Drivers of inequality

Academic hiring committees play a role:

Robert Oprisko of Butler University found that half of the jobs in university political science programs went to graduates of the top 11 schools. That is to say, if you have a Ph.D. from Harvard, Stanford, Princeton and so on, your odds of getting a job are very good. If you earned your degree from one of the other 100 degree-granting universities, your odds are not. These other 100 schools don’t even want to hire the sort of graduates they themselves produce. They want the elite credential.

That is from David Brooks.


Isn't this a natural outcome of
(1) quality is hard to measure, so relying on the elite school signal is probably as accurate as anything
(2) the people doing the hiring are probably graduates from an elite school
(3) only 10% or so of Ph.D. graduates are going to get a position, so only the elite need be hired, making less known school graduates *really* rare (i.e. weird)
(4) It's much easier to justify to outsiders (administration) why you made the hire when it's based on qualities that they understand
(5) It may be easier to get funding & recognition if the hiree already has some personal connections with members from his Ph.D. cohort already employed as faculty members, etc.
(6) Students and potential students will likely value the program more because of the enhanced signaling of quality by the faculty.

I vote 4. I'm constantly amazed by the extent to which people are motivated by CYA.

Nobody ever gets fired for buying IBM.

I vote 1, with a twist.

The credential will be important in fields of "study" where there is no actual knowledge produced. In sciences, where research actually matters and thus can be evaluated, the perceived quality of the candidate's school will be less important.

In "pure" (as opposed to applied) math and theoretical computer science - my fields - essentially the same phenomenon occurs. These are fields where I would say "actual knowledge" is produced. There may be more exceptions in these fields - the excellent PhD student at a low-ranked university with a *really* amazing result can still get a top-tier job - but those exceptions are still quite rare. This suggests either that the brand name contains more actual signal than one might like to think, and/or that the other factors listed by Tom are significant. I, for one, think that (3) is the most important factor here, mixed in with some (1) and some additional subconscious biases on the behalf of hiring committees.

However, I can also strengthen (2) as follows: students from top tier schools get recommendations from professors at top tier schools. These are likely the same professors that advised a previous round of students who now have positions everywhere and are doing the hiring, so the hiring committee (consciously or not) puts more weight on the recommendations from their former advisors and colleagues at the top tier school.

only 10% or so of Ph.D. graduates are going to get a position, so only the elite need be hired, making less known school graduates *really* rare (i.e. weird)

This arithmetic is critical to understanding if there is a problem. Unless we have some idea of the ratio of openings to graduates there really is no way to tell what's going on.

And this is surprising? I imagine that this is pretty much the same in Econ. What's the breakdown for the GMU Econ department. Certainly on observes that many of the leading economists come from a small pool of schools (Chicago, Harvard, MIT leading the pack).

Economics faculty come from a much larger pool of schools. For instance, GMU has professors with PhDs from Case Western, GMU, Auburn, and the University of Wisconsin at Madison.

Well, first two caveats to your statement: UW Madison is considered a nearly top tier school. Not MIT good, but if you were to construct a list of 10 to 20, it'd likely be there. Secondly, GMU is something of an odd case, in that it has a weird ideological slant to its hiring. This is also related to why they hire their own: they tend to produce exactly that sort of economist, and tend to value these traits way more than other schools do.

Counter-argument to your point: my alma mater, Georgetown, is a pretty mid-tier place. And yet look at the profs, and you see Rochester, Cambridge, Oxford, Princeton, Michigan, Berkeley, Yale, Minnesota, MIT, Columbia, Brown, NYU, Cornell, Chicago, Wisconsin, Harvard, UCLA, Johns Hopkins, Ohio State, Penn State, UCSD, and Stockholm. That wasn't cherry-picking, it was a complete list.

Admittedly, more than 11. But using USW&N, those US schools are all in the top 30, and the international ones are of equivalent quality. Georgetown is tied for 46th/47th. So that means they did not make a SINGLE faculty hire that wasn't a 15-spot jump. Almost all of them are actually 30 spot jumps in ranking. I think you'll see similar trends elsewhere.

No one ever got fired for buying IBM.

Having worked with or listened to some of the persons who graduated from "elite" schools, what you are purchasing, as a research institution, is the network and contacts, and not teaching ability. With so many papers being jointly authored, being networked with other researchers increases your publication rate and the transmission of the writing across the network. In addition, the institutions from which they graduated are also recipients of large federal research grants, so that when you graduate from one of them, you are more likely to be connected to a funding network in the future as you could work with or propose a grant with a person from your alma mater.

So, you get networks of writers and networks of funding when you hire someone from these institutions. (You can visually see some of this if you use network visualization software like Gephi to examine co-authoriship or grant applications.)

What do you get when you hire someone from a school that is not known for research or does not graduate researchers?

How do you create more open networks might be the question to ask if you want to have more variety or tap more talent?

This isn't paradoxical at all. If it is true that half of the jobs go to graduates of the top 11 schools, the only way for a college to boost their job placement rates are to become a top 11 school in the future, and to do that you have to hire professors from the current top 11 schools. This suggests that qualified candidates from all schools are difficult to differentiate, and that, as everyone knows, that signaling is more important in higher ed than knowledge gained per tuition dollar spent.

