Subtleties of the Crappy Job Market for Scientists

By Julianne Dalcanton | July 9, 2012 3:22 am

Recent reports and articles have generated a lot of buzz about the difficulty of finding employment in the sciences. These articles mirror the anxieties of the young astronomy community with whom I am most familiar. Scientists are not stupid and are pretty good with data, so they can look at the number of graduate students, the number of postdoctoral positions, and the number of faculty ads, and correctly assess that the odds of winding up with a long-term academic position are not good.

However, difficulty finding a “long term academic position” is not the same thing as difficulty finding a job. Buried in those same articles is the fact that the unemployment rate for physicists (which likely mirrors that of astronomers) is between 1-2%. In contrast, the lab-based biologists and chemists (which are the focus of the articles) are not finding employment at all, or if they do, it’s frequently in a position that makes no use of their technical skills.

To me, what this implies is that most of the skills mastered by PhD-level lab-based scientists are not readily transferable to other jobs, and are not easily generalized (or at least, are not perceived as generalizable by employers).  The ability to work well in a lab setting is only valuable if the economy supports large numbers of labs.  Industry used to host these, but the era of corporate research is largely over.

In contrast, a typical astronomy postdoc has experience with software development, image processing, filtering, large data volumes, experimental design, data visualization, project management, proposal preparation, and technical writing — all of which are generic skills that can be applied to a wide variety of technical positions outside of astronomy.  Jobs that use these skills do not require large infrastructure overheads, and thus can be found in start-ups, and in almost any region of the country. Moreover, the typical astronomy or physics postdoc has had much more autonomy and freedom to lead projects, whereas lab-based biology appears to be far more pyramidal, giving postdocs far fewer venues in which to demonstrate their initiative and leadership.

In short, while few astronomy and physics PhD’s are explicitly educated for positions outside of academia, their training actually transfers quite well.

The problem in astronomy and physics is therefore not employment, but expectations. The fact remains that many PhD students do not fully understand that they are unlikely to ever have the equivalent of their advisor’s job, and leading to (completely understandable) fear and frustration when discovering that one’s goal is not likely to ever be achieved. This misconception is primarily a failure of mentoring and education. First, astronomy and physics have never had a 1-to-1 ratio between people earning PhDs and the number of faculty jobs, so while the ratio may be particularly unfavorable now, there was never a golden era. Second, there is no reason that the routine business of being a faculty member should be an appealing job description to every single person who is interested in astrophysics. As such, students should never be made to feel that they’re failures for not getting a particular flavor of academic position, and should instead always be encouraged to explore other avenues that could use their talents while bringing them greater day-to-day satisfaction. And based on the studies, I’m grateful that those options appear to exist for the physicists and astronomers who change their direction.

CATEGORIZED UNDER: Academia, Science and Society
  • Bee

    The German Physical Society has an eye on the unemployment rate among physicists in Germany, and they consistently find the same: It’s far below the population average.

  • David Derbes

    I earned a PhD in theoretical physics in 1979, and went straight into high school teaching. Except for a two-year stint at Tulane, I’ve done nothing but high school teaching. It’s been a good fit for me, and may be for others. The country needs the best math and science teachers it can get. More young PhD’s should consider this, if they like teaching and like young people.

  • Phillip Helbig

    In many cases, the frustration is not failure in getting a permanent academic job per se but rather seeing people who—by any criterion one chooses—are not as well qualified getting permanent academic jobs. It’s not just not being the right person in the right place at the right time, but rather than non-academic criteria are too often deciding factors.

  • Moose

    In my department there’s definitely an air of ‘if you don’t stay in astronomy, you’ve failed’. I’m finishing my PhD and leaving to work in the tech industry – and I couldn’t be happier about it. But all the way through, I’ve had people subtly and not-so-subtly pushing me to stay in academia, despite the fact that it is evidently not a good fit for me. This ranges from a general lack of information and skills training for careers outside academia, to being actively discouraged from going on courses to develop my transferable skills, to the head of department and my supervisors telling me I shouldn’t be searching for jobs elsewhere and that my thesis is the only thing that matters for my future.

    Thankfully, I’ve got a good few transferrable skills that secured me a position – but one bugbear is the fact that academia tends to use programming languages that are not used in the commercial world (Fortran, anyone?) – a fact that automatically rules out a lot of the kinds of jobs I’d like to apply for. It p’s me off because my department KNOW that 9/10ths of us will not get faculty positions, and yet they don’t even seem willing to acknowledge that there is a world outside academia, let alone attempt to help prepare us for it.

  • Scott

    It’s interesting to note that the crappy job market for lab-based scientists in academia reported in the US press may not be as awful here in Europe (there’s a brief mention of this here: It would be interesting to see the statistics for the European chemistry/biology market to compare with those from the US; perhaps US-based lab-based scientists should look further afield for research positions?

  • Julianne Dalcanton

    In my department there’s definitely an air of ‘if you don’t stay in astronomy, you’ve failed’.

    That kind of behavior just ticks me off. Given the crap involved in an academic career, I’m amazed that -anyone- chooses to succeed in it. Nobody is surprised when people don’t wind up marrying the boyfriend/girlfriend they chose at the age of 20 — why should people be surprised when they discover the career they chose at 20 isn’t the right fit either?

    non-academic criteria are too often deciding factors.

    But most of what’s involved in successfully carrying out a faculty job is not just “research excellence”. It’s not like a prize for being awesome — it’s a complicated job description and a lot of different skills go into it. Again, this is a mentoring/training problem, where people are led to believe that high-quality research is the only factor.

  • Julianne Dalcanton

    Scott — might be interesting to cross-correlate with the importance of the manufacturing sector vs “knowledge” sector.

  • romain

    I’m a former theoretical physicist working in a biology lab. Most PhD students and postdocs there would be horrified to read that they are not involved in “experimental design, data visualization, project management, proposal preparation, and technical writing”. I’m afraid this reflects some inherent physicists’ contempt to biologists.
    The only thing is that -in general- physicists are better at maths and programming than chemists/biologists. These skills are thought to allow you to work in many different jobs.

