A few things about the academic job market have caught my eye recently, but don’t really add up to a big coherent argument. I’ll note them here, though, to marginally increase the chance that I’ll be able to find them later.
— First, this piece at the Guardian got a lot of play, thanks in part to the dramatic headline Science careers: doomed at the outset but even more thanks to the subhead “Has it become harder for graduate students to thrive, and are our best potential scientists giving up on academia?” Most of the people I saw re-sharing it used basically just that last clause, often stripping out the punctuation to make it a statement: Our best potential scientists are leaving academia.
Only, it doesn’t really show that. That’s not to say it’s not a major indictment of the bad state of graduate education, just that it doesn’t really connect the dots to show that “our best potential scientists” are leaving. There’s an argument to be made there, to be sure– you might reasonably believe that the best potential scientists are also the students most likely to have other options, and thus the most likely to get out while the getting is good. That argument remains largely implicit, though. And it bugs me to see it asserted as fact.
You could, with only slightly less plausibility, make a counter-argument that “our best potential scientists” are those with such focus and drive that they will pursue science no matter what, even in the face of all sorts of unreasonable hardship. In fact, that’s the standard argument in favor of the hazing-like structure of graduate education in a lot of places– we need to weed out the unfit, to select the best.
If you want to make a definitive that “our best potential scientists” are leaving, I’d like to see some data showing that the “best” students are disproportionately discouraged. I’m not sure how you’d really generate that, though, given that there isn’t a generally agreed-upon measure of the quality of scientists, let alone the potential quality of students.
My own guess, if you put a gun to my head and made me stake out a position, would be that the discouragement process is basically random. At least in terms of “quality” and “potential”– it’s probably selecting for something, but I suspect that something is pretty much orthogonal to actual scientific ability. that’s just a guess, though, and I might add that if you’re going to go around pointing guns at people over stupid shit like this, you need to re-think your life choices.
— Somebody on Twitter– I’m pretty sure it was Ben Lillie, but can’t confirm because Twitter– linked to this PDF report on underrepresented minorities in STEM fields a week or so ago, and I’ve had it open in a tab for a while. This is a little old– I think this version is from 2010– which you can tell by the fact that it doesn’t use “STEM” over and over and over. Ah, those idyllic days…
Like all responsible investigations in this area, this is kind of a Rorschach blot of social science, in that you can find more or less whatever you want. If you’re a glass-half-empty sort, you can point to the summary table showing that, in physics, 5.6% of Ph.D.’s granted in the ten years leading up to 2005 were earned by students from underrepresented groups, while only 2.5% of faculty in 2007 were from those groups. There’s also a small decrease in diversity from 2002 to 2007, with the total fraction of faculty from underrepresented groups falling from 2.6% to 2.5%, and the assistant professor fraction dropping from 5.2% to 4.4%.
On the other hand, if you’re a glass-half-full sort, you can reasonably point out that the discrepancy is less bleak when you look at more direct comparisons. After all, looking at percentages over all faculty ranks sort of mushes together the stratigraphy inherent in academic hiring– assistant professors are mostly drawn from recent Ph.D.’s, while full professors can date back to the disco era, when Ph.D. cohorts were much less diverse than they are now.
If you compare the numbers for recent Ph.D.’s and assistant professors (the larval stage of tenured faculty), the gap is much, much smaller– 5.6% of Ph.D. recipients are from underrepresented groups, and 4.4% of assistant professors. Broken out by subgroup, you see a similar picture– 2% of Physics Ph.D. recipients are black, compared to 1.2% of assistant professors; 2.9% of Ph.D. recipients are Hispanic, compared to 3.3% of assistant professors.
They also have gender statistics, which are much the same– women are 14.3% of Ph.D. recipients up to 2005, and just 9.1% of faculty. But women are 16.8% of assistant professors, slightly better than their representation in the graduating cohort.
So, you know, half-full, half-empty. Whatever picture you want to present, you can pull numbers to bolster your argument from this report.
— That by itself wouldn’t really justify a post, but the interesting wrinkle in this report, to me, was that it actually tries to look at the question of prestige. The argument about comparing Ph.D. demographics to assistant professors is not new, and the usual response is that yes, women are hired into new faculty positions at a slightly higher rate than their representation in the Ph.D. cohort, but they mostly get lower-prestige jobs.
This report gets at that a tiny bit by looking at the top 100 departments in each field (by some ranking), and breaking out the statistics for the top 50 and numbers 51-100. Interestingly, this shows the opposite effect usually claimed– the overall comparison is 14.3% of women in the Ph.D. class to 16.8% of assistant professors, but for the top 50 it’s 14.3% to 17.5%, and the bottom 50 it’s 14.3% to 15.6%. The fraction of women in new faculty ranks is actually slightly higher at the more prestigious institutions.
