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

Journal of Criminal Law and Criminology

Volumn 101, Issue 2, 2011, Pages 493-527

  Laudan, Larry ^a   Allen, Ronald J ^b  

^a UNIVERSITY OF TEXAS AT AUSTIN   (United States)

^b NORTHWESTERN UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

EID: 84857508876     PISSN: 00914169     EISSN: None     Source Type: Journal    
DOI: None     Document Type: Article

References (66)

1. Michelson v. United States, 335 U.S. 469, 486 (1948).
2. Rules 413-415 also create more particular exceptions to the exclusion of prior crimes. FED. R. EVID. 413-415.
3. See text accompanying note 47. It is important to stress that the focus in this Article is entirely on prior convictions. Accordingly, little will be said here about whether prior arrests and prior acquittals or alleged bad acts that never led to convictions should be admitted. Interesting as such questions are, the data utilized here do not illuminate them. For more on this issue, see notes 56 and 80.
4. As it turns out, this is not a dilemma for the defendant with priors but it is for the defendant without priors, as the failure to testify is highly associated with guilty verdicts. This provides empirical support for the relative plausibility theory of juridical proof, as we discuss note 71. On the relative plausibility theory, see Ronald J. Allen, Factual Ambiguity and a Theory of Evidence, 88 NW. U. L. REV. 604, 609 (1994).
5. The eminent legal scholar John Henry Wigmore once claimed: "The natural and inevitable tendency of the tribunal-whether judge or jury-is to give excessive weight to the vicious record of crime thus exhibited, and either to allow it to bear too strongly on the present charge, or to take the proof of it as justifying a condemnation irrespective of guilt of the present charge. " JOHN HENRY WIGMORE, EVIDENCE § 194 (3d ed. 1940). Along the same lines, the U.S. Supreme Court in Michelson v. United States, 335 U.S. 469, 475-76 (1948), commented: The state may not show defendant's prior trouble with the law, specific criminal acts, or ill name among his neighbors, even though such facts might logically be persuasive that he is by propensity a logical perpetrator of the crime. The inquiry is not rejected because character is irrelevant; on the contrary, it is said to weigh too much with the jury and to so overpersuade them as to prejudge one with a bad general record and deny him a fair opportunity to defend against a particular charge. Andrew Morris goes so far as to opine that the admission of propensity evidence under Rule 404(b) "produces grave consequences for thousands of criminal and civil defendants. " Andrew J. Morris, Federal Rule of Evidence 404(b): The Fictitious Ban on Character Reasoning from Other Crime Evidence, 17 REV. LITIG. 181, 184 (1998). The issue surely is not whether the consequences are grave but whether they are inferentially appropriate.
6. Calvin Sharpe claims, on the strength of no cited empirical data whatever, that "the probative value of all character evidence, including evidence of other crimes, is often not very great, while it usually will have substantial prejudicial effects. " Calvin W. Sharpe, Two-Step Balancing and the Admissibility of Other Crimes Evidence: A Sliding Scale of Proof, 59 NOTRE DAME L. REV. 556, 560 (1984). He continues: "Whatever the true probability that commission of a crime will be followed by other criminal acts, it is highly unlikely that the aggregate intuitions of a jury will produce an accurate assessment of the worth of such evidence. Furthermore, even if statistics indicating the probability of a second theft, given a first, were available, a jury's ability to ascribe to such evidence only its properly proportioned weight is highly questionable. " at 562 n.27.
7. Supreme Court Justice Robert Jackson insisted that "all practicing lawyers know to be unmitigated fiction" that "prejudicial effects can be overcome by instructions to the jury. " Krulewitch v. United States, 336 U.S. 440, 453 (1948). Justice Stewart concurred, insisting that certain kinds of evidence "are at once so damaging, so suspect, and yet so difficult to discount, that jurors cannot be trusted to give such evidence the minimal weight it logically deserves, whatever instructions the trial judge might give. " Bruton v. United States 391 U.S. 123, 138 (1967). Judge Learned Hand held that asking jurors to lay aside such information as prior crimes requires an effort that "is beyond, not only their powers, but anybody's else. " Nash v. United States, 54 F.2d 1006, 1007 (1932). Michael Saks confidently notes that "[w]hen informed about a defendant witness's prior crimes for the permissible purpose of evaluating credibility, jurors use the information for the impermissible purpose of inferring the likelihood that the defendant committed the currently charged crime. " Michael J. Saks, What Do Jury Experiments Tell Us About How Juries (Should) Make Decisions?, 6 S. CAL. INTERDISC. L.J. 3, 26 (1997).
