Eyewitness identification reforms not always supported by data

The reliability of eyewitness identification can be problematic, as the recent exoneration of a Los Angeles County man convicted of sexual assault demonstrates. Three eyewitnesses identified Luis Lorenzo Vargas as the man who raped them in 1999, and he was sentenced to 55 years in prison. DNA evidence in one of those cases exonerated Vargas in November and implicated the so-called "teardrop rapist," who remains unidentified.

Motivated by hundreds of false convictions like Vargas's, many states and law enforcement agencies have reformed eyewitness identification procedures in order to increase the accuracy and reliability of identification evidence. California's resistance to enact some reforms has been criticized by advocates of procedures adopted in other states. As it turns out, reforms the state did adopt were based on solid science while some it resisted were on shaky ground, psychologist Steven E. Clark writes in a paper published in Policy Matters, a journal of the UC Riverside School of Public Policy.

Clark notes that one of the reforms that California did adopt was a rewording of the pattern criminal jury instructions. Prior to 2006, juries were instructed to consider the certainty of the witness in assessing the accuracy of that witness's identification. In 2006 the instructions were changed to instruct jurors to consider the certainty at the time the witness made the identification. This small change in the wording is consistent with research suggesting that a witness's initial expression of certainty is a good predictor of that witness's accuracy, whereas the witness's certainty expressed much later at the trial is a less useful predictor of accuracy.

The difference between the certainty expressed at the initial identification versus the certainty expressed later has important implications for the Vargas case, Clark says. The victims who identified Vargas at trial had not been as certain of their identifications during earlier lineups. The victims' lack of certainty at the time they made their identifications may have been a mark of error and a clue to Vargas's innocence.

A reform that California resisted would have required police to present lineups sequentially, rather than presenting everyone in the lineup simultaneously. New research in the last three years suggests that California's resistance was warranted. Sequential presentation of the lineup, once thought to increase identification accuracy, may actually decrease identification accuracy.

"The link between false eyewitness identification and false convictions of the innocent has motivated a national movement to reform the police procedures that are used to obtain eyewitness evidence as well as the legal procedures that regulate the use of that evidence," Clark writes in "Eyewitness Identification: California Reform Redux." "California appears to have lagged behind and to have resisted reforms that have been adopted in other states. However, new social science data, theory, and analysis have changed the research landscape, suggesting that the reforms that have been adopted in California were on solid scientific footing whereas some of the reforms that California resisted were indeed on shaky ground."

In "Eyewitness Identification" Clark describes the techniques most commonly used to obtain eyewitness evidence and underlying assumptions about the reliability of these procedures that may not be supported by data. He recommends that lawmakers attempting to apply new research to evolving social policy:

Mandate the use of best practices, but not mandate specific procedures as a matter of law Include sunset provisions that require an evaluation of those best practices and the research upon which they were based Strengthen the research foundation for policy reform, basing policy changes on the body of research rather than one or two studies Clarify the relationship between science and policy. When policymakers are asked to consider reforms based on scientific research they "should not rely on research summaries or opinions of others," but "should respond with four words: Show me the data."

The report can be found at: http://policymatters.ucr.edu/wp-content/uploads/2015/12/pm-vol7-1-eyewitness-reform.pdf