Jury Trials, Impeachment and Otherwise

A pretty major trial concluded on Saturday in the US Senate, as you might have heard. Despite a 57-43 majority voting to convict President Trump, he was officially acquitted, the Senate requiring a two-thirds majority in impeachment trials. The reality that a simple majority in the Senate isn’t enough to get something passed (most of the time) is nothing new. But it is still shocking:

But rather than dwell on the antimajoritarian features of American lawmaking—many of my Broadstreet coeditors are eminently more qualified than I am to do so—the Senate trial pushed me to think more about trials than the Senate.

Let’s start with the two-thirds majority required for conviction in an impeachment trial. This isn’t the standard in jury trials either, but the difference is in the other direction. In criminal trials, conviction by jury must be unanimous. Obviously, no juries have as many members as the Senate trial—or as many jurors with such strong conflicts of interest. But still, unanimity is a high bar. As Henry Fonda in Twelve Angry Men and Pauly Shore in Jury Duty taught us, that standard can be important and push juries to make the right choice. (I have no regrets about the previous sentence.)

Puck, making the argument against jury unanimity (I think)

Though unanimity has always been required for federal trials, at the state-level unanimity is a new standard, following a 2020 Supreme Court decision. Prior to Ramos v Louisiana, two states, Louisiana and Oregon, both allowed criminal convictions with less than unanimous juries; 10-2 in both states. The histories of these standards are dark.

In Oregon, the 10-2 standard dates to a 1934 state constitutional amendment. Concern over Catholic and Jewish migration to the state prompted the change which effectively meant a single holdout or two on the jury could not stand in the way of the majority.

In Louisiana, the standard is even older and dates to the white backlash to Reconstruction (for more on race and political violence, see these two posts from Trevon Logan). As of 1880 (entering the state constitution in 1898), Louisiana required a 9-3 majority to convict. I first encountered this law thinking about the convict leasing system in the South of this era. Reducing conviction standards or the difficulty of getting a conviction is one direct way to increase the supply of convict labor. Clearly, as Louisiana was the only Southern state with such a provision, it was not a requirement for convict leasing, but it perhaps played a role.

What are (were) the effects of the laws in Oregon and Louisiana? There are some quite elegant theories about how juries work (see for example Neilson and Winter IRLE 2005 and Bouton, Llorente-Saguer, and Malherbe JPE 2018 who feel strongly that unanimity is bad), but quantitatively I don’t think we know (as always, I’m open to twitter-flaming when I’m wrong about something). 

One key constraint is that quantitative work on courts historically is really difficult. Naidu (2010) studies anti-enticement laws in the postbellum South but can’t show exactly (or roughly) how many people were prosecuted for breaking anti-enticement laws as “[m]uch of the relevant archival evidence remains dispersed in county courts across the South.” Instead, he relies on the cases where local court verdicts were appealed to higher courts, cases that are digitized. If Suresh couldn’t figure out how to collect some hard to find data, there’s not a lot of hope for the rest of us.

There’s even a “nice” natural experiment in non-unanimous rules in Louisiana: In 1974, a constitutional amendment changed its standard from 9-3 to convict to 10-2 to convict. Obviously, your DD alarms should be going off. How did YOUR FAVORITE OUTCOME HERE change in Louisiana versus its neighbors after this change? Reading my notes from when I first encountered this: “This feels like a pretty clean and useful policy change for understanding one of the big rules of the game in criminal trials.” Well said, past me. But it turns out the right kind of historical outcomes are really hard to collect (or at least they were for me) so we have to leave it at we don’t know (yet… I’ve been carrying around this question for several years, so if anyone wants an enthusiastic but bandwidth-constrained coauthor, let me know!)

Even the direction of the effect of unanimity is complicated. Anecdotally, the unanimity requirement in other states played a role in some high profile cases. In Race Against Time, Jerry Mitchell reports that the plotters behind the killings of James Chaney, Michael Schwerner, and Andrew Goodman—klansmen Sam Bowers and Edgar Ray Killen—were both originally saved by holdout jurors (though Bowers was involved in the Mississippi Burning cases, his holdout juror was actually in a different Civil Rights era Klan murder, that of Vernon Dahmer). It seems likely that many other (less high profile) Civil Rights era trials of KKK murderers, thugs, and plotters ended with similarly hung juries, justice delayed or denied with by only a few (or a single) member holding out and refusing to convict.

The Louisiana non-unanimity rule (and to a lesser extent the standard in Oregon) are smaller pieces of a bigger picture of how the legal system worked and did not work in the postbellum post-Reconstruction South (and the rest of America). As Higgs notes in Competition and Coercion, “[t]o appeal to the law was to seek aid in the enemy’s stronghold” (page 132). Disenfranchisement led to laws that were discriminatory in both letter and administration. And of course, the exclusion of African Americans from the voter pool excluded them from the jury pool as well.

In Washington last week, the makeup of the jury pool also drew some notice, from Senator Leahy sitting in as judge and juror to the potential co-conspirators (or at least enthusiastic fans of the defendant) sitting in judgment. Clearly, who was on that jury mattered. But that isn’t unique to impeachment either. In a series of papers, Anwar, Bayer, and Hjalmarsson have shown that jury racial mix really (really) matters, older jurors are more likely to convict, and juror politics matters

Pro-suffrage argument for women to get the vote and serve on juries

But all of those papers from Anwar, Bayer, and Hjalmarsson, nice as they are, use data from years starting with 20XX. So I want to flag my favorite paper in this set that draws on data from the Old Bailey from 1918 to 1926. As of 1921, women in England gained the right to serve on a jury. What happened with juries of her and his peers? Though overall conviction rates don’t move, women on juries did increase the conviction rates for sex offenses and violent offenses with female victims. They also show fewer convictions in abortion-related cases.


  • I am an Assistant Professor in the Boston University Department of Economics. I am also a Faculty Research Fellow at the NBER in the Development of the American Economy program. My primary research interests are in labor economics and economic history.

Leave a Reply