Eyewitness Identification, Democratic Deliberation, and the Politics of Science

• Author: Andrew E. Taslitz
• Position: Professor, Howard University School of Law
• Pages: 271-326

Page 271

Professor, Howard University School of Law; former Assistant District Attorney, Philadelphia, PA.; BA., Queens College, 1978; J.D., University of Pennsylvania Law School, 1981. Professor Taslitz thanks the Howard University School of Law for its support of this project and Debbie Kim, Leah Aden, and Keri Fiore for their expert research assistance.

Introduction

Most of the articles in this symposium focus on the science of eyewitness identification: on how to make lineups, photospreads, and other identification procedures more accurate, thus helping to convict the guilty while acquitting the innocent. In this article, I am more interested in the politics of how the legal system, primarily via the prosecutor, uses social science to craft legal policy in this area. How science is used can be as important as what it teaches.¹ Political goals and scientific teachings may coincide, but they may also conflict, and the political system may choose to draw different implications from scientific research than might the majority of relevant social scientists. This result makes sense, because turning science into law frequently involves normative judgments.²For example, newer recommended identification procedures may lower the risk of convicting the innocent but may also raise the risk of acquitting the guilty. Whether greater protection for the innocent outweighs the dangers to public safety posed by more of the guilty roaming free is a political, not a scientific, judgment, however much some scientists might disagree.³

The procedures by which the legal system translates science into public policy also have much to do with building the legitimacy of, andPage 272 political support for, that policy.⁴According to the dominant political theory, the modern American political system strives to be a deliberative democracy. Deliberative democrats favor open institutions consisting of diverse members engaging in an informed give-and-take about the common good.⁵The reality may be noisy, messy, and unpleasant, but nevertheless, under the right conditions, it can approximate many of the goals of deliberative democrats. The benefits of democratic deliberation are many, including error-correction, catharsis, informed value-articulation, restraint of governmental abuses, and promotion of a shared sense of mission and common peoplehood.⁶Not every institution need be open and diverse in its deliberations. The need for quick action, for effective bargaining skills (which require hiding your real agenda from your opponent) in competitive arenas like the marketplace, and a host of other reasons counsel in favor of using purely internal deliberations. This is true for the early stages of the policy process as well as in case-specific applications of the rules generated by that process-at least by some of our political institutions. Nevertheless, absent a second, more open, diverse, and publicly inclusive deliberative stage, many of the benefits of democratic deliberation will be lost.⁷

This article defends a vision of ongoing deliberative democratic institutions providing policy-creation and oversight in the area of eyewitness identifications. Many institutions can play a role in such schemes. In what follows, however, I focus primarily on the prosecutor and secondarily on the police. A thorough, multi-institutional analysis would be too lengthy for a single article, and as a former prosecutor, that office is what I know best. More importantly, prosecutors play a particularly important role in the eyewitness identification process. Prosecutors are the only lawyers with an ethical obligation to take affirmative steps to avoid convicting the guilty.⁸Moreover, if they lose on suppression motions or face acquittals at trial because of faulty police procedures, they have a strong adversarial incentive to pressure the police into adopting better future investigative methodologies. Furthermore, in many locations, prosecutors play an important role in educating the police. Although it is usually the police who administer lineups, photospreads,Page 273 and other identification procedures, prosecutorial policies and practices can help to shape police behavior.⁹

Indeed, though the police-prosecutor relationship can be adversarial, it is more often cooperative. Thus, it is impossible to talk about the prosecutor's role in the identification process without discussing the police. Furthermore, the pressure for reform is not one-way: it can come from progressive police leadership instructing and persuading prosecutors.¹⁰Accordingly, in this Article I also discuss the role of the police.

My discussion of deliberative democracy in eyewitness identifications uses the American Bar Association (ABA) House of Delegates' August 2004 Resolution on Eyewitness Identification Procedures¹¹(ABA Eyewitness Resolution) as a springboard. The ABA Eyewitness Resolution was designed to be a model of basic principles to guide law reform efforts in individual jurisdictions and to place the nation's most prestigious national bar association's stamp of approval onto the innocence reform movement.¹²

The ABA Eyewitness Resolution recommended specific procedures-which were identified by social scientists-that would be superior to current practice.¹³But it also contained provisions specificallyPage 274 focused on democratic deliberative processes,¹⁴and it is these latter provisions that I want to address. In summary form, those deliberative provisions were, and their significance is, as follows:

1. "[I]nternal mechanisms should be created within police departments and prosecutors' offices periodically to update. . . [sound identification] procedures to incorporate advances in social scientific research and in the continuing lessons of practical experience."¹⁵Properly understood, this statement mandates the creation of procedures for informed, continuing, internal deliberations within police departments and prosecutors' offices. Such an internal form of deliberation is justifiable at an early stage to allow for the free exchange of ideas within law enforcement, an exchange that might be inhibited if the defense had access to those ideas. At the same time, however, studies reveal that closed procedures by like-minded individuals may magnify error. Although informed leadership can minimize these problems, not all leaders are properly skilled, nor can purely internal deliberations achieve all the political benefits of an open, inclusive process. A second public deliberative stage is needed.¹⁶

2. Fortunately, a companion resolution, the ABA Resolution on Innocence and Systemic Remedies (ABA Systemic Remedies Resolution) broadly recommends the creation of institutional methods for improving all procedures in a jurisdiction that affect the rate of convicting the innocent while releasing the guilty, including improving eyewitness identification procedures.¹⁷Although the Systemic Remedies Resolution leaves much to local discretion, the accompanying report lists a variety of options, including one that I will particularly advocate: the creation of Criminal Justice Coordinating Councils structured in ways that promote broadly inclusive democratic deliberation.¹⁸

3. The Eyewitness Resolution further recommends that where practicable, jurisdictions use larger size lineups, thus decreasing the chance that an eyewitness guess will identify an innocent man.¹⁹However,Page 275 no specific number of foils is recommended, and in the past, the potential transition costs of larger lineups meant that police and prosecutors resisted acting in the face of vague, aspirational rules. The more public Criminal Justice Coordinating Council proceedings will, however, put that resistance to a new test.²⁰

4. Although most researchers favor replacing simultaneous identification procedures (the witness views everyone in the line at the same time) with sequential ones (the witness views each suspect and foil individually), persuasive scientific dissenters worry that we still do not know enough about the circumstances under which each technique may be more accurate.²¹Furthermore, some of the research can be read as raising the risk of acquitting the guilty along with lowering the risk of convicting the innocent-a question turning on a normative judgment of how to balance these costs and benefits.²²The ABA thus favors comparative experiments in which one police district uses sequential methods, while another retains simultaneous ones, to promote more informed and continuing deliberation on this subject.²³

5. Finally, the ABA seeks to promote deliberation not only over identification policy, but over its case-specific implementation as well. Thus, the ABA favors increased use of experts on the factors affecting eyewitness accuracy and enhanced jury instructions on that same point, thereby embracing a vision in which informed jury deliberations are the surest path to justice in a particular case.²⁴

Part I of this article defines democratic deliberation, explains its virtues and weaknesses, and analyzes some of the social forces that make effective governmental deliberation harder and that transfer decisionmaking power in the area of criminal justice to the branch that is traditionally the least deliberative branch-the executive. Part I also emphasizes that...