This article (Clauset, Arbesman, and Larremore. “Systematic inequality and hierarchy in faculty hiring networks”) has been making the rounds lately. The article uses a network method to extract prestige rankings from the set of graduate degrees and faculty hires. It shows “that faculty hiring follows a common and steeply hierarchical structure that reflects profound social inequality.”
Blog posts, tweets, and stories about the article (e.g., this one from the Monkey Cage) have mostly picked up on the idea that the fact that prestigious departments generally hire Ph.D.s from other prestigious departments must mean that “academia is not a meritocracy.” While I would certainly not claim that academia is a meritocracy, I don’t think the Clauset et al. paper demonstrates that.
The basic story is that, across three very different disciplines (computer science, history, and business), the prestige of the department issuing the Ph.D. is the strongest predictor of hiring. I don’t think anybody will be surprised at that. But there are (at least) two possible explanations for that observation:
- Prestige of Ph.D. is an — even the — independent cause of prestigious hiring; or
- Prestige of Ph.D. is an important signal of scholarly quality and promise.
Under explanation 1, scholarly quality must be assumed to be entirely exogenous; under explanation 2, scholarly quality is endogenous to the process of selection and education toward the Ph.D. I think explanation 2 at least needs to be treated as a legitimate null hypothesis if the alternate hypothesis — that prestige drives hiring separate from scholarly quality — is going to be portrayed as the main finding. Here’s how the article tries to address that issue:
Strong inequality holds even among the top faculty producers: the top 10 units produce 1.6 to 3.0 times more faculty than the second 10, and 2.3 to 5.6 times more than the third 10. For such differences to reflect purely meritocratic outcomes, that is, utilitarian optimality of total scholarship (13), differences in placement rates must reflect inherent differences in the production of scholarship. Under a meritocracy, the observed placement rates would imply that faculty with doctorates from the top 10 units are inherently two to six times more productive than faculty with doctorates from the third 10 units. The magnitude of these differences makes a pure meritocracy seem implausible, suggesting the influence of nonmeritocratic factors like social status.
Now, I’m not sure what p-value is associated with “seem[ing] implausible,” but be that as it may, there are at least two other issues to be considered here. First, scholarly quality must be equivalent to scholarly quantity; “total scholarship” must be assessed in terms of the number of items published, not the inherent quality, importance, or even influence of those items.
Second, and a lot more important, there’s a lot riding on the word “inherent” in the phrase “inherent differences in the production of scholarship.” I think they mean something like “exogenous” — that is, that the scholars’ productivity is a property of the scholars themselves, so a department’s productivity is simply the aggregation of those individual scholars’ productivity. But there are lots of reasons to think that’s not true. Prestigious doctoral training programs may actually teach their students something, not just about the substance of the field but also about professional behavior and publication. Such programs also practice very selective admissions, which, if effective, would select for Ph.D.s with strong work ethic, scholarly habitus, etc. And prestigious hiring departments may successfully apply resources (lower teaching loads, better graduate students themselves, more institutional research resources, better colleagues, etc.) to enhancing scholars’ “inherent” productivity. In short: better Ph.D. programs may produce better scholars, who may be more productive due to better departments’ more fertile environment. I’m not saying this account is true, but I don’t think it’s effectively rejected by this paper.
The article editorializes:
How many meritorious research careers are derailed by the faculty job market’s preference for prestigious doctorates? Would academia be better off, in terms of collective scholarship, with a narrower gap in placement rates?
These are reasonable questions, but they’re not particularly well informed by this article. For this question to be answered, we’d need to know how broadly distributed equivalently meritorious scholars are across less-prestigious Ph.D. programs. The article assumes they must be, both in terms of Ph.D. admissions selectivity and in terms of learning and professional development accomplished while in those programs. And they don’t ask the obverse question: how many meritorious research careers have been enhanced or fostered by high-quality jobs being awarded mostly to high-quality scholars?