Northwestern University statistics professor Bruce Spencer concluded in a recent paper that juries  provide accurate results 80% of the time. The paper, "Estimating the Accuracy of Jury Verdicts” was recently released online prior to a planned publication in this month’s issue of the Journal of Empirical Legal Studies (if you do not want to wade through Spencer’s statistical analysis and discussions, Northwestern created a very detailed press release that is a good start). Spencer considers two studies of 271 criminal juries in which the presiding judge filled out a survey explaining what the judge believed the correct verdict to be prior to hearing the jury’s verdict. Approximately the juries agreed with the presiding judges approximately 80% of the time. Spencer is careful to point out that a jury that disagrees with a presiding judge is not necessarily wrong. But he explains that jury-judge agreement is a reliable indicator of jury accuracy. Spencer also argues that his study is based upon a relatively limited data set, and that his study should not be a basis for broad conclusions so much as additional research.

As you might imagine, Spencer’s study has generated lots of blogosphere commentary. Anne Reed of the Deliberations blog (an excellent blog and my source for this post — thanks Anne) has an interesting prospective. She argues that the study highlights the tension between the jury’s job – applying the facts to the law and determining whether the party with the burden has met that burden – and what the jury thought its job would be – seeking truth and justice:

In short, the relevance of Prof. Spencer’s work to trial lawyers may be not in what it says about juries, but in what it says about jurors.  From high-school dropouts to university professors, they come to court with a clear, and often mistaken, idea of what their task is.  If your case depends on correcting that idea, you may need to work harder than you imagined. 

I think Reed makes an interesting point. But I took something a little more positive away from Spencer’s work. I was very impressed that juries and judges agreed 80% of the time. As a child, my dad (a criminal defense attorney) routinely asked my family and me to predict the outcomes of his trials. We were usually correct. My dad was not. At some point during law school, I stopped being able to predict his case outcomes. The law changes how you think. Perhaps lawyers become too clouded with burdens of proof and rules of evidence to appreciate how a jury sees a trial. As a federal district court law clerk, I had a similar experience. I saw a number of trials and as we waited for the jury, we would often try to predict the results in chambers. The only people who reliably predicted the results were those without law degrees. So, the 80% accuracy rate suggests to me that juries are doing a good job of putting aside their Law & Order notions of the courtroom and making the parties meet their burdens.