The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent


When Can Defendant Pleading Self-Defense Introduce Evidence of Alleged Attacker's Past Violent Acts?


From People v. Guerra, decided today by New York's high court:

We are asked once again to discard the rule recognized in People v. Rodawald (N.Y. 1904) and People v. Miller (N.Y. 1976) that "preclud[es] the admission of prior violent acts of victims in cases where a claim of justification is made" unless the defendant was aware of the specific acts at the time of the assault. We decline to do so.

Defendant stabbed the victim in the chest with a small knife, causing life-threatening injuries. At trial, the court determined that defendant was entitled to raise a justification defense. Defendant sought to introduce evidence of the specific violent conduct underlying four of the victim's prior youthful offender adjudications to prove that the victim was the initial aggressor with respect to deadly physical force. Supreme Court, in accordance with Miller, prohibited the jury from considering that evidence for that purpose….

"Youthful Offender status provides youth four key benefits: relief from [a] record of a criminal conviction, reduced sentences, privacy from public release of the youth's name pending the Youthful Offender determination on misdemeanor offenses only, and confidentiality of the Youthful Offender record." Youthful offender designations are given to those who have "a real likelihood of turning their lives around," and the protection gives these individuals "the opportunity for a fresh start, without a criminal record." Given these policy concerns, we see no reason to revisit the Miller rule in this case.

{[Under the dissent's proposed rule, a] defendant accused of murdering a 17–year–old could, if the victim happened to have a prior Youthful Offender determination, offer direct evidence of specific conduct committed by the victim as a child to show the killing was justified. Our exclusion of such evidence is neither "archaic," "obsolete," nor "out of step with other jurisdictions." To the contrary, defendant seeks to offer evidence of prior bad acts that would not be admissible under the Federal Rules of Evidence [404] or in nearly any state that has adopted those rules.}

Defendant's additional challenge to the constitutionality of the Miller rule is without merit.

Judge Rowan Wilson, joined by Judge Jenny Rivera, dissented; the dissent is long, but here are some excerpts:

Imagine for a moment that you are a juror in a criminal case. On trial is a young man with no history of violence. He is charged with stabbing another young man, one who on four prior occasions has, without provocation, assaulted and beaten up strangers. The defendant says he stabbed the victim with a penknife attached to his keys because the other man wielded a broken beer bottle as a weapon, hit him and was attempting to cut him with the beer bottle. The victim says he never had a beer bottle, never threatened the defendant, and it was the defendant who pulled out a knife and stabbed him following a minute of name-calling back and forth.

As a juror, would you feel better able to determine who was the initial aggressor if you knew of the victim's history of violence, or would you be better able to determine the truth without any information about the victim's prior violent attacks? Under the doctrine the majority leaves in place today, no court can ever allow you to consider that information in deciding who was the initial aggressor.

Santino Guerra stabbed Dylan Pitt with a penknife, after a verbal altercation between strangers turned violent. Mr. Guerra claimed he was acting in self-defense, and the trial court concluded that he was entitled to an instruction as to justification, to which no challenge is made. It is the People's burden to prove lack of justification beyond a reasonable doubt. Included in that burden is the requirement that the People prove that Mr. Guerra, not Mr. Pitt, was the first aggressor. Certain evidence suggested that Mr. Guerra was the first aggressor; other evidence suggested that Mr. Pitt was.

To assist the jury in determining that question, Mr. Guerra sought to introduce the facts underlying four of Mr. Pitt's prior arrests, each of which led to a criminal conviction replaced by a youthful offender adjudication. The trial court unsealed two of the youthful offender adjudications and permitted Mr. Guerra to introduce the facts concerning those two offenses, but issued a limiting instruction that the evidence could be used only for the purpose of evaluating Mr. Pitt's credibility and not to be considered in determining who was the initial aggressor.

The sole issue on this appeal is the propriety of the limiting instruction. There is no challenge to the unsealing of the youthful offender records and no challenge to the presentation to the jury of the underlying facts of Mr. Pitt's violent acts leading up to those adjudications. The majority says, in essence, that we should uphold the limiting instruction to protect Mr. Pitt's confidentiality and give him a chance to turn his life around.

Whatever force that position might have in a different case, it has none here, because that very evidence was exposed to the jurors (and anyone who attended the trial) in this case. Moreover, a defendant's right to put the People to their burden to prove guilt beyond a reasonable doubt—including proof that the defendant was the initial aggressor—is guaranteed by the U.S. Constitution. Mr. Pitt is not on trial; his statutorily protected interest in confidentiality pales in comparison and cannot be asserted to deprive Mr. Guerra of a fair trial….

Innocent people go to prison and guilty people go free when we exclude relevant evidence. Our blanket prohibition on using any evidence of the victim's character or prior bad acts in considering a dispute as to who was the first aggressor undermines our "truth-seeking function" by barring the consideration of facts relevant to a material issue. Excluding that evidence is especially concerning because it offends "our basic philosophic belief that in criminal cases there is to be greater latitude in admitting exculpatory evidence than in determining whether prejudicial potentialities in proof offered to show guilt should result in its exclusion." …

The exclusionary rule articulated in Rodawald and Miller has been justified on three bases, none of which is consistent with the truth-seeking function of courts nor our rules governing propensity evidence generally. The primary rationale has been that "the worst man has the right to live the same as the best, and no one may attack another because his general reputation is bad." The second reason, articulated most clearly in Miller, "is the need to carefully limit and narrow the issues that the jury must decide." Miller and Rodawald also suggest a third reason, similar to the second but relating only to the admission of specific acts as opposed to general reputation: "[g]enerally, character and reputation may not be proved by reference to specific acts, except to impeach the credibility of character witnesses."

