. The Task Force notes that no erroneous conviction case in Boston was based on an inaccurate showup identification, which provides some anecdotal evidence for the reliability of showup identifications.

(e) Mug books (collections of photos of previously arrested persons) should be viewed only if all other reliable sources and leads have been exhausted. Viewing repeated images has a tendency to blur the witness’ memory of the actual perpetrator’s face. Results of a mugbook identification should be evaluated with caution.

(f) In selecting fillers for a photo lineup or live lineup, complete uniformity of features is not required. Avoid using fillers who so closely resemble the suspect that a person familiar with the suspect might find it difficult to distinguish the suspect from the fillers. Selecting a filler whose appearance is nearly identical to the suspect’s focuses the witness’ attention on the similar photos or persons, thus creating a suggestibility factor.

In summary, the DOJ recommendations adopted by the Boston Police will mean that investigators are developing and documenting reliable information from eyewitnesses, preparing fair and effective photo and live lineups and, through standard printed forms, providing optimal, scientific conditions for eyewitnesses involved in an identification procedure.

Sequential Photo Lineups and Live Lineups.

The common current practice for photo lineups and live lineups is to present to the eyewitness all photographs or lineup members at one time. This practice is known as a simultaneous array. Scientific research has demonstrated that, under simultaneous conditions, eyewitnesses tend to compare photographs and lineup members with each other to determine which one most closely resembles the perpetrator relative to others, a process called relative judgment. Professor Wells devised an alternative procedure — a sequential presentation — that would reduce the tendency of eyewitnesses to rely on relative judgments. In the sequential procedure, the eyewitness is presented with one photograph or one lineup member at a time, and the eyewitness must decide on each photograph or person before viewing the next photograph or person. Research demonstrates that sequential procedures would result in significantly fewer misidentifications. The Task Force recommends that photo lineups and live lineups be conducted using the sequential procedure.

Blind Administration of Arrays and Lineups.

With blind administration of identification procedures, the person conducting the photo lineup or live lineup is not aware of which photograph or person is the suspect. The importance of blind administration procedures stems from what researchers call the confidence malleability of the eyewitness. Confidence malleability describes an effect that has been repeatedly demonstrated in scientific research – that the eyewitness’ confidence in his lineup pick can increase in a variety of different ways. For example, increased confidence can come from confirmation (“You picked the right suspect”), repeated questioning about the incident or the identification, or any number of non-verbal and inadvertent cues.

There is significant research demonstrating confidence malleability, and the confidence of the eyewitness is often an important factor at trial. The eyewitness often will testify about his degree of confidence, and judges specifically instruct jurors to consider the degree of confidence expressed by the eyewitness. An eyewitness’ confidence, therefore, should be based on his own memory rather than on external information. The importance of blind administration is well-accepted in psychological scientific research of all types. It is standard scientific practice to conduct scientific research using administrators who do not know specific information about the test they administer. The scientific technique should be used for actual identification procedures to eliminate the possibility that a witness’ confidence will be influenced inadvertently by the administrator. After nearly 25 years in eyewitness scientific research, Professor Wells describes blind administration as the single most important step a department can take to improve its identification procedures.

V. Analyzing Eyewitness Evidence The reforms recommended in the previous section will remove many conditions that lead to eyewitness misidentifications. Those reforms will not, however, eliminate all eyewitness error. As demonstrated by scientific research and erroneous conviction cases, some eyewitness misidentifications will occur even when the best procedures have been followed. The Task Force, therefore, devoted considerable thought to the following question: Are there new ways of analyzing identification evidence that would give prosecutors greater ability to expose misidentifications? Answering that question led the Task Force into uncharted territory; eyewitness evidence reform undertaken in other jurisdictions and in the DOJ Guide has not included prosecution practices.

The Task Force concluded that charting this territory for prosecutors was critically important. As one Task Force member described it, “How do we ask police to accept important levels of change without asking the same thing of prosecutors? In this era of erroneous convictions, if any serious prosecutor does not think that lawyers can do a better job with identification cases, then we have a problem, caused by naivete or arrogance, that has to be addressed.” Having made the decision to expand its focus to include prosecutors, the Task Force makes the following recommendations:

1. Adopt instructions for prosecutors for investigating and prosecuting identification cases.
2. Adopt a policy that an attorney be provided for each suspect who participates in a live lineup.
3. Establish in the District Attorney’s office an Eyewitness Evidence Committee of senior prosecutors to review investigations of identification cases and charging decisions, and to direct training efforts.
4. Continue the practice and require as written policy that every identification case brought in Superior Court be approved for prosecution by either the First Assistant or the Chief of Homicide.
5. Continue the policy of the District Attorney’s office not to oppose post-conviction requests for DNA testing of relevant evidence.
6. Continue the District Attorney’s DNA Committee, an existing committee of the office’s senior prosecutors to review, monitor, and make decisions on post-conviction motions that involve requests for DNA testing.

Instructions for Prosecutors.

