A Revealed-Preferences Ranking of Law Schools

In March, my co-author Christopher J. Ryan, Jr. and I posted a draft article to SSRN, describing a law school ranking system based on the revealed preferences of law students. The article got some attention from David Bernstein at the Volokh Conspiracy, Paul Caron at TaxProf Blog, David Lat at Above the Law, and the Scholastica blog. I'm pleased to say that the article will appear in a forthcoming issue of the Alabama Law Review. In the meantime, we have posted a revised draft, titled "A Revealed-Preferences Ranking of Law Schools." Here is the revised abstract:

The U.S. News & World Report “Best Law Schools Rankings” defines the market for legal education. Law schools compete to improve their standing in the U.S. News rankings and fear any decline. But the U.S. News rankings are controversial, at least in part because they rely on factors that are poor proxies for quality, like peer reputation and expenditures per student. While many alternative law school rankings exist, none have challenged the market dominance of the U.S. News rankings. Presumably the U.S. News rankings benefit from a first-mover advantage, other rankings fail to provide a clearly superior alternative, or some combination of the two.

In theory, the purpose of ranking law schools is to provide useful information to prospective law students. Rankings can provide different kinds of information for different purposes. Existing law school rankings seek to provide information that will help prospective law students decide where to matriculate. However, objective rankings can provide useful information only if they measure factors that are salient to prospective law students, and different factors are salient to different students.

This article provides the first subjective ranking of law schools. It describes a method of ranking law schools based on the revealed preferences of matriculating students. Law school admission depends almost entirely on an applicant’s LSAT score and undergraduate GPA, and law schools compete to matriculate students with the highest possible combined scores. Our method of ranking law schools assumes that the “best” law schools are the most successful at matriculating the most desirable students. Accordingly, this article provides a “best law schools ranking” based exclusively on the LSAT scores and undergraduate GPAs of matriculating students. In contrast to objective rankings of law schools, which attempt to tell prospective law students which law school they should attend, this article provides a subjective ranking of law schools, by asking which law schools prospective law students actually choose to attend. This “revealed preferences” method of ranking law schools may help identify which factors are actually salient to prospective law students.

Of course, comments and suggestions are very welcome.

Comments

[Mike Spivey]

Brian,

Thanks for this. I was very much interested as I have long wanted to overlay a ranking of "top choice" (i.e. desirability) schools with USNWR. Also, I went to Vanderbilt Peabody for my higher ed work, so I already like your co-author.

here is my concern, hopefully you can help me here. Every year thousands upon thousands of matriculants choose a school they find less desirable than other school(s) they were admitted to because of tuition remission, "merit scholarship." Do you control for this in any way?

Thanks!

Mike

[Brian Frye]

Mike,

Thanks for your comment & glad you found the paper interesting! No, we intentionally do not control for tuition remission/"merit scholarships" - and I appreciate your use of scare quotes there! - because we consider price one of the (many) factors that may be salient to matriculating students. The goal of the paper is quite limited - to ask whether & when objective ranking systems align with the revealed preferences of matriculating students. When they don't, it suggests that students are responding to factors not measured by objective ranking systems. We speculate about what some of those factors might be, but there any many different possibilities, surely including price. The point is that some schools seem to deploy those unmeasured factors more effectively than others.

For what it's worth, CJ has a forthcoming paper with some interesting and counterintuitive empirical observations on the effect of "scholarships" on matriculation rates. Watch for it!

BLF

[Al Brophy]

Mike, not sure we should be correcting for the tuition discounts that schools offer to attract students. Isn't this the point of Brian's and Christopher's paper: which schools are recruiting the best students?

Brian, one thing that interests me about your paper is the way you combine the 25th, 50th, and 75th percentiles of both UGPA and LSAT. This is different from how US News figures student quality. I'd be curious how different your ranking is from US News'. Maybe a table would be helpful here. My guess is that these will be pretty similar, but the way schools have adapted to US News' measures (I take it they've not cared much about 25th/75th in recent years) may mean that there are some non-negligible differences. I suspect that you capture some data that schools aren't gaming, or aren't gaming to the same extent they game the median LSAT/UGPA.

[Brian Frye]

Al,

Thanks, that's a great point. I think our ranking system does incorporate marginally more data about student "quality" ("desirability"?) than the US News rankings. In theory, that should help bring its picture of students choices even more clearly into focus than the US News rankings, which at least echo student choices.

Of course, our ranking system is really only a preliminary effort, intended to stimulate conversation of student choices and demands. We hope to do follow up studies in order to tease out additional information.

BLF

[Mike Spivey]

Thanks all, this was very helpful! I am a little slow on the uptake but I get it now. Again, I enjoyed reading and look forward to the next round from Chris.

[Brian Frye]

Mike,

My pleasure! CJ & I would both be delighted to hear your expert insider take on this project & our other projects relating to law school admissions.

BLF

[concerned_citizen]

UT-Austin notwithstanding, the T14 as set out by USNWR otherwise pretty much remains and the top 20 not much disrupted, either (UM and especially BYU being notable new entrants).

It seems your methodology will inherently be tangled up with the USNWR's halo effect on what are viewed as the most elite schools - many more of the students with the very highest credentials tend to go to those schools, in large part because USNWR's ranking told them those were the best schools.

If so, I don't see a way to "fix" that particular issue. But for those students looking primarily toward the other 180 or so schools, your ranking should be another useful bit of information to take into their considerations.

If they don't already, that is, in a more manual sort of way. Many I talk to already use lst.com to sort by GPA and LSAT in any event - sometimes to ensure they're including schools in their application pool where they'll be assured of being above average. :-)

[Brian Frye]

Concerned Citizen,

Thanks for your comment, I largely agree with your assessment. Our working assumption is that the market for the most elite schools is national & the market for the rest of the schools is regional, with other as-yet-unidentified factors coming into play. I think you are right that students are taking the "manual" approach - hence the discrepancies! Our hope is to provide some additional information & encourage discussion of what matters to prospective students & why. While I think there are plenty of things about legal education that need "fixing" - or at the very least, could be improved - our purpose was not to "fix" ranking, but to ask what it is for & how it can provide the most useful information to all of the relevant parties. For example, I think that school that underperform in our rankings might ask why prospective students unexpectedly find them less appealing than their competitors? There are many potential reasons, some of which are probably more easily addressed than others.

BLF