Last week in Hempstead, New York, a young woman accused five men of gang rape on the Hofstra campus. In 2006, in Durham, North Carolina, a young woman accused members of the Duke lacrosse team of gang rape.
Both accusations turned out to be false. Both cases were eventually dismissed. The Hofstra defendants spent three nights in jail before prosecutors dismissed charges. The Duke defendants spent nearly a year under indictment and reportedly millions of dollars in legal fees before charges were dismissed.
Why the difference? The apparent credibility of the accusers? The relative strength of the exculpatory evidence? I doubt it. The difference between three days and twelve months lived under the long shadow of accusation was simply luck of the draw. The Hofstra defendants drew one set of prosecutors, and the Duke defendants got Mike Nifong.
Nearly seventy-five years ago, the Supreme Court wrote that the prosecutor has a "twofold aim of which is that guilt shall not escape or innocence suffer." A comment to the Model Code of Professional Responsibility mandates that a prosecutor's "duty is to seek justice, not merely to convict." At the very least, the notions of protecting innocence and seeking justice suggest that ethical prosecutors ought not charge defendants who are innocent.
But we are just starting to explore the hurdles to this seemingly incontrovertible aim in light of what psychologists have known for years about cognitive biases. Psychologists, for example, know that people do not approach decision-making as good scientists. While scientists subject their working hypotheses to rigorous testing - proof through the failure to disprove - the natural human tendency is for people to search for information that serves only to support their working theory. Tell a person to find out whether a stranger is extroverted, and they'll ask, "What do you do to liven up a party?"
Prosecutors who approach cases as scientists might test the defendant's guilt by searching for evidence of innocence. But the realities of human cognition might intervene. They might ask questions that can only sink the defendant. Search for witnesses who can only seal his fate. Employ flawed identification processes.
The Duke accuser, for example, identified the defendants after viewing photographs of the lacrosse team with none of the decoy faces that are usually added to the mix. It was a test she could not fail and the circle of potential suspects could not pass.
Prosecutors who think like scientists would also evaluate potentially exculpatory new evidence neutrally, asking themselves what the case would have looked like had they confronted this evidence from day one. But cognitive research demonstrates that people interpret new information through the lens of their existing beliefs. Prosecutors who believe the defendant is guilty may discount the value of potentially exculpatory evidence, minimizing it as fabricated, unreliable, or, at most, a wrinkle to be straightened away before the jury. Meanwhile, they continue to search for, and accept as valid, additional evidence against him.
We've seen this phenomenon before. In how many of the growing number of DNA exoneration cases has the defendant's release been accompanied by a statement from the original prosecutor that the defendant might nevertheless be the culprit? Because of selective information processing, beliefs in a defendant's guilt are naturally sticky.
Ironically, the desire to be ethical might actually hinder a virtuous prosecutor's ability to give proper weight to exculpatory evidence. People process information as necessary to avoid "cognitive dissonance" between their beliefs and their conduct. Ethical prosecutors embrace their role as ministers of justice. They insist they would never prosecute an innocent. So when potentially exculpatory evidence comes to light, how do they avoid the discomfort of dissonance between their conduct and their deeply held beliefs? Cognitive science suggests they might unintentionally distort their analysis of the new evidence. If it's not exculpatory, then the defendant isn't innocent, and the prosecutor has lived up to his ethical charge.
Despite the shortcomings of human decision-making, prosecutors sometimes get it right. I suspect reasonable minds will eventually differ on whether Nassau County prosecutors responded quickly enough to inconsistencies in the Hofstra accuser's initial statements, but the accuser ultimately recanted when prosecutors confronted her with the existence of a video. Prosecutors who might have enabled an inculpatory explanation for the tape instead went searching for exculpation, even when doing so required cross-examining their own witness.
As the number of exonerations of wrongly convicted defendants continues to grow, prosecutors should take the lead in creating a culture devoted to innocence. But doing this requires more than a commitment to ethics. It requires prosecutors to approach their cases as skeptics, to test their own cases, continually and rigorously, and to search for innocence. It takes open discovery policies that acknowledge that defense attorneys might recognize exculpatory evidence where prosecutors miss it. It takes institutional changes like "fresh look" committees in which neutral attorneys evaluate innocence claims. It takes an admission that prosecutors -- even the ethical ones, even the ministers of justice and the protectors of innocence -- are only human.