Robots Versus Aliens

By Mike Dorf


To my amazement, the phrase "robots versus aliens" yields about 86,000 hits on Google, but none of them, I suspect, in the sense I have in mind, which is not as a proposed feature film.



No, I take the inspiration for the title of today's post from a recent NY Times story about how sophisticated computer software does a better job at finding relevant documents than lawyers and paralegals do, and at a fraction of the cost.  The article implied that such discovery bots would put thousands of lawyers out of work.

Color me skeptical, mostly because those lawyers were already losing their jobs.  In recent years, as clients have become increasingly cost-conscious, law firms have outsourced a substantial portion of document discovery to English-speaking lawyers and paralegals in lower-cost locales: in a word, to aliens.  (The term "alien" is sometimes deemed offensive when used to refer to non-citizens rather than extra-terrestrials.  I mean no offense, but titling this post "Computers Versus Non-Citizens" just wouldn't have cut it.)  So, if lawyer jobs are being lost to computers/robots, those are mostly jobs that would otherwise have been lost by American lawyers anyway.

Moreover, there never was a good justification for paying law firms several hundred dollars per hour to have young associates sort through hundreds of thousands of pages of documents looking for the few that were relevant to some massive litigation.  This work could have been done as well or better by paralegals, or at least by "contract" lawyers.  It often was done by both in the days before outsourcing and computer review picked up steam.  So, why did the litigation departments of law firms pay their junior associates to do document review?

One possibility is that the work really does demand high-level legal training.  There is a kernel of truth here.  If you're looking through thousands of documents for evidence of, say, antitrust conspiracy, it surely helps to know exactly what does and does not constitute such a conspiracy.  But this nonetheless still strikes me as a weak justification.  It may take sophistication to distinguish a document that tends to prove antitrust conspiracy from a document that only shows parallel behavior, but it takes little legal sophistication to distinguish completely irrelevant documents from arguably relevant ones.  For an interesting discussion of the economics, quality, and ethics of document review when performed by lawyers versus non-lawyers or domestic versus foreign lawyers, click here.  For myself, I doubt that the old system of using high-paid junior lawyers for document review was cost-justified.  I tend to think that law firms did this because they could get away with sending clients large bills for work that could have been done more cheaply.

It's worth noting that law schools spend almost no time whatsoever training students how to do document review.  I realize that this fact needs unpacking.  After all, it's possible to get a JD without learning lots of skills that are essential to much legal practice.  But at least law schools typically offer electives in such subjects: E.g., trial advocacy classes; negotiation workshops; advanced litigation courses that teach students how to take depositions.  I am not familiar with any law school course anywhere in which students are taught the actual nuts and bolts of reading through thousands of documents to see which ones reference matters in some dispute.  And that's because this is not really a distinctively lawyerly skill.

Indeed, under the old system, most new litigation associates regarded document review as something on the order of a hazing ritual.  You prove to the partners that you have the necessary drive to stay on the payroll by slogging through this work and then you are gradually given more responsibility: second-chair a deposition; then first-chair a deposition; argue a simple motion; write a brief; and eventually you're exclusively doing these more demanding and interesting tasks, while the next classes of associates have moved into document review.

Thus, from the perspective of a top-notch new lawyer looking to work in big-firm litigation, the computerization and/or outsourcing of document review should be very welcome news.  It means that she can start off doing the sort of work that better utilizes her distinctive skills and training.  Of course, that's not to deny that on net, computerization and outsourcing will reduce the total number of person-hours of labor being performed by American lawyers.  But if this is work that didn't need to be performed by American lawyers (or even by human beings) in the first place, it's hard to argue that the profession can in good faith mourn its passage.