In 1987, about four months after a brutal rape and kidnapping, the 20-year-old Tulsa victim asked to hear men in a police lineup speak because, she said, she would never forget the voice of her attacker. She was denied.
Arvin Carsell McGee Jr. was in the mix because some police officers thought he resembled a sketch. He had been on a deferred sentence agreement for possessing a stolen credit card and vehicle.
The victim positively identified him, and McGee would face three trials, the first two ending in a mistrial with single jurors holding out.
With each trial, the victim became more convincing in her testimony, and McGee became more adamant in his denials. The only science was from blood testing that matched 19% of the black male population, and his alibi was his girlfriend.
It’s difficult to not believe a victim swearing with full, emotional detail that a specific person committed such an intimately violent act. It makes sense a person being violated would have an ironclad memory.
That’s wrong, according to decades of science, social research and exonerations.
Arvin McGee didn’t commit that crime. DNA testing led to his release in 2002 and a civil settlement against the city of Tulsa for $12.5 million.
He’s not alone.
Since 1989, at least five other Oklahomans have been released after wrongful convictions that relied heavily on eyewitness identification. After litigation, those five cases cost $24 million in civil settlements and state compensation.
In addition, another five cases involving false confessions led to court settlements.
In total, taxpayers have paid $42 million for those 11 exonerations based on eyewitness misidentification and false confessions.
Oklahoma lawmakers passed two laws to elevate protocols statewide around eyewitness statements and confessions. Both go into effect Nov. 1.
Senate Bill 798 requires evidence-based practices used to prevent eyewitness misidentification. Senate Bill 636 requires law enforcement to record suspect interrogations for murder or rape cases as a protection against false confessions.
Vicki Behenna, executive director of the Oklahoma Innocence Project, praises the legislation as criminal justice reforms.
“These are preventative measures to ensure the integrity of the criminal justice process,” Behenna said. “For each step we take in the state to ensure the integrity of the investigation and prosecution, the better we will be on the back end (of appeals).”
Behenna spent 25 years as a federal prosecutor and says the evaluation of eyewitness statements has evolved.
“As young prosecutors, we were trained that eyewitness identification was the best,” she said.
“Over time, social scientists and researchers have found cross-cultural identification is no good. And, we find people in traumatic situations — as much as they are trying to remember details of the person committing the crime — their ability to remember a sequence of events, details and actions aren’t accurate.
“Based on this research development, I think now those people believing eyewitness identification understand it should be taken with a big grain of salt. Now, I wouldn’t stand alone on someone’s witness testimony, and I think that’s the approach of most prosecutors.”
Similar laws have been adopted by other states and received backing of the U.S. Department of Justice, International Association of Chiefs of Police and the American Bar Association.
Oklahoma becomes the 25th state to require certain suspect interviews to be recorded in their entirety. Federal policy requires law enforcement record interrogations for all crimes.
Best practices in eyewitness identification have been implemented in 24 states and federal law enforcement. Four key components in the Oklahoma law are:
• Blind administration of lineups, meaning the officer in charge doesn’t know the suspect’s identity. If that is not possible, a blinded administration can be used. In these, the officer in charge is prevented from seeing which lineup member is being viewed by the witness, preventing intentional or unintentional coaching.
• Telling eyewitnesses that the perpetrator may or may not be in the lineup. This way they do not feel like they have failed if they don’t make a selection. It also removes pressure to make a choice.
• Selecting non-suspects for a lineup that generally match the eyewitness description of the perpetrator. This prevents a suspect from noticeably standing out.
• Asking eyewitnesses to state in their own words their level of certainty in their identification.
After McGee was exonerated, the Tulsa Police Department began evaluating how to improve its interview techniques and eyewitness protocols.
The practices were adopted, and the Oklahoma City Police Department also put them in place.
The new laws mostly affect small police departments and sheriffs.
The state police chiefs’ and sheriffs’ associations backed the standards and are offering training on implementing them.
“These policies and procedures are things that need to be implemented and taught in departments and adopted as practices,” she said. “There are organizations that are going to help smaller departments to make sure procedures are fully implemented. This is not something far off the grid.”
The biggest pushback came from recording of interviews because of the potential cost of digital equipment or storage.
It’s a weak argument considering cellphones now have even advance movie-making ability, and storage options include the digital cloud and servers.
It may be hard to understand why someone would confess to something they didn’t do. But research shows suspects with low IQs or who have been in abusive relationships may be susceptible to suggestive techniques.
“It is to protect officers and suspects; it protects both sides,” Behenna said.
The Oklahoma Innocence Project was founded in 2011 and is affiliated with the national Innocence Network. The Oklahoma group is funded solely by donations and grants and relies on the work of law students. It has five active cases, five to 10 in review and 700 pending
“Wrongful conviction costs everyone,” Behenna said. “It costs not just the person wrongly convicted, but there is cost to the person’s family. There is a societal cost with the real perpetrator out there. It costs in the financial awards and compensation.
“These kinds of best practices and measures being implemented are really a cost savings to society in a bigger way than the person wrongly convicted.”
Tulsa City Councilors offered a forum recently on the Equality Indicators report, which uses 54 equality measures that compare outcomes of groups likely to experience inequalities.