Yes, science produces too many PhDs

I came across a blog post this morning entitled Does Science Produce Too Many PhDs? I think the answer is an obvious “yes” but I’ll use the question as an excuse to rehash an argument I have presented before, which is that most analyses of the problems facing yearly career researchers in science are looking at the issue from the wrong end. I think the crisis is essentially caused by the overproduction of PhDs in this field. To understand the magnitude of the problem, consider the following.

Assume that the number of permanent academic positions in a given field (e.g. astronomy) remains constant over time. If that is the case, each retirement (or other form of departure) from a permanent position will be replaced by one, presumably junior, scientist.

This means that over an academic career, on average, each academic will produce just one PhD who will get a permanent job in academia. This of course doesn’t count students coming in from abroad, or those getting faculty positions abroad, but in the case of the UK these are probably relatively small corrections.

Under the present supply of PhD studentships an academic can expect to get a PhD student at least once every three years or so. At a minimum, therefore, over a 30 year career one can expect to have ten PhD students. A great many supervisors have more PhD students than this, but this just makes the odds worse. The expectation is that only one of these will get a permanent job in the UK. The others (nine out of ten, according to my conservative estimate) above must either leave the field or the country to find permanent employment.

The arithmetic of this situation is a simple fact of life, but I’m not sure how many prospective PhD students are aware of it. There is still a reasonable chance of getting a first postdoctoral position, but thereafter the odds are stacked against them.

The upshot of this is we have a field of understandably disgruntled young people with PhDs but no realistic prospect of ever earning a settled living working in the field they have prepared for. This problem has worsened considerably in recent  years as the number of postdoctoral positions has almost halved since 2006. New PhDs have to battle it out with existing postdoctoral researchers for the meagre supply of suitable jobs. It’s a terrible situation.

Now the powers that be – in this case the Science and Technology Facilities Council – have consistently argued that the excess PhDs go out into the wider world and contribute to the economy with the skills they have learned. That may be true in a few cases. However, my argument is that the PhD is not the right way to do this because it is ridiculously inefficient.

What we should have is a system wherein we produce more and better trained Masters level students  and fewer PhDs. This is the system that exists throughout most of Europe, in fact, and the UK is actually committed to adopt it through the Bologna process.  Not that this commitment seems to mean anything, as precisely nothing has been done to harmonize UK higher education with the 3+2+3 Bachelors+Masters+Doctorate system Bologna advocates.

The training provided in a proper two-year Masters programme will improve the skills pool for the world outside academia, and also better prepare the minority of students who go on to take a PhD. The quality of the  PhD will also improve, as only the very best and most highly motivated researchers will take that path. This used to be what happened, of course, but I don’t think it is any longer the case.

The main problem with this suggestion is that it requires big changes to the way both research and teaching are funded. The research councils turned away from funding Masters training many years ago, so I doubt if they can be persuaded to to a U-turn now. Moreover, the Research Excellence Framework provides a strong incentive for departments to produce as many PhDs as they possibly can, as these are included in an algorithmic way as part of the score for “Research Environment”. The more PhDs a department produces, the higher it will climb in the league tables. One of my targets in my current position is to double the number of PhDs produced by my School over the period 2013-18. What happens to the people concerned seems not to be a matter worthy of consideration. They’re only “outputs”…

35 Responses to “Yes, science produces too many PhDs”

  1. Phillip Helbig Says:

    As usual on this (and almost all other) topic(s), I agree completely.

    The solution (for a fixed amount of spending) is obviously to have more permanent jobs and fewer temporary jobs, with people getting a permanent job earlier in their career. This would help solve another problem, namely that some people leave the field not because they are not good enough but due to lack of job security. Who is the last person you know who is not independently wealthy and who has children who got a permanent job in astronomy? Even apart from the social aspect, it should be a goal for world astronomy to retain the best people, not (at best) the best people still around after 10–15 years on soft money.

    Once at dinner at an astronomy conference, I happened to be sitting across from a well known senior astronomer. The conversation turned to this topic (not on my initiative) and I remarked that there are good people who leave the field not because they don’t want to stay and not because they aren’t good enough, but due to lack of job security (note: in almost all cases, not lack of money; even if they make 3–5 times more after leaving academia, even if lack of money is the problem, 20 per cent more would be enough in most cases to solve it). He looked around the dining hall and