Now that there's a new SALT Salary Survey out, I want to take a look at the relationship between law faculty salary data and U.S. News ranks. By way of background, SALT (the Society of American Law Teachers) requested information on salaries and on summer research grants from 200 law schools in the United states and Puerto Rico. The survey was conducted by email. Non-responding schools were followed up with another email and if necessary by telephone. Sixty-six schools responded -- a response of 33% . Many thanks, by the way, to Eric Janus, President and Dean of the William Mitchell College of Law, who conducted the survey.
In my recent paper on the relationships between U.S. News law school rankings and racial diversity (and several other variables), I divided the ABA-approved law schools into four groups to obtain a finer set of categories than the current two-tier U.S. News system. I’m using that four-way group again in this post, which deals with some results of the 2011-2012 salary survey published by SALT this month and with the 2013 U.S. News ranks.
Here is the classification of the 195 ABA-approved law schools:
Group. N Ranks Schools
1 50 1 - 49 Yale thru American, Pepperdine (tie)
2 50 51 - 99 Baylor thru Rutgers-Camden, Tulsa
3 46 101 - 145 St. Louis + 4 other schools thru Montana, North Dakota
4 49 Rank Not Published
Total 195
Here are the responses by tier:
Response
Tier Total Yes No % response
1 50 8 41 16.00
2 50 21 29 42.00
3 46 19 27 41.30
4 49 18 33 36.73
The response rate differs across tiers. (Chi-square = 9.41, df = 3, p = .02). Schools in Tier 1 were less likely than other schools to respond to the survey. None of the 28 highest ranked schools in the current U.S. News rankings responded. The highest ranked school that responded is the University of Iowa (rank 29), followed by North Carolina (38), Maryland (39), Ohio State (39), Arizona (43), Colorado (44), and UC-Hastings (44), which – perhaps incidentally – are all public schools.
Peer assessment is an important component of U.S. News ranks, so we would expect the above schools to be included among responding schools ordered by peer assessment (a 5-point scale), and most are. The responding schools with the highest peer scores are North Carolina (3.6), Iowa (3.4), Ohio State (3.3), UC-Hastings (3.3), Arizona (3.2), Florida (3.2), and Colorado (3.1).
Assessment of schools by lawyers and judges (another 5-point scale) is also a component of U.S. News ranks. The 7 responding schools with the highest cores on this dimension are North Carolina (4.0), UC-Hastings (3.9), Iowa (3.8), Arizona (3.7), Ohio State (3.6), Florida (3.6), and Colorado (3.4).
Now, as to the salary data ...
Salary of Tenured Professors by Tier
Tier N M SD
1 8 169,475 6,204
2 21 147,500 18,558
3 19 139,078 16,282
4 18 142,441 14,194
Salaries differ by tier: F(3, 62) = 7.52, p < .001
According to the Tukey multiple comparison procedure, Tier 1 salaries differ significantly from those in the other tiers, and salaries in the other tiers do not differ significantly.
Summer Research Stipend by Tier
Tier N M SD
1 8 15,473 5,026
2 21 11,651 2,789
3 19 10,476 3,142
4 16 10,133 3,056
Summer research stipends differ by tier: F(3, 60) = 5.41, p = .002
As with salaries, the Tukey multiple comparison procedure found that Tier 1 summer stipends differ significantly from those in the other tiers and that stipends in the other tiers do not differ significantly.
Next up ... some correlations between peer assessment scores and median salaries at the reporting schools.
Interesting stuff. Is it worth trying to account for variations in cost of living?
Posted by: Jonathan H. Adler | May 12, 2012 at 07:51 PM
I think that's a very good idea, Jonathan. I hope to work this data up a little more over the next week or so. As I began plotting some of this I noticed that some urban schools have high salaries relative to their U.S. News peer assessment scores. (Drexel and Touro come to mind right now.)
As you may have seen, the response rate isn't the best, so I'm also hoping to work through some of the publicly available data on individual faculty members, which will allow for a lot more precise conclusions. That's a longer-term project, obviously.
Posted by: Alfred Brophy | May 12, 2012 at 09:16 PM