  • Julianne Dalcanton

    romain — I have a ton of respect for biologists, and did not mean to imply otherwise. My views were shaped by the lab-focused biologists I’ve known, who spent the dominant fraction of their time dealing with classic “wet” lab stuff, not coding. Neither is “better”, but one is more transferrable. In addition, the young biologists who blog (albiet a biased subsample) consistently complain about not being allowed to direct their own research.

  • Derek

    one bugbear is the fact that academia tends to use programming languages that are not used in the commercial world (Fortran, anyone?) – a fact that automatically rules out a lot of the kinds of jobs I’d like to apply for

    While I’d certainly encourage any department to further the usage of more modern and versatile languages (Python, anyone?), and more importantly, software development practises, the purpose of a PhD education is not training in specific programming languages (which are bound to change over the time of a few postdoc generations anyway), but precisely the higher level skills described in the article, which will always let people adapt more easily to a new working environment. Unfortunately employers do not always seem to realise this, or at least they tend to look first for someone who has both the overall experience and the exact specific training required for the job.

  • gameswithwords

    Some grief seems to come from failure to realize just how different different departments are. When I was researching graduate schools, I realized quickly that a small number of schools (maybe a half-dozen) where essentially everybody who wants to gets a job in academia, usually at a pretty decent university. There are plenty of schools where essentially nobody gets a job in research (they are more likely to either leave the field or be a teaching adjunct somewhere).

    I don’t know about physics, astronomy or biology, but I wouldn’t be surprised if it were similar, especially since this isn’t especially unique to science. The legal profession, for instance, is very similar. All the hand-wringing about the fact that so many students are taking on $200k in debt and never land that Big Firm job tends to neglect the fact that if you go to a top 3 law school and want a Big Firm job, you’re almost guaranteed to get one (ignoring 2008-2009, which was hell for the legal market). If you go to 80% of the law schools out there, you’re almost guaranteed not to get one. So is law school a good deal? If you go to Harvard, Stanford or Yale, definitely. Otherwise, maybe, maybe not.

    Whose fault is it? When that many people are making uninformed decisions, it’s hard to blame individuals. But it’s not obvious that the problem is with the system, either!

  • secretseasons

    My dogwhistle interpretation skills are rusty, but I wonder what #4 Phillip Helbig was really getting at?

  • Brett

    I think the big problem is that learning does not stop when you get your PhD. You could have the best thesis in the history of physics, but if you’re a prick like Fritz Zwicky, then Vera Rubin is going to get all the credit decades later. Despite Physics being a field based solely on what’s correct and what is incorrect; getting a job is still a social display. Most scientists have a problem with empathy; which is why Sheldon Cooper from the Big Bang Theory is so funny, because it’s so true. Yes, you’re smarter than a majority of the population, but there’s a right way and a wrong way to go about asserting that. Sometimes it’s better to let your team screw something up than to get in a fight about it just because you know the correct answer. The world is full of petty people with undeserving egos, but you just have to deal with it. The most perplexing challenge you ever face while at your job may be understanding how someone as stupid as your boss is in charge of you; but you still have to tip toe around their ego and treat them with respect.

    Science is unadulterated knowledge with no ego. Business is the exact opposite.

  • Tommy

    Julianne, the reason I take issue with the “non-academic” criteria comments, is that a quick glance at the data Erich Poppitz has compiled ( )on high energy theory hires shows that one of the main criteria is simply having gone to Harvard or Princeton. While those are good schools, I seriously doubt they are 20+ times better than many of the others listed. Simply put, the way decisions are being made isn’t really supported by data for the complex role you have. It’s generally about the same few people pushing their students through every time.

    I’ve successfully made the jump into tech, and I’m quite happy. That said, it was a very, very difficult jump to make. In fact, I thought much as you did that I had all these great skills, and employers would snap me up. That’s just not the case *at all*. Employers do not care about “transferable” skills for the most part. They want to know you can code in C++, or Java. I got consistently dinged for lack of experience working with a large code base on maintainable code. Those things are generally not present in academia. My department also had little to no formal help (some people privately helped me more). As a postdoc, I was literally unable to even get someone official to look over my resume for tips. I received no suggestions about networking organizations, basically no support at all. I had to do all the legwork myself, which became it’s own full time job. If departments want to seriously take the approach that there is life outside academia, and it’s a viable career path, then the support structures need to be put in place. As of now, they are doing a huge disservice to their students and postdocs.

  • Phillip Helbig

    “My dogwhistle interpretation skills are rusty, but I wonder what #4 Phillip Helbig was really getting at?”

    One example, of many: special help for dual-career couples, which usually involves giving someone a job they would not otherwise have been offered. “But it doesn’t take anything away from anyone; it’s paid for out of extra money!” Yeah, right. “But it’s not a permanent job, it’s just a five-year fellowship to tide me over until a real job opens up!” Wouldn’t that be something that everyone else who wants to stay in academia would welcome? Most people would agree that offering a job to one’s spouse who is otherwise not qualified is unethical, but when a “trailing spouse” gets offered a job because the “leading spouse” gets offered a job, that is not only supposed to be ethical, but one is inhumane if one doesn’t support this policy. Where is the difference? “Where’s the problem? We were up front and said take us both or take neither of us?” That might be fair between the couple and the university, but not to other people looking for a job. One of the great accomplishments of civilization and progress is hiring people because they are qualified for the job, not because they come from the right family, are of the right race, are willing to bribe those making the decisions etc. This is a step backwards.

    Another example: Blackmail. This perhaps happens elsewhere as well, but is probably more prevalent in academia. In the non-academic world, hiring the wrong person can mean a big loss of money for the company, but in academia, the risk to those making the wrong decision is small. Also, there is often no official reason given as to why a certain person was hired. (Not everywhere; in some countries, external referees rank the candidates and write justifications for who is ranked where on the list, and this information is either public or provided to the other applicants.)

    Another example: Job descriptions are tailored to a specific person.

    Another example: Jobs are not widely advertized.

    While there might be some justification for the last two, the people should at least be up front about the fact that they might have been forced to advertize the job for legal reasons, but actually it’s a done deal. (Yes, I did see one job announcement which said “Members of the institute will probably apply”.)

