Have you ever had the same advertisement chase you around the internet? So you looked at that new suitcase once, and now it is popping up on every page you view. That effect is one part of our current world, where marketers and other persuaders seem to know more and more about our individual traits and preferences. Just had a life change? Graduation, new job, marriage, baby, or separation? That information is available and for sale, allowing salespeople to tailor their appeals to an unprecedented degree. It's a part of "big data,” which refers to the complex, growing, and interlocking systems of recording consumer and citizen data for the purposes of assessing credit, targeting purchases, and persuading voters. While the person trying to sell you a suitcase knows the uses of big data, chances are your jury commissioner doesn't. Jury pools are still assembled using lists that are often old, incomplete, and inaccurate. And juries are still generally assessed using the attorney's perceptions and, to the extent it's allowed, oral voir dire. University of the District of Columbia Law Professor Andrew Guthrie Ferguson suggests that it might be time for that to change. The article appears in the current Notre Dame Law Review, and is covered in a recent piece in The ABA Journal, and can be downloaded (for free) from SSRN.
Ferguson notes that despite its commercial ubiquity, big data has not meaningfully found its way into the courtroom yet, and “this institutional ignorance is purposeful, puzzling, and soon to be challenged by ever-expanding 'big data' technologies which are currently collecting billions of bits of personal data on American citizens.” After a thorough discussion of the opportunities and risks of incorporating big data into jury selection, professor Ferguson concludes that courts will have to take a balanced approach. The discussion, however, is more favorable than critical, with Ferguson believing that many of the threats (for example, privacy, court system legitimacy, real and perceived separation between government and corporations, and equal protection) can be addressed through careful implementation. He also shares the opinion that the more targeted information made possible by big data leads to less discriminatory selection: “A large-scale Batson workaround.”
There might be some role for Big Data in modernizing the jury pool management process. After all, if marketers know whether an individual has moved or died, the jury comissioner ought to know that as well. But I believe there is only a small role for actually assessing and picking that jury. In this post, I'll take a look at Dr. Ferguson's qualified argument for using big data in jury selection, noting where I agree and where I disagree.
A Good Idea for Calling a Jury Pool
Dr. Ferguson criticizes the method courts use in order to pull venire panels as one that is a “purposely shrouded and opaque process.” Current federal law requires that such lists be drawn randomly from government lists, typically looking at age, race, gender and zip code only. To improve the show rate and to make the results more representative, big data companies could be paid by court administrators to compile lists that would be more accurate in meeting the “fair cross section” requirement of law and state and federal constitutions. Those lists could do better job of removing deceased jurors or those who due to age, disability, or language will be disqualified once they show up in court. Beyond that, the more specific data could also be used to allow administrators to send out batches of summonses that would proportionally match the demographic characteristics of the venue. That step that could hold potential as a remedy for districts that have faced successful “fair cross section” challenges.
A Bad, or at Least Incomplete, Idea for Assessing a Potential Juror
While big data may have clear value of improving the quality of the mix showing up at the courthouse, Ferguson takes the argument further arguing that big data could be “incredibly helpful for litigants trying to pick a favorable jury.” Ferguson’s main argument is that allowing access to big data evens the playing field, and promotes distributional equality. “Instead of knowing a juror is a 23-year-old, white woman who works as a nurse, litigants in a big data world might know that the juror also reads parenting magazines, but not news magazines, recently went bankrupt, votes Republican, owns a licensed gun, and gives to religious charities.”
Access to that kind of information, he argues, is currently denied to many because only wealthy parties can afford expensive jury-selection consultants. To balance the scales, court systems should provide court approved profiles to all litigants at public expense.
While the fact of selective access to jury consulting based on wealth is a fair criticism, I see a difference between big data and the kinds of social media research that consultants provide. What we do relies on either public records or conscious social sharing. Big data, in contrast, is dominated by commercial and financial information that is, by nature, proprietary. The scope and form of this data exists only because most of us now use cards rather than cash to make our way through the world. Somewhere buried in that credit card agreement, I can only assume, there is a permission to store, use, and transfer that information. But without an option to say “No thanks” and still use the card, it is not a real choice. In short, we don’t have the equivalent of Facebook’s privacy setting on our bank and credit card agreements. Big data’s collection of consumer data is automated and systematized, not situational. It is not a case where a business says, “we need something on Mr. Jones, let’s look and see what we can find.” Rather every time Mr. Jones does something that affects his electronic finances, it is recorded somewhere, and these systems talk to each other in order to form, at all given times, a pretty complete picture of Mr. Jones. That has the potential to make the information more of a shotgun, when what is needed for jury selection is more of a scalpel.
To share one timely example, I receive many advertisements, emails, and other funding appeals from Donald J. Trump. I’m not sure why, since long-term readers of this blog probably understand that I’m on the liberal end of the spectrum. My theory is that I show up in Trump’s data trawling probably because I read articles and follow links about Trump, perhaps because I’ve blogged about him. The big data can count the times I’ve used “Trump” online, but cannot (yet) really understand what I’ve said about him.
Ferguson offers big data, not as a supplement but as an alternative to oral voir dire. It shouldn’t be. Instead of spending the court resources on distributing big data profiles, courts should instead invest those same resources into written questionnaires that are particular to the forms of bias that are most likely to matter in this case. Of course, oral voir dire and survey responses aren’t perfect, but there are many ways of making that information much more reliable, and in any case, the potential jurors know what information they are providing, and that serves as a very reasonable check on what could otherwise be a wide-open fishing expedition.
Dr. Ferguson’s most important point is that some role for big data is an inevitability, so courts and attorneys should start thinking about the intersections of the data that court systems use and the data the commercial world relies on. That point is well taken, and big data services might be worth checking out. There is at least one (Jury Mapping) that specializes in the litigation use of big data, and that might be good material for a future review in this blog.