Penn Political Review

Grad Schools Probably Shouldn’t Have a Lottery

Update 9:30AM: I forgot to link to this post from earlier in reference to professor Jackson. Mistake corrected. Also, to be clear, the kind of market I refer to at the bottom is a labor market.

Having just started work on an (awesome) research project on the history of College Houses at Penn, my mind seems increasingly drawn to higher ed politics when catching up on my RSS feeds. If such a thing were possible. Don’t worry, this will come back to national politics at the end! Albeit very tenuously.

The residential life of undergraduates – at Penn, at any rate – occupied a great deal of university institutional planning in the late 20th century (from about 1970). Penn, being a research institution, is the kind of place that might be expected not to care about undergraduates, so its undergraduate development focus probably indicates a larger trend. “Intellectual labor should be seen to merge seamlessly with intellectual life,” proclaimed a commission on undergraduate education in 1989.

Yet the life of the mind has been under some pressure lately, not least from the appearance of William Pannapacker’s infamous Chronicle of Higher Education piece, “The Big Lie.” Accordingly, the higher ed focus has shifted from the mental life of the undergraduate to the career prospects of the PhD.

Among the concerns about PhD programs is their competitiveness – how can a department possibly choose two from among hundreds of applicants? Pannapacker’s answer would be to discourage all but the topflight from applying in the first place. Another popular answer is to institute a lottery. Penn’s own John L. Jackson, Jr (whom you may remember from this panel), is the latest person on the internet to float such an idea. Ironically enough, contradicting Pannapacker, he says there is an increase in the graduate applicant pool as “further credentialization becomes the reasonable response to a decimated job market.”

I’m OK with a lottery in principle. But once you poke at the idea, you find lots of caveats relating specifically to graduate schools in the humanities. First, while applying to grad school is not the same as applying to college, the pressure to spread apps around still holds. I don’t see a responsible way to institute a lottery system of any kind without a high degree of coordination among different universities. If every school has its own lottery, eventually every student will apply to every school. That simple. So the first principle has to be that schools coordinate their lotteries. In fact, every subject would have to coordinate a lottery among its departments in various schools. Because Wharton and Annenberg will be damned if they’re going to let a lesser Ivy League graduate business or communication program poach from their applicant pool just because both universities have good law schools (for example).

So what is the effect on the top tier? In undergraduate education, one can easily get a well-respected degree from 30 schools. If one does not intend to go into a PhD program, 50 schools. But as Brian Leiter notes, you need to go to a very strong PhD program (top ten or fifteen) if you hope to get a tenure-track position afterward. Would a lottery system take place only in these top tiers? If so, how sticky would that make departments’ reputations? How difficult would it be to jump up (or down)?

Maybe, let’s say, you don’t care about the top tier. Maybe we should resist the kind of reputational gamesmanship that prevails in the top echelons of academia, and just admit that at the far end of the bell curve things look kind of the same. At any rate, how good your department is at what you do may mean more than how good your department is generally. But this brings up a further point: departments have to match the research interests of their applicants with the research interests of their faculty. If you’re into the history of Renaissance Italy, and the big-deal Renaissance Italy professor at Top-Tier History U. just retired, nothing will make that program a good match for you. Ditto if the professor already has a full advising load. Oh, and this same concern means that a top-tier department in one subspecialty may not be top-tier in another.

So: in order to effectively run a PhD program lottery, different departments would have to coordinate lotteries; different tiers of quality would have to be recognized, perhaps down to subspecialties; and students would have to be matched pretty well with the advising capacity of a department. I don’t see how this can happen without a ranking system of some sort (which is what we wanted to avoid). The closest thing we have to a grad school lottery right now, that I know of, is the matching system of MDs to residency programs. That lottery does, in fact, involve a ranking of schools by applicants, and a ranking of applicants by schools. But here we return to the original problem, which is that departments have a hard time ranking applicants. Perhaps students could rank departments, and departments could put students into tiers? Perhaps the program could involve rounds of lotteries at different tiers, looking a little like this scheme?

When a centrally-administered matching system gets this complicated, it may be best just to leave things to an individually-negotiated market. I know markets are unfashionable these days. But I thought I’d throw that out there.

Like this Article? Share it!

About The Author

John Gee is a senior in the College majoring in Intellectual History. In addition to PPR, he is a member of the Residential Advisory Board and the Philomathean Society.


  1. Ett May 25, 2010 at 5:37 pm

    Markets are unfashionable?

  2. John Gee May 25, 2010 at 8:29 pm

    Well, it's probably better to put it the other way, and say that active regulation is fashionable. But, actually, yes: lots of people on the left want strong, central, active regulation; and lots of people on the corporate right want weak or no regulation (of whatever they happen to be talking about). So that leaves actual markets in a fairly unfashionable position.