8. This conventional analysis is neatly summed up in the observation by Theodore Eisenberg and Valerie Hans that "testifying is risky for defendants with prior records because the records may be revealed on cross-examination. " Theodore Eisenberg & Valerie Hans, Taking a Stand on Taking the Stand: The Effect of a Prior Criminal Record on the Decision to Testify and on Trial Outcomes, 94 CORNELL L. REV. 1353, 1373 (2009). As our subsequent analysis will make clear, this claim is half right: testifying does indeed make it more likely that defendant's prior crimes will be revealed but their revelation, should it occur, does little to increase significantly the (already high) risk that the testifying defendant with prior crimes will be convicted.
9. Daniel Givelber, Lost Innocence: Speculation and Data About the Acquitted, 42 AM. CRIM. L. REV. 1167, 1190 (2005). It should be noted here (and below where the NCSC data are referenced) that the NCSC database divided trial results into "acquittals, " "convictions, " or "hung juries. " Analysts of this data set generally adopt the convention that "hung" results are counted as half acquittals and half convictions. Others exclude the hung trials altogether. These differing conventions explain why there are small discrepancies in the results reported by various authors. For descriptions of the data set, see note 71,.
10. That compares with the mean acquittal rate in all the trials in the NCSC study (including those of defendants without priors) of about three-in-ten. See note 71.
11. Obviously, the magnitude of change in the acquittal numbers is larger than in convictions because the absolute numbers of acquittals are smaller. Still, the acquittal rate for those with unknown priors is only about 17.5% higher than for those with known priors. In absolute numbers, in 15 out of 74 cases there were acquittals when the priors were known. Givelber, note 9, at 1190. Had there been only three more acquittals in this subset, the rates would have been almost identical. Thus, even the larger disparity does not suggest that knowledge of priors has a dramatic affect on juries.
12. Specifically, it is 184% higher. Very similar data are reported by Givelber, note 9, at 1190 tbl.1 (reporting an acquittal rate 216% higher).
13. Givelber, note 9, at 1190. So far as we are aware, Professor Givelber was the first analyst to draw attention to the fact that conviction rates for defendants with prior crimes in the NCSC study were significantly higher than conviction rates for defendants without them, regardless of whether the jury was informed of those priors. "It is whether or not the defendant has a criminal record-not whether the jury learns about it-that has the greatest influence on the acquittal/conviction decision. "
14. Anthony N. Doob & Hershi M. Kirshenbaum, Some Empirical Evidence of the Effect of s. 12 of the Canada Evidence Act upon an Accused, 15 CRIM. L.Q. 88, 90-91 (1973).
15. W. R. Cornish & A. P. Sealy &, Juries and the Rules of Evidence, 1973 CRIM. L. REV. 208, 218.
16. Valerie Hans & Anthony Doob, Section 12 of the Canada Evidence Act and the Deliberations of Simulated Juries, 18 CRIM. L. Q. 235, 243 (1976).
17. William Thompson et al., Inadmissible Evidence and Juror Verdicts, 40 J. PERSONALITY & SOC. PSYCHOL. 453, 455 (1981).
18. Thomas R. Caretta & Richard L. Moreland, The Direct and Indirect Effects of Inadmissible Evidence, 13 J. APPLIED SOC. PSYCHOL. 291, 297 (1983).
19. Roselle Wissler & Michael Saks, On the Inefficacy of Limiting Instructions: When Jurors use Prior Conviction Evidence to Decide Guilt, 9 LAW & HUM. BEHAV. 37, 43 (1985).
20. Bruce Rind et al., The Effect of Crime Seriousness on Simulated Jurors' Use of Inadmissible Evidence, 135 J. SOC. PSYCHOL. 417, 418 (1995).