  • Julianne Dalcanton

    a quick glance at the data Erich Poppitz has compiled ( )on high energy theory hires shows that one of the main criteria is simply having gone to Harvard or Princeton.

    While some of this effect is certainly a networking issue, it’s not clear how much is “value added” by the institution and how much is the student population they’re working with. Some fraction of the graduate admission at the 90% of not-top-tier institutions is about finding diamonds in the rough — students with potential who may have faced some struggles in the past, but who show promise. In contrast, the students going into top-tier schools are almost entirely those that had their act together from the get-go (whether from privilege, good-fortune, drive, and/or just being ridiculously talented). I’m not too surprised that a population of students who had everything going for them at the time of admission also winds up ahead of the game down the road — there’s a good chance that these students would have excelled no matter where they wound up. That said, it’s not even a given for those coming out of the top 3 programs.

  • Sarah

    “But most of what’s involved in successfully carrying out a faculty job is not just “research excellence”. It’s not like a prize for being awesome — it’s a complicated job description and a lot of different skills go into it. Again, this is a mentoring/training problem, where people are led to believe that high-quality research is the only factor.”

    @ Julianne – perhaps being a successful faculty member is not just down to “research excellence”, but it seems like becoming one in the first place is. I’d love to teach, work with students, do dotAstronomy stuff, outreach etc, as well as do research or build instruments, but my skills or experience or potential in those areas seem to count for nothing at all in the faculty jobs market. That is a big frustration. But perhaps I’ve not met the right hiring committee yet :-)

    For the rest, totally agree with you.

  • Eric G. Barron

    What Moose describes is all too common. In astronomy, the academic elitism is quite widespread and it hurts students and postdocs. From what I’ve observed, the “either you’re an academic or you’re a failure” attitude originates with those academics–primarily mid-to late career–who know *they* would not be able to succeed outside of academia due to their inability to handle a more rigid work schedule, their failure to stay current with the necessary skills, or the thought of missing the perks (“conferences” held at luxury resorts are not as common outside of academia). The problem is that those people are often the ones who are the loudest voices in their departments due to seniority or politicking…so the students and early-career scientists suffer because of the insecurities of old men.

    Many departments are also hurting in the area of providing transferrable skills. As the importance of computing in astronomy increases, computing education in astronomy is actually falling further and further behind. (Note that there is a difference between being able to program and being able to program *well*.) For example, handing students a sheet of IDL commands and telling them to type the commands in to see if their plots match what is on the sheet should not pass for computer programming education…but it does, even at some large or “prestigious” institutions. Then there are the cases of grad students or postdocs being handed code by their advisors (who may even have been given the same code by *their* advisors) and just using that code without ever knowing how it actually works or what it really does…or even how to build it when current compilers stop supporting the deprecated features that were all the rage when the code was originally written in the 1980s. I’ve spoken to astronomers whose complete knowledge of “databases” begin and end with comma-separated text files. I know there are some astronomy departments (mainly the ones where computing itself is one of the research focus) that do a good job with computing education, but overall I’m not encouraged by the direction in which the field is moving in this area.

  • Brett

    I think you’re right about that Julianne :

    “While some of this effect is certainly a networking issue, it’s not clear how much is “value added” by the institution and how much is the student population they’re working with. Some fraction of the graduate admission at the 90% of not-top-tier institutions is about finding diamonds in the rough — students with potential who may have faced some struggles in the past, but who show promise. In contrast, the students going into top-tier schools are almost entirely those that had their act together from the get-go”

    but I do see that shifting. I would use the University of Michigan as an example. Colorado state, The Perimeter Institute in Canada, etc.

  • Julianne Dalcanton

    Brett — Totally agreed that there are plenty of students at non-“top tier” schools who have had their act together from the time they were infants, and plenty of non-“top tier” schools who provide sufficient “value added” that their students do better than they would have elsewhere.

    Sarah — agreed that research excellence is a threshold condition, but there are places that do look at the bigger picture. However, it’s not always easy to tell where they are from the job ads, and it’s certainly not 100%.

  • Lord

    I think students tend to focus on the subject rather than employment so providing them with an understanding of what options are available, what they are like, and the odds of reaching them can be enormously helpful. The options can and will change but at least knowing what they are currently can give them a glimpse and get them thinking about what they want and what they should do for it. Often I think this is avoided as discouraging, but usually other options are even more discouraging, and it is certainly better than later disappointment.

  • Jasper

    I would just like to say that I disagree with most of what is being said in the above. Here in Europe I see around me it is much easier for lab-based, specific skilled people to get a job. I’m currently doing a PhD in theoretical physics and when I talk to most tech companies it is hard for them to see the benefit in hiring me. I agree with Tommy that most employers don’t care about transferable skills and are only focused if you can program in C++ or Java.

    I don’t see where you get the data to say: “In contrast, the lab-based biologists and chemists (which are the focus of the articles) are not finding employment at all, or if they do, it’s frequently in a position that makes no use of their technical skills.”
    Afterwards you clearly say that physicists and astronomers also frequently get a position that makes no use of their technical skills (apart from the obvious programming skills) and should be encouraged to get such a job.

    Also the above article seems to state that companies would like to employ postdocs. I would like to strongly disagree with this point. I’ve asked the big tech companies in my country and postdocs are generally viewed to be as overqualified and to old and thus not worth to invest in.

    In today’s market it seems to me there is only one condition to get a tech job: good programming skills = job. And this is independent of whether you do physics, biology or chemistry.

    Thus my advice to any PhD students out there:
    – Do a PhD (it has value in that it makes you more independent, stress-resilient, efficient and helps to get you in top tech jobs)
    – If it turns out that you do not like academia that much: don’t go and do a postdoc, whatever the people around you might tell you
    – Learn to program in C++ and Java

  • Tommy

    I couldn’t agree with Jasper more. The “transferable” skills idea is
    largely a product of people who haven’t actually gone job hunting.
    When I started, I was shocked at how different and more difficult I
    was led to believe it would be by academics. Yes a company like
    google wants smart, independent problem solvers. But you better also
    know C/Java really well, know advanced algorithms, data structures
    etc. I doubt most PhDs know about swap sorts and the advantages of
    linked hash maps in big-O speak. There is one field that cares about
    these transferable skills: finance. Money isn’t the only reason people
    end up there. Until departments implement more support structures and
    reach out themselves to local industry, this will not change.

