Prior Convictions at Criminal Trials

In latest FindLaw column (available here) I describe and react to a paper (forthcoming in the Cornell Law Review and available in draft here) by two of my colleagues at Cornell, Valerie Hans and Ted Eisenberg. The paper persuasively argues that the admissibility of a defendant's prior criminal record has several consistent effects: 1) it deters defendants with a record from taking the stand in their own defense; 2) it significantly reduces jury reluctance to convict in marginal cases; 3) it does not affect jury assessment of the defendant's credibility, despite the fact that its admissibility is specifically premised on its relevance to witness credibility. I propose a number of possible reactions one might have to these data, including the radical (or reactionary) notion that we might disqualify criminal defendants from testifying at own trials.

In this post, I want to consider a different (and surprising) fact that emerges from the data: juries do not appear to count prior convictions as "evidence" supporting the likelihood of a defendant's guilt. That is, though juries are more likely to convict a defendant with a prior record, they nonetheless suggest (in rating the strength of the evidence) that the proof against the defendant is apparently no stronger in such cases than it is in prior-record-excluded cases in which juries acquit. That juries would not count a prior conviction as evidence of a defendant's guilt of the crime charged is important (and surprising) in two respects.

First, a major reason that prior convictions are ordinarily thought to pose a risk of unfair prejudice against a criminal defendant is the fear that, regardless of instructions to the contrary, a jury is likely to draw the following inference: the defendant committed crimes in the past and is therefore more likely to have committed the crime for which he is currently being prosecuted. If Hans and Eisenberg are correctly interpreting the data (and my review of their paper suggests that they are), then this fear is not well-founded: juries apparently do not fall into the trap of considering prior bad acts in deciding the likelihood of a particular bad act. Juries understand, in other words, that a person's apparent inclination to commit robbery does not tell us very much about whether it was he or some third party who robbed a particular bank three months ago. This suggests a level of sophistication on the part of the jury about which the evidence law has often been quite dubious.

On the other hand, the second important (and surprising) aspect of juries' ability to discern the relatively low relevance of prior convictions to guilt and innocence in a particular case, is that we are left to conclude that the jury is unable (or unwilling) to apply the standard of "guilt beyond a reasonable doubt" to defendants who have a prior record. The jury, in other words, is not confused by the evidence; it is instead repelled -- in the case of prior felons -- by the demanding standard of proof. If this is true, then juries appear far more willing than we might have thought to take the law into their own hands. For an ordinary criminal defendant, it is acceptable to allow ten (or a hundred or a thousand...) guilty people go free rather than incarcerate (or execute) one innocent person. But for a habitual criminal, perhaps, this permissive approach to what we might call "wrongful acquittals" is harder for juries to swallow. The stakes may simply feel too great. To put this differently, the downside of a wrongful acquittal, in the case of a defendant with a record, is that a habitual offender is free to offend again, while the downside of a wrongful conviction is that a habitual offender who happens not to have committed the particular crime charged spends time behind bars.

If this is the cost/benefit analysis in play, then the jury is rejecting the fundamental structure of a criminal trial as an assessment of guilt or innocence of a specified act (rather than the suitability of a particular person for preventive detention). This brings to mind the "war on terror" theory of detention with which we have lately become very familiar and may pose a far greater threat to criminal justice than the comparatively benign (but apparently not-so-tempting) inference that a prior offense sheds light on the odds of a presently charged crime.

Posted by Sherry Colb