Thursday, December 11, 2008

The Job Market

Bill's post Monday of a Michigan post-doc position garnished quite a few comments, generated by legitimate worries about the academic job market.

1992 was the single worst year so far in the CS faculty job market. Many of our students applied to 100-200 schools. Very few got good jobs and several others either left academics or took jobs at departments at levels well below their capabilities.

That bad job market discouraged many from getting Ph.D.s in computer science. But those who did graduated in the middle of the dot-com boom when CS jobs were relatively plentiful. The moral: Don't let today's job market determine whether you consider graduate school. By the time you finish your Ph.D. and a postdoc position, the recession should be well over and the first wave of theoretical computer scientists will start retiring.

That's little comfort for those searching for a faculty job this year. I expect this year's academic job market to be much worse than 1992, not just in theoretical computer science but in all academic disciplines. At Northwestern "hiring into new positions will be deferred as much as possible, while hiring into vacated positions will be carefully scrutinized to ensure that the prospective candidates present significant opportunities that might not be available in the future" and we are one of the healthiest financially. Harvard is freezing both hiring and faculty salaries. I'm hearing of many schools that plan to cut salaries.

Best suggestion is to ride it out. Take another postdoc or other temporary position. Consider jobs outside of North America. The disappearing slots will have to be filled once the economy rebounds and no one will hold it against you that you were unable to secure a US faculty job in 2009.

26 comments:

  1. It may just be a false impression, but it seems that 1) there are much fewer people (than last year for example) on the job market this year and 2) there are many more postdoc positions.

    ReplyDelete
  2. Another useful piece of advice for theoreticians (probably not so useful for those who have already or are going to graduate imminently) is the following.

    Try to pick up skills in another area such as economics or biology. In that way you can probably wrangle a job as a researcher in another department in a good university -- then you would be able to attend the colloquia and seminars in the mathematics/cs departments, talk to people, have access to the library and can continue to do research in a pure area on the side. Many (very) famous ex-Soviet mathematicians did not have appointments in mathematics departments, because the number of such positions were very few, but they did excellent work (in mathematics) while formally working in Institutes for biology, genetics, economics, geophysics etc.

    In fact some theoretical CS people have already started do this and are making their students learn about auctions and other micro-economic topics as well as biology.

    ReplyDelete
  3. "In fact some theoretical CS people have already started do this and are making their students learn about auctions and other micro-economic topics as well as biology."

    People are doing this because they are really interested in the topics and how they are related to TCS, not to "wrangle" a position in another department.

    ReplyDelete
  4. To the advisors here:

    Can one realistically get back to academia after years working in industry assuming one has been publishing in good journals?

    How likely is it to get a tenured position with enough quality publications but no teaching experience?

    ReplyDelete
  5. Many people went from industry to academia after many years in, for example, the great AT&T Exodus. If you don't publish, it would likely be hard to go back (unless you are famous). But if you don't teach (and are a good speaker), then there should not be a problem.

    ReplyDelete
  6. What about the conversion from tenure-track to tenure -- have universities been "freezing" that too?

    ReplyDelete
  7. to anon #4: the following examples from theoretical cs should help: madhu sudan, david williamson, moshe vardi, mihalis yannakakis, michael mitzenmacher, howard karloff, joan feigenbaum, dimitris achlioptas, suresh venkatasubramanian, the list goes on... and this doesn't include people who had short stints at labs (fortnow, rubinfeld, sampath kannan, et al).

    ReplyDelete
  8. Try to pick up skills in another area such as economics or biology. In that way you can probably wrangle a job as a researcher in another department in a good university...

    I can't imagine a department hiring someone without either a degree in the relevant subject, or a significant body of work related to the subject. (Say, a mathematician whose thesis work was on mathematical biology.) Just "picking up skills" in another area is not enough -- it's not like it's any easier to get a faculty job in any other discipline. And could you imagine a CS department hiring a biologist because they knew how to program?!

    ReplyDelete
  9. Can one realistically get back to academia after years working in industry assuming one has been publishing in good journals?

