Police lineups due for reform

By Marissa Boyers Bluestine

Witness misidentification is a leading cause of wrongful conviction, having contributed to 75 percent of the 273 convictions that have been disproved by DNA evidence nationwide, and nine of Pennsylvania's 11 such convictions. The problem got worldwide attention with this week's execution in Georgia of Troy Davis, which went forward even though seven of nine key witnesses against him had recanted.

Research over the past 30 years has consistently shown that simple reforms can greatly reduce the misleading influences inherent in traditional eyewitness identification procedures. But Philadelphia has yet to reform most of its identification procedures. With the exception of having a dedicated detective for postarrest lineups, Philadelphia's identification procedures are out of line with best practices.

Experts agree that photo and live lineups of suspects should be conducted by officers who are not involved in the cases at hand to avoid any inadvertent improper clues as to the "right" person. But pre-arrest photo arrays here are routinely conducted by investigating detectives who know the identity of the targeted suspects.

Moreover, while best practices require that photo lineups be created using a victim's description of the perpetrator, photo arrays here are created by a computer based on the characteristics of a given suspect - even if the suspect differs from a victim's description in significant ways.

Unless witnesses are warned that a perpetrator "may or may not" be present in a lineup, they often feel the perpetrator must be there and that it's their job to find him. When the perpetrator is not in the lineup, that leads to mistaken identifications. And decades of studies have shown that once a witness makes an inaccurate identification, his memory of the event - and the perpetrator - is irreversibly altered.

District Attorney Seth Williams' office now carefully reviews all eyewitness cases, particularly when there are major discrepancies between a description and the suspect, before deciding whether to prosecute. While this does not address issues that may arise during the identification procedure itself, it is a significant step in the right direction - one that other prosecutors should consider.

As officials in other jurisdictions have realized, other relatively simple and inexpensive changes can ground procedures in science and assure accurate, reliable identifications. Resistance to modifications of police practice is common and understandable, considering that officers are often being asked to change the way things have been done for decades. But experience shows that over time, most officers come to embrace these reforms, because they make it easier to arrest and convict the right people. This is particularly true when reforms are presented not as a matter of blaming the police or questioning their integrity, but of updating practices to reflect the latest science.

Last month, New Jersey's Supreme Court issued a landmark decision requiring major changes in the way courts evaluate and instruct juries about witness identification evidence. Before reaching its decision, the court appointed a special master to review more than 30 years of social science research on memory and identification. The decision is a shining example for the rest of the country, and there are heartening signs that Pennsylvania will follow suit.

After spending years examining Pennsylvania's criminal justice system, the Advisory Committee on Wrongful Convictions, created by the legislature to explore the causes of wrongful convictions, issued a report this week with a series of science-based recommendations on identification procedures. These recommendations are being followed by law enforcement agencies across the country, some for more than a decade, and not one has decided to abandon them.

There's always room for improvement when it comes to ensuring the safety of citizens and upholding the integrity of the justice system. And misidentifications not only threaten the innocent; they also derail investigations.

Moreover, while police focus on building cases against innocent people, the real perpetrators are free to commit other crimes. In 31 cases in which misidentified suspects were exonerated and actual perpetrators were identified, those perpetrators went on to commit other crimes. These include 52 rapes, 17 murders, and eight other violent crimes that took place after the wrongful convictions. And these numbers are quite conservative because they include only violent criminal convictions.

Change is often unsettling, but for the sake of the public good and the many people who have been wrongfully convicted because of witness misidentification, it's time for Pennsylvania to acknowledge the scientific research and embrace policies that will improve public safety while preventing more horrible miscarriages of justice.