On jury duty

I've been thinking about jury duty lately, in part because I narrowly avoided it earlier this week. The particular topic I've been considering is why this particular in-kind tax bothers me.

First, let me be clear that I am generally a fan of jury trials, my concern is with the institutions used to select jurors.

For non-US readers, the general set-up is that potential jurors are chosen at random, usually from voter registration rolls. In some jurisdictions it is pretty easy to get out of serving on occupational or hardship grounds. In others, like Ann Arbor, it is not, though Ann Arbor will let you essentially pick a date when you want to serve, within some time frame. Jurors are paid a very minimal amount and most employers do not pay for jury duty days - you have to take leave - though they are forbidden from otherwise punishing those who end up serving on juries. Montgomery County Maryland, where I did one day of jury service, rubs the low pay in your face by suggesting that you donate your tiny check to a charity of *their* choice. The length of service depends on whether or not the potential juror gets assigned to a trial and, if so, how long the trial lasts. Thus, the size of the in-kind tax has a very skewed distribution when measured in terms of number of days, as most trials are short but some are very long.

Four things bother me about the current regime. First, it is tied to voter registration. Second, the tax it imposes is random in size due to the different lengths of trials. Third, jurors are essentially slaves. Fourth, the system makes inefficient use of the available labor. Let me consider each issue in turn.

1. Why should people who register to vote be punished for doing so? This link should mainly deter people with high values of time from voting. Those people are likely to cast more informed votes.

2. Random taxes are a bad idea in general if, as one expects, people are risk-averse. One of the lessons of economics is that it is often good to convexify. Paying jurors their wage, rather than a low flat fee, and funding the payments out of general revenues, would spread the cost of jury duty more broadly. It would also make plain the true cost of jury trials, and so provide an incentive for using them only when the costs exceed the benefits. The use of coerced labor disguises the social costs of jury trials and so likely leads to their overuse.

3. One wonders if volunteer jurors would do a better job than coerced jurors. It would be worth funding some research on this if, indeed, the literature does not already offer up evidence on this point. We moved the military from slaves to mercenaries (i.e. from a draft to an all-volunteer force) partly for this reason.

4. The current system implicitly assumes that someone whose opportunity cost is $1000 is a ten times better juror than someone whose opportunity cost is $100. This seems unlikely. If it is not the case, then it is inefficient (in the economic sense of that term) to choose jurors without reference to their opportunity costs. Sensible institutions would obtain the desired level of juror quality at minimum cost, not without reference to cost, as in the current system.

Some possible policy changes:

1. Use SSNs to pick jurors rather than voter registration records.

2. Allow those selected for jury duty to pay a fee rather than giving up their time. The fees could be used to finance higher compensation for those who serve. I would happily have paid a couple of hundred dollars to avoid the risk of serving. In a sense, this is like the system of "substitutes" used under the union draft in the civil war.

3. Have all-volunteer juries. Some people probably really like doing this. Perhaps it could be something like the CCC or TFA where you sign up for a year or two, go through both screening and (gasp) training, and then serve on a number of juries. Having jurors with both training and experience might substantially improve the quality of the decisions, in addition to whatever improvements arise from using volunteers rather than slaves.

Finally, as an aside, why is the link between jury duty and voter registration never mentioned in discussions of lower voter turnout in the US? It would be interesting to see if cross-state variation in turnout is related to the leniency of the jury system (e.g. how easy it is to get out of serving).

Addendum: not one but two alert readers - thanks to Sasha and Julie - have pointed out that voter registration rolls are no longer used to select juries, it having been ruled unconsistitutional to do so about a decade ago. So, ignore the first policy recommendation. Most states, including Michigan (look under "how"), now apparently use driver's license lists, which should spread the burden reasonably well other than places like New York City.