Patrick Bayer, department chair and professor of economics at Duke, is known for his research in areas such as racial inequality, segregation, social interactions, housing markets, education and crime. Last week, Bayer, along with a team of Duke-led researchers, published a study examining more than 700 non-capital felony cases in two Florida counties from 2000 to 2010. The study found that juries with no blacks convicted blacks 81 percent of the time, and whites were convicted 66 percent of the time. This gap was almost entirely eliminated when at least one member of the jury pool was black. The Chronicle’s Arden Kreeger spoke with Bayer about the study, citizens’ right to a fair trial and possible policy solutions to address the inequalities.
The Chronicle: What compelled you to engage in this research?
Patrick Bayer: This is a severely understudied question, so getting data from the courts and from various jurisdictions on the composition of juries and on trial outcomes and especially what we needed for this study... is really hard to come by. We had essentially very limited evidence to date on how the racial composition of juries affects trials, and, despite the fact that a Sixth Amendment right to a fair trial is one of the fundamental bedrocks of our criminal justice system, there’s been kind of longstanding anecdotal evidence and concerns about representation on juries.
TC: Can you clarify why you chose to study jury pools?
PB: We look at the jury pools in this case because the variation in who happens to be called for jury duty... is effectively random. That means the set of trials that occur when it’s an all-white jury pool versus a jury pool that includes some black members look a lot alike. We’re able to isolate that impact in just that random change in who happens to be called for jury duty that day. It’s a lot like trying to approximate a clinical trial in medicine, for example, where the two groups are randomly assigned to either the drug in question or a placebo, and you can assess the impact of that drug by comparing the outcomes in those two groups because the membership in those two groups was completely random. If you actually tried to look at the seated jury, depending on who the defendant is, the attorneys may select jury members systematically in a way that shows some correlation, but it may not be causal.
TC: Are these the results you anticipated?
PB: The results are larger than I anticipated. The results are actually very large, when you think about the comparison we’re making. The population of the area that we’re studying is two counties, about 5 percent black. A lot of the time, there’s not a black member of the seated jury, so a lot of times there are a couple black members in the pool, but they don’t get seated, but we’re getting a change of about 17 percentage points in going from no black members of the pool to at least one or two. It’s that big, despite the fact that in many of these trials you’re still getting an all-white seated jury. That was the most surprising thing to me—the magnitude of the effects are really large. We wanted to approximate as close to an experimental design as we could, so what’s random here is the composition of the pool. As soon as you start looking at who is selected to be on the jury, that’s not at all random: Each attorney is systematically is trying to figure out who’s going to be best for their case. We focus on making comparisons across the set of cases where there are no black members of the pool.
TC: What are the implications of the study? What do you feel should be done to address this inequality?
PB: The main implication is that racial composition on the jury matters a lot. This is something that we need to look into a lot more. Our study is really the first to use data from real trials to look at this question, and it finds a big impact. On one level, our analysis argues for a lot more transparency on the part of local jurisdictions in terms of these types of data and what the outcomes are. There’s a broader range of policy or institutional changes that you can imagine considering. [One] way you can increase representation is by increasing the size of juries.... [In the United States] over the last several decades, in order to save money, these local jurisdictions are going to smaller juries, so, in our analysis, we’re looking at counties... where the size of the juries is six people instead of 12.... The chances of representation by any minority member of a community on the jury is much smaller as a result, so, in this case, you get all white juries way more often than you would if it was a 12-person jury.
TC: What about other minority groups? Would you expect to see similar results in the case of Latino defendants, or even female defendants?
PB: Women are not a minority in this sense, the jury is about 50 percent male [and] 50 percent female. That’s an important point here. If both attorneys have an equal number of jurors... there’s not that much room for the two sides to distort things. We really worry about situations where there’s a group in question, whether it’s on the basis of race or religion, that is a small minority. Then it’s possible for one attorney to completely exclude that group from their representation. We have looked at the effects of age and gender in the same data set. Gender is a protected category, and the attorneys are not supposed to use gender as sole basis of excluding a potential juror. We do see that men and women are seated in roughly the right proportion. The attorneys are allowed to use age, and they do discriminate on the basis of age. The prosecutors tend to prefer an older jury, so they strike relatively young jurors. Defense attorneys do the opposite, and this does seem to have an impact. When the jury pool is older, conviction rates are higher.... It’s constitutionally legal at this point to use age as a basis for selecting jurors.
TC: A recent New York Times article discussed a North Carolina case in which an inmate was taken off death row after it was determined that racial bias played a significant factor in his death sentence. Can you see your research being used in cases like this?
PB: I’m not 100 percent sure, in terms of my particular study. If attorneys use race to strike jurors, it’s pretty easy to keep black jurors from ever being on the jury, if the local population has a small enough fraction of blacks. In our data set, each attorney typically has six strikes that they can use, and there’s never that many black members of the jury pool. It would be possible to exclude all black members of the jury. As people look at the evidence in different jurisdictions, in our analysis our attorneys selected [jurors] in the right proportion, and there was no direct evidence of attorneys striking black defendants too much. There does appear to be that kind of evidence in other places.
TC: Do you have any plans for future research on this subject?
PB: Part of my hope is that this project opens up a lot more data across a wide variety of jurisdictions in the U.S. We’re trying to look more directly at the evidence in the cases. We can see in our data set that this random variation in the jury pool is having a big impact on the outcomes of the trials, but we can’t quite say if it is the quality of the evidence or the quality of the cases.... Are those similar, or are those very different? We’re looking to get data from the transcripts from the trials to take any mention of race out of those transcripts and have people evaluate the quality of the evidence, just to see what kind of a reasonable distribution outcome should be in these cases. Right now, we can’t really say which juries are getting things right.