What Advocates Can Do
What Advocates Can Do
While states regulate prior conviction impeachment through rules or statutes that vary in their particulars, all states but Montana cling to the notion that prior convictions can be admitted to impeach the credibility of at least some witnesses. This is different from admitting prior convictions to show that a witness is biased, for example if that witness has agreed to testify in exchange for a lower recommended sentence. Prior conviction impeachment is justified by the notion that simply having a prior conviction says something about the general moral character of the witness, not about their bias or motivation to lie in the particular case. When prior convictions are admitted through impeachment rules, most states assign judges the role of balancing the probative value of these convictions against the unfair prejudice that they inflict. Below we offer some thoughts on how you can argue on these two fronts.
Most jurisdictions require judges to assign a probative value to prior convictions before admitting them for impeachment purposes. This begs a question-- what are we measuring with this assessment of probative value?
The answer in most jurisdictions is that the prior conviction must be probative of a witness’s “propensity for truthfulness.” Thus, what judges must determine in conducting balancing on prior convictions is what those convictions tell us about a witness’s propensity for truthfulness. This is essentially a probabilistic question: Is someone with a prior conviction likely to lie on the witness stand?
The scientific answer to this question is no: someone with a prior conviction is not necessarily more likely to lie when testifying. Researchers have found that human character traits are an amalgam. While people do have behavioral traits, those traits are very context-specific. A person may tend to respond with a lie in one specific scenario and with the truth in a slightly different one. People adjust how they act depending on factors like whether they think they will be caught in a lie or their own cost-benefit assessment of lying in a particular instance. In order to predict someone’s “propensity for lying,” we would need very detailed information about how they acted in a nearly identical context in the past.
The science does not support the legal view of people as having straightforward traits of character, such as honesty, that might be indicated by a prior conviction. Prior convictions are too amorphous an indicator of the way a person behaves in response to situational inputs to offer meaningful information about whether they are more likely to lie as a witness. Given the level of specificity needed to predict human behavior, the very breadth of the claim that having a prior felony conviction predicts lying on the witness stand makes it suspect. In the United States, the criminal law targets actions so diverse that there is no unifying theory of criminal behavior, other than that the person has knowingly or unknowingly broken a criminal law. Even if we accept that a prior conviction tells us about past behavior, that information is far too general to help predict future lying.
So one way to argue the balancing tests is to assert that a prior conviction has minimal probative value on the question of a witness’s honesty or dishonesty. If that position is challenged, advocates might push for any proof that prior convictions have actual probative value on a witness’s character trait of being a liar.
Although many states do not permit balancing when a prior conviction is thought to involve dishonesty or false statement, the arguments canvassed here also pertain to those convictions. Extrapolating from existing personality research suggests that a conviction for fraud, for example, is too imprecise an indicator to reliably predict a witness’s likelihood of lying on the witness stand. Rather, we would need to know whether the witness has lied on the witness stand in the past in a situation in which they had similar motives for lying and a similar risk of being caught in the lie.
A second approach to arguing the lack of probative value of prior convictions is to point out that criminal convictions cannot be assumed to have a close connection to events on the ground. For this reason, we cannot assume that prior convictions offer accurate information about a defendant. And this assumption of accuracy is crucial to the idea that a prior conviction tells us about a defendant’s character trait.
We can’t assume prior convictions reflect events on the ground because prior convictions are not necessarily the outcome of a well-functioning criminal legal system. Trials have been replaced by an assembly line of negotiated guilty pleas, and this is facilitated by systemic inequalities that burden a defendant’s ability to take matters to trial. A defendant may choose to forego trial for reasons that have little or nothing to do with guilt or innocence. Those reasons include lack of financial resources, threats of higher sentences, lengthy waits for trial due to overloaded dockets, detention prior to trial when defendants cannot post bond, and uncertainty about the strength of the prosecution’s case against them. For many of these reasons, one defendant may have no prior convictions to be impeached with while another may, even if they have engaged in similar behavior. And too often, that distinction may be drawn along lines of race. The disfunction and inequity in the criminal legal system means that prior convictions should not be used as a hallmark of personality.