Investigators and prosecutors need training and experience to analyze an eyewitness identification. The traditional analysis focuses and ends on the ability of the witness to observe and perceive the perpetrator and includes the following factors: the duration of the assault, the witness’ focus on the perpetrator’s face, the lighting conditions, the confidence of the witness, and any aspect or condition of the victim that might affect his or her ability to make observations (such as age, intoxication, or eyesight). These factors continue to be important for investigators and prosecutors, and jurors are instructed that these are the factors they should consider in evaluating eyewitness identification. In addition, however, police and prosecutors need to perform a broader analysis that includes more than the conditions under which the observations were made.

The need for proper analysis is sufficiently critical that the Task Force recommends instructions for the prosecutors who, through their work with the police, assist in the investigation and, through their work in the Grand Jury, develop and document the investigation. It is not sufficient to change the way police conduct identification procedures without also setting instructions for prosecutors to analyze the resulting identifications and integrate the eyewitness identification with the other facts of the investigation. It is expected that these instructions will be used in training, supervision and, most importantly, the daily work of prosecutors who are investigating identification cases. The recommended instructions are as follows:

INSTRUCTIONS FOR PROSECUTORS:

INVESTIGATING AN IDENTIFICATION CASE

The police investigator is responsible for the investigation of a case. The prosecutor, however, has ultimate responsibility for the cases that are prosecuted. The prosecutor, therefore, should participate in the investigation in every way that is productive. These instructions are designed to assist prosecutors in developing facts in identification cases, analyzing the facts, and making decisions on which suspects to charge.

1. First and foremost, approach each case objectively and with a critical view of the evidence. Historically, if a witness says he is confident in an identification, the investigators and the prosecutor have adopted a similar confidence in the strength of the case. The confidence of the witness should not be the primary reason for the prosecutor’s confidence in the case. Always be open to the possibility that an identification is mistaken, and develop the evidence with that possibility in mind. The witness’ identification is only one fact in the investigation; it must be weighed and considered with and against all the other facts. The value of the identification is determined by the other facts and circumstances established in the investigation.

2. The prosecutor must be vigilant and adhere to the strictest standards in the following six areas: (1) definitively establish the description which is initially given by the witness; (2) definitely establish every fact concerning the identification procedure; (3) know the crime scene and physical evidence inside and out; (4) thoroughly develop and investigate all circumstances — positive and negative — which are relevant to the perpetrator’s identity; (5) establish and document all this information as early as possible in the investigation; and (6) use the Grand Jury to develop and document all the evidence concerning the description, the ID procedure, the crime scene, and the circumstantial evidence relevant to the identity of the perpetrator. At all stages, remember that you are working not to build a case against a particular suspect, but to develop facts that identify the actual perpetrator.

To accomplish these goals, take the following steps:

1. It is essential to speak as soon as possible with any police officer who had contact with the identification witness. Examine, but do not rely solely on, the Incident Reports and Supplemental Reports. Make it a priority to interview all officers who spoke with the witness, including those officers who did not identify their involvement by filing reports. The goal is to know everything about what happened at the scene, the description given and the ID procedure (including the exact words used by the investigator and the witness).
2. In establishing the description given by the witness, be careful to talk to both uniformed officers and detectives. Both will probably have relevant information.
3. With a show-up identification, assume that several officers participated. For example: two officers met the witness at the scene, two different officers transported her to the show-up site, and three other officers conducted the show-up. From those seven officers, there exists a single Incident Report. It is up to you to work with all the officers involved to get all the information in an accurate and documented form. A critical part of the information is the actual words used by the officer and the witness (and the suspect).
4. With a photo lineup identification, it is often not immediately apparent (and sometimes not documented) how the suspect’s photograph got into the array. You must know the answer to that question. The decision to include a suspect’s photo is often based on good circumstantial evidence, and sometimes based on something less probative. Including a photo without articulable reasons for doing so can be the key point in turning an identification case into a misidentification case.
5. Get the 911 tape and turret tape immediately. Make sure the turret tape includes all channels that the officers used in communicating with each other. (The specific information on channels needs to be obtained directly from the officers involved.) These recorded conversations are frequently the most accurate source for the initial description, especially when the identification witness made the 911 call.
6. Interview the identification witness as soon as possible. Open-ended questions are critical. Do not try to lead the witness into echoing what the police report states. After the witness provides a narrative, focus the witness on specific areas in order to get the important and necessary details.
7. The initial interview is the time to examine the variables that affect the witness’ ability to perceive the perpetrator. These variables — time, distance, positioning, focus, lighting, etc. — must be established.
8. Go to the scene.
9. Pay close attention to the crime scene (and the physical evidence contained therein) as described by witnesses and police and as documented by reports and photographs. Get photographs developed as soon as possible. If no scene photos were taken, get some taken. Review Crime Lab evidence receipt. Discuss the evidence with the criminalist at the Crime Lab. Know the scene inside and out.
10. Consider what evidence needs to be tested for the presence of DNA. Consult with the investigator and criminalist concerning testing of physical evidence.
11. Consider what evidence needs to be fingerprinted.
12. Evaluate all ballistics evidence for possible testing.
13. Take good notes of all your interviews, observations, and impressions. Notes will be invaluable in your Grand Jury and trial preparation.
14. Corroborate the identification as thoroughly as possible through circumstantial evidence. This can take many forms — too many to describe specifically — but its importance cannot be overestimated. The strengths and weaknesses of the circumstantial evidence are key factors in assessing the accuracy of the identification. The circumstantial evidence always informs, and sometimes directs, the charging decision.
15. Consider whether an investigation would benefit from having a witness who identified a suspect in a photo lineup also view a live lineup containing the suspect. This additional step would be taken when a second identification procedure might yield critical information about the witness’ ability to make an in-person identification.
16. Consider what information a “non-identifying eyewitness” (one who fails to pick the suspect or picks a filler) is providing. The non-identification may mean that the witness did not have a sufficient opportunity to observe, but it might also mean that the suspect is not the perpetrator.
17. Follow up on all alibi information: interview witnesses, put them before the Grand Jury, obtain documents, evaluate circumstantial evidence. If defendant is represented by counsel, ask counsel for all information concerning the alibi so it can be evaluated and presented to the Grand Jury. Be aware of the possibility that a witness who is biased (e.g., suspect’s mother or girlfriend) may be telling the truth.
18. If he’s not yet charged, make sure suspect gets interviewed.
19. Establish the facts, don’t try to change the facts.
20. Follow alternative paths and investigate alternative suspects as far as the evidence leads.
21. In going through the process of gathering and documenting information and evidence, continually evaluate and analyze the facts. Consider different scenarios, including ones consistent with misidentification. Try to poke holes in your own theories about the case. Compare each witness statement to the statements made by others and the physical evidence.

