The devastating impact of prior crimes evidence and other myths of the criminal justice process.

• Author: Laudan, Larry

We concur in the general opinion of courts, textwriters and the profession that much of this law [concerning exclusion of evidence of prior crimes] is archaic, paradoxical and full of compromises and compensations by which an irrational advantage to one side is offset by a poorly reasoned counterprivilege to the other.... [Nonetheless] [t]o pull one misshapen stone out of the grotesque structure is more likely simply to upset its present balance between adverse interests than to establish a rational edifice.

--Justice Robert H. Jackson for the United States Supreme Court (1)

1. INTRODUCTION

   1. ANALYSIS AND BACKGROUND

      In the majority of legal systems in the developed world, triers-of-fact are routinely made aware of the prior convictions of the accused. In the United States, the admission of evidence of prior crimes is much more difficult, facing a multitude of hurdles. According to Federal Rules of Evidence (FRE) Rule 404, prior crimes categorically cannot be used to show character in order to prove "action in conformity therewith" or to show a propensity to illegal acts (unless defendant triggers a discussion of character), and they must also pass the Rule 403 balancing test, which rules out such evidence if the judge concludes that it is substantially more unfairly prejudicial than probatory. If a defendant takes the stand, prior convictions can be used as evidence that he is a liar under Rule 609--subject to the exceptions spelled out in Rule 404(b)--but generally not as evidence that he committed the crime with which he is charger. (2)

      The exclusion of prior crimes is hot an occasional quirk of the American legal system. It is the rule rather than the exception. In nine out of ten jury trials of defendants with prior convictions in which the defendant does not testify, the jury never learns about the prior convictions through evidence. (3) Even in trials where the defendant takes the stand, the jury learns of his priors only about half the time. Since more than half of defendants who go to trial have prior convictions, we are talking about massive exclusions of prima facie relevant evidence.

      Certain aspects of American evidence law--and considerable American evidence scholarship--are founded on a widely-shared set of beliefs concerning the admission of prior crimes of the defendant. Most of these beliefs are predicated on hypotheses about the perverse inferential psychology of jurors when it comes to evaluating prior crimes information. Others are founded, not on presumed juror psychology, but on dubious epistemic hypotheses about the probatory strength of prior crimes evidence. Moreover, there is substantial handwringing about the difficult choice defendants face about taking the stand in their own defense: if a defendant takes the stand, he risks being destroyed by his prior convictions; if he does not take the stand, he is destroyed by his silence in the face of plausible accusations. (4)

      Consider a few of these hypotheses:

      [H.sub.1]: Telling jurors about the prior crimes of a defendant dramatically increases their disposition to convict him and thereby puts at unnecessary risk many innocent defendants with criminal records. (5)

      [H.sub.2]: Information about prior crimes is, in most cases, only marginally relevant to rational decisions about defendant's guilt in the instant case. (6)

      [H.sub.3]: Hence, in the Rule 403 balancing test between probative value and unfair prejudice, prior crimes evidence should generally be excluded as being substantially more unfairly prejudicial than probative.

      [H.sub.4]: Whenever prior crimes evidence is introduced for a specific purportedly non-propensity purpose allowed under Rule 404(b) (e.g., motive, opportunity, intent, and so on), and accompanied by a limiting instruction, juries generally ignore the instruction and construe the prior convictions as evidence of a criminal propensity. (7)

      [H.sub.5]: The defendant with prior crimes faces a Hobson's choice in selecting his trial tactics: either (a) he testifies, in which case his prior crimes are more likely to be revealed and thus, by [H.sub.1], his prospects of a conviction will be greatly enhanced or (b) he does not testify, in which case he faces a negative inference from his silence, despite judge's instruction to the contrary. Defendants--especially those with prior convictions--will often elect silence, believing that the potential adverse inference from silence would be less incriminating than the revelation of their prior crimes would be. (8)

      One way or another, all of these hypotheses undergirding the conventional wisdom about prior crimes evidence are empirically testable. More than that, they have already been tested and most stand refuted or, at least, rendered highly implausible. That notwithstanding, many judges and legal scholars have been largely indifferent to, or unaware of, the empirical evidence, apparently persuaded that their own intuitions, grounded in decades of judicial experience, provide ample basis for the status quo, however clumsily cobbled together it may be.

