Wednesday, March 19, 2014

German Ph.D. programs: their products and their wasteproducts

Let’s imagine that you have decided, contrary to the advice of every responsible adult in your vicinity, that you want to pursue a Ph.D. in German. Of course you have scanned the list of the programs whose Ph.D.s have done the least badly recently in finding tenure-track jobs. Of course you know not to go to grad school without a decent financial aid package, and you’re prepared to walk away if you don’t get funding. But you also want to know which programs do a good job of turning grad students into Ph.D.s rather than washouts. What should you be looking for?

Imagine a Ph.D. program (Able University) with ten students. All of them eventually finish their Ph.D.s, taking five years on average. Consequently, the program graduates two Ph.D.s in an average year.

Now imagine a second program with ten students (Baker University). All of its students eventually finish, but it takes them twice as long—ten years. So there is only one Ph.D. graduate on average each year. You don’t know why it takes the students so long, but ten years seems like a long time to finish a Ph.D. If you could choose between them, Able University seems like a better option. If you finish your Ph.D. in average time at Able and get the TT job you’re convinced is waiting for you, you’ll be preparing to go up for tenure when the average member of the same Baker cohort is just finishing a Ph.D. Attending Able will add five years to your career and mean a half-decade less spent living on grad school income (*cough* food stamps *cough*).

Now imagine a third program with ten students (Cutthroat University). Cutthroat moves its students through the program with all due haste, but a large fraction fail out of their quals (you can’t translate humanistic Latin verse without a dictionary? too bad, you fail), and another big chunk reach ABD status but never finish their dissertations (this chapter is terrible, but I can’t quite put my finger on why; just rewrite it). Of the doctoral students that start, half finish in five years, but the other half never finish. Consequently, Cutthroat also graduates only one Ph.D. per year on average. A fifty percent chance of finishing grad school dismays you. All else being equal, attending Able seems like a better choice than Cutthroat U.

Now it’s not a bad thing at all if students choose to drop out of a grad program or abandon a dissertation. In even the best program, some students will discover that their talents or interests lie elsewhere, or the long road to the Ph.D. just isn’t worth it. But in general, prospective students will want to steer away from programs whose students are so unhappy that they are bailing out in high numbers, or whose qualifying exam requirements are so unrealistic that many students are failing them, or whose faculty mentoring is so unhelpful that few students are finishing their dissertations.

In the real world, how do you tell which programs are more like Able? You would need to know the enrollment of every Ph.D. program in North America and how many Ph.D.s they are graduating each year. If you knew that information, you could divide the average enrollment by the average number of Ph.D.s granted and arrive at a figure that is close to a reasonable number of years for someone to spend earning a Ph.D.: five, in the case of Able, compared to ten for Baker and Cutthroat. You won’t be able to tell the difference between Baker and Cutthroat—that is, you won’t be able to say if a program’s students are taking a very long time to finish, or if they are dropping out—but you’ll just try to avoid them both anyway. In general, the lower the number, the more efficient a program is at turning grad students into Ph.D.s, and that’s what you want. If only someone published that information somewhere…

Here is where the annual “Personalia” feature of Monatshefte comes to the rescue. It does in fact publish grad program enrollments and the number of degrees granted. For master’s degree enrollments and degrees awarded, the results are unsurprising, with the number of degrees generally falling close to half the number of students enrolled, about what you would expect from two-year M.A. programs.

For Ph.D. programs, the results are more revealing. If you compile all the numbers for the years 2007-8 through 20012-13, you could rank all the doctoral programs from most to least efficient. And this is what you would find:

Ranking of Ph.D. programs in German according to Ph.D.s earned by average enrollment, 2007-8 to 20012-13 (rank, average doctoral enrollment, total Ph.D.s awarded, average Ph.D.s awarded per year, and average enrollment divided by average number of Ph.D.s awarded)

Note that I’ve used my own counts of Ph.D.s, which I believe are more accurate than those in “Personalia” (and since they’re higher, my numbers cast the grad programs in a better light than they would otherwise appear).

There is a lot of interesting information in this table. The average program has an enrollment of 18 doctoral students, while Wisconsin, Berkeley, and Washington University in St. Louis are twice that size or larger. The average program graduates 2.3 Ph.D.s per year, or 85.7 among all programs. The three schools just named produce the most Ph.D.s, followed by Texas, Princeton, and Harvard at three Ph.D.s per year.

But none of the programs mentioned so far fare well in terms of how efficiently they move students towards completion. The top program in those terms is Cincinnati, whose graduates have also done relatively well at finding TT jobs through competitive searches. The example of Cincinnati shows that grad programs can prepare doctoral students for academic employment at a range of schools without requiring many years in a grad program or a string of failed comps and unfinished dissertations. (I’m not associated in any way with Cincinnati, but now I wish I were.)

Florida, in second place, is the smallest program included in the ranking, and with less than five students on average it was almost too small to include. But its size fits an unmistakable pattern: Every single one of the ten most efficient programs are smaller than average, with an average of just 12.5 doctoral students as a group. Small size is not necessarily a sign of efficiency, however; the lowest rankings, places 31-38, are also smaller programs, with an average of 14.1 students. The second tier (places 11-17) also tend to be smaller, with 13.9 students on average, while the largest programs, with 24.6 students on average, are found in places 18-30. Smaller enrollments might indicate more personalized attention from faculty, or it might be a sign of institutional neglect.

* * *

Dear DGS’s and department heads: If you think my figures are in error, then please submit accurate “Personalia” questionnaires, because the figures here are based on what you yourselves reported, and improved in your favor by my own research.

Above all, the next time a discussion of “How to Improve the Ph.D. in German for the Twenty-First Century” comes up at GSA, please take the microphone away from people whose programs massively overproduce Ph.D.s, have shamefully low TT placement rates, or take the better part of a decade to create grad school failures instead of turning grad students into Ph.D. holders. I’m tired of hearing people who hide their failure behind their prestige telling the rest of us how to be just like them. Let’s hear instead what the folks at schools like Cincinnati have to say about doctoral education in German and how to prepare graduates for today’s academic careers.


  1. Great job ANS. I have been looking for transparent doctoral statistics like this for years. Useful for incoming students. I'm curious why you prefer Personalia for the data instead of IPEDS when every university is mandated to report all the doctoral graduates in IPEDS.

  2. Another great thing about the data is that it gives lie to the notion
    that all elite doctoral programs (Ivy + Berkeley + Michigan) are safe.
    Some of the elite programs have dismal graduation rates and job
    placement statistics too. Who knew about Cincinnati?

  3. I haven't used IPEDS because it doesn't provide discipline-specific data (unless I'm missing something really basic). Since I'm specifically looking at German as a discipline, "Personalia" is as good as it gets, at least for some questions.

    Some people treat admission to an "elite" grad program like a good luck charm that will spare them from the fate that is in store for other, less qualified applicants. It simply isn't so, but people cling to that myth anyway.