I recently related some figures, originally collected and published by the American Association of Law Schools (AALS), indicating that women and minorities do better on average than men and non-minorities at landing academic jobs through the AALS's Faculty Recruitment services. Some commentators explained that result by supposing that women and minority job candidates might on average tout better credentials than their counterparts. I promised to look into the matter. The AALS has asked me to refrain from conducting the necessary research, however. Allow me to explain.
The AALS collects various types of information about all candidates seeking law teaching jobs through its recruitment services. It then makes that data available in an online searchable database—the Faculty Appointments Register (FAR)—to law school deans and the members of law school appointments committees. I am one of the latter. I thus thought that I might run some searches on the FAR database to figure out whether candidates' credentials correlate with their sex or minority status.
Specifically, I proposed to run searches on the FAR database allowing me to fill the 24 spaces in a table four columns wide by six rows tall. The column headings would refer to types of candidates, and would comprise these categories: Female; Minority; Female Minority; and All. The row headings would refer to candidates' searchable professional attributes, and would comprise these categories: Clerkship, Major Published Writing; Law School Teaching; Top 10% of Class; JD from Top 17 "Feeder" School (as defined by Brian Leiter's research); and All (a category referring to candidates with all of the aforementioned attributes).
Knowing the critical bent of my readers, I anticipated a number of questions. Why those credentials? (Because they seem like fairly widely recognized ones and because the FAR database's interface allows searches for them.) Why not run searches for male or non-minority candidates? (Because the FAR database does not allow searches of that sort; it allows searches only for female, minority, female minority, or all candidates.) Why think that running such a search on the extant pool of candidates would teach us anything about the make-up of past candidate pools? (Because it seems reasonable to assume that the present pool does not radically differ from prior ones and, at any rate, we might at least get data allowing us to eventually learn something about the role that sex, race, and academic credentials played in the present pool's search for law teaching jobs.) Subsequent events have rendered those questions moot, however.
Recognizing that the AALS offers the FAR database for recruitment purposes, I thought that I ought to seek its permission. I dutifully contacted the AALS and explained my research proposal. After several polite email exchanges, I found the AALS unwilling to approve of my plans. Why? For pretty good reasons, I must admit.
It seems that a similar request earlier this year led the AALS to realize that it had never actually won the consent of past candidates to do the sort of statistical research that it has long done on the FAR database. The AALS now apparently plans to ask for such consent in the future, thus clearing the way for it and at least some third parties to conduct statistical research that does not risk violating candidates' confidentiality. It remains for the AALS Executive Committee to decide, however, about the propriety of allowing third parties such as myself to conduct research on the contents of FAR databases from this and prior years' hiring seasons.
Conceivably, given that candidates to-date have not been notified about the possibility of such research, the Executive Committee might decide to withdraw even the research that the AALS has conducted and published. Or it might simply deny all third parties access to the data. Given that the AALS has the stated goal of "the improvement of the legal profession through legal education," however, I think it more likely that the Executive Committee will open the FAR database to certain types of statistical research—namely, research that stands to both preserve candidates' privacy and reveal how legal academia works.
[Crossposted to MoneyLaw.]