Below is an article I wrote, published this month in the Utah Bar Journal. It might not be of much interest to anyone but the nerdiest of evidence law aficionados, but here you have it:
Keep Calm and Argue the Facts: A Pragmatic Approach to the Doctrine of Chances
Last year the Utah Supreme Court handed down State v. Verde, 2012 UT 60, which explicitly embraced a shadowy and oft-maligned exception to the character evidence prohibition known as the Doctrine of Chances. The Doctrine of Chances (herein “DOC”) has been around since at least 1915, and I suspect it has been confusing attorneys and judges the whole time.
As explained in Verde, the DOC may allow evidence of other events and circumstances outside the charges in question, based on “the objective improbability of the same rare misfortune befalling one individual over and over.” Id. at ¶47 (internal cites omitted). Specifically, the defendant in Verde was charged with sexual assault, and the court remanded the case back to the trial court to consider whether evidence of other uncharged sexual assault allegations from other complaining witnesses could be admissible under the DOC to disprove fabrication on the part of the current complaining witness. Id. at ¶62 (the remanded case is still pending retrial as of this writing).
In other words, being accused of three similar sexual assaults at three different times by three different witnesses may be an “improbably rare misfortune” which is unlikely to befall an innocent individual, making it more likely that the current witness is not fabricating his story. That’s the logic of the DOC.
Many defense attorneys are dismayed by the implications of State v. Verde. Understandably so. At first blush, the Doctrine of Chances threatens to completely swallow the enshrined prohibition against character and other bad acts evidence offered by the prosecution to show propensity. The obvious complaint against it goes: “Can’t the prosecution now offer evidence of any prior crimes or accusations and call it DOC evidence instead of character evidence?”
It’s a valid question. I think the answer is definitively no, which I explain later, but, perhaps more importantly, I think defense attorneys would be better served by preparing to distinguish DOC evidence or argue for its exclusion under Rule 403 rather than pushing back against the doctrine itself, as tempting as it may be. There are four reasons:
Section I: The Doctrine of Chances is Probably Here to Stay for a While
While an idea’s longevity is certainly no reason to accept it on principle, it is good reason to doubt the idea can easily be put to rest. As mentioned above, the DOC has been around for 100 years, and, though there has been disagreement during that time[i], several of the major authorities in the evidence world have considered and accepted it (E.G. Wigmore, Inwinkelreid).[ii] Many state and federal courts are also on board, see Verde, n.27 (noting other jurisdictions that have adopted the DOC), including, most importantly, our own. Like it or not, the DOC has some momentum, and there’s Newton’s law of motion working in its favor now. Further, as explained in Section II below, there might be an important practical reason why the doctrine persists.
Section II: The Doctrine of Chances May be Good for Defendants as a Whole, If It Preserves the General Character Evidence Prohibition
It’s worth remembering that the general character evidence prohibition may have Constitutional dimensions[iii], but there is not any clear Constitutional mandate for it. Evidently, it can be modified at will by the legislature, which is not famous for its friendliness to criminal defendants. The 2008 modification allowing evidence of other bad acts in cases of child molestation is a good example. See Utah Rule of Evidence 404(c). The legislature decided that evidence of past child molestation is simply too important and probative to be restricted by the character prohibition. There is no reason the exception couldn’t be widened, or the general character evidence prohibition done away with entirely. Some scholars have argued for such a step.[iv]
In fact, as Edward Inwinkelried argues in a recent article, the Doctrine of Chances might be the release valve that keeps the whole character evidence prohibition from imploding. See Edward J. Inwinkelried, An Evidentiary Paradox: Defending the Character Evidence Prohibition by Upholding a Non-Character Theory of Logical Relevance, the Doctrine of Chances, 40 U. Rich. L. Rev. 419 (2006). His argument is long and detailed, and I certainly wouldn’t do it justice here, but as a very brief suggestion: consider a famous DOC case in America, United States v. Woods, 484 F.2d 127 (4th Cir. 1973). The defendant was charged with murder of her 8-month-old foster child, who died of asphyxiation. Id. at 128-30. The testifying physician explained that he thought homicide was the likely cause of death, but that the asphyxiation might be explainable by some previously unidentified disease. Id. That testimony probably left room for reasonable doubt, were it not for other supporting evidence that the judge allowed in under the DOC. Namely, over the preceding 25 years, at least 20 of the small children that passed through the defendant’s care had suffered episodes of unexplained asphyxiation. Id. at 130. 7 of those children had died. Id. Unsurprisingly, the defendant was convicted of homicide. See id.