Of course, on the other side of the scales in the balancing test is the risk of unfair prejudice. And here, prior convictions fail balancing as well. No study of how fact-finders actually use prior convictions has found any evidence that they are, in fact, used to assess truthfulness. Rather, prior convictions have the pronounced—yet wholly impermissible—effect of lowering the burden of proof in close cases, making it easier to convict those with prior convictions. In one study, for example, jurors’ discussions of a defendant’s credibility did not differ depending on whether they were informed of the defendant’s prior convictions. Significantly, however, “jurors interpreted the case evidence differently as a result of knowing the defendant’s record.” The researchers concluded that “determinations of the defendant’s credibility are not the prime method by which criminal record influences guilt judgments.” Instead, “[t]he evidence against a defendant with a prior record appears stronger to the jury.” These empirical results suggest that prior convictions carry an extraordinarily high risk of unfair prejudice.
This has implications for judicial balancing of the probative value of a prior conviction in assessing a witness’s propensity for truthfulness as weighed against the risk of unfair prejudice. Even if we hypothesized some marginal increase in our ability to predict lies based on the existence of a prior conviction, that would never outweigh the proven near certainty that a fact-finder would instead use the information impermissibly to derive greater moral comfort when convicting the defendant. Indeed, the risk of unfair prejudice almost certainly substantially outweighs any slight boost in our ability to predict lying from prior convictions.
Because prior convictions lack probative value on the question of truthfulness or untruthfulness, Professor Jeffrey Bellin, has argued that if courts were to correctly apply already existing balancing tests within the rules, they should almost always prohibit such impeachment of defendants.
Relatedly, Professor John Blume has highlighted the risk that prior conviction impeachment contributes to wrongful convictions. In an important study he focused on a group of people who had been exonerated after their conviction on serious charges. He investigated why it was that so few of these people had testified in their own defense, despite having stories of innocence to tell. He discovered that the fear of prior conviction impeachment was the primary reason given by their attorneys for their clients’ waiving their right to testify. He thus concluded that “threatening a defendant with the introduction of his or her prior record contributes to wrongful convictions either directly—in cases where the defendant is impeached with the prior record and the jury draws the propensity inference—or indirectly—by keeping the defendant off the stand.”
As Blume’s research indicates, one form of unfair prejudice brought about by prior conviction impeachment is the deterrence of testimony by those facing criminal charges. Early case law highlighted the “importance of the defendant’s testimony” as the most important consideration for trial judges determining whether to permit prior conviction impeachment. Many jurisdictions continue to include this as a factor for courts to consider, but fail to recognize its importance, sometimes even inverting the factor, so that they conclude that the more important the defendant’s testimony the more the court ought to err on the side of permitting impeachment. Defense attorneys should attempt to reassert the centrality of this factor as a reason to prohibit prior conviction impeachment.
Jury instructions are commonly relied upon as a means of controlling the risk of unfair prejudice. It is important to note, however, that in this context they do not work. Deterrence of testimony occurs even though litigants are on notice that jurors will be instructed that they are to use the convictions only on the issue of credibility. Hawai’i’s Supreme Court rejected this form of impeachment for defendant-witnesses in part because of the inability of jury instructions to control the unfair prejudice. And research shows that jurors use this evidence for prohibited purposes even in contexts when they are instructed that they must not do so.
Finally, the racial and economic disparity that prior conviction impeachment inflicts is a form of unfair prejudice that judges should weigh. Prior convictions are imposed disparately along racial and economic lines, and their use for impeachment compounds existing disparities that pervade the criminal system. After the death of George Floyd, many state supreme courts (or their chief justices) voiced their commitment to promoting racial equality. Vigilance about the impact of prior conviction impeachment represents one means by which courts can give those commitments some substance.
Some states have provisions that appear to make at least some convictions mandatorily admissible. There are at least two strategies that defense attorneys can consider here. First, arguing for the narrowest possible interpretation of the scope of the category. For example, if prior convictions that involve dishonesty or false statement are admissible without balancing, it is important to argue that that is in reality a very narrow category that only includes classic crimes of dishonesty like forgery or false statement. You can find an example of this form of argument in an appellate brief from Washington State. Second, you might argue that judicial discretion to exclude prior convictions should be read into the rule. Some state courts have already found that such discretion exists, even where the language of the rule appears to require admission of a prior conviction. And litigation is pending in the Oregon Supreme Court on the issue of whether Due Process requires that judges be able to consider the risk of unfair prejudice before admitting a prior conviction.