  • Julianne Dalcanton

    If it turns out that you do not like academia that much: don’t go and do a postdoc, whatever the people around you might tell you

    This is superb advice.

    As is learning to program in C++ and Java. And really, once you know how to program in one language, it’s just not that hard to learn another.

  • Julianne Dalcanton

    Also, I don’t have any way to assess how difficult it is to move out of academia beyond anecdata (grad students at UW who have decided they don’t want to leave Seattle, and prepared themselves to make the switch before graduation, have found interesting jobs in a few months — and none of these were in finance) and the global data that the overall unemployment rate for physicists is significantly lower than other fields.

  • Charles

    Ever since that Washington Post article I think we’re finally heading to the crux of the issue. Minted PhD students not finding a long-term job in academia is one issue. Those carrying PhDs looking for a job outside academia is another issue altogether.

    The Washington Post article made it seem like going from post-doc to post-doc isn’t the norm. I read a blog that said if you’ve been in a post-doc for more than 5 years shouldn’t you give up by then?

    But this, sadly, is the status quo these days for those with a PhD. Lack of job security. The Academia space puts a priveledge on those that constantly uproot themselves and go work in different labs in different countries. And we ask ourselves why there is a huge rate of attrition for those with PhDs leaving academia.

    Then there’s the problem of those with a PhD as being seen with no marketable skills for the general job market. This is merely a perception and not a reflection of the real state of affairs. As your article rightly points out “experimental design, data visualization, project management, proposal preparation, and technical writing”. There are many PhDs that do not realise they have these skills. Even worse is that there are many employers that do no realise it either. The ability to learn and self-teach is given no weight in a job market outside academia.

    Perhaps more effort should go towards educating employers of the vast untapped potential in the newly-minted PhDs that roll out of the ivory towers on a constant basis.

    Another issue is one of there not being enough clear avenues for PhDs to exit academia. We’ve all heard everyone talk of how they want more scientific-minded people in other avenues of life (like politics for example). But where are the clear paths to other non-academic options? Few exist. I know many that really had to start at the bottom with internships etc once they left academia to get a real foothold in another line of work.

    Many problems, few solutions.

  • LM

    @#27 Well, part of the reason you won’t see many scientists in politics is because of justified disdain in academia for politicians. Perhaps this hurts the funding for basic science because of the lack scientifically literate people in Congress, but ultimately it comes down to the population as a whole being grossly scientifically illiterate.

    Politics is a dirty business where the truth takes a back seat to talking points, money, tribalism and party loyalty, both in Europe and the US– but it’s especially dirty in the US and in the developing world.

    Some parts of academia are pretty political and gamey, like certain theoretical subfields in physics, but it’s not close to being comparable to what politicians have to do.

  • Neal J. King

    27, Charles:

    “Perhaps more effort should go towards educating employers of the vast untapped potential in the newly-minted PhDs that roll out of the ivory towers on a constant basis.”

    This is a bit unrealistic in a buyer’s market.

    What would be more to the point would be cluing in these new PhDs on what they specifically have to offer that might be of interest to potential employers.

  • anon

    If students don’t seem to understand that they are unlikely to wind up with a job like their advisor’s, the fault doesn’t lie with bad advising, but with the understandable reactions of the students themselves. Most students who go to graduate school and successfully complete a PhD are used to having been the smartest person in their class for most of their lives. They understand full well that most students don’t wind up as professors (we are talking about people who are very good at math, right?) They just assume that they will personally be an exception. Most of them went to grad school specifically because they wanted a job like their advisor’s, and there is a strong element of wishful thinking. Don’t place all of the blame on the advisors.

  • David Nataf

    If astronomy were like medicine, it would be much harder to get into graduate school and much easier to stay in the field.

  • R

    I agree with Jasper and Tommy — especially given the current unemployment rate, employers don’t care about “transferrable” skills, because there is always someone out there who has the exact skill that they want.

    The jump to the tech world can be made, but you need experience with C++, Java, and SQL.

  • Jon Claerbout

    Imaging the earth’s interior differs from imaging the heavens, but 45 years teaching the former (seismic imaging), I’ve found all (50) the PhD graduates got industrial jobs immediately and a handful (6) got academic jobs. The industrial jobs divide into oil companies and seismic survey contractors. Current salaries for new PhDs range about $110-130k.

  • exPostdoc

    David Nataf hit the nail: the number of PhD candidates needs to drop almost by an order of magnitude for academia to become a sane environment. Stop pretending that PhD programs should prepare 1/10 to 1/100 of its members for professorships and the rest for programming jobs, and make admission in PhD programs harder. That way retention in academia would get closer to that 1-to-1 rate that Julianne mentions (something like 1-to-5 could be acceptable; right now it’s much much lower than that).

    So then: Why is the system interested in producing many more PhDs than are needed in academia? Somebody mentioned it above: buyers’ market, but also: cheap labor to produce and analyze data for the avisor’s glory and publication list.

    In the current model, all big players win: Universities get tuition fees for more PhD candidates and overhead from NSF/NIH grants; tenured faculty get cheap labor for their labs and groups; google & Co. get to hire the stream of ‘failures’ that decide not to pursue an academic career after the PhD (people used to such underpaid jobs that they literally flip when they hear what google is gonna pay them).

    The losers in this story: those that enter their PhD programs dreaming of becoming a tenured professor and have to leave after 3 postdocs, with the feeling of being a failure. Those that have to move across the globe every 2 years for another meager temporary contract. Too naive, they should’ve known better? Maybe, but the system is built to deceive them, and they’re not the only ones to be blamed.

  • Walter

    I have had many discussions about this and I had myself thought that limiting the number of PhDs students accepted will solve the problem, but that would be closing the door to people to explore by themselves and chose their own path. It is natural that some people will leave the field, the only thing that prospective graduate students need to know is what are the typical career paths of scientists, so they are aware that there are other alternatives. Some freedom to explore other areas can also be good, maybe even by encouraging PhD students to take a minor on a more practical and applicable field.