It is also consistent with there being skill differentials between schools. I don't think it 'suggests' that they're difficult to differentiate. It can be consistent with both signaling and actual skill/training quality differences.

Ah yes, of course, "quality" *has* to differ. When experienced economists can misunderstand each other, perhaps they can misunderstand the quality of a candidate.

I knew I shouldn't have gone to school #12. A mistake I did not make when picking a law school.

I knew there was something about you I liked. :)

Columbia, 84

I wonder what the data looks like for high demand fields like finance or nursing. At my school we tend to go for the best of the lower ranked schools because we just can't compete for the elite candidates on salary and other requirements.

You are on the right track here. PhD in political science is a bit like acting, in terms of the classic superstar effect. Inequality follows. Needless to say, calling this a "driver of inequality" is silly. Most people don't participate in superstar-type labour markets.

I wonder how significant attracting graduates of elite colleges to teach will be in the future as we move to online education and as degree granting by institutions like the University of Wisconsin award bachelors degrees based on tests along, allowing those with job experience, an assortment of past classes and what you learned on a MOOC serve as the basis for your degree.

What may matter as a "teacher" is your acting ability and how well you teach to the test.

What's interesting is that they will take the least promising person from a top 12 school over a more promising person from a lower ranked school. Maybe they are hiring for the name brand or the network.

But then you have a lot of professors who are pompous asses and thing that all their students are morons because the students couldn't or didn't get into the school their professor came from. And those pompous asses make up the hiring committees and tenure committees.

And most of them are liberals, so guess who they hire.

We should only care about this because professors are so liberal, and liberals "care" about inequality (the heart bleeds). Unless you are going to throw this is the face of a liberal professor, it really is of no consequence.

After all, somebody need to fill all the part time adjunct positions at community colleges, not to mention barista at Starbucks.

I imagine a study of what law schools most law professors graduated from would produce approximately the same results--most are from about ten or twelve law schools--and probably a significant majority of those are from the top five or six. This is likely true even for the law schools that are in the bottom half of the rankings by U.S. News.

Nice piece! However, David Brooks starts talking about spatial inequalities. But then he seems to confuse income and spatial inequalities.

1- "Pumping a few dollars into San Joaquin, Calif., where 2.9 percent of the residents have bachelor’s degrees and 20.6 percent have high school degrees, may ease suffering, but it won’t alter the dynamic."

2 - "It’s simply to say that the liberal agenda is not very good at addressing the inequality problem it seeks to solve. The meritocracy is overwhelming the liberal project."

1- “Pumping a few dollars into San Joaquin, Calif., where 2.9 percent of the residents have bachelor’s degrees and 20.6 percent have high school degrees, may ease suffering, but it won’t alter the dynamic.”

What would help drive up wages in San Joaquin, Calif., which is 95.6% Hispanic, is less unskilled immigration.

The point to me is that phd candidates in political science at those other 100 schools are wasting their time. Most of them don't even know that they are already doomed. That isn't a meritocracy; that is a scam.

The real problem seems to me that we're producing way too many Ph.D.s in political science.
If the only jobs for political scientists are as teachers of political science, that all seems rather circular. Shouldn't every Poly Sci professor only graduate one replacement for himself every generation in that case?

But then he might have to actually pick up a nontrivial teaching load or grade a stack of papers on his own.

There is signalling value, but there is also selection bias, which is related to but different from signalling value. Perhaps, the best future faculty candidates are more likely than less promising candidates to be admitted and go to the elite universities for graduate school. Of course, if signalling value exists, it exists precisely due to a widespread belief, rightly or wrongly, that this is true.

I suspect that a high proportion of faculty also have very good high school grades and SAT scores. If no faculty hiring committee ever looked at those criteria, however, then there could not have been any signalling value, and we could not really say that the hiring committees "wanted the elite credential" of good high school grades and SAT scores.

There is a subtle but important distinction between saying that a particular characteristic is a good predictor that someone will get a particular job vs. saying that those with that characteristic have a better chance of getting that job. That distinction seems to be essential to many of our discussions about inequality.

There is no meritocracy. The vast majority of college admit decisions are based upon legacy, diversity, or social profile. His example, the Obamas, are both legacy and diversity admits.

Academic admits are less than 10% of the Ivy League.

This is merely a self-perpetuating oligarchy like that of any other empire, with an admission process that co-opts just enough rising talent to keep them from being too restless.

If it was a meritocracy, it would be full of Asians.

According to the Unz piece, it would be more flyover white considerably less Jewish and about the same Asian.

Does anyone else find it ironic that, while writing this kind of article, David Brooks is teaching a course at YALE of all places? Why isn't Mr. Brooks sharing his brilliant insights at the community college?

I suppose it's a little mean to take Mr. Brooks to task for preaching something he doesn't seem to practice - antielitism. The truth is that many people from elite backgrounds like to rant against elitism, until its time to donate to a university or teach as an adjunct professor.

Lets be honest - elitism is imperfect, awful, and all the rest - but it's a fact of life.

Very, very good post. Excellent

Comments for this post are closed