What really happens to new CS PhDs? A glut of PhDs, even in Engineering.
May 4, 2016 at 8:32 am 8 comments
I was surprised to see the numbers quoted below. PhD unemployment is that high? Aman Yadav just pointed me to an article in The Atlantic, with even more depressing news about the number of years to PhD, the debt after PhD, and the percentage of unemployment — see here.
CS is grouped into Engineering, so I tried to find the stats just on CS PhD’s. The 2014 Taulbee survey (see link here) says “The unemployment rate for new Ph.D.s again this year was below one percent.” But goes on to say, “The fraction of new Ph.D.s whose employment status was unknown was 19.7 percent in 2013-14; in 2012-13 it was 20.8 percent. It is possible that the lack of information about the employment of more than one in six graduates skews the real overall percentages for certain employment categories.” It’s not clear that we know what happens to new CS PhD’s, and what the real unemployment rate is.
Percent of Doctorate Recipients With Job or Postdoc Commitments, by Field of Study
Field 2004 2009 2014
All 70.0% 69.5% 61.4%
Life sciences 71.2% 66.8% 57.9%
Physical sciences 71.5% 72.1% 63.8%
Social sciences 71.3% 72.9% 68.8%
Engineering 63.6% 66.8% 57.0
Education 74.6% 71.6% 64.6%
Humanities 63.4% 63.3% 54.3%
Entry filed under: Uncategorized. Tags: computing education, CS education, graduate education.
8 Comments Add your own
Leave a comment
Trackback this post | Subscribe to the comments via RSS Feed
1. Bonnie | May 4, 2016 at 8:43 am
Well, we are having an impossible time hiring one of those fresh PhDs, as are many schools.
One thing – many graduating PhDs in CS need sponsorship to work in this country. That may mean it takes longer for them to find a job. Also, how many go back to their home countries? That could skew the “unknown status” category
2. Mark Urban-Lurain | May 4, 2016 at 9:01 am
This is consistent with findings from other fields such as biology. See for example, this from Nature (http://www.nature.com/news/the-future-of-the-postdoc-1.17253)
After a 2014 STEM Central Plenary talk by Jo Handelsman, Associate Director for Science at the White House Office of Science and Technology Policy, who reiterated the call for more STEM graduates/workers, I wrote this blog post about the STEM crisis (https://stem-central.net/blog/stem-crisis-whom-sky-falling/#.VynwInpQvm4) in which I noted:
“However, there is not unanimity about the “crisis” in STEM workers. While the data analyzed are complex, there are several analyses that note that, IF there is a shortage of STEM workers, salaries in these disciplines would rise, when, in fact, they are flat and even declining in real terms. See for example (Xie & Killewald, 2012). A recent analysis in the IEEE Spectrum summarizes much of this research (Charette, 2013). From this perspective, the drumbeat of “crisis” is about keeping the cost of STEM workers low for employers.”
3. Mark Guzdial | May 4, 2016 at 9:44 am
Mark, I just blogged (at Blog@CACM) on the same Handelsman talk — I just focused on a different part: http://cacm.acm.org/blogs/blog-cacm/201925-using-other-stem-disciplines-to-predict-future-computing-education
4. Jeff Forbes | May 4, 2016 at 11:05 am
Last I checked, PhD employment for CS was quite different than for other STEM fields. We could check with Betsy Bizot, but I’d expect that the unknown employment status doesn’t imply unemployed in any meaningful way.
5. Mark Guzdial | May 4, 2016 at 11:08 am
Jeff, does Betsy have data other than from Taulbee? It’s the Taulbee study that I’m quoting in the post that says that the low return rate leaves the question unanswerable right now. I bet that CS PhD employment is different than in other STEM fields, but I don’t think we have the data right now to know.
6. Jeff Forbes | May 4, 2016 at 1:43 pm
I meant that Betsy would have good intuition on whether the Taulbee survey data might be missing unemployed PhDs by misclassifying them as unknown employment.
7. kirkpams | May 4, 2016 at 11:07 am
Part of this is just bad reporting. The Atlantic uses the caption, “Percentage of Doctoral Recipients Without Employment Commitments, 2014.” The NSF report uses, “Definite commitments at doctorate award…” The Atlantic’s dropping of the timing quantifier adds hyperbole to the discussion. It bends the statistics to fit into their snappy tab title “Jobs are Scarce for Ph.D.s.”
There are many reasons a new Ph.D. might not have a job commitment immediately at graduation. Many were hoping for an R1 tenure-track faculty position and didn’t get it. Since those jobs often don’t have campus visits until March or April, their job search gets pushed back considerably. Bonnie also pointed out that many Ph.D. awardees may be working through visa issues and many may be in the process of returning to their home country. I would add that many on CS and engineering may be experiencing delays relating to getting a security clearance. In addition, none of these figures seem to account for entrepreneurs wanting to create a start-up. Also, the Ph.D. process is grueling; some people become jaded and burnt-out, and may be voluntarily choosing to take a break, to reevaluate their goals, or to leave the field entirely.
All things said, there’s a fundamental problem of definitions here. What constitutes “unemployment rate” in regard to this issue? Are we discussing short-term or long-term unemployment? How are we considering those who abandon their field and become employed somewhere else? Also, economists typically define unemployment in relation to those who are seeking work. However, all of the statistics above are in relation to the total number of Ph.D. awardees without regard to post-graduation intent; some may not be pursuing immediate work for family, medical, or other personal reasons.
8. What’s going on with CS PhD Enrollments: Guest Blog Post from Betsy Bizot | Computing Education Blog | May 20, 2016 at 7:32 am
[…] Values in the Engineering column are from Doctorate Recipients from U.S. Universities: 2014 (for the 2004 and 2009 figures) and Doctorate Recipients from U.S. Universities: 2013 (for the 2013 figure), Table 42, National Center for Science and Engineering Statistics, available from the “data” tab at http://www.nsf.gov/statistics/srvydoctorates/ These were reported in Mark Guzdial’s Computing Education Blog https://computinged.wordpress.com/2016/05/04/what-really-happens-to-new-cs-phds-starving-the-beast/ […]