  • B

    We need to institute some kind of PhD birth control. If we can get a cultural change whereby no tenure-track prof. can have more than X students over a Y year career, it means they’d have to be choosy about their students, rather than generating a pipeline of them.

  • Tommy

    I will state again, departments need to be more honest and also actually help people transition out of academia if they realize only 1/50-1/100 of their students will be able to stay in. Have an actual department specific career services advisor, with actual contacts in interested sectors like tech, finance, consulting, private labs etc. Help students and postdocs make these transition. The current system just literally casts them out to fend for themselves after benefiting from said cheap labor. My line above about resume help was not an exaggeration. I contacted my university’s career services office and was told no help could be given to postdocs, even after I offered to pay for it. That’s a very sharp contrast from say business schools that have networking events, and tons of help for every step of the job search, even back to helping students determine where they might fit best.

  • bowlweevils

    @12, I would like to point out that lawyers can’t do math, aside from counting money, so there are actually 6 Top 3 law schools and at least 15 top 10 law schools. You need to have a few drinks in you to properly visualize the data.

  • gameswithwords

    @38. There are just the 3:) Anyone who says otherwise went to school #4.

  • I.P. Freeley

    First, astronomy and physics have never had a 1-to-1 ratio between people earning PhDs and the number of faculty jobs, so while the ratio may be particularly unfavorable now, there was never a golden era

    This glosses over the critical change in the job market–namely the explosion of postdoc positions. In the past, PhD astronomers were overproduced at around a 2-1 ratio, and now that’s probably closer to 4-1. Now, rather than having new PhDs leave the field in their late 20’s, we have seasoned postdocs forced out in their mid 30’s! Check the rumor mill, in the US there are now more postdoctoral prize fellowships than faculty hires every year.

    As others in the thread have pointed out, it’s a much better idea to leave the field right after grad school than to try during a postdoc. As a student, you have access to your university’s career center and can probably take classes from other departments to help make a transition.

    I think junior scientists should simply boycott postdoc positions. Most postdoc positions are dead-end under-employment.

    There’s a simple solution to the job shortage–move funding from PI-grants and prize fellowships to create permanent positions at federal labs. This is an allocation problem, not an overall funding issue. Both the astronomy and biology job shortages were caused by increases in funding that were poorly allocated. Unfortunately, senior scientists like their grant money, and thus no one with clout will lobby for this solution. That’s why I think change has to be forced from the bottom-up with junior scientists refusing to take postdoc positions.

    I’m making a long rant about this because I think this issue is hurting the field. With a job market this bad, we are not able to recruit and retain top talent. There was an article in Science a while back documenting how top undergrads are no longer majoring in science, and this is attributed to the job market. If a PhD in astronomy doesn’t lead to a viable career, it becomes the equivalent of a PhD in the humanities–a vanity degree sought by the very dedicated but not particularly talented students. We are entering the second age of the gentleman astronomer.

    Imagine if Starbucks hired 1000 new employees and paid them for a month of barista training, after which they fired 750 of them. And then they did this every year. While simultaneously laying off experienced baristas. It’s pretty clear how this is wasteful, and how this would drive away otherwise qualified applicants. The way we run science in the US is similarly inefficient, and is driving away talent. We have plenty of questions left in astronomy (WTF is dark matter? HTF do you make a globular cluster?). We need to attract the best students and train them efficiently if we are ever going to answer them.

  • Gary

    Julianne Dalcanton: “The fact remains that many PhD students do not fully understand that they are unlikely to ever have the equivalent of their advisor’s job, and leading to (completely understandable) fear and frustration when discovering that one’s goal is not likely to ever be achieved.”

    et al;

    Lesson learned: …many PhD candidates are not as smart as they believed their tuition was purchasing the antidote.

    Lesson learned: “… discovering that one’s goal is not likely to ever be achieved.”

    Good to know that your professor isn’t telling you, up front, that your tuition is not well spent.

    Hey, your professor has your best interest in mind.

    Lesson 1.

  • Stephanie

    As a recent astronomy PhD who is no longer in astronomy, I don’t think making grad school admissions tougher is necessarily a good idea. I had a very happy 7-year relationship with Astronomy, which ended in an amicable breakup in which we were able to remain friends (I still sometimes read the ApJ abstracts, and Astronomy occasionally cites my papers). Had I been dumped at the end of college when I was still passionately in love, it would have meant serious heartbreak, rather than gradually learning we weren’t meant to spend a lifetime together.

    One of the most important things that departments and professors can do is to encourage grad students to lead fulfilling lives while working on their PhDs – have babies and hobbies and not feel that they have to put their lives on hold. (Thanks, Julianne.) Otherwise the realization that a tenured faculty job is not in the cards can lead to resentment over years of pointless indentured servitude.

    @40 It’s really too bad that there aren’t more professional research staff positions in astronomy, as there are in biology (my current job). While not everyone can be a faculty member, the work currently being done by short-term postdocs could be done more productively by people who aren’t spending much of their time thinking about or applying for the next in a series of 2-year jobs. Longer-term employment = happier scientists with stable family lives = more science getting done.

  • Razib Khan

    a minor note: i think one issue is that biologists are on average less intelligent than physicists. less intelligent people have fewer options.

  • Erik

    @42, 35 – this is a very important point, I think. It is true that the PhD produced-to-tenured ratio is quite low, but that’s not a bad thing on its own – students just need to be told very clearly what their chances are every step of the way, just as Julianne pointed out very nicely in the original post. And sorry @30, but the on-the-ground truth is that about the only people that grad students are willing to hear this from are their advisors or other faculty mentors. So regardless of whether or not it’s their “fault,” the current culture means that faculty are the only ones who can really do this.

    Regardless, it’s completely contrary to the ideal of education to say “You can’t learn here because we’re not willing to make you leave eventually.” And slamming the gates shut at the PhD level ends up causing major diversity problems, as well. It’s more problematic at the postdoc level, though, for the reasons @40 pointed out quite clearly.