21. Edith Greene & Mary Dodge, The Influence of Prior-Record Evidence on Juror Decision Making, 19 LAW & HUM. BEHAV. 67, 72 (1995).
22. Sally Lloyd-Bostock, The Effects on Juries of Hearing About the Defendant's Previous Criminal Record: A Simulation Study, CRIM. L. REV. 734, 743 (2000). Lloyd-Bostock claims her results are a "clear" confirmation of a "prejudicial effect, " when all the evidence apparently shows is an increase in the probability assigned to guilt by mock jurors when they learn of prior crimes. Id. at 753. That increase that may or may not be justified by the probative weight of the prior crimes evidence. Id.
23. Most of the mock jury results summarized above depend on the analysis of "close" cases.
24. Another myth about criminal justice that is called into question by data is the micro-economic explanation that most cases that go to trial tend to be close. If they are not close cases, the argument runs, a plea bargain or dismissal will often be entered, leaving factually close cases to predominate at trial. This is disconfirmed by at least three sources. First, if cases were close, the requirement of unanimity should produce substantial hung juries, yet the rate for hung juries on all counts hovers around 7.5%. PAULA L. HANNOFORD-AGOR ET AL., NAT'L CTR. FOR STATE COURTS, ARE HUNG JURIES A PROBLEM? 43 (2002), available at http://www.ncsconline.org/WC/Publications/Res_Juries_HungJuriesProblemPub.pdf. Second, close cases should be difficult to decide, yet the average length of jury deliberations in noncapital felonies is three hours, two for misdemeanors, and a rather brief six for capital cases. GREGORY E. MIZE, ET AL., NAT'L CTR. FOR STATE COURTS, STATE-OF-THE-STATES SURVEY OF JURY IMPROVEMENT EFFORTS: A COMPENDIUM REPORT 38 (2007), available at http://www.ncsconline.org/d_research/cjs/pdf/SOSCompendiumFinal.pdf. Third, if cases were close, there should be substantial judge-jury disagreement, but there is not; and with regard to what does exist, "[w]e find little evidence that evidentiary complexity or legal complexity help explain rates of judge-jury disagreement. The judges' lower conviction threshold seems to be driving most of the difference." Theodore Eisenberg et al., Judge-Jury Agreement in Criminal Cases: A Partial Replication of Kalven & Zeisel's The American Jury, 2 J. EMPIRICAL LEGAL STUD. 171, 173 (2005).
25. HARRY KALVEN, JR. & HANS ZEISEL, THE AMERICAN JURY (1966).
26. HANNOFORD-AGOR ET AL., note 30.
27. Martha Myers, Rule Departures and Making Law: Juries and Their Verdicts, 13 LAW & SOC'Y REV. 781 (1979).
28. KALVEN & ZEISEL, note 31, at 160-61. This pair of statistics is not quite what it purports to be. For a set of arcane reasons, Kalven and Zeisel defined, for purposes of this table, a "defendant without priors" as one who (a) had no priors or (b) took the stand or (c) managed to hide his priors from the jury. Any defendant who failed to satisfy at least one of those conditions was counted as a "defendant with priors. " at 159 n.17. Although their questionnaire solicited information directly about which defendants did and did not have prior convictions (in the usual sense of that term), they did not include that data in this computation.
29. Givelber, note 9, at 1190. Our own analysis of the NCSC data suggests that these figures are closer to 73.5% and 52% respectively. Still, the gap in conviction rates between defendants with and those without priors is similarly striking on either method of tallying the results.
30. The Chicago Jury Project showed a 25% acquittal rate for defendants with priors versus 42% for those without, and the NCSC study's rates were 23.9% versus 43.9%.
31. Daniel Givelber & Amy Farrell, Judges and Juries: The Defense Case and Differences in Acquittal Rates, 33 LAW & SOC. INQUIRY 31, 38 tbl.1 (2008).
