According to the Innocence Project, mistaken identifications are the leading factor in wrongful convictions, accounting for almost 70 percent of the wrongful convictions overturned by post-conviction DNA evidence.
Thomas D. Albright, director of the Vision Center Laboratory at the Salk Institute for Biological Studies, listed some of the many variables affecting eyewitness testimony in a 2017 paper published in the Proceedings of the National Academy of Sciences, entitled “Why Eyewitnesses Fail.”
The variables include: viewing conditions (lighting-distance-duration), distracting stimuli (weapons-loud noises) and internal states of the observer (attention-motivation-skill-prejudice).
Albright explains that witness misidentifications are essentially failures of visual perception (seeing things accurately) or failures of memory (lack of precision in storage and recall).
I can vouch for the truth of his observations, having been involved in several cases dealing with witness misidentifications while representing both plaintiffs and defendants.
One of my clients, Cleve Heidelberg, was convicted in 1970 of killing a Peoria County, Ill., sheriff’s deputy who was responding to a call about an armed robbery at a drive-in movie theatre. After killing the officer, the shooter fled the scene and a high-speed car chase ensued.
The shooter wound up crashing the car and escaping on foot. The shooter had borrowed Heidelberg’s car, and when Heidelberg went to retrieve his car later that night he was arrested by officers at the scene. Several police officers reported and later testified at Heidelberg’s criminal trial that they observed him driving the subject vehicle as it sped past them at various points during the chase.
The reliability of these alleged identifications was suspect, since the officers would have had only a fraction of a second to observe the driver whizz by them at a high rate of speed at night. Yet, based on, among other things, this powerful purported eyewitness evidence, Heidelberg was convicted.
Incredibly, prior to Heidelberg’s sentencing, the real shooter, James Clark, confessed to the crime, but the confession fell on deaf ears and Heidelberg was sentenced to 99 to 175 years in prison. He served 47 years in prison before his conviction was vacated in 2018.
Perhaps one of the most striking examples of the flaws in eyewitness testimony is the case of Jerry Miller, who was convicted of rape, robbery and kidnapping in 1982.
In the Miller case, a woman was raped in her car in a parking garage in downtown Chicago. After the rape, with the victim now locked in the trunk, the rapist attempted to drive the woman’s vehicle out of the parking garage but was prevented from doing so by the two parking lot attendants who recognized the victim’s car and noticed that she was not driving it.
The perpetrator had a brief conversation with the parking lot attendants and then fled the scene.
The parking lot attendants provided a description of the driver to a police sketch artist and the sketch was published in a daily police bulletin. Two police officers thought the sketch resembled Miller, a man they had questioned a few days earlier. Miller was placed in a lineup and one of the parking lot attendants positively identified Miller; the other parking lot attendant tentatively identified him.
Miller was ultimately charged with rape, robbery and kidnapping. At his criminal trial, both parking lot attendants identified Miller as the perpetrator. Miller was convicted and sentenced to 52 years in prison.
But in 2006, DNA testing on the victim’s slip exonerated Miller and established the identify of another man, Robert Weeks, as the real perpetrator. At the time of his release, Jerry Miller had served over 26 years in prison for a crime he did not commit.
What the cases of Miller and Heidelberg highlight is that witnesses are often wrong.
The person they think was the perpetrator in fact had nothing to do with the crime. This can be true even in situations where the witness has an ample opportunity to observe the offender. In Miller’s case, the two parking lot attendants were immediately suspicious when the victim’s car approached the ticket booth. The victim, a white female, parked her car in that garage every day and the attendants noticed that as the car approached there was an unknown black male driving the car.
The attendants were also able to have a brief conversation with the driver before he fled on foot. Nonetheless, the attendants mistakenly identified Miller in a subsequent police lineup.
(Sometimes there is an issue of racial bias connected with eyewitness misidentification, but this was not a factor as the parking lot attendants were also African American.)
In other cases, the circumstances of a purported witness identification may be flimsy; yet the prosecutors still rely on that identification, which is then presented as rock-solid evidence at trial.
Improper lineup procedures can also lead to witness misidentifications. The other individuals placed in a lineup with the suspect (known as “fillers”) are supposed to reasonably resemble the suspect in terms of age, height and weight so that the suspect does not stand out.
This is not always the case. When proper procedures are not followed, the potential for error exists.
The ‘Starved Rock Murders’
I am currently involved in another case in which the failings of witness identification are starkly evident.
Chester Weger was convicted in 1961 of the notorious “Starved Rock Murders,” in which three women were brutally killed and their bodies left in a park located in LaSalle County, Illinois about two hours southwest of Chicago. He served 61 years in prison before being released on parole in February 2020.
Until he was released on parole, the 81-year-old Weger ranked as the second longest-held inmate in Illinois’ prison system. Weger’s attorneys, Celeste Stack and I, continue to seek to prove Chester’s innocence and are seeking to review and analyze the physical evidence and are working to be able to conduct forensic testing on some of that evidence.
Police said they believed Weger was a suspect in a rape that had occurred a few years earlier. He was 19 at the time of that attack. In an effort to overwhelm Chester and break him down, he was placed in a lineup regarding two incidents that took place in 1959 in a nearby State Park called Deer Park, one of which was that rape case and the other was a purse snatching.
The lineup photo from one of these lineups shows how suggestive it was: The other four men in the lineup were much older and bigger than Chester. This created an inherent bias that made the lineup unfair, and likely affected the victim’s subsequent identification.
Witness misidentifications can affect both plaintiffs and defendants. A witness may mistakenly identify a criminal suspect. Similarly, a witness may mistakenly perceive the circumstances of a crime or event, such as a car accident.
Chester denied having anything to do with these other crimes but was identified by the victims based on these suggestive procedures. Thus, Chester was charged with three separate felonies as a way to overwhelm him and break him down.
Law enforcement officers, prosecutors, attorneys, and jurors need to realize that just because a witness says so does not make it so. A full and fair examination of all the circumstances of the subject crime, and all the evidence involved, is necessary.
Additional Reading: Truth, Lies and Police Lineups, The Crime Report, May 5, 2021.
Andy Hale is a civil rights attorney, specializing in wrongful convictions in Chicago, Illinois. He is the founder of Hale & Monico, America’s Justice Attorneys. Hale is an Emmy-nominated executive producer for several true-crime documentaries highlighting wrongful convictions.