I’m curious if anyone has ever attempted to quantify how much variation you can get in convictions from jury to jury. I’m imagining staging the same trial, as close as they can get it, for many different juries. The same lawyers making the same arguments, with the same judge, the same expert testimony, etc. And just seeing how often one jury would convict and one wouldn’t.
Maybe analyzing the jurors themselves - perhaps whether having a defendent of the same race/sex/age as them makes them more favorable, or their ideas about justice, or anything like that.
Has there ever been a serious attempt to quantify jury variability like that?
I’ve never heard of it, but I like the idea. Are you suggesting something that amounts to a stage play repeated many times, but with each new audience as a jury?
Jury consultants sometimes conduct mock trials on behalf of their clients, which is the closest thing I can think of to what you’re describing, but on a much smaller scale, obviously.
In the cases I’ve participated in, you take a large pool of “jurors” and provide them all with the same presentation. It can be as simple as a single person reading them the facts of a case in summary fashion, or as complex as putting on a true mock trial with “prosecutors,” a “defendant” and “witnesses.” After the presentation is complete, the large group of jurors is then split into smaller focus groups. Each group is then asked the same set of questions and given the opportunity to debate the issues of the trial.
In this way, it’s possible to see how one or two people with a different way of thinking can have a big influence on the outcome of a trial.
About 20 years ago a friend who was in law school showed me some studies like that except they were for civil trials, not criminal. I remember one phrase the jurors kept using to describe the plaintiff in exit interviews was, “She won the lottery!”
The British politician, novelist and noted jailbird Jeffrey Archer did write such a play about 15 years ago. At the time he was facing a perjury trial himself. (He was later convicted; got 4 years.) The play was about the trial of an accused murderer. The role of jury was assigned to the audience, who voted on to convict or acquit the protagonist.
I don’t know if any record was kept of the pattern of convictions and acquittals, but two factors might make it difficult to draw any firm conclusion from such a record. First, after the vote the play proceeded with one of two alternative endings (depending on the outcome of the vote), and both endings revealed whether the protagonist was, in fact, guilty. Once the play had been running for more than a short time, word of this might have reached at least some of the audience, and this might have influenced how they voted. Secondly, in the principal London production of the play, the protagonist was played by Archer himself, and votes might have been influenced by peoples’ (frequently strong) feelings about Archer rather than about the character as written.
I don’t think it works very well if you know it’s staged. If it’s going to be up to scientific research standards, the subjects can’t know that they’re actually part of the research since it will affect their decision making.
I work in the trial consulting field as a graphic designer, and I have participated in some of these “jury research exercises”. I put that in quotes, because they are, as mentioned, also called mock trials, but there’s more to it behind the scenes. In my experience, as the mock trial is being conducted, the trial team and their client(s), and the trial consultants are in an adjacent room with video feeds of what is going on with the focus groups (or, if it’s a small enough exercise, behind a one-way mirror).
As the lawyers present their cases, jurors are given handheld devices with which they can rate their reactions to what has been said and they also respond to questionnaires related to the case, on which they rate “strongly agree, somewhat agree, somewhat disagree, strongly disagree”, or similar. The results are transmitted to a video feed that displays a continuous line graph that can show general trends in juror opinions. And, as Asimovian points out, the participants are split into smaller groups to discuss the case.
Of course, the depth of analysis provided by these jury consulting firms depends on clients’ budgets. There are 1-day mock trials, 2-day exercises, and now, online jury research is beginning to emerge as an alternative to a full blown focus group.
The thing I like to remember is that you obviously can’t just tell your client what they want to hear. There is a lot of social science, psychology, and statistics behind jury research. It goes beyond “54 year old white male, electrician, 3 children…” It depends on the venue as much as the profile of the juror. An electrician from Detroit might have a completely different reaction than an electrician from Phoenix. At any rate, if the research shows a potential bad outcome for the client, then it may be time to consider settlement, or some other option.
Gotta stop there. There are millions of things to discuss in relation to this field. For a reasonable understanding of what the trial consulting field is all about there are numerous resources, BUT… take the upcoming CBS series “Bull” with a giant boulder-sized grain of salt, because, from what I saw of the trailer, that ain’t the way it works.
But there are exceptions when you can argue that the experiment requires them not to know. It usually involves allowing them to think they are being tested for something else.
Note, these are formal exceptions, not just something someone can argue, They have to present their experiment design to an ethics board. And they push things like it being the minimum required and such.
I wouldn’t say that every mock trial needs it, but I do think it should be done, if only to see how much thinking the trial is real influences the results. That way such discrepancies can be built in.
I suppose part of the question is the variability of people’s views. It may be difficult to quantify. Your stereotypical electrician from Detroit or even Phoenix may be white, middle class, stable, but might still have a deep skepticism of the truthfulness of police or a distrust of forensic scientists. Can you ask a prospective juror “do you think police lie?” Your civil case juror might have a dislike of big business, or a dislike of people “who won the lottery”. (Just start a debate here about the MCDonalds hot coffee case).
The question is whether basic profiles are generally accurate enough, or whether you are allowed ask the right questions to determine prospective jurors’ views. There’s the old joke that the prosecutors on the OJ Simpson case were happy to add women to the jury, thinking they’d identify with the battered and stalked and murdered wife. The defense went along, because they believed - more correctly? - that black women had an antipathy to white blonde bimbos who “stole” available black men.
I just wanted to say that I’ve only participated in one matter where the technology was being employed at this high a level, but my experience precisely matches with your description. It was quite the operation.
Thanks for providing a better explanation than I was able to.
One thought about this is that it will depend on the case: something that comes down to one person’s word against the other’s is going to have more variability than one with a smoking gun, surveillance footage, multiple witnesses, compelling motive and signed confession. And, our current legal system is designed so that the second category of slam-dunk verdicts almost never go to trial.
So cases that actually go to juries are, by their nature, going to be more variable than cases overall.
This is an important point to consider. In federal criminal cases, over 90% of defendants plead guilty rather than seek a trial. These cases, if they had undergone a full trial before a jury, would likely have led to a guilty finding because the facts are likely to be straightforward.* As to the remaining 10% of cases, these are presumably those where the Defendant (and likely his/her counsel) believe that the facts are such that they are willing to risk trial. So a jury empaneled before these cases is more likely to reach a more variable result than a jury in the other 90% of cases where the Defendant pled out.
I work in a largely appellate practice, where our cases are decided not by juries, but by panels of three appellate federal judges (11 judges when sitting en banc). We have a fairly good expectations of whether we will win or lose a case based on the makeup of the panel, but there are always cases where the outcome is still a surprise.
*Of course, defendants often enter a plea not because they are guilty, but simply because they made a reasonable calculation that taking a plea bargain is a better deal than pursuing a trial where they might see a longer sentence.
Also useful would be an analysis of whether the jurors are hungry or not:
The study looked at judges in Israel, but the results nonetheless suggest that there are lots of variables that you’d have to consider if you’re setting out to do a study like that.
I don’t think you can quantify it - too many variables.
But one thing seems obvious. If you look at media coverage of celebrated trials nearing conclusion, with comments from legal experts and very experienced trial observers, they tend to be extremely wary of making any sort of predictions - far more so than when handicapping legal rulings from judges. And some say explicitly the reason for that - with a jury anything is possible.