Now consider if such a case happened today, in Utah, and imagine the public attention it would receive. Consider what would happen if the evidence of those other dead children was not allowed at trial, because of the general character evidence prohibition, and the defendant was then acquitted. Consider the public outrage, the “Miscarriage of Justice” headlines, the television pundits and letters to the editor. As Inwinkelried put it, most laypersons would consider the exclusion of such obviously relevant evidence “an affront to common sense.” Inwinkelried at 421. And it likely wouldn’t be long until the legislators started looking at the whole character prohibition sideways, thinking “do we really need this thing?”
With the general character prohibition gone, judges would be left doing Rule 403 balancing for every case with prior bad acts evidence, and doing it without the scale automatically tipped away from bad character evidence. It would be very time consuming and probably result in the admission of more such prior bad acts evidence. I would argue that the general character evidence prohibition serves defendants better, even with the shadow of the DOC over it.
Section III: Doctrine of Chances Evidence Is Distinguishable from Character Evidence
To explain the distinction, I’ll present two cases. The first is the “original” DOC case of Rex v. Smith, 11 Cr. App. R. 229, 84 L.J.K.B. 2153 (1915). In that case, the defendant’s new wife died in the bathtub, leaving the defendant a substantial inheritance. See Inwinkelried at 434 (citing to case). The defendant claimed it was an accidental drowning, but you can likely guess the evidence allowed in under the DOC: two of the defendant’s previous wives had also died in his (startlingly accident-prone) bathtub. See id.
The second case is a hypothetical, and, I hope, typical character evidence case: the defendant is charged with bank robbery. He claims mistaken identity, but has a shaky alibi. The prosecution wants to offer evidence that the same defendant was convicted of robbing two other banks in the previous five years.
The distinguishing question, then, is whether the facts of each case present an “improbably rare misfortune”–a statistical anomaly that tends to rule out the possibility of innocent bad luck. See Verde, 2012 UT at ¶47.
So, in the Bathtub Case, the underlying, uncontested facts of the case–that three of the defendant’s wives drowned in his bathtub–present a statistical anomaly that tends to rule out the possibility of innocent bad luck. In the Bank Robbery Case, the underlying, uncontested facts of the case–that the defendant has been convicted of two prior bank robberies and now a third bank has been robbed–do not present a statistical anomaly tending to rule out the possibility of innocent bad luck. The fact that another bank robbery happened is not at all unusual; after all, banks get robbed all the time.
That, I think, is an easy, short-hand way to distinguish the two archetypal cases, but maybe it begs the question. How do you define an “improbably rare misfortune”? Let’s say the prosecution in the Bank Robbery Case is especially clever, and they say, “Ok, we don’t want to offer the prior bank robbery convictions, we just want to point out to the jury that the defendant has been accused of three different bank robberies. What are the chances that an innocent man gets accused of three separate bank robberies? It’s an improbably rare misfortune!”
Framed that way, the situation is more akin to the actual Verde case, which deals with uncharged accusations, and Verde itself provides the answer: it sets out four foundational requirements which must be met for DOC evidence involving uncharged misconduct: (1) materiality, (2) similarity, (3) independence between the accusers and (4) unusual frequency. ¶57-61. Here, the prosecution’s theory is easily shot down for lacking independence among the accusers. As the court points out, “the existence of collusion among various accusers would render ineffective the comparison with chance repetition.” ¶60.
It is hard to imagine a situation where accusations made by law enforcement agents would be sufficiently independent to represent an “improbably rare misfortune.” After all, if someone is under suspicion (or has been convicted) of past bank robberies, they will be top on the list of suspects for the next unsolved bank robbery.[v] Modern law enforcement, with its instantaneous background checks and inter-agency databases, involves collusion by its very nature. That fact alone should help assuage the fear that the DOC would be relevant in most criminal cases.
