Eyewitness identification. In a tale of two competing versions of the facts, it can be the best of evidence and the worst of evidence. Best for an aggressive prosecutor. Worst for an innocent person accused of a capital offense.
The wrongful conviction case of Anthony Dansberry certainly illustrates the persuasive power of eyewitness identification and the potential it has to lead to an unjust outcome if it is not free of influence and bias. Dansberry’s case — at one time a capital case — cries out not only for his exoneration, but also for change in the investigation process. That is if we truly want to protect the innocent.
A clemency petition has been filed by lawyers at Northwestern University’s Center on Wrongful Convictions on behalf of Dansberry, who is currently serving a 75-year sentence for the murder of 77-year-old Edna Abel of suburban River Forest. Among the issues raised by the 1996 conviction is the reliability of eyewitness identification.
In 1991, an eyewitness picked out Dansberry’s image from a photo array and then pointed him out in a police lineup. For a jury sitting in judgment — particularly in judgment of an African-American man identified by a white woman as an assailant — little more needs to be heard.
There is quite a bit more, though, that should be considered. That is especially true when the photo is included in a group and the review is administered by an officer who knows which person already has been singled out as the suspect.
The first problem is what social scientists call the cross-race effect. It is a behavioral thing. It is common. It affects all racial groups when identifying members of other racial groups. Because, well, they all look alike. It is why such eyewitness identification is unreliable. Vanderbilt University psychologist Daniel Levin, a proponent of the concept of cross-race recognition deficit, has written that people tend to “emphasize information specifying race” more than individual characteristics across racial groups. So, when a white witness is shown an entire page of photos of African-American men, it is difficult for the witness to select the correct individual. That is, unless there are other factors introduced that might influence her.
One of those factors is the officer administering the process. That officer’s awareness of the targeted suspect can be communicated to the witness, sometimes inadvertently. The mere suggestion that one member of the standard simultaneous photo lineup is the assailant can sway a witness. Often, a “positive” ID is met with some kind of confidence boosting confirmation by the administering officer — again, even if only inadvertent — making it even more likely the eyewitness will pick out the targeted suspect in the actual lineup. Sometimes the process is not so inadvertent.
In Dansberry’s case, two witnesses who did not identify Dansberry in a lineup reported police attempted to steer them in his direction. According to a video produced by CWC lawyers, one eyewitness, Donald Walsh (now deceased), asserted an officer pointed to Dansberry and asked him, “Are you sure he’s not the one?” Another female eyewitness was a bit uncertain when she identified someone other than Dansberry. According to the video, an officer told her she had made the wrong choice. Neither of these witnesses was called to testify at Dansberry’s murder trial.
Whether due to the cross-race effect, alleged police misconduct or just plain unconscious steering, mistaken identity accounted for 73 percent of wrongful convictions that eventually were overturned based on DNA evidence, according to the New York-based Innocence Project.
What has been proposed as an alternative is what is called the “double-blind sequential identification” process. There are two key components. The first is the officer administering the photo review and the later lineup must be unaware of the suspect’s identity — or even whether there is a suspect in the lineup. Arguably, then, no steering possibilities.
The second component is the sequential review. Photos are examined by the witness one at a time. Same for the live lineup. When combined with the double-blind method, according to the Innocence Project, there is a greater chance of accuracy.
A handful of jurisdictions have adopted this method. Not Illinois. That is because of a 2006 report to the General Assembly on a year-long pilot program conducted in Chicago, Evanston and Joliet. According to the report, the standard simultaneous identification process was superior to the double-blind sequential process.
A number of commentators argue the Illinois study was flawed and the outcomes affected by some police who didn’t buy into it conceptually. They also point to programs in other jurisdictions, like New Jersey and North Carolina, which have been effective in preventing false positives, like the one lawyers say landed Anthony Dansberry in prison and has allowed a crime to go unsolved.
Thus, they argue on their evidence that a system of justice that is not just blind (and, for that matter, not even just colorblind), but also double-blind will go a long way toward preventing what might amount to criminal irony. Where innocent people remain behind bars. Where guilty people remain free. Where victims remain without redress.
Christopher Benson is an associate professor of journalism at the University of Illinois, Urbana-Champaign.