Thanks to Batson and associated cases, we now have an uneasy working rule on voir dire in U.S. courtrooms: In exercising peremptory strikes, you can pick and choose on any basis…other than discriminatory ones. Basing strikes on race, ethnicity, gender, nationality and some other demographic categories is not allowed. That prohibition, however, is widely considered a paper tiger because for the attorney exercising the strike, all that needs to done if that strike is challenged is to come up with a plausible non-discriminatory basis for the strike. And that need not involve outright lying to the court, it could simply involve implicit bias. For many critics, that means that, in practice, as long as a reason passes the “delivered with a straight face” test, the strike it is allowed.

This focus on targeting only admitted discrimination is unusual, because in most other contexts, discrimination can occur in two ways: discrimination based on purpose and discrimination based on disparate impact. If a given action has a disproportionate impact on a specific group, then it is potentially actionable discrimination. The original Batson case targets affirmative intent to discriminate, not disparate impact. Subsequent decisions seem to have largely upheld that focus on intent rather than effect. However, some (Baldus et al., 2012) have argued for a broader perspective, including the greater use of statistical evidence in proving Batson claims. Instead of looking at the issue legally, my goal is to look at the normative question: Should disparate impact be used as a reason to challenge a party’s use of strikes? In other words, what should we do with questions that target a legitimate bias while also having a disparate impact based on race or some other protected category? In this post, I’ll unpack what that means and consider some arguments on both sides of the question.

What Would a Disparate Impact Standard Look Like in Jury Selection?

It helps to make a clear distinction up front. I see three levels of voir dire that could potentially be discriminatory:

Overt Discrimination: The prosecutor decides, “I want to minimize the number of African-Americans on my jury” and uses their strikes accordingly.

Pretextual Discrimination: The prosecutor has the same goal, but uses a reason (like, “Not making eye contact,”) to justify those same strikes.

Disparate Impact discrimination: The prosecutor legitimately is trying to avoid actual bias, but focuses on a form of bias that skews toward one race or category over another.

The difference between the last two forms of discrimination is not just the goals of the party using the strikes, but also the foundation for the bias. In a criminal context, the best example is probably negative experience with law enforcement. For the prosecutor relying on police testimony, negative experience with law enforcement is a real issue. The juror who feels they’ve been unfairly targeted or harassed will have a tougher time listening to and believing police witnesses. At the same time, African-Americans are far more likely to have had that negative experience. So relying on a real source of bias has the effect of disproportionately targeting a particular race. In a civil context, a less dramatic but still real example is probably anti-corporate bias. It is a real bias, but younger Americans are more likely to possess it. So a civil defendant basing their strikes on that bias could also be reliably discriminating against younger jurors.

Is a Disparate Impact Standard for Voir Dire a Good or a Bad Idea?

I can see arguments on both sides of this question. So for that reason, I want to frame this as an exercise I used to assign my debate students, by contrasting the arguments on both sides:

Good Idea

The bottom line is that disparate impact is discrimination. The effect is the same. If overt discrimination threatens the right to a fair trial and more broadly deprives citizens of the experience of jury duty, then disparate impact questions do the same. But there is no right to a completely bias-free jury. Both sides simply do their best within the rules to target the extremes of bias. That doesn’t mean getting everyone with an opinion off the jury.

Case law would have to establish how much disparate impact a factor has before it becomes an impermissible basis for a strike. That might sound onerous, but it probably isn’t much different from the distinctions courts have made in applying anti-discrimination law in other contexts. Closing argument: Discrimination is discrimination, and it shouldn’t matter if it is purposely based on a protected category or based on something that strongly correlates with a protected category.

Bad Idea

But there is a distinction. A question that ends up having different impacts on different races can still be aimed at a real bias. A factor that actually makes potential jurors predisposed to the other side is a factor that parties should be able to target in voir dire. True, but a disparate impact standard could easily become an open-ended limit on the ability to target bias. Few attitudes and experiences are going to be evenly divided across races, ages, genders, etc. They will skew one way or another. That raises the possibility that just about any good question in voir dire could end up having a disparate impact. That would make the task of discovering and addressing bias even more difficult and uncertain than it already is.

Closing argument: The purpose of allowing strikes is to provide parties with a practical way of addressing forms of bias that might not rise to the level of a cause challenge. As long as litigators are aiming at actual bias, their choices should be protected because that is a big part of what makes a jury system fair and rational.