    Something important is missing from most of this thread, though: teaching. Many astronomy graduate students and even postdocs I know who would consider or even prefer a teaching-oriented job if they can’t get an R1 slot… But they’re smart enough to know that being an adjunct is as crappy as being a graduate student (in the employment sense). So an alternative tack is to make teaching jobs more appealing: make lecturers 1st-class academic citizens! Many (most?) research-oriented professors would much rather have someone else teach many of the classes anyway, and then both groups could focus on what they are better at…

  • Phillip Helbig

    “- Learn to program in C++ and Java”

    I heard about one job applicant who put such skills on his CV and was asked at the interview “Why do you list these ancient programming languages on your CV?” :-)

    I’m a Fortran man myself. :-)

  • Phillip Helbig

    Whether one wants to put more money into academia is another issue. For a given amount of money, what is needed is a) more permanent jobs and less temporary jobs, which means b) deciding who stays and who leaves will happen earlier. In the case of almost all people who ended up with a permanent job based on merit, their superiority was evident already at the time of their first degree. Also, more temporary jobs should be tenure-track jobs.

    Why doesn’t this happen? Primarily because it means more longer-term commitments on the part of those funding jobs.

  • Phillip Helbig

    While it might be difficult for academics to adjust to life in the non-academic world, some folks have it even worse:

  • anonymous

    An ironic comparison: the last two sentences of #14, and the claim of #43 (or the rest of #14)…

  • Pingback: More on jobs & Ph.D.s | Gene Expression | Discover Magazine()

  • realta fuar

    I’ve always told my undergrads that a) you should only go after a ph.d. in astronomy if you don’t think you can be happy with yourself if you don’t and b) you’re probably not going to get the research job you think you want NOW. If that leads to well adjusted “gentleperson astronomers” (who have also followed Stephanie’s superb advice @42) then both they and the world in general are better off than they would be otherwise.

  • Lab Lemming

    One thing about lab-based chem and bio PhD’s is that in many cases, the PhD positions are funded by industry specifically to provide low cost R&D labor. I don’t know if there is an equivalent class of drudge student in Astronomy, but it is not surprising that people who are just there to be used don’t have great future prospects.

  • Penguin

    I think Tommy hit the nail on the head:

    “Have an actual department specific career services advisor, with actual contacts in interested sectors like tech, finance, consulting, private labs etc. Help students and postdocs make these transition.”

    Universities expect to help undergrads get jobs, but they have nothing for grad students and postdocs. Grad students and postdocs are the backbone of research productivity for minimal financial compensation. The least universities can do is offer genuine career support to those who want or need to leave the academic system. This support will never come through faculty — faculty never left the system so they don’t have the right connections or expertise.

  • James

    Of course, if you’re a physics PhD whose closest approach to programming was just plugging things into mathematica, and has never done the slightest bit of data analysis in your life, then you’re no good to anyone. Being told at finance (and other) interviews that you’re great at maths, but useless at the rest, gets a little tiring.

  • Dan Schroeder

    Actually there was a “golden era”, at least for physics and presumably for astronomy. From about 1870 until 1970, the number of physics PhDs awarded in the U.S. grew exponentially with a doubling time of about 10 years. During the era of rapid exponential growth of academic physics departments, the jobs-to-PhDs ratio was much higher than it is now (though still not 1-to-1, of course). The average physicist at a PhD-granting institution, over the course of a career, had to train about 10 “replacements” who would go on to become academic physicists at PhD-granting institutions.

    But exponential growth is not sustainable. The golden era ended about 40 years ago, and today we live in the era of sustainability, when the average physicist at a PhD-granting institution needs to train only one replacement. That’s a fundamental change from the way things used to be, but unfortunately, many of the attitudes from the past era still persist.

    I learned all this from a talk by David Goodstein in 1999:

  • Nicholas Suntzeff

    As Vice President of the American Astronomical Society, we on the Council are trying to get this issue into wider view in the AAS. Our Employment committee is actively recruting speakers from non-academic positions to show the process of transferring PhD skills into industrial skills. What are the main advantages that PhD astronomers should be aware of in their use to industry? What skills should they work on in their PhD or postdoctoral program in order to make them more employable?
    We will have a workshop on this issue at the 1/2013 AAS meeting in Long Beach and I encourage anyone interested in participating to come and meet with non-academic astronomers and learn some of the keys to their successes. You may want to send your thoughts to the Employment Committee of the AAS, which is a very active group trying to help create AAS policy to make astro jobs more attractive to industry and government.

    One thing I disagree with Julianne is that lab-based experience is not useful in getting jobs. You will find from the companies that hire astronomers that the lab-based experience is VERY useful. In our astro labs, we do not learn very specific tasks. If you are going to build a spectrograph, you need to learn LabView, python or Perl, C++, maybe Java in software, and learn to talk to instruments via your laptop to do motor control. You learn about fibers, gratings, lasers, termal control, … The breadbox way in which we design and prototype instruments is a vital skill needed in industries that do not do pure software, and there are very few places students are trained in this.

  • I.P. Freeley

    @Stephanie41, @Walter34
    I know Sean Carroll has blogged similar concerns about limiting grad school admissions in the past (i.e., PhD birth control vs abortion via job market). It seems like splitting hairs to some extent–are you really OK with grad schools rejecting 85% of applicants, but 92% is just too far? Would it make that much difference if grad school admissions transitioned from hyper-selective to ultra-selective? One obvious compromise would be to expand masters programs, so you can accept the same number of grads per year, but only fund a fraction beyond 2-years. In any case, I think the first priority should be killing off postdoc positions. I guess I’m in favor of limiting grad admissions, but I could be swayed away from that easily.

    I am curious how medical schools have managed to not suffer a bubble burst. It seems like the biological sciences massively overproduced PhDs in the late 90s, while physics, astronomy, and law all burst in 2008. How has the job market for MDs stayed so strong? Are med schools intentionally limiting the number of admissions? Are the aging baby boomers driving expansion? Do I just have to wait 5 more years and then the med school pipeline will burst too?

  • AcademicLurker


    I believe the original post was referring mainly to lab based biologists and chemists. IMO, I think she’s correct. Doing synthetic chemistry involves a very specific set of skills that don’t translate readily to doing anything else. Ditto for culturing cells and doing Western blots.