      If these familiar hypotheses were true, they would constitute a triple whammy directed against the admissibility of prior crimes evidence. [H.sub.2] would have it that in most criminal cases the evidence of prior crimes scarcely qualifies as relevant. That notwithstanding, insists [H.sub.1], jurors are prone to inflate their genuine probative value. Finally, [H.sub.4] undercuts any hope that jurors' inferential mischievousness with prior crimes evidence can be kept in check by a sternly worded instruction. If all this were true, it would make sense to admit prior crimes evidence only under compelling circumstances. But if, as we believe can be plausibly inferred from existing empirical studies, none of these hypotheses (save [H.sub.4]) is true, then a serious reconsideration of current exclusionary practices concerning prior crimes evidence should be in order.

      This Article will show that the extant empirical evidence about how real juries respond to evidence of prior crimes raises grave doubts about the soundness of two of these five claims (specifically, [H.sub.1] and [H.sub.3]). As for [H.sub.2]--the alleged marginal probative value of prior crimes information--we are not persuaded that the relevance or its probative value of a prior crime is chiefly what should be at stake. Finally, the tactical dilemma, [H.sub.5], founded as it is on the soundness of [H.sub.1] to [H.sub.4], is likewise badly flawed.

   2. THE INITIAL CONUNDRUM

      Let us begin with a mildly surprising and seriously troubling pair of statistics. Both come from a relatively recent large-scale study sponsored by the National Council of State Courts (NCSC) of 358 criminal trials by jury in four different urban jurisdictions of the country. The first set of statistics is this:

      (1A) The acquittal rate for those defendants with prior crimes, whose prior crimes were not admitted as evidence to the jury was 23.9% (9), while

      (1B) The acquittal rate for those defendants with prior crimes, whose prior crimes were admitted as evidence to jurors, was 20.3%. (10)

      The first point worthy of note, although it is by no means the most surprising one, is that juries, fully aware of a defendant's prior crimes, are prepared to acquit in about two cases out of ten. (11) This statistic would appear to give the lie to the idea, entertained in some circles, that admitting evidence of prior crimes is the kiss of death for a defendant. The more intriguing point about this pair of statistics is that, among those defendants who do have prior crimes, the conviction rate among those whose prior crimes never emerge explicitly is only modestly lower than the conviction rate among defendants whose prior crimes become known. Specifically, juries convicted defendants whose prior crimes were unknown to them about 76% of the time and they convicted defendants with prior crimes known to them slightly less than 80% of the time. This appears to reinforce the point that a jury's learning of prior crimes directly through the evidence is not the inflammatory, unfairly prejudicial, conviction-ensuring information it is often depicted as being. While the conviction rates in the two sets are not identical, there appears to have been only a modest increase in the conviction rate when jurors learned directly through evidence of the defendant's prior crimes. (12)

      If that is so, the strenuous efforts of legal experts and defense attorneys to restrict the admissibility of prior crimes evidence seem misplaced. Admitting evidence of prior crimes apparently leads to few additional convictions. Under such circumstances, railing against the admissibility of prior crimes on the grounds that they unfairly disadvantage defendants with criminal records is unnecessary hyperbole.

      Now consider a second and rather more salient statistic, again from the NCSC study:

      (2) The acquittal rate for defendants with no prior convictions was almost twice as great as (13) the acquittal rate for defendants with prior convictions. (14)

      Not only is a defendant with no prior crimes about twice as likely to be acquitted as a defendant with priors, but in addition, as we shall see below, having prior crimes turns out to be one of the strongest predictors of a guilty verdict that we have available, stronger even than the testimony of an eyewitness to the crime who fingers the defendant. (15) And it remains a powerful predictor of the jurors' verdict even when the jurors have not been informed of its existence. To put it mildly, this result is puzzling, especially given that (1A) and (1B) entail that a jury, upon learning of the prior convictions of a defendant, is not much more likely to convict him than if the priors had not appeared. In short: jurors are much more likely to convict defendants with priors than those without, even while the jury's being informed of the priors does relatively little to increase the conviction rate.

      The...