Groundbreaking research: One out of every 10 rape convictions wrong?

As a young man, Michael Jones pleaded guilty to back-to-back attempted molestations of two girl strangers. However, he adamantly maintained his innocence while in prison and on parole. He said his lawyer had coerced him into pleading guilty by threatening him with life in prison if he went to trial. Michael was one of a handful of Black people in a rural white community; both of the little girls were white. He was identified when police brought him to the station and showed him to the girls. There was no lineup procedure with foils; he was the only choice the girls were given. On the basis of his two convictions, government evaluators diagnosed Michael with pedophilia and recommended civil commitment.

As a teenager, Paul Smith tried to molest a younger boy. He was arrested at the scene and confessed. He disputed only one point in the victim’s statement: that he had threatened the younger boy with a gun. Police searched his home and found no gun. Pre-conviction polygraph testing indicated he was being truthful when he denied having a gun. Over the ensuing years, however, clinicians in sex offender treatment programs hammered at him to admit that he had used a gun. Government evaluators said Paul’s “denial” and “minimization” of his gun use influenced their recommendation for civil commitment.

In cases such as these, I am consistently struck by the naïveté of clinicians and forensic evaluators alike, who accept police reports and especially victim accounts as the gospel truth. From my former career as a criminal investigator, I can attest to the fact that even impartial observers with no conscious motivation to distort are never 100 percent accurate in describing events they have witnessed. As Daniel Schachter so clearly articulates in Seven Sins of Memory, distortion is the nature of the human animal. It is even more likely to occur in situations involving high levels of stress, fear and emotionality.

So I was happy to see that the issue of false convictions for sex offenses is getting some much-needed and long-overdue attention. Or, let me qualify that: Happy about the empirical research, but less than thrilled with a theoretical article on the psychological dynamics underlying false accusations. Let me take those up one at a time.

Dredging old cases for DNA matches

The most methodologically rigorous study to date, released in June, suggests that somewhere between 8 and 18 percent of men convicted of sexual assault may be innocent. The federally funded research project randomly sampled convictions in Virginia between 1973 and 1987, before DNA testing was widely available, and compared preserved physical evidence with the DNA profiles of convicted men.

After poring through more than half a million cases, researchers found 422 sexual assault cases in which DNA evidence was preserved. In 8 percent (33) of those cases, the DNA evidence was exculpatory and supported exoneration. Because many of the DNA comparisons were inconclusive, this amounted to 18 percent of the cases in which it was possible to make a definitive determination one way or the other based on DNA analysis. (The data and the analyses are complex and not without flaws, so I recommend reading the study itself before relying on these numbers.) Noted the researchers:

"Even our most conservative estimate suggests that 8 percent (or more) of sexual assault convictions in a 15-year period may have been wrongful. That means hundreds, if not more than a thousand, convicted offenders may have been wrongfully convicted. That also means hundreds (if not more) victims have not received the just result, as previously believed. Therefore, whether the true rate of potential wrongful conviction is 8 percent or 15 percent in sexual assaults in Virginia between 1973 and 1987 is not as important as the finding that these results require a strong and coordinated policy response."

Bennett Barbour. Photo credit: 

Joe Mahoney, Times-Dispatch

Unfortunately, the researchers ran out of money before they could do more exhaustive analyses of the cases in which innocence was suggested. In the project’s wake, the government is battling with false confession activists who want access to the data, reports the Richmond (Virginia) Times-Dispatch. Police and prosecutors want to restrict access; exoneration activists argue that people have a right to know when their DNA does not match that collected in the crime for which they were convicted.

The project has led to the exoneration of at least four men. Putting a face to them is Bennett S. Barbour, who served a prison sentence for a 1978 rape. He had moved and did not receive the 2010 letter notifying him that the DNA specimen cleared him and matched a convicted rapist instead. A volunteer lawyer finally tracked him down and broke the good news by phone 18 months later.

Research into wrongful convictions has pinpointed several leading causes. These include:

Top sources of wrongful convictions. The Innocence Project

• False witness testimony (including mistaken identification and lying codefendants) 
• Faulty forensic evidence (especially comparisons of hair and bite marks) 
• False confessions 
• Police being influenced by prior knowledge of a suspect 
• Brief jury deliberations 

These problems are compounded by racial bias both in the criminal justice system and in society more broadly. African American men make up far more than their share of those who were convicted and later exonerated based on DNA evidence.