32. We say "explicitly" because other rules of evidence facilitate such inferences by allowing in evidence of, or relevant to, prior crimes evidence. Rules 413-15 admit prior bad acts which could be prior crimes, as does Rule 404(b). It should be noted that many of the grounds for admission of prior bad act evidence under Rule 404(b) actually involve a propensity inference even though they are supposedly admissible on some other ground than propensity. For a discussion, see RONALD J. ALLEN ET AL., EVIDENCE: TEXT, PROBLEMS, AND CASES 244-57 (4th ed. 2006) and Thomas J. Reed, Admitting the Accused's Criminal History, 78 TEMP. L. REV. 201 (2005). As Reed puts the point: "Someone... might jump to the conclusion that courts admit uncharged misconduct in a backhanded, under-the-table way that conceals the real agenda of the criminal justice system. That view might lead... [them] to conclude that a judicial conspiracy exists to convict the accused on the basis of bad character while seeming to prohibit conviction on account of bad character. " at 248. There are other rules of evidence that have similar effects, such as the common law res gestae rule. See, e.g., Edward J. Imwinkelreid, The Second Coming of Res Gestae: A Procedural Approach to Untangling the "Inextricably Intertwined" Theory for Admitting Evidence of an Accused's Uncharged Misconduct, 59 CATH. U. L. REV. 719 (2010). It may be the case that these other rules largely overwhelm the exclusionary rule of Rule 404.
33. Alex Stein has offered a quite different analysis of the situation of the innocent defendant with prior convictions. Knowing that his priors may be revealed if he testifies (and aware that such a revelation may make his conviction more likely), he will nonetheless be disposed to "tell his story. " According to Stein, "the defendant's readiness to risk impeachment by prior convictions might signal credibility. " ALEX STEIN, FOUNDATIONS OF EVIDENCE LAW 165 (2005). He adds: "By testifying about his or her innocence, the defendant conveys to fact-finders that his or her exculpatory testimony is true and that there are no specific reasons for disbelieving it.... Because the defendant's signaling is costly, it indicates the defendant's credibility as a witness.... " Moreover, adds Stein, "many guilty defendants prefer to separate from the pool by not testifying in their defense, " because the true stories told on the stand by the innocent defendant are more likely to stand up to prosecutorial cross-examination than the fictitious stories concocted by the guilty: "Reasons discrediting the defendant's self-exonerating testimony predominately appear in cases featuring guilty defendants. " Stein's argument amounts to the claim that innocent defendants with prior convictions have a strong incentive to testify since their willingness to run the risk of revelation of their priors betokens to the jurors the truth and sincerity of their testimony, whereas those who are guilty will not take the stand because their false stories would likely be shattered by cross-examination. This analysis is directly disconfirmed by the data. According to the NCSC data, (a) 45% of defendants with priors testified; (b) most defendants with priors who chose to testify were convicted (77%) and thus were presumably guilty; and, most tellingly, (c) a testifying defendant with priors was more likely to be convicted than was a non-testifying defendant with priors (77% versus 72%). The inconsistency of this data with Stein's argument is obvious. First, if testifying signals innocence, the signal is not being received by the jury that convicts 77% of those with priors who testify. Second, according to the anti-pooling argument, those who do not testify are predominantly guilty and those who testify predominantly innocent, yet in the NCSC study, juries convicted more of those who testified than those who did not. In other words, it does not follow from any of this data therefore that the guilty avoid taking the stand and the innocent signal their innocence by testifying unless defendants make irrational choices and jury verdicts are highly unreliable. Thus, his argument amounts to an explanation of massive irrationality and a prediction of enormous error rates, all within the confines of an analysis predicated upon rational choice.