    Slightly related to the OP, is Astronomy as prestige driven of HEP theory seems to be?

    I recall seeing some data on another physics blog a few years ago indicating that about 10 graduate programs are responsible for essentially all of the TT job placements in HEP theory.

  • Pingback: The Finch and Pea()

  • Pingback: Being a science career realist | The Finch and Pea()

  • Neal J. King

    It would seem that APS could usefully operate a “physics outplacement” center, to provide the employment market for PhDs and postdocs that universities routinely provide for undergraduate seniors.

  • Pingback: Links 7/10/12 | Mike the Mad Biologist()

  • Sarah

    A few thoughts from the middle of the fray, many along the lines of what others have said.

    1) It would be helpful to have departments support the idea of external work. Some already do, and some may pay it lip service but emotionally punish those who float ideas outside of academia. Knowing the numbers, we should be supportive and realistic – not everyone who goes to grad school, or finishes a PhD, is going to stay in academia. And “staying in academia” comes in lots of flavors. I think that the status of research scientists is improving/stabilizing – but again, it is very dependent on institutions. And others have mentioned the disrespect towards teaching and other activities. Attitudes could certainly improve.

    2) Medical schools throttle hard at the med school level, to keep the demand balanced, and salaries level. To be honest, I think we’d lose out by this sort of throttling. The skills required in graduate school and throughout academia are often not those that get judged around an admissions committee table. It seems like an awful lot of the people who drop out early-ish in grad school do so because… they were great at the schooling side, but the independent research is not what they had imagined. Depth in admissions keeps our field full of diverse and interesting people who find their niche in a lot of surprising places.

    3) We should probably issue everyone a copy of “Oh, the places you’ll go!” by Dr. Seuss. Sometimes I feel like in all our fretting about the job market we forget – life isn’t such a certain thing, and academia isn’t so special when it comes to job stability, methods of hiring, or the fact that people are actually involved. Sometimes I think we forget that faculty jobs aren’t for everyone, that it is *not* first past the post, and that life is not always fair (whatever that means). Could we do better? Certainly. Should we talk about it? Absolutely. Should we tear out our hair about how we’re letting down the youths and taking terrible advantage? Perhaps we can hold off on that just a smidge longer.

    We are all grownups here (even the grad students), and anyone who can’t do the cost benefit analysis probably deserves the three postdocs waiting for them ;>

  • Shawn Kolitch

    A few of you have made comments about law and lawyers. I have a PhD in theoretical gravitational physics from UCSB and I am now a partner in an intellectual property law firm in Portland, Oregon. Just a few weeks ago I gave a colloquium to the UCSB physics department about intellectual property law as a career option for physicists.

    My job is extremely interesting, as I deal with the cutting edge of just about every conceivable area of technology and I am constantly required to become an expert in fields that are new to me. For this reason, intellectual property law is a field where a science or engineering background is not just desirable but legally required, and the deeper your science background the better. A physics background is particularly useful.

    I did not attend a first tier law school, yet I have advanced quickly in my career as an attorney and I feel that I could get a job just about anywhere, partly because the number of physics PhDs with law degrees is so small. However, some jobs for physicists at intellectual property firms, such as patent agent, technical consultant or technical writer, do not require a law degree. I encourage any physicist looking into alternative careers to consider intellectual property law, regardless of your interest in attending law school. However, I will add that I personally loved law school and that scientists generally make very good law students.

  • Lisa

    This is a fascinating thread of comments, and I appreciate all of your insights. I come at this from the perspective of a homeschooling parent of two bright, science-oriented children (one in physics, one in biology) who already envision a life of research and academia in the science world. Of course, things could change dramatically between now (at their ages of 10 & 12) and when they graduate, but it is interesting to get some perspective on what their future might hold if they persist with their current passions and plans.

    It seems to me that one of the problems is the pressure on so many young people to attend university at all. We have decimated our manufacturing and skilled labour jobs in the western world (I’m in Canada) and have created a situation where virtually everyone believes they need to attend university to be employable at all (and they’re probably right). Unfortunately, many of these people are ill-suited to post-secondary education, as evidenced by the number of remedial courses needed for basic math and essay-writing skills at this level.

    This, in turn, creates the situation where employers seeking a bright/educated work force won’t even look at a candidate without a masters degree, and many people who would never have gone beyond a BA in earlier generations now stay in school longer. Maybe I’m romanticizing the past, but it used to seem (25 years ago) like only those with professorial ambitions stuck around to do their PhDs and they generally found their place in academia when they’d put in their time.

    It seems a shame to have so many bright young people jump through hoops for years and years to set themselves apart from the masses, only to be stuck in low-paying, low-prestige post-doc positions that never lead where they had hoped.

    As a former trial lawyer myself (and married to a corporate lawyer), I had always intended to disuade my kids from that choice of career, but I appreciate reading Shawn Kolitch’s comment, above, about how he married the two fields (although from the perspective of Future Tuition Payor, I confess it frightens me a bit).

    Anyway, I had imagined that my kids would have little trouble pursuing their dreams with the right support system, encouragement of their passions, and their natural talents, so thanks for disabusing me of that notion. The better approach might be to ensure their toolbox of skills is diverse and readily adaptable, and to keep their bedrooms ready for them when they leave for university, as they’ll probably be back.

  • Lisa

    *dissuade* — sorry for not proofreading better!

  • Phillip Helbig

    @63: Why homeschooling?

  • truly anomalous

    To Moose (who wrote)

    “Thankfully, I’ve got a good few transferrable skills that secured me a position – but one bugbear is the fact that academia tends to use programming languages that are not used in the commercial world (Fortran, anyone?)”

    I’ve got good news for you. The commercial world (at least the part of it I know) does use Fortran and for good reason: nothing better for mathematical calculations!