Section IV: Borderline DOC Evidence Can be Excluded Under Rule 403
Certainly the archetypal DOC case still looks like a propensity argument–that is, it still seems to rely on assumptions about repeated (bad) behavior over time, which the character evidence prohibition ostensibly prohibits. In the Bathtub Case, the fact of the previous drownings only seems relevant if the jury uses it to assume that the defendant tends to murder his wives by drowning them and that he did the same thing in the present case (IE: propensity). Likely this is another reason why the DOC is so troubling to practitioners.
Inwinkelried argues that pure DOC cases do not rely on character propensity, because the chain of inferences stops short of character judgment. See Inwinkelried at 448-457. In the Bathtub case, for example, the statistical improbability of the multiple drowning may help a jury determine only that one or more of the drowning “was not accidental.” Id. Verde explicitly adopts Inwinkelried’s reasoning, where the court says that propensity inferences “do not pollute” the DOC reasoning. Verde ¶50 (going on to explain that DOC inferences may help a jury reach “a conclusion that one or some of the occurrences were not accidents or false accusations”).
Certainly the DOC at least invites (if not begs for) a character propensity judgment (as do many types of admissible 404(b) exceptions, such as Modus Operandi). So, regardless of how convinced you are by the logical distinction above[vi], the most effective way to use the DOC’s obvious association with character propensity is under a 403 balancing argument. After all, the DOC’s arbitrary stopping point–the “conclusion that one or some of the occurrences were not accidents or false accusations”–significantly limits its relevance. And that stopping point is very likely to be lost on the jury, who will almost certainly take the obvious next step to character judgment and propensity, regardless of any limiting instructions.
Therefore, defense attorneys can argue that proffered DOC evidence has to be especially probative of a contested fact for the limited permissible relevance not to be substantially outweighed by the obvious risk of unfair character judgments. Verde invites such an argument, in fact, having remanded the case to the trial court to consider not just the DOC itself, but “the weighing called for under rules 404(b) and 403.” See 2012 UT at ¶62.
Of course, the distinction between DOC evidence and character evidence would not always be as easy as in my examples. Close cases will be difficult, but the four factors outlined in Verde significantly restricts the doctrine’s general applicability under 404(b), and the logic which distinguishes it from impermissible character evidence limits it even still–making it susceptible to 403 exclusion in all but the most unusual cases. My point is that the DOC need not be the exception that swallows the rule, but just another fact-specific and somewhat nebulous exception in an area of law already thick with them. And arguing these exceptions might well be the price we pay for the general character prohibition’s continued existence.
[i] See Edward J. Inwinkelried, An Evidentiary Paradox: Defending the Character Evidence Prohibition by Upholding a Non-Character Theory of Logical Relevance, the Doctrine of Chances, 40 U. Rich. L. Rev. 419, 444 (2006)(listing multiple evidence scholars that believe the DOC is not a valid exception to the character evidence prohibition).
[ii] See 2 John Henry Wigmore, Evidence in Trials at Common Law § 302 (James H Chadbourn revisor, 1979); Inwinkelried, supra note i, at 434-457.
[iii] Inwinkelried, supra note i, at 434.
[iv] See, e.g., Richard B. Kuhns, The Propensity to Misunderstand the Character of Specific Acts Evidence, 66 Iowa L. Rev. 777, 803-804 (1981)(“ The first step in a rational approach to the admissibility of specific acts evidence would be to abolish the purported distinction between character and noncharacter or propensity and nonpropensity evidence, and to state simply and directly that the admissiblity of specific acts evidence depends on a careful balancing of probative value against the concerns with prejudice, time consumption, and distraction of the fact-finder.”).
[v] Evidence scholar Paul Rothstein makes a similar point in an argument against the DOC. See Paul Rothstein, Intellectual Coherence in an Evidence Code, 28 Loy. L.A. L. Rev.1259, 1263 (1995)(“In the movie Casablanca, the police, having heard a crime had been committed, respond with the classic expression, ‘Round up the usual suspects.’ As in real life, a person who has been charged before commonly is charged again any time a vaguely similar crime is reported. Thus, contrary to the doctrine of chances, it is not so unlikely that an innocent person would be repeatedly charged falsely.”). Obviously, Rothstein’s version of the DOC did not include the foundational requirements adopted in Verde.
[vi] Personally, I am not quite convinced. But, as I noted above, I think the DOC is persistent and that there is a pragmatic reason for its existence, so, practically speaking, it’s a moot point.