We recently had
a discussion about the issues surrounding the process of getting a PhD in the US and thanks to the
AIP's Statistical Research Center, here's an illustration of a strong argument in favor of fixing the system. First, let's look at the trends in enrollment of 1st year grad students at PhD-granting departments:
Note that while there was a small drop-off in enrollment in 2009, the numbers have been fairly consistently around 2,800 for most of the past decade after a big dip in the late 90's thanks to the dot-com boom, which siphoned away a lot of potential grad students.
Now look at the number of PhD's granted over a much longer time period, although let's focus on the last ~15 years:
Here again there is evidence of the late 90's dot-com boom, delayed by about the average length of a PhD program, but note that the average is somewhere in the ballpark of 1,300, or roughly half of those entering grad school.
Now let's look at the number of tenure-track faculty hires over the last decade:
Interestingly these numbers are extremely constant. But again the average is somewhere in the ~360 phase, or only a third of the number of physics PhD's produced. This is roughly in line with the
anecdotal evidence from my department that showed about 30% of PhD's from 2000 to 2005 had tenure-track faculty positions.
That means that of those student's starting a PhD program at your university this year only roughly half will get their PhD and only about 13% will get a PhD faculty position. Of course all programs are not created equal, but the averages don't lie.
So in the grand scheme of things, we are (a) doing a poor job getting people through the existing PhD programs and (b) preparing people for jobs that a large fraction of those that do make it through will never have.
Nick, you conclusion spells out the two biggest problems well: people are taking too long to get PhDs and the vast majority are aiming with their PhD for jobs they will never get.
ReplyDeleteI guess the biggest way to bring down both these numbers is just cut federal and state funding for funding graduate PhDs requiring departments to accept less to begin with. You can always make the argument that one reason the supply (graduate students) is so different from the demand (# of jobs requiring a physics PhD) is because the government is intervening.
However, I guess there is already a panic that we don't invest enough in the physical sciences as it is. However, I would be curious if the problem with science is there aren't enough graduate students or if the problem is more like things like LISA get cut because there just isn't enough money.
Well, then I have a question. If, after 5 years or so, roughly 2/3 of new physics PhDs and 5/6 of entering PhD students aren't getting jobs in academia, what are they doing? I understand that unemployment among PhDs is actually really low, so are they mostly working in industry, national labs, secondary education, or what? I would say that we shouldn't be so much cutting the number of incoming PhD students as we should be being more honest with them and helping them prepare for the many other options out there.
ReplyDeleteI don't think that tenure-track jobs in academia are the only ones requiring a PhD. Roughly 1/3 of the PhD's from my program are in research positions -at national labs, university research institutes, or soft-money institutes like JPL or SwRI. The bigger story here is that we somehow lose half of all students between admission and graduation.
ReplyDeleteTo agree with your point, of 43 people that got PhD's from my department between 2000 and 2005, not a single one was unemployed.
ReplyDeleteOkay, 13% go tenure track and 1/3 go to things like National Labs... So why is it not the case that many if not most people getting PhDs in physics will get jobs that don't require them? .13 + .333 ~< 0.5.
ReplyDeleteRoughly 1/3 of those that get PhD's get a tenure-track job - but you're right that only 13% of those that start a PhD program will get one. So I think it's really more like we're training 33% more PhD's than we really need - at least if my department is typical.
ReplyDeleteI see. Thanks Nick. You always have great data.
ReplyDeleteOf course Universities love grad students and post-docs because they get a lot of research done, which helps professors to get grants, which is how many universities stay afloat. So perhaps it would help if funding went directly to the universities, rather than indirectly through overhead on grants.
ReplyDeleteTotally unrelated question - how do I get my picture to appear here? I can't seem to figure it out.
ReplyDeleteNick, next you your name in the comment box does it say edit profile anywhere?
ReplyDeleteBlah
ReplyDeleteNope
ReplyDeleteSorry about that - I was trying something out and now it seems I can't delete it.
ReplyDeleteNN & JS,
ReplyDeleteMany PhD in Physics find other careers in finance, business, marketing, and, many change fields. Most of you will explore five disciplines in your lifetime, if not more. The grants to professors are much more effective in doing something in the field than a block grant to an university which may choose to not fund sciences. However, I do not like universities are basically running research labs where grad and post-docs are cheap labor, and profs are but the first line supervisors. But, the competition for these positions is high, hence cheap labor. We should not forget that many go in industry with a bachelor's in physics and most find meaningful careers.
Perhaps a better model is to have substantial funding from state governments? CU-Boulder is slated to receive about $80 million of it's $1.1 billion operating budget for FY2011 (that's about 7%) from the state - everything else is tuition, grant money, or fundraising. State funding encourages the educational aspect of a university's mission as well as the concept of public, accessible higher education. Effectively this is the European model, except that here we do it at a state level rather than a federal one.
ReplyDelete