The difference is more striking for black Ph.D.’s and faculty. Overall, the split from Ph.D. to associate professor is 2.0% to 1.2%, but for the top 50 it’s 2.0% to 1.6% and for numbers 51-100 it’s 2.0% to 0.5%. Again, the numbers are better at the more prestigious institutions.
Again, though, there’s something for everyone, because if you move up a rank, the fractions reverse– women are 12.6% of associate professors at Top 50 institutions, and 14.3% at numbers 51-100 (this probably ought to be compared to an earlier Ph.D. cohort; they give a value for the ten-year period up to 1995 of 10.8%). For black Ph.D.’s and faculty, it’s 0.4% associate profs at the top 50, and 0.8% at numbers 51-100 (compared to 1% of Ph.D.’s up to ’95). So, there’s an argument to be made that the top institutions are paying attention to diversity in hiring, but not tenuring those folks. I suspect this is muddled, though, by the tendency of people who don’t get tenure in the Ivy League (and equivalent) to end up with tenured positions at schools a little farther down the prestige scale.
You can also argue that by looking only at the top 100 institutions, this is working in very rarified territory and ignoring the community colleges and lower-level state universities that serve the majority of students, etc. Which, yes, it is. But then, the usual argument is that things are much worse for women and underrepresented minorities in elite academia than at the lower levels, and these numbers suggest that the “much worse” of elite academia isn’t all that terrible in an absolute sense.
So, you know, social science continues to be messy, and analysis of the academic job market is complicated and confusing. Also, the Sun rose in the east this morning, and my kids are super cute.
It seems to me the academic job market is based on ever increasing need for faculty. Each professor in a graduate school has more PhD students than needed to replace them. Of course without the slave labor the current research process would grind to a halt.
Do we have statistics on the average number of Phd students a research professor has in his career?
If we were to envision a steady state size of the research enterprise in the economy what would the number of Phds per research professor be. It seems that today a new hire in a research Phd is competing for an at most constant sized grant funding pool.
The most important thing about that article in The Guardian is that it was trivial to guess that the author was from a field that has BIO in its name. Actually, two that each have bio in their names. That alone explains most of items 1, 2, and 4. More importantly, a different version of #4 (sustainability) is that students in the bio sciences SHOULD discover the reality of a drone-like future ASAP, because the dream of being a professor and a P.I. at a research university is most likely just that. Shades of physics grad school in 1969, only 10 times worse.
#3 is just a variant of “kids these days”. It has always been almost impossible to keep up with the literature. (The joke about the Physical Review growing in length at greater than the speed of light was old when I was in grad school, and I am old.) That what good upper-level special-topics grad classes and a good mentor are for.
Completely off topic, but I’d love to see the 2015 version of that oil and gas workers data (fig 1) from that article about minorities in science, just to get a sense of how dated it is!
Yeah, all of the problems of academic science in general are magnified in the biomedical fields, because biologists and doctors are apparently huge assholes…
The oil and gas workers graph is somewhat reminiscent of the “coming wave of retirements” thing that’s been promised in academia for at least a generation before me. It would be interesting to see an updated version of that, but I suspect that those graphs are only generated when trying to create a sense of impending crisis.
Another thought that occurs to me is that, while there have always been some students leaving the field (either just after getting their degree, or after a postdoc), I didn’t have a reliable way of quantifying it among students of my generation. Now, we have social media such as LinkedIn (presumably also Facebook, but since I have not yet joined the latter I do not know this firsthand), so I know that the brilliant student I met a few years ago is now working in Silicon Valley or Wall Street or wherever. There are certainly good reasons why students may be more likely to abandon the field today: funding has been especially tight the last few years (not in physics to the extent that it is in biomedical fields, but enough to be noticeable), and many students are making the rational calculation that they are unlikely to become a professor (and even if they do, the payoff may not be worth it–they may not be concerned about their own salary, but they are expected to feed their army of postdocs and grad students) and need to consider other lines of work. But it could be that I’m seeing more of this just because I have ways of seeing it that I didn’t have even five years ago.
Lyle, why are you assuming that the only acceptable or desired career path for a PhD student is an academic position? Not all grad students want to be professors, nor should all grad students want to be professors.
Good point @#5.
Data from the AIP show that, over a very long period of time, only about 1/3 of physics PhDs end up in academia. The period in the 1960s was an anomaly, but there is quite a long memory of that era in the structure of grad programs.
I hadn’t thought about what LinkedIn offers as insight into where people, go as mentioned @4. That has a possibility of making it easier for departments to track the pathways of their grads who follow the non-academic route and adjust what they do in PhD programs accordingly.