The first rationale functionally extends our evidentiary doctrine in Molineaux beyond its jurisprudential foundations. In brief, our Molineaux rule prevents the prosecution from using evidence that a defendant has committed past crimes or bad acts to prove that the defendant probably committed the charged crime. We explained that our exclusion of such evidence was "the product of that same humane and enlightened public spirit which … has decreed that every person charged with the commission of a crime shall be protected by the presumption of innocence until he has been proven guilty beyond a reasonable doubt": introducing evidence of past bad acts "would lead to convictions, upon the particular charge made, by proof of other acts in no way connected with it, and to uniting evidence of several offenses to produce conviction for a single one."

By contrast, in civil law, liability is often proved by propensity evidence. In the criminal context, we generally break from our norm of permitting past conduct to demonstrate a propensity in order to protect the presumption of innocence and high bar of proof beyond a reasonable doubt—not because such evidence has no probative value. But neither the Molineaux rule nor its underlying purpose has any application to the use of prior bad acts of the complainant, who is not entitled to the same procedural protections as criminal defendants because the state is not criminally prosecuting the complainant.

Adapting our law to conform to the general approach in almost all other jurisdictions would not impinge on the reasons for eschewing propensity evidence to prove guilt. A modified rule would apply only in cases where a defendant was entitled to a justification charge, and only in the subset of those cases in which first aggressor status was disputed. In such cases, allowing a defendant to introduce exculpatory propensity evidence (whether reputational or specific relevant bad acts) poses no risk to erosion of the presumption of innocence or weakening of the burden to prove guilt beyond a reasonable doubt; those protections, after all, belong to the defendant. Nor would a modified rule harm victims of crimes by introducing needlessly prejudicial evidence.

The second and third rationales are largely undercut by Miller itself. There, we altered the rule in Rodawald by allowing the introduction of specific violent acts known to the defendant, even if those violent acts were not directed at the defendant. In permitting the admission of such evidence, though "mindful of the danger that the principal issues to be resolved may be lost in an endless maze of collateral matters," we resolved that concern by noting that questions of relatedness and extent of the proof would be sufficiently constrained by the trial court's "exercise of its sound discretion" to exclude prejudicial, cumulative, or collateral evidence. Having concluded that trial courts are fully capable of regulating violent acts about which the defendant knew so as to avoid unfairness to victims and miring juries in irrelevant mini trials, it is difficult to conclude that trial courts would be incapable of doing the same for prior violent acts unknown to defendants….

The Miller restriction puts juries like the one here in a peculiar position. On the one hand, if, prior to their altercation, Mr. Guerra happened to see reports of Mr. Pitt's prior bad acts in the news, or heard about them from neighbors, the jury could hear evidence of those bad acts to help decide whether Mr. Guerra was justified in stabbing the complainant. On the other hand, we say that the jury may not make that same deduction when deciding who it thought started the fight. In deciding the first aggressor question, the jury was—because of the Miller rule—admonished that it must not prejudge the complainant based on his prior St. Patrick's Day brawl. The Miller rule defies common sense and is out of line with much of the rest of our evidentiary law. We should update it, much as Miller updated the antiquated regime of Rodawald.

The "policy concerns" (majority op at 3) referenced by the majority should not prevent us from revisiting this doctrine in this case. Applying a revised rule would not have resulted in any additional harm to Mr. Pitt. Supreme Court independently decided to unseal Mr. Pitt's youthful offender records and permitted Mr. Guerra to ask Mr. Pitt about two of the convictions for the purpose of indicting his credibility. Under the Confrontation Clause of the Sixth Amendment to the U.S. Constitution, Supreme Court was required to permit Mr. Guerra to ask Mr. Pitt about those convictions. The only question facing Supreme Court with respect to these two adjudications was whether it should permit Mr. Guerra to use them as evidence about who was the first aggressor (it did not) and whether it should issue a limiting instruction to the jury admonishing them only to consider the convictions for the purpose of Mr. Pitt's credibility (it did). As for the two wholly excluded offenses, Supreme Court would be free at a retrial to continue to exclude all evidence about them. Modifying Miller as Mr. Guerra requests would not require the court divulge any new information about Mr. Pitt's prior bad acts….

The majority speculates that a defendant accused of murdering a 17–year–old victim might offer direct evidence of a prior youthful offense to show the killing was justified, "but one cannot simply assume that the checks [in the system] fail." In the majority's hypothetical, a trial judge would have found the evidence relevant, insufficiently prejudicial, and noncumulative at the very least, and a jury would weigh it along with all other relevant evidence. Suppose—to alter the majority's hypothetical a bit—the 17–year–old victim had several youthful offender adjudications resulting from attempted knifepoint robberies over the past year in the exact location where the killing took place and was found shot by the defendant, knife in hand. Surely in such a context the previous attempted robberies would help a jury decide whether the defendant was the first aggressor. Although hypotheticals can help us think more clearly and design better rules, the majority's textureless hypothetical does not help us grapple with the limiting instruction at issue in this case….

Congratulations to T. Charles Won, who represented the state.