34. Lloyd-Bostock, note 28, at 745. In her study of more than 200 mock jurors, presented with an initially weak case, and then told (in most cases) of prior crimes with varying characteristics, the results that emerged are quite contrary to standard intuitions. Here are some of the more salient data: Note that her study does not suffer from the methodological problems of the mock jury studies trying to determine if the use of priors, standing alone, is unfairly prejudicial. She is varying the type of priors. The most curious result on this study is the "Told 'no priors'" data. One would think that conviction rates in that setting should be considerably lower. One possible-although also possibly ad hoc-response is that in the study the jurors were told only that the defendant "has no previous convictions, " and in the judge's instructions that "you have heard that the defendant has no previous convictions. " at 740. With no exploration of the defendant's good character, jurors may very well have speculated about conduct that did not result in convictions, and indeed that inference might have been motivated in part by the unelaborated information and instruction. The instruction, "you have heard, " seems to us particularly likely to be viewed as ambiguous. In any event, the complex interaction of decision makers and instructions is well known. See, e.g., Saul M. Kassin & Samuel M. Sommers, Inadmissible Testimony, Instructions to Disregard, and the Jury: Substantive Versus Procedural Considerations, 23 PERSONALITY & SOC. PSYCHOL. BULL. 1046 (1997); Nancy Steblay et al., The Impact on Juror Verdicts of Judicial Instruction to Disregard Inadmissible Evidence: A Meta-Analysis, 30 LAW & HUM. BEHAV. 469 (2006).
35. In subpart III. C, we explore one possible systematic difference below with regard to state behavior in the face of priors.
36. One possible explanation for the data involving defendants with priors is that motions in limine will provide the defense attorney with information about the admissibility of priors, allowing him to make the calculation of the likelihood of a guilty verdict with full information prior to the defendant testifying, and that the predictions about odds of conviction are well calibrated. There are numerous difficulties with this explanation. First, the attorneys' predictions are not just "well-calibrated, " they are astonishingly well-calibrated, and thus the story is implausible. Moreover, they lose this astonishing calibration ability when it comes to making decisions about whether defendants without prior records should testify. Second, this explanation requires not just amazing calibration ability but full knowledge. Unfortunately, the latter condition does not exist. It is common practice for judges not to give definitive resolutions to the motions in limine that are essential for the explanation. The third problem is that this assumes that the only knowledge of priors will come from their formal revelation at trial. That is false. There are various mechanisms for this information to get out. See note 56. Fourth, and perhaps most devastating to this speculation, is that it is hard to make sense of a claim that the lawyers are good at calibrating odds of guilty verdicts when most people who go to trial are convicted. In most of those cases, the defendant would have been better off with a plea bargain.
37. Kalven and Zeisel conjecture that "[t]he jury's broad rule of thumb here, presumably, is that as a matter of human experience it is especially unlikely that a person with no prior record will commit a crime.... " KALVEN & ZEISEL, note 31, at 179.
38. See Eisenberg & Hans, at 1386. Eisenberg and Hans similarly observe that the NCSC data suggest that "[t]he conviction threshold appears to differ for defendants with and without criminal records. " Id.
39. And in this case the common knowledge happens to actually qualify as knowledge by virtue of being true. See, e.g., Larry Laudan & Ronald J. Allen, Deadly Dilemmas II: Bail and Crime, 85 CHI.-KENT L. REV. 23 (2010).
40. This argument is explored at length in Larry Laudan, The Rules of Trial, Political Morality and the Costs of Error: Or, Is Proof Beyond a Reasonable Doubt Doing More Harm Than Good?, LAW & PHIL. (forthcoming July 2011).
41. The case for the interpretative flexibility of "beyond a reasonable doubt" has been made most cogently by Erik Lillquist in his Recasting Reasonable Doubt: Decision Theory and the Virtues of Variability, 36 U.C. DAVIS L. REV. 85 (2002); see also LARRY LAUDAN, TRUTH, ERROR AND CRIMINAL LAW ch. 2 (Gerald Postema ed., 2006).
42. We share this view, although we should note that it is not universally embraced; one concept of "prejudice" includes the effect of evidence on juror utility functions. See ALLEN ET AL., note 56, at 164-66 (4th ed. 2006). Those who do not share this view might try to change outcomes by instructing the jury as to the utility functions to use-such as do not take into account the possibly deadly consequences of your decisions. The Supreme Court, in In re Winship, held that due process requires proof beyond a reasonable doubt in criminal trials. 397 U.S. 358 (1970). This appears doubtful to us if it means a unitary standard no matter what the case is about. The more plausible due process argument is to the effect that the standard of proof in all trials must reflect, as the Court likewise noticed in In re Winship, the respective costs of the mistakes that can be made. Since it is plausible to believe that falsely acquitting a serial felon is more costly than falsely acquitting a first-time offender, it follows that a less exacting standard of proof is plausible in the former case, depending upon the relationship between the changed costs of a false acquittal and a false conviction.