    But I’d recommend that in addition learn C++/Java. My philosophy is to use the best tool (language) for the job at hand, as not one programming language is the best fit for everything. Do the heavy-duty computations in Fortran, then build a GUI on top of it using C++ and/or Java, for example…

  • Lisa

    @Phillip Helbig — “Why homeschooling?” When our kids were in school, it seemed that they did most of their real learning at home, as they were always researching things (usually science-related) and following interests (organizing their natural history collections and finding things to add to them was a particular favourite at a young age). Unfortunately school took up the best hours of the day, and wasn’t really productive for them, other than as a place to visit with other kids. We cut out the middle man and now let them learn at home, at whatever level (generally much higher and deeper than school) and on whatever topics they need or want. They still get to hang out with friends, but it isn’t the central focus of their day and they are much happier now. Thanks for asking!

  • Pingback: Academia vs. the Private Sector for Science Jobs | Technoccult()

  • Unemployed

    This type of propaganda really needs to stop. The job market for physicists is truly dismal right now, and the 1-2% figure is certainly flawed. The same article you mentioned points at that only 4.6% of chemists are unemployed. But among recent graduates that number is 62%. The same is almost certainly true in physics. Speaking from personal experience, after doing two prestigious postdocs I have now been unemployed for a year. Over 50% of the physicists in my group are now unemployed 4 years after the PhD. And the typical physics fallback jobs (quants, tech companies, etc…) are not hiring like they used to.

    We really need to face this problem.

  • Pingback: The Scienceblogging Weekly (July 13th, 2012) | Stock Market News - Business & Tech News()

  • Alberto

    Here is a cheap yet elegant solution. Each department should simply make public the statistics about their alumni. What fraction of alumni got faculty positions? Just by knowing this, graduate applicants will make much better informed decisions about whether to apply and pursue a career. Why these stats are not normally available (this is a rethorical question)?

  • Anon observer

    The problem with discussions like this is that the people who have the power to change the structure have absolutely no motivation to do so. But with funding headed in the direction in fields like astrophysics it is I suspect sooner rather than later young faculty in some places will have trouble pulling in enough grants to justify their existence. *That* is when it will go beyond people in ivory towers paying cheap lip service.

    Both the science and the associated funding structure of astrophysics is heading more towards the life sciences model of big projects with fewer people holding the purse strings. And some of the same problems are inevitable. Young researchers are able to direct less of their own research in Astronomy than they used to and be less independent.

    I do not mean to pick on Julianne, but when I read the bit about junior researchers with independent projects the first project that naturally came to mind is a huge multi cycle project to map M31 directed by her. Useful no doubt, but it does suck up a fair bit of HST time, and there are severe restrictions on observing sources in the vicinity each year. Inevitably it will impact quite a few smaller projects that might have trained some young and maybe not so young researchers to direct their own projects. Again not picking on the blogger. But this is the biology model. One PI on a large project. It is the trend across the board in astronomy. I don’t see the beneficiaries clamoring to give up this benefit for the good of the field. Neither should they! But it would be good to be frank about the realities instead of the widespread sense of superiority over the biological sciences.

  • Julianne

    Anon observer — It’s definitely true that astronomy has more large bio-scale projects than it used to, and it would be a bad thing for the field if that’s all there were. But, I think we’re very far from that point. There is clear value in having -some- number of projects that tackle problems that are too large for small individual investigation — even more so when they leave behind enormous legacy data sets with high archival value (SDSS, GOODS, WISE, and hopefully the M31 obs). So, I’d argue that a mix is actually not a bad thing in terms of the science. It’s not obviously a bad thing for young researchers as well — these larger projects have allowed junior scientists to launch their careers with high impact results. However, the risks are also there, as plenty of larger projects have chewed up junior scientists as slave labor and spit them out.

  • Phillip Helbig

    I agree; one needs a mix. If anything, I see a problem in the other direction. With short-term grants the only source of funding, and even no tenure in some places which used to have it, it becomes difficult to do long-term projects. However, as Paul Schechter said, surveys are the life-blood of astronomy. Many surveys are used for much longer and for different things than they were originally planned for.

  • Anon observer

    I agree about the mix. Sound scientific reasons to do both. That’s the calorie free soundbite that everyone can agree on.

    But it is trending towards larger programs, and not all of the reasons are scientific. Larger programs are cheaper to administer per unit time, and funding crunches push it that way. Also, many senior people in the field were never happy about sharing resources when this explosion of small programs and short term grants came about. They have always pushed in the direction of larger programs, and still are. I remember one interesting case in which a major facility polled its user base to ask if there was a demand for larger scale programs. The answer was there wasn’t. But a little later such programs appeared anyway.

    Somewhat similar situation with Kitt Peak which might shut down due to funding reasons. But there has always been a vocal subset at universities with private facilities that have never wanted it to exist and have worked towards that goal.

    We can talk about training people to make them independent but we are also taking away the resources to do so. In many cases we are creating data slaves. And nearly all of them were pretty good at it out of grad school. This useful post-doc training bit is largely a self-serving myth. It is more and more like biology where people who worked in labs in grad school are getting “trained” further to work in labs for the benefit of fewer and fewer senior people. I am not offering an opinion on the morality or scientific merit of it, just a criticism of the lack of transparency about it.

  • From Physics To Finance

    As a recent graduate who successfully transitioned from Physics to Finance, here’s my $0.02 for grad students:
    1. The job market is bad in general, not just for Physicists.
    2. If you plan on leaving academia, do so immediately upon graduation. It becomes harder once you’re a postdoc.
    3. Develop other useful skills much before you graduate. Eg: Join a programming club, take a couple of courses in your university’s business school, apply for internships etc.
    4. Physics/Astronomy research *does* give you transferable skills, but it’s your job to sell yourself and market these skills to employers. Prepare a good resume that highlights skills employers might find useful

  • Phillip Helbig

    @76: Good advice. Remember, even Newton went from physics to finance. :-)

  • Another Anon

    Julianne, I’m with Anon Observer 72. Yes, we need a mix of both large and small projects, but the trend is that the large projects are quickly starting to suck up more and more time on our facilities. And I’m sorry, there’s no point in you trying to stand up to that point because you ARE the PI on one of those big projects. Look, now HST is trying to justify going for even larger projects, even when the oversubscription rate is 8!


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Cosmic Variance

Random samplings from a universe of ideas.

See More

Collapse bottom bar