When it is time to make final charging decisions, accept the evidence you have developed and give the evidence the weight it merits. If the investigation has established no circumstantial evidence to support a one witness photo identification in a stranger situation, you must proceed with extreme caution given what we now know about the possibility of misidentification. If the prosecutor thoroughly investigates the case, thinks critically about the facts, and bases charging decisions solely on the facts, the prosecutor will have taken the most important steps to reduce the risk of an identification case becoming a misidentification case.

Attorney For Live Lineup.

Because the live lineup requires the suspect to participate, the Task Force concluded that the suspect would benefit from the presence and advice of an attorney. The suspect’s attorney, of course, would have no authority to direct investigators and prosecutors in the performance of their duties. The attorney would, however, be able to give the suspect information and advice, and would be able to make observations that might be of value when filing a motion to suppress the identification or cross-examining witnesses at hearing and trial.

Applying Consistent Standards.

The Task Force considered whether, in addition to giving instructions for use by individual prosecutors, there could also be changes in management and supervision that would ensure that uniformly high standards were met in every identification case. Toward that goal, the Task Force recommends that the District Attorney establish an Eyewitness Evidence Committee of senior prosecutors to review investigations of identification cases. The Eyewitness Evidence Committee would review identification cases that involved complex investigations, difficult issues and close calls on charging decisions. The Committee will combine the skills and experience of several prosecutors and apply that expertise to the most challenging cases.

As a complement to the Eyewitness Evidence Committee, the Task Force also recommends, as a management policy, that every identification case brought in Superior Court be approved for prosecution by either the First Assistant District Attorney or the Chief of Homicide. Centralized approval will mean that every identification investigation and every decision to charge a suspect will have to meet certain high standards. Given what is at stake in identification cases, the District Attorney should require that cases are reviewed and approved at the highest level.

Post-Conviction Practices.

A prosecutor’s pursuit of justice does not end with a conviction. Erroneous conviction cases around the country have shown prosecutors using a variety of approaches to post-conviction cases involving untested evidence. These practices run from adamant opposition, to willful inaction, to cooperation and pursuit of possible exculpatory evidence. The Suffolk District Attorney’s office has conducted its post-conviction work ethically and professionally in the erroneous conviction cases. Two practices from recent years have been particularly effective and should be adopted as permanent, written policies.

First, the District Attorney’s policy not to oppose post-conviction requests for DNA testing of relevant evidence gives every convicted defendant the chance to use DNA testing when it might have a chance of yielding exculpatory results. This policy should be permanent and uniformly applied.

Second, the District Attorney’s DNA Committee, comprised of senior prosecutors with significant experience in DNA issues and post-conviction cases, should be made permanent and continue to review, monitor and make decisions in post-conviction motions that involve requests for DNA testing. The DNA Committee has provided in recent years an important internal forum for analyzing challenging issues, evaluating new evidence, and determining the fairness of a conviction in light of new information. Its work is of sufficient important that the DNA Committee should be a permanent part of the agency.

VI. Training and Accountability

The Task Force recommendations will be most effective if implemented through proper training and monitored with meaningful oversight. The Task Force, therefore, makes the following recommendations directed towards implementation and accountability.

1. Initial and ongoing training of police officers, detectives, and prosecutors on identification procedures, investigation of identification cases, and prosecution of identification cases, including joint training where appropriate.

2. Annual internal review, for at least the next two years, of these recommendations to evaluate the implementation and the effectiveness of new procedures and policies.

3. Continuation of this Task Force for the purposes of (a) monitoring the adoption of these recommendations and (b) suggesting any revisions or additional recommendations that would improve the investigation and prosecution of identification cases.