43. Given the analysis in this paper, the mystery is why a defendant with no prior convictions-and his counsel-would deliberately forego the opportunity to make that fact known to jurors. One explanation offered by one of our helpful critics during the presentation of this paper at the Northwestern workshop, goes like this: "Assuming no perjury, there is often very little a guilty defendant could say on the witness stand that will help him. A positive correlation between testifying and acquittals among prior-free defendants reflects that the guilty testify less frequently and thus does not imply causation. " The problem is that in the real world, the assumption should be that a defendant is quite willing to commit perjury. After all, most who testify, with and without priors, are found guilty; and if they are not committing perjury, there is an amazing error rate at trial. Second, a guilty defendant for whom the Government has made a plausible case of guilt really has no choice but to take the stand and commit perjury; therefore the prediction should be that they would readily. By taking the stand, not only do they get to lie, but they increase the probability of demonstrating to the jury their clean record, which looks like it has probative value to fact-finders. Another difficulty with this explanation is that it is entirely unclear why it would only apply to prior-free defendants. It should also apply to those with priors, and the results there are quite different.
44. These results emerge from our own analysis of the five data sets that constitute the NCSC data. (Our thanks to Jorge Luis Silva for invaluable help integrating the sets.) See HANNAFORD-AGOR ET AL., note 30. The data also provide empirical support for the relative plausibility theory of juridical proof. The empirical hypothesis concerning proof beyond reasonable doubt of the relative plausibility theory is that guilt is determined by whether there is a plausible story of guilt and a plausible story of innocence. If there is a plausible story of guilt and no plausible story of innocence, convictions will ensue. It seems likely that cases where defendants do not testify involve cases of essentially no presentation of a defense, thus leaving the prosecution's case largely unrebutted. As most cases would not go to trial without a plausible story of guilt, in most of the cases with no defense offered there should be convictions, as the data show. On the relative plausibility theory, see Allen, note 4, at 609. Another potentially confounding variable is that defendants without priors have more to lose with a conviction than those with priors. Standing alone, this suggests that the chance of conviction should be higher than for those with priors, as those without priors may take riskier cases to trial rather than plea. That is not what the data show, however. Perhaps this is because those without priors, having more to lose, may also invest more in the effort to avoid conviction. This point seems to cut in both directions, thus suggesting the better explanation is the one we give in the text.
45. With apologies to Casablanca.
46. Donald Dripps, in comments on a previous draft of this Article, pointed out that another possible confounding variable is the ethical dilemma of defense counsel if they urge clients to tell them the truth, are told a story of guilt, but the client still wants to testify. Theoretically, a lawyer cannot sponsor perjured testimony. This may be operating to some extent, but its scope is somewhat questionable given the rate at which those both with and without priors who testify are convicted.
47. FED. R. EVID. 404(a)(1).
48. Cf. Josephine Ross, "He Looks Guilty": Reforming Good Character Evidence to Undercut the Presumption of Guilt, 65 U. PITT. L. REV. 227, 231 (2004).
49. For instance, here is the jury instruction for the U.S. District Court for the Western District of Oklahoma, promulgated in 2006: The defendant has offered evidence of his reputation for good character or testimony in the form of opinion as to his good character.... Evidence of good character may be sufficient to raise a reasonable doubt whether the defendant is guilty, because you may think it improbable that a person of good character would commit such a crime. Evidence of a defendant's character, inconsistent with those traits of character ordinarily involved in the commission of the crime charged, may give rise to a reasonable doubt. Jury Instruction, United States v. Good, 2006 Jury Instr. Lexis 1938 (W.D. Okla. 2006) (No. CR-06-0070-F).
