This article originally appeared in The Recorder on November 8, 2018.

Few topics provoke as much consternation among lawyers and law students as jury nullification. It’s a legal problem child — serving as a vital check on government power while simultaneously undermining the very premise of our adjudicative system. And what can trial courts tell juries about this unruly creature, which fits so poorly into the rubric of the judiciary? That’s precisely the question faced by the Ninth Circuit Court of Appeals in the recent case of United States v. Lynch

Charles Lynch ran a marijuana dispensary in Moro Bay, California. At the time, California had already decriminalized medical marijuana. And Lynch claimed that he had called the DEA before opening the dispensary and had been informed that “it was up to the cities and counties to decide how they wanted to handle the matter.” 

Lynch’s dispensary was financially successful, bringing in about $2.1 million in revenue. But that success was short-lived. Less than a year after the dispensary opened, federal authorities obtained a warrant and raided Lynch’s home and the dispensary. The federal government then indicted Lynch on five counts related to his possession and distribution of marijuana.

During voir dire, Lynch’s counsel told prospective jurors that the ultimate decision whether to convict Lynch rested with them alone. The government objected to those statements, arguing that they seemed to call for jury nullification. The district court cautioned Lynch’s counsel to avoid questions suggesting jury nullification. Lynch’s counsel continued to ask prospective jurors if they agreed that “whether to find a person guilty or not guilty is your decision.” And one of the prospective jurors responded that he believed “there is something called jury nullification, that if you believe the law is wrong, you don’t have to convict a person.” 

The district court then stopped voir dire and instructed the prospective jurors that: 

Nullification is by definition a violation of the juror’s oath which, if you are a juror in this case, you will take to apply the law as instructed by the court. As a . . . juror, you cannot substitute your sense of justice, whatever it may be, for your duty to follow the law, whether you agree with the law or not. It is not your determination whether the law is just or when a law is unjust. That cannot be and is not your task.

The court also asked each prospective juror whether he or she could follow the court’s instruction not to engage in nullification. The two prospective jurors that indicated they would have difficulty doing so were excused for cause. And in its closing instructions, the district court reiterated that “You must follow the law as I give it to you whether you agree with it or not. . . . You will recall that you took an oath promising to do so at the beginning of the case.” The jury convicted Lynch. 

On appeal, Lynch argued that the district court erred in instructing the jurors regarding nullification. A divided panel affirmed. 

The majority opinion — penned by Judge John M. Rodgers, a Sixth Circuit judge sitting by designation — concluded that the district court’s admonitions were proper. In doing so, it relied on the Ninth Circuit’s 2017 decision in United States v. Kleinman, which distinguished between instructions that permissibly admonish a jury to follow the law and those that impermissibly imply that jurors could be punished for nullification or lack the power to nullify. The panel also concluded that the instructions were particularly justified here because Lynch’s counsel provoked the initial discussion of jury nullification. 

Judge Paul J. Watford offered a succinct but thorough dissent. Tracing the long history of jury nullification, the dissent described the power as “[o]ne of the fundamental attributes of trial by jury . . . .” From the perspective of the Founding Fathers’ generation, “the jury’s independence from control by the judiciary provided assurance that application of national law would rest in the hands of local citizens attuned to the concerns of their community, not in the hands of officials beholden to a distant central government.” The Constitution’s twin guarantees of trial by jury and trial in the state in which the crime was committed were intended to serve as a check on federal power — a check that, by its nature, can be truly effective only if jurors are empowered to nullify. 

The dissent also identified several flaws in the majority opinion and in the court’s prior decisions on nullification. 

First, although the Ninth Circuit has approved instructions that tell jurors they cannot substitute their sense of justice for doing what the law requires, the dissent suggested that such instructions might in fact be improper because they affirmatively misstate the nature of the jury’s power. Jurors have no “right to substitute their sense of justice for what the law requires, or to determine whether a law is just or unjust, but they unquestionably have the ability to exercise that power — in fact, doing so is the very essence of nullification.” 

Second, the dissent argued that, even assuming the correctness of those prior decisions, the admonition here that “[n]ullification is by definition a violation of the juror’s oath” crossed the line. That was so because it carried an “implicit threat” of punishment for violating an oath made in open court. Indeed, the court in Kleinman had concluded that a very similar instruction was improper on that basis, affirming only because it concluded that the error was harmless. 

Third and finally, the dissent noted the difficulty inherent in applying harmless error analysis to improper nullification instructions. Identifying the basic constitutional guarantee at issue as the defendant’s right to trial by jury, the dissent observed that “[a]n instructional error of this nature would appear to defy analysis for harmlessness, since ‘the effects of the error are simply too hard to measure.’” The dissent thus implied that improper nullification instructions should in fact be considered structural error, despite the Ninth Circuit’s contrary holding in Kleinman. Moreover, even assuming the error was not structural, the burden was on the government to show that the error was harmless beyond a reasonable doubt. And given the popularity of medical marijuana in California, the government would have been hard-pressed — and indeed, did not even attempt — to show that the error had no effect on the outcome. 

The Lynch case thus highlights several outstanding tensions. If a jury has the power to nullify but no right to do so, can it be told that it is forbidden from nullifying? And if a defendant has a right to trial by jury but no right to ask that jury for nullification, is trial by a jury stripped of its nullification power fundamentally unfair? How can a reviewing court tell when an improper instruction is harmless, and should harmless error analysis even apply under these unique circumstances? Just as jury nullification fits uneasily into our process for adjudicating guilt, trial errors springing from the abridgment of that power do not lend themselves to straightforward appellate review. 

What is clear is that, to the extent that federal and state policies diverge — and the law fails to reflect the prevailing attitudes of a particular geographical area — jury nullification will become increasingly salient. And given the historical importance of the nullification power and the problems with existing case law identified by the dissent, this issue may well warrant rehearing en banc. 

On Appeals is a monthly column by the attorneys of the California Appellate Law Group LLP*, an appellate boutique with offices in San Francisco and Los Angeles. Susan Yorke is of counsel with the firm. She served as a law clerk for two Ninth Circuit judges and in the Appellate Division of the Oregon Department of Justice. Find out more about Susan and the California Appellate Law Group LLP at

*In July 2022, the California Appellate Law Group was renamed the Complex Appellate Litigation Group.

This article is copyright © in the year of publication above.