Arguments based in state and/or federal constitutional provisions are a promising avenue in contesting impeachment with prior convictions. Advocates have invoked a variety of constitutional rights in the prior conviction impeachment context, such as the right to testify, to a fair trial, to put on a defense, to Due Process, to an impartial jury, and—when denied the opportunity to cross-examine prosecution witnesses—the Confrontation Clause. And sometimes this works. The Supreme Court of Hawai’i, for example, was persuaded by defense litigators that the state’s prior conviction impeachment regime violated both the state and federal constitutions, and that ruling set in motion the redrafting of the rule. An intermediate court in Oregon was persuaded that that state’s prior conviction impeachment regime violated the federal Due Process Clause. The Oregon Supreme Court is currently reviewing that case, in which the Prior Conviction Impeachment Reform Coalition submitted an amicus brief.
We recommend that advocates consider state and/or federal constitutional claims as part of the toolbox in arguing against prior conviction impeachment. As with most of our recommended strategies, creating a record at the trial level will aid the litigation of your arguments (both constitutional and non-constitutional) on appeal. At this time, it is not clear whether courts will be more responsive to constitutional-level arguments or arguments about the lack of probative value of prior convictions for impeachment. For this reason, consider making both types of argument. One recent appellate brief from Washington takes this approach, drawing on prior conviction impeachment scholarship to advance non-constitutional and constitutional (state and federal) arguments for change.
Evidence rules can change as a result of judicial opinions, but they can also be changed by the entity responsible for their amendment, such as a court-appointed rules committee responsible for the state code of evidence. For example, the rules drafters in Kansas and Montana changed their states’ rules, moving from permissive regimes to very restrictive ones. Figure out how the rules get changed in your jurisdiction—or reach out to us and we can advise. And think about proposing a rule change. Note that changing rules on prior conviction impeachment would not affect defense counsel’s ability to impeach witnesses with evidence of bias. For example, if a witness testifies after a plea deal with prosecutors, nothing in these rules would preclude making that fact known to the jury because it shows the witness’s bias. Bias is not encompassed by the rules on prior conviction impeachment and is always available as an argument for why a witness should not be believed.
The NACDL report on prior conviction impeachment suggests avenues that states might consider, each of which has support in either current practice or scholarly writing:
1) Declare this form of evidence inadmissible. This is what Montana’s rule does.
2) Permit impeachment of non-defendant witnesses with evidence of lying under oath. Professor Simon-Kerr has suggested a rule that would permit the impeachment of repeat courtroom players who are not the defendant with evidence that they have lied under oath; this might encompass police officers, for example, or expert witnesses.
3) Permit impeachment of non-defendant witnesses with dishonesty or false statement convictions. Hawai’i and Kansas have a version of this rule, and the NACDL report on prior conviction impeachment contains a model provision that aims to preserve judicial discretion, to narrow this category as much as possible, and to minimize room for disputes about which convictions are included. While convictions of all sorts lack probative value and inflict unfair prejudice, we include this proposal out of a sense that states may insist on keeping at least some form of prior conviction impeachment on the books.
4) Prohibit impeachment of defendant witnesses. This umbrella approach could involve regimes 2) or 3) or some other arrangement but would prioritize the protection of those facing criminal charges from this form of impeachment, in recognition of their constitutional rights and their unique vulnerability to unfair prejudice. Again, Montana, Hawai’i and Kansas are all potential models, and Professor Roberts has drafted a sample provision.
5) Make explicit the possibility that the defense could use this tool. A rule of this sort might declare that impeachment by prior conviction is prohibited, except where the exclusion of such evidence would violate the defendant’s constitutional rights. Such a provision is not technically necessary, because of course the constitution trumps a rule of evidence, but it would serve as a reminder that the defense can assert a constitutional right to use prior conviction impeachment to confront witnesses. Precedent for this kind of reminder of constitutional protections exists in rules like Federal Rule of Evidence 412.