In the public eye, Bill Cosby's goose is pretty much cooked. As a legal matter, however, his conviction on sexual-assault charges stemming from an incident in 2004 will likely rest on whether the trial judge decides to admit what is known as "prior bad acts" evidence. This question, now being heavily mooted in the press and on legal blogs, is a tricky one.
The "prior bad acts" in question are of two forms: Cosby's own testimony, in a deposition unsealed last northern autumn, that he obtained quaaludes to facilitate sex acts with women, and the accusations of several dozen women that what Andrea Constand, a former Temple University employee, claims the entertainer did to her, he also did to them. Most of us by now believe that there's fire behind the smoke, and that Cosby, a man the world once revered, is a deeply troubled or perhaps deeply evil human being.
Reaction to Cosby's sex assault charge
The most expensive TV shows of all time
Australia's biggest film festival opens in Melbourne
Sally Obermeder expecting surrogate baby
Karl Stefanovic says sorry
Tony Delroy leaves ABC Nightlife
Sesame Street axes human characters
Karl Stefanovic uses transphobic slurs
Reaction to Cosby's sex assault charge
In the shadow of a famous mural of Bill Cosby, people in Washington react to the news of his arraignment in a sex assault charge.
But the jury will never hear everything the public does, and for good reason. In the courtroom, unlike in the real world, the fact that a defendant has acted in a particular manner on another occasion cannot be taken as evidence that he acted that way on the occasion in question. This is what is known as propensity evidence, and it has long been forbidden in US courts.
Let's divide the "bad acts" evidence into two categories. Cosby's admission in the deposition should come into evidence easily. As a general rule, a party's own past statements are fair game. The tougher question involves the testimony of other women about what Cosby allegedly did to them. I think it's likely the court will admit the evidence, but I'm not sure the court will be right.
No matter what the rest of us think, it's a hallowed rule of procedure in the US that past behaviour can't be used to show guilt. Pennsylvania, like nearly every state, has adopted what is known as Rule 404(a): "Evidence of a crime, wrong, or other act is not admissible to prove a person's character in order to show that on a particular occasion the person acted in accordance with the character."
The rule forbids the introduction of what's known as "propensity" evidence – evidence intended to make the jury think, "Well, he's done this other bad stuff, so he probably did this too." What this means in practice is that if A is on trial for selling drugs, the jury likely will never hear that A has a conviction for the same crime on his record.
The hallowed rule naturally has hallowed exceptions, and the prosecution will doubtless try to take advantage of them. In news reports, lawyers have suggested that the testimony of other women about what Cosby allegedly did to them should be admissible under the exception for "modus operandi" evidence. But the argument isn't as simple as it may seem.
The "modus operandi" exception has been around for a long time, but isn't actually found in the rule. The traditional definition is rather technical, but the way commentators on the Cosby case are using the term, it's a way of summarising several important parts of the exception under Rule 404(b) for evidence tending to show "motive, opportunity, intent, preparation, plan, knowledge, identity, absence of mistake, or lack of accident".
Like most evidence professors, I teach my students that these exceptions are narrowly construed. Otherwise they swallow the rule. The question, in much-quoted language, is "whether the characteristics relied upon are sufficiently idiosyncratic to permit an inference of pattern". The subtle but important point is that the exception exists for cases in which the pattern of past crimes is so distinctive and unusual that it's highly unlikely that anyone else committed the charged crime. The courts often refer to this pattern as a "behavioural fingerprint".
How strict is the requirement? Consider the notorious case of Kristen Gilbert, the nurse convicted in 2001 of murdering four patients at a Veterans Affairs facility in Massachusetts. Even though the prosecution contended that Gilbert killed the patients by injecting fatal drug doses, evidence that she'd tried to do the same to her husband was excluded. The risk that the jury would engage in forbidden propensity reasoning was too high.
Why then do I think the Pennsylvania courts will admit the evidence? Because of a case on point decided just last June. In Commonwealth versus Tyson, defendant was on trial for rape. According to the prosecution, the accuser felt weak after donating blood and decided to leave work early. Resting at her apartment, she asked the defendant, an acquaintance, to bring her food. He did so. She fell asleep, and awakened to find the defendant having intercourse with her. She made him stop, then fell asleep a second time, and the same thing occurred.
At trial, the prosecution sought to introduce evidence that defendant had a prior conviction for rape. The trial court said no. An appellate panel of the Pennsylvania Superior Court reversed, pointing to the similarity of the facts as charged. In both cases the defendant was casually acquainted with the victim. In both cases the victim was black. In both cases the defendant was in the victim's home by invitation and found her to be in a defenseless state, unable to give legal consent. Wrote the Superior Court: "The relevant details and surrounding circumstances of each incident further reveal criminal conduct that is sufficiently distinctive to establish ... a common plan or scheme."
One could reasonably argue (as an earlier Superior Court panel had found in the same case) that the "relevant details" on which the court rested its opinion "do not amount to more than insignificant details common to many instances of sexual assaults". But the court went the other way – part of a recent trend towards a broader, pro-prosecution reading of 404(b) – and in so doing, I would think, doomed Cosby's chances of keeping the testimony out.
At this point the wise non-lawyer reader will object that in the Tyson case, the defendant had actually been convicted of the previous rape, whereas Cosby has never before been charged. But that's irrelevant. If the offered prior acts are such that a reasonable jury could find a modus operandi or common plan to exist, then they are admissible as long as the jury could find by a preponderance of evidence – that is, a 51 per cent chance – that Cosby committed them.
Let me emphasise, lest the point be lost, that I am not defending Cosby, and I share the general outrage at what he's evidently gotten away with for all these years. But when we stretch the rules of evidence to make sure we get all the bad guys, we reduce the protections that the rest of us are due.
Stephen Carter is a Bloomberg View columnist and a law professor at Yale.