In a recent post, co-blogger Orin Kerr offers a thoughtful critique of my and Glenn Reynolds’ recent defenses of jury nullification. Orin’s points have some validity. But, on balance, I still think that jury nullification is likely to be more beneficial than harmful relative to presently feasible alternatives.

Orin argues that it would be better to leave discretionary decisions to prosecutors rather than juries, because the former have greater knowledge, and are more democratically accountable. Jurors are more likely than prosecutors to make mistakes out of ignorance.

I certainly agree that jury ignorance is a genuine problem, and have even written an entire article on the subject. It is probably true that prosecutors are more knowledgeable than jurors are, and therefore less likely to err out of ignorance. However, in a world of rampant overcriminalization, this virtue of prosecutors is counterbalanced by serious defects. Prosecutors are a self-selected group disproportionately drawn from those who believe that the current scope of criminal law is reasonable.

People who believe that a large percentage of the laws on the books should not be there are unlikely to take a job that involves numerous prosecutions they consider to be unjust, and also unlikely to be hired even if they choose to apply. By contrast, jurors are likely to be a more representative sample of the community (or at least not skewed in the same way). In addition, prosecutors are often rewarded based on conviction rates. That may incentivize them to pursue cases that are morally dubious, but easy to win, because there is little question that the defendant did in fact commit the offense in question. Jurors don’t suffer from this type of perverse incentive. Unchecked discretion by unrepresentative prosecutors is particularly dangerous when the scope of the criminal law is so broad that almost everyone violates it at one time or another.

In addition, while juror knowledge is far from perfect, jury nullification can partially offset the effect of voter ignorance. Extensive data shows widespread ignorance about most public policy issues, including crime. Today’s massive overcriminalization is in part the product of public ignorance, including the public’s tendency to believe that crime rates are higher than they actually are, and rising when they are actually falling. By contrast, jurors have stronger incentives to think carefully and rationally about the issues presented to them, and therefore are likely to be less influenced by ignorance than voters. If the alternative to jury nullification is unquestioned juror enforcement of laws enacted at the behest of ignorant voters, the net result will be greater influence of ignorance on criminal justice, not less.

Orin worries that a just conviction might be blocked by a single ignorant or erratic juror. But, in reality, it is rare for a jury to be “hung” by a single dissenter because such solo dissenters tend to give in.

Public ignorance also weakens Orin’s point about democratic accountability. While prosecutors are subject to democratic control in theory, in practice they are rarely disciplined for even the most egregious abuses of their discretion. Things might be different if voters paid close attention to prosecutorial performance, and incentivized elected officials to remove or discipline misbehaving prosecutors. But rationally ignorant voters rarely pay attention to such matters, except perhaps in a few highly publicized cases. And, obviously, the public’s ability to monitor prosecutors is further undermined by the vast scope of the criminal justice system, and the enormous discretion this gives prosecutors.

Jury nullification is far from a panacea for the dangers of overcriminalization and prosecutorial discretion. And it certainly has its downsides. As I noted in my previous post, I used to be opposed to nullification myself. I am still much less enthusiastic about it than its more aggressive proponents. In a well-structured criminal justice system, there might be little or no need for jury nullification. In the system we actually have, it’s a lesser evil compared to the realistically feasible alternatives.