The composition of ranked programs

By Michel Antoine-Xhignesse

There's been quite a bit of talk on the internet recently about the role of institutional prestige in philosophy, especially in hiring decisions. The Cocoon has led part of that discussion, with Marcus tallying a whole whack of job market stats (see here and here). And every once in a while someone posting in a job market thread will advise job seekers to have a look at the faculty roster wherever they're applying to see whether people with pedigrees similar to their own have managed to snag jobs there. (Brian Leiter used to keep some information of this kind—see here).

A few weeks ago, I decided to start gathering the information for myself. I started looking at the faculty lists of departments ranked by the PGR as well as departments in Canada, and keeping track of which PhD-granting institutions were represented. It didn't take long before some interesting trends started to show up. Since quite a few people seem interested in this kind of information, I've decided to share my findings with you all. I'm not trying to push any particular conclusions here: you can draw your own. My data is here.

Please note that there are mistakes in this data. Originally, I was just doing this to satisfy my own curiosity, so I wasn't too bothered by it. I guess the mistakes matter a bit more now that I'm making it public, but given the trends involved, I don't think they're likely to be too significant.

The first sheet will show you all the data for the I53 (sorted so that the stronger placers come first), plus 20 other ranked/evaluated departments that I haven't made it to quite yet. The second sheet shows you the same info for Canadian programs. The third sheet tallies up the sums.

Finally, if you turn to the fourth sheet in the Excel document, you'll see that for the I53, I compared their Leiter and placement ranks. The first set of columns shows just that. The second set shows the same information, but sorted by the differential between their ranks, so that the departments that do best are at the top. The last set of columns is just the Gourmet ordering vs. this new ordering, to give you a sense of how things change. It's not a particularly informative way of looking at things, but I was curious, so I did it.

A note on methodology: I tried to count only T/TT faculty; no post-docs, emeriti, VAPs, adjuncts, etc. Doubtless, a few slipped through the cracks. I'm also much less familiar with job titles in the UK, so I'm sure I made more mistakes with respect to those departments. Finally, I have not updated departmental numbers to reflect this year's hires and retirements.

So far, I've counted the PGR International top-53 (I53) and 32 Canadian departments (every PhD- and MA-granting institution in Canada). That's a total of 82 departments, with 1700 faculty members. I do plan to eventually finish counting the rest of the departments ranked by the PGR, as well as all of those recommended by SPEP. That should be about 137 departments total, plus as many American MA programs as I can stomach. That, however, is work for the weeks and months to come. In the meantime, here's what I've got.

I53

In the I53, five departments have 50 or more placements (out of 1212 total, across 129 departments). They are:

• Oxford - 116
• Princeton – 96
• Harvard – 84
• Pittsburgh - 57
• MIT – 55

These five departments account for nearly 34% of all I53 placements. 101 of these 129 departments have fewer than ten I53 placements. Collectively, those 101 departments account for almost 20% of I53 placements. The other 28 account for 77% of them (about 3% are unknown).

Canada

Interestingly, Canada's state is very similar to that of the I53. 122 departments are represented, with a total of 488 faculty. The top five  placers (well, six: there's a tie for fifth) are a little different, though:

• Toronto - 63
• Oxford - 27
• Western Ontario - 27
• Pittsburgh – 21
• Princeton - 15
• Sorbonne - 15

Together, these 6 account for about 34% of placements. Only 7 departments (add Université de Montréal to the above) have 10 or more placements, and these 7 account for 37% of the placements. The other 115 departments make up 60% of placements, with about 3% unknown.

I53+Canada

If you add together the I53 and Canada, then 35 out of 175 departments (20%) have 10 or more placements. The top ten are as follows:

• Oxford - 143
• Princeton - 111
• Harvard - 91
• Toronto - 89
• Pittsburgh - 78
• MIT - 62
• Berkeley - 52
• Cambridge - 47
• Stanford - 43
• UCLA – 43

So if we add Canada to the I53, Toronto and UCLA find their way onto the map of the top ten departments, and Rutgers and Yale drop out. Together, those ten tippy-top departments account for 45% of all placements. 67 (38%) out of all 175 departments represented have just one placement in those 82 PhD-granting programs; that means that 38% of departments have not quite 4% of all placements.

Two departments (2%) have 100 or more placements, and these two account for nearly 15% of all placements between them. In fact, that 2% has as many placements as the bottom 124 placers together (these placed between 1 and 6 people). Oxford (the top 1%) has as many placements as the bottom 101 departments combined.

Discussion 

The prospects look kind of grim for anyone not coming from the most elite departments where that very roughly seems to correspond to the I10 (NYU and USC aside; USC's showing is dismal compared to the other I10 programs). That does seem to lend some weight to the thought that the discipline's research agenda is heavily influenced by the tippy-top departments, even if only because they seem to staff the bulk of the other research-oriented departments. The one major exception (and surprise) that I can see is Western, which doesn't acquit itself at all badly in the I53 and does fantastically well in Canada.

That's enough blather from me. What do you all think?

Comments

[Christopher Stephens]

Hi Michel,
Thanks for this. I have a question about your conclusion that "the prospects look grim for anyone not coming from most elite departments..." While I don't doubt that there is some truth in this, I'm a bit puzzled about the extent to which your data supports this conclusion.

If I understand your data correctly, you don't make any effort to control for the size of the graduate programs. So while it may be true that the probability of having gone to Prestigious University X is high given than one has a job, don't applicants really want to know the reverse probability: what is the chance that I'll get a job given that I've gone to University X?

So small programs that place a high percentage won't do well on this metric, and large programs that don't place a very high percentage can still do well by your measures. Is that right? or have I misunderstood the data here?

[Clayton]

Hi Michael,

There's a lot here to think about. Thanks for the post. One small thing about USC worth noting is that their department has climbed the ranks very quickly over the past decade. (If I recall correctly, they were barely in the top 50 in the first Leiter report. (I might be mistaken about that.)) I think it's worth bearing that in mind when thinking about the correlation between appointment and ranking. Lots of highly ranked schools have had that status for longer.

Also, in terms of UK programs, Lecturer is somewhere in between Assistant and Associate Professor. (You'll have something akin to tenure once you pass an initial probationary period and passing this is nothing like getting through the tenure review process.) Senior Lecturer or Reader is the next stage up. Professor comes next and last.

[Michel X.]

Christopher: You're right, I haven't controlled at all for the number of graduates per program: I've focused purely on the brute number of graduates employed at various departments. It would be great information to have, if it was widely and easily available, but collecting and sorting it is more work than I'm up for myself. What we *can* still do is eyeball things, and I think eyeballing bears out the grimness I mention: the best placers are pretty uniformly tippy-top departments with significant halos. They are also (as you indicate) among the larger and more established departments.

Clayton: Thanks for the clarification. I'd figured it was more or less like that, but the UK numbers probably still warrant double-checking.

That's an interesting point about USC--to be honest, I'd forgotten about that entirely. I think it points up a not insignificant trend, too: the departments that do best by this measure are departments that have been around for a long time, have enjoyed tippy-top status for a very long time (perhaps explaining why USC, NYU, Rutgers and so on don't seem to do as well), and have produced a very large number of graduates.

One of the things to remember is that because I only counted T/TT faculty, a large number of those faculty members got their PhDs quite some time ago, when the discipline itself had a very different composition. So it's not at all surprising to see Oxbridge and the Ivies doing so well. That makes the success of some of the newer players (e.g. Western) that much more surprising. (Incidentally, I'm a Michel, not a Michael!)

[Clayton]

Sorry Michel!

I try not to wear glasses when on the computer, but now I know that I'll need to rethink my font sizes.