The Signaling Problem of Law School Recommendation Letters

As a member of a law school admission committee, Mark Graber says he would like to read recommendation letters specifically tailored to his school. In the comments, Kevin Heller demurs:

...I wonder about your statement that professors should "write letters that are specifically geared to the law schools to which those students have applied," not "cliches promising that your students have the ability to succeed in some random law school." Is that really a fair expectation? Most students apply to more than two dozen law schools – from reaches to safeties. Do you really believe that professors should tailor their recommendation letters to each school? Even if they have the time – which they don't – do they have the knowledge? ... I could spend some time on the internet finding out more, but again – my students will be applying to 23 other law schools, so that's not exactly a practical possibility.

In an update to his post, Graber agrees with Heller, but also notes that most letters he reads “could do a lot better job indicating approximately where a student would be appropriately placed.”

This is a difficult issue that has become more difficult because of the standardization of the application process. In the old days, a professor had to send a letter to each law school his student was applying to. This was inconvenient, but it had the upside of allowing the professor to tweak the letters to fit the targeted schools. Now, the Law School Admission Council has made it possible to write just one letter of recommendation that will be forwarded to as many law schools as the student chooses. This is far more convenient for both professor and student, but it means the professor must write a single letter that will suffice for every law school.

The need to write a single letter has upset the old signaling equilibrium. Under the old equilibrium, any negative comment could sink an application. Likewise for a failure to say wonderful things, also known as “damning with faint praise.” Any admission committee, whether at Yale or Podunk State, would recognize these sleights. So if you had a student who was not first-tier material, but who could be expected to perform quite well at a lower-tier school, you could send a tepid letter to the former and a glowing letter to the latter.

The standardized application process has foreclosed that option. Any negative comment or failure-to-praise intended for committees at upper-tier schools will be read by committees at lower-tier schools as well. This means letter writers have to find a way to convey the appropriate placement for their student in a single letter. Ideally, selection committees should recalibrate their evaluation of recommendation letters in light of this fact. Committees at schools below the top tier should be prepared to accept lots of students with less-than-glowing letters – and the lower the school’s rank, the more negative comments they should be willing to tolerate.

But I wonder whether we’ve reached the new equilibrium yet. I worry that many admission committees consider any recommendation with negative comments to be a “bad recommendation,” regardless of the school’s quality. Since I teach a course on Law & Economics, I usually have at least one or two students each year asking me for law school recommendations. Consequently, I’m very curious to know how actual admission committees approach this issue, and whether their approach has changed as a result of the standardized application process.