The Gladwell rankings, which are based on tuition as well as other factors, highlight a matter that I think people haven’t much discussed: the two functions of law school rankings.
Law school rankings are useful to would-be students, who want a rough sense of which school is the best bet for them. But they are also useful to employers, who want a rough sense of how they should compare applicants from different schools. For instance, is someone who’s in the 75th percentile at USC roughly comparable (considering for a moment only school performance) to someone who’s in the 75th percentile at Vanderbilt? Employers who don’t know the schools well will likely use the school rankings to determine that; since USC and Vanderbilt are ranked very close together, they’ll probably treat the two students’ law school performance as roughly comparable.
Some factors matter a good deal for both reasons, especially since employers’ evaluation of the school should generally affect students’ evaluation of the school; as a student, you want a school that will produce a credential that employers will value highly. On the other hand, some factors matter only to students; tuition is one such. As an employer, I will generally not care much about how much an application spent for an education. Instead, I’ll want to know how good an education he likely got, plus also how competitive the school was (since that is relevant to evaluating the student’s grades or class rank). It wouldn’t be reasonable for me to evaluate a 75th percentile student at BYU the same as a 75th percentile student at Harvard (the Gladwell rankings rank BYU as #2 and Harvard as #3); the second student had to compete against a pool that was probably considerably stronger intellectually, so his 75th percentile performance should be more impressive to me as an employer.
Of course, in some indirect way the cost of an education might be relevant to employers, in either direction. Perhaps one could infer that a very expensive education likely in some measure reflects a higher-quality school, since it suggests that the market is willing to pay a lot for that education. Or one could infer that a less expensive education will likely mean someone who is cost-conscious, which is often a good skill for a lawyer. Or one could infer that a very expensive education means the student likely has a lot of debt, and is thus likely to especially value a high-paying job at this firm, and to stay at the firm for a long time. But all these connections are, I think, pretty indirect and highly contingent on a lot of variables (for instance, people who go to expensive schools might have a lot of debt, or only a little, if they or their spouses or their parents paid for the education out of pocket) — even more indirect and contingent than many other factors, such as the school’s average grades, average LSATs, and so on. So on balance I expect that an employer should care fairly little about education cost in evaluating an applicant’s school.
This suggests that it might be helpful to have two different ranking systems. One, for students, would try to somehow measure return on investment, likely focusing mostly on financial factors — tuition, average starting salary of graduates, some honest measure of employment shortly after graduation, perhaps some longer term measures of salary some years after graduation, and so on — though perhaps also adding some other quality of education factors (to the extent that one thinks they aren’t adequately captured in return-on-investment). Another, for employers, would try to somehow measure the quality of the education that the students have received, and the inherent intellectual quality of the student body as a whole, so that one can compare performance of applicants from different schools.
Right now, the U.S. News rankings work in some measure, albeit highly imperfectly, for both purposes. This is because, as Ilya points out, students can easily determine the tuition at various schools, and can combine that information with the U.S. News rankings; employers can exclude the tuition information, and rely entirely on the U.S. News rankings. But it might be better to have two different ranking systems (even setting aside the question of how each ranking system could be better than the U.S. News system), precisely because the needs of the ranking system consumers — students and employers — are different.
A few specific caveats:
1. I realize that each element of the ranking system can only be a rough proxy for what employers or students care about. But in a world of limited information, we have to rely on such proxies. No student can just try out twenty schools for a semester each, and see which one gives the best results for his particular learning style; no employer can hire twenty applicants for one job for three months, and just keep the best one. That’s why employers look at grades and the school from which a student has graduated — and, in the process, consider (directly or indirectly) the school’s ranking.
2. I also have lots of quarrels with the existing U.S. News rankings; I’m not trying to praise those rankings as such, but just suggest that a ranking system that heavily weighs tuition costs will, all else being equal, be less useful to employers than the U.S. News system.
3. It may be better to have more different ranking systems, for instance a comprehensive system focused on students who want to go into criminal law, one for students who want to go into non-criminal-law public interest law fields, and so on. But the problem is that many students don’t really know where they’ll end up, and having more ranking systems may thus increase confusion and information costs for the students more than they decrease them. The advantage of having a system for prospective students and one for employers is that people know quite confidently whether they’re prospective students or employers.
4. I use “Gladwell rankings” here just as shorthand for any ranking system that heavily relies on factors such as tuition; I realize others could come up with other such systems, and that Gladwell himself might have been criticizing rankings generally more than proposing a particular alternative.
5. I also assume here that law school grades (perhaps combined with the average incoming test scores of the school) predict, to some material degree, a graduate’s likely abilities as a lawyer. If they don’t, then reasonable employers should start looking to radically different predictors.