50. Another is concern that excluding such evidence by defendants would violate their right to present a defense.
51. FED. R. EVID. 404.
52. Except for his character for lack of truthfulness. FED. R. EVID. 609.
53. This raises another possible explanation for the data we have been exploring: that the promise of excluding propensity evidence highly associated with prior crimes is not redeemed because of rules such as Rule 404(b). See note 56. We do not explore this in the text for two reasons. First, the studies we have do not control for the subversion of Rule 404(a) by the other rules. Second, if the other rules essentially let evidence in equivalent to prior crimes, again it is obviously silly to be worried about excluding that very evidence that other rules admit.
54. According to Bureau of Justice Statistics figures, of the 3,265 convicted rapists released from prison in 1994, some 46% were re-arrested for a felony within three years. PATRICK A. LANGAN & DAVID J. LEVIN, BUREAU OF JUSTICE STATISTICS, RECIDIVISM OF PRISONERS RELEASED IN 1994, at 9 tbl.10 (2002). However, only 2.5% of them were arrested within three years for a sexual assault. By contrast, 13% of those who served time for robbery were re-arrested for robbery; 22% of assaulters were re-arrested for assault; and 23% of burglars were re-arrested for burglary. Perhaps some defender of current practices can explain why an assaulter's similar priors are routinely excluded from trial and a rapist's are routinely admitted, despite the fact that the latter's priors are both less relevant and more prejudicial than the former's (for instance, in the NCSC study, jurors learned of an accused rapist's prior convictions 71.4% of the time while, in cases of aggravated assault, they learned of priors 27% of the time). See Eisenberg & Hans, note 8, at 1374 tbl.5.
55. See generally Laudan & Allen, note 64. Of the dozen or so studies of exoneration familiar in the literature, there are none that suggest a false-positive rate of more than 5% and many make it much smaller than that.
56. 380 U.S. 609, 623 (1965) (Stewart, J., dissenting).
57. For an empirical study of the matter, see VALERIE P. HANS & NEIL VIDMAR, JUDGING THE JURY 124-27 (1986).
58. Arguably, admitting prior convictions as a matter of course would also have the desirable effect of undermining the principal rationale for the Griffin rule, Griffin v. California, 380 U.S. 609 (1965), thereby re-enabling jurors to draw adverse inferences from defendant's silence, which would remove one further obstacle to the defendant participating in the trial itself.
59. Plea bargaining would be facilitated by eliminating an important ambiguity concerning the admissibility of what some will continue to think is important evidence; for the same reason, more defendants may decline jury trials, again lowering costs. Of course, some may not think these are advantages, but we do.
60. Edward J. Imwinkelreid, An Evidentiary Paradox: Defending the Character Evidence Prohibition by Upholding a NonCharacter Theory of Logical Relevance, the Doctrine of Chances, 40 U. RICH. L. REV. 419, 433 (2006).
61. As we are about to discuss in the text, one needs to worry about unintended consequences. For all the reasons we have discussed, it does not seem likely to us that changing admission rules would have a dramatic effect on the mix of litigated cases, but we wish to acknowledge the possibility.
62. We are aware that the point we make in the text generalizes and that any change in the rules may change the incentive structure in myriad unpredictable ways some of which may be irrational-perhaps defense lawyers will not believe the data presented here, for example, and thus perhaps will feel the need to engage in more compromises during plea bargaining if prior records are more readily admitted. One can make one's own judgments about the significance of such concerns.
63. The manner in which information about prior convictions would be related to juries, the amount of detail about them, and so on, would need to be worked out. There would be much to argue about here, but these hardly present insurmountable problems. Judges are aware of prior records at bench trials, and no one seems to object to that. The data we have presented here suggest that there is no good reason to fear how juries handle this type of data, thus suggesting that the details of admission should be reasonably resolvable. A more difficult issue would be what instructions should be given about the use of the evidence. We would favor an instruction to the effect that the jury may consider the evidence on whatever issues to which the jurors think it is pertinent, but with a strong admonition about, first, its inferential limits, and second, that the issue is not whether the defendant is a serial offender but instead whether he committed the act of which he is accused.
64. We are indebted to Michael Pardo for encouraging us to think about this.
65. Griffin v. California, 380 U.S. 609, 614 (1965).
66. ROBIN AULD, A REVIEW OF THE CRIMINAL COURTS OF ENGLAND AND WALES 564 (2001).