False accusations: A role for psychology?

Flat-out false accusations of rape -- like that depicted in To Kill A Mockingbird -- are rarely the cause of exonerations. But they do occur. Now, a prominent forensic psychology professor and his student propose 11 pathways to false allegations, and suggest that psychology could play a role in helping to sort reliable from unreliable reports. Write Jessica Engle and William O'Donohue in the Journal of Forensic Psychology Practice:

"[W]e suggest that some psychological disorders may increase the likelihood of believing a sexual assault occurred when it did not. Additionally, some psychological disorders may be related to an increase in motivation to fabricate an allegation of sexual assault in an effort to achieve what may be believed are the positive consequences of a false report…. [P]sychological evaluations may inform forensic evaluators of psychological processes by which a person may either intentionally or unintentionally file a false allegation of sexual assault."

The motivational and information processing pathways they propose lean heavily on psychiatric disorders -- including antisocial personality disorder, borderline personality disorder, histrionic personality disorder, psychotic disorders and intellectual disability -- as causes of false allegations. For example, here’s how they suggest that a histrionic personality style could lead to a false allegation:

"[A] person who is histrionic may, after a co-worker complements her clothing and accidentally bumps into her during the day, construe these actions as intentional communications of sexual interest. This misperception can lead her to feel that if the individual had touched her chest while bumping into her, it was an intentional action of unwanted assault. Thus, a pathway to false allegations of sexual assault may be through individuals with a diagnosis of histrionic personality disorder who for reasons of attention and misinterpretation may knowingly or unknowingly make a false allegation of sexual assault."

Okay, I’m not saying that people don’t lie, or make mistakes. Other research suggests that anywhere from 2 to 10 percent of all sexual assault reports may be false. But some of the examples provided in this article stretch credulity, and reek of sexism. I don’t know too many women, histrionic or not, who don't know the difference between an innocent compliment and a sexual assault.

A classification system based largely on pathologizing women runs the risk of reifying the mythology of so-called “rape myths,” in which only “good,” virtuous women can be raped. It seems especially problematic to disbelieve women with psychiatric problems when -- as the authors acknowledge -- they are the ones most likely to be sexually victimized.

More broadly, it is improper for clinicians to wade into the waters of truth-telling or lie detection. We weren’t there, and we don’t know what happened. It's problematic enough when we use character traits to predict the future. Stating that people (read: women) with this or that disorder are more likely to be lying or distorting reality opens the door for yet more improper use of psychiatric diagnosis in court.

Rather, as suggested by the Virginia data, we need to be skeptical at all times, and to keep our minds open to competing hypotheses based not on psychiatric stereotyping, but on the individual case facts. Maybe an assault happened, maybe it didn’t. Maybe the witnesses have their facts straight, maybe they don’t. Maybe the person who was convicted is the real culprit, and maybe he isn’t.

It’s clear that false convictions and false allegations are two separate beasts. And if that’s not complicated enough, there are true cases that are falsely recanted! For example, in a recent Welsh case, “Sarah” was repeatedly raped and forced into prostitution by her husband. When she recanted her report, she was convicted for perverting justice.

So, did Michael Jones (top of post) try to molest the two little girls? Maybe. Maybe not. The point is that we will never know for sure, and we should embrace -- rather than avoid -- that uncertainty. Present the competing scenarios, and analyze the case both ways, so that the trier of fact has all of the information.

The complexities in understanding sexual assault patterns are mind-boggling, and can make your head spin. False convictions, false accusations, false retractions. And then there's the other end of the spectrum: A vast proportion of sexual assaults – probably somewhere between 85 and 95 percent – are still going unreported altogether. And when victims do come forward, prosecution is rare, and convictions even rarer.

It's one gigantic mess, all around.

The U.S. Department of justice Study is: Post-Conviction DNA Testing and Wrongful Conviction by John Roman, Kelly Walsh, Pamela Lachman and Jennifer Yahner.