    Are you talking about a research lab position or a real industry job? In the former case, one can definitely move back to academia if one is doing well. In the latter case, the issue is not just publishing in journals but being a visible presence in a community, attending conferences, etc. Concretely, think who your letter writers will be; your industry manager's letter will be ignored by the hiring committee (unless you are in a research lab).

    How likely is it to get a tenured position with enough quality publications but no teaching experience?

    Lack of teaching experience has no bearing on being hired, though having significant teaching experience can help.

    ReplyDelete
  10. I can't imagine a department hiring someone without either a degree in the relevant subject, or a significant body of work related to the subject.


    I think what was meant in the original comment was that one should also publish a few papers in the secondary field which is often easier to do than in thoretical CS or other areas of pure math.
    Since there is more research funding (say in bio) I would think that there are plenty of "soft" positions in those areas. It does no harm for theoreticians or pure mathematicans to target some of these as stop-gap employment before they prove that really big theorem that makes their career take off.

    ReplyDelete
  11. Re: jobs in other departments, has this actually happened? (Not counting closely related departments like operations research, EE, etc.)

    Are there people with PhDs in computer science working in economics and biology departments?

    ReplyDelete
  12. I have been a recruiter for a number of years and have watched a lot of new graduates struggle with setting themselves apart. They rely on their degree to get a job. The degree is important, but it's just a ticket to be considered. If you have 100 people, all with the same degree and similar grades applying for a job, you need to find a way to separate yourself. Otherwise, you create an image of just being a commodity. The key is providing specific examples of your accomplishments and showing how those examples demonstrate your ability to provide value to a company.

    Gary Capone
    Palladian International, LLC
    http://blog.palladiancr.com

    ReplyDelete
  13. "I think what was meant in the original comment was that one should also publish a few papers in the secondary field which is often easier to do than in thoretical CS or other areas of pure math."

    How does this strike you for arrogance? Here we go again ... Pure math people often look down their nose at TCS because its just applying stuff, it's not "real" "hard" math. And apparently we have the same (dangerous) attitude towards other fields ...

    ReplyDelete
  14. Biological/medical sciences will always need good programmers to do bioinformatics, to work on an interdisciplinary research team.

    Depending on your contribution in the research itself, you may or may not be included as a co-author in publications. So make sure to ask at the start of a project.

    ReplyDelete
  15. The average age of NIH "First Award" recipients is now 42, and has been trending steadily upward for more than 20 years.

    These data suggest that it is not (at present) reasonable for young PhDs in biomedicine to plan on achieving independent investigator status much before age 50.

    ReplyDelete
  16. which schools are rumored to cut faculty salaries? If somebody knows what lance is referring to it would be great if they could give us more info.

    ReplyDelete
  17. "Don't let today's job market determine whether you consider graduate school."

    This is dangerous advice. You definitely should use the job market to help make an informed decision about graduate school. Don't base your decision entirely on today's job market, of course, but look at some sort of average. Be aware that when the job market improves competition will be tougher---lots of people are taking postdocs.

    ReplyDelete
  18. Which schools are rumored to cut faculty salaries?

    The State of Maryland is imposing a 5-day "furlough" on all state employees (including faculty at the University of Maryland), which is effectively a 2% pay cut. (I am simplifying things a bit here.)

    ReplyDelete
  19. In fact some theoretical CS people have already started do this and are making their students learn about auctions and other micro-economic topics as well as biology.

    ReplyDelete
  20. Speaking as an engineer who assiduously scouts the mathematical literature for new algorithms and new ideas ... aren't there at least some mathematicians who perceive, like White House Chief of Staff Rahm Emanuel, that:

    "You never want a serious crisis to go to waste. And what I mean by that is an opportunity to do things you think you could not do before."

    Aren't there at least some mathematicians who---rather than "riding it out"---perceive new creative opportunities for the mathematics community in the present crises (where by "crisis" is meant, both the short-term economic crisis and the longer-term global stewardship crisis)?

    After all, it's the holiday season ... folks could use some well-founded optimism.

    So if you've got some Emanuel-style creative optimism, please share it! :)

    ReplyDelete
  21. I'm a bitter and jaded grad student facing this miserable job market. I think quite a few students are delaying their graduation because of it. I think some of the advice of looking for more applied work in economics, biology, statistics, or even policy is really reasonable.

    However, in the spirit of optimism for the holiday season, and to humor John Sidles, I will say that things can't get much worse... (right? or am I just deluding myself?)

    ReplyDelete
  22. To "the bitter and jaded grad student"

    Even if you do take a job in another area,
    you can think of it as a "day job" and continue to do what you would like to do during other times. Remember that some of the greatest mathematics were often produced under extremely trying conditions -- the experiences of Leray, Grothendieck, Ramanujan and a host of Soviet mathematicans come to mind. Even in modern times, people working in industrial research labs often have to devote half their time doing company related non-research work. So bad as it might be your situation is not hopeless.

    Btw, the primary reason why the theory job-market is so bad is the failure of NSF to fund pure-ticians at the same level as applied practitioners. This is not news of course, but unless this is remedied this will lead to trickle down effect, with department heads getting biased again theory. Why is it that mathematicans are "encouraged" to collaborate with the engineers, but almost never the other way around ?

    ReplyDelete
  23. Anonymous deplores: "the failure of NSF to fund pure-ticians at the same level as applied practitioners"

    Uhhh ... at present, the engineering directorate receives about 15% of the total NSF budget ... which is why practicing engineers tend to believe that the "pure-ticians" already get 85% of the pie!

    That's why the use of terms like "pure-ticians" risks igniting academic arguments that, historically, have not led anywhere good, at the NSF or anywhere else.

    If common-sense is to be found anywhere, it seems to me that the most urgent need is for mathematicians, scientists, and engineers to work together ... to combine good theorems, with good physical insight, with good engineering design ... in service of solving the many urgent problems of our nation and our planet.

    For this to happen, maybe we all have to leave our cubicles, more than we do at present. Which might even be fun.

    That's my 2¢, anyway ...

    ... and Happy Holidays to all ...

    ... and special thanks to Lance and Bill, for hosting this fine weblog! :)

    ReplyDelete
  24. There's been massive overproduction of PhDs in this country for years, if not decades. Yet every once in a while, the comments section at one of the TCS blogs gets filled with people who are shocked, shocked, at how bad the job market is for theory grad students. I understand the overall economic situation is especially bad these days, but this has little to do with the fact that the creation of PhDs is greatly outpacing the creation of academic jobs. This is how its been, and this is how it will be for as long as we have such a fecund system of higher education.

    There's nothing wrong with lots of PhDs - we need people with advanced technical training more than ever. But it should be clear to my fellow grad students by now that unless you're the second coming of Alan Turing you need to prepare for the highly likely event that you won't be able to score a decent academic job. This is not news. And frankly, if you're spending 5 years in grad school, with all the resources of a major research university at your fingertips, and are failing to acquire any marketable skills, then you're doing it wrong.

    ReplyDelete
  25. It seems to me that some of you younger folks have only a hazy notion of what a technological boom feels like, or how it works ... this 1957 Time Magazine article titled The New Age may help.

    It's very easy to criticize this Time article (and others like it) for being shallow ... but ask yourself, are Ramo and Wooldridge telling all they know and all they think?

    Also, it's too easy to (superficially) decide there were no "pure-iticians" involved ... but deeper historical reading will reveal that this is very far from being the case.

    ReplyDelete
  26. And frankly, if you're spending 5 years in grad school, with all the resources of a major research university at your fingertips, and are failing to acquire any marketable skills, then you're doing it wrong.

    The basic "marketable skill" one hopes to acquire in a Phd program is really the ability (in the long run) to produce more of ones kind. In order to do so effectively one needs a position in a doctorate granting institution. This is the nature of academia and is true in any discipline -- from philosophy to mathematics and computer science. This is also why most people want positions in academia after they graduate, and there is absolutely nothing wrong with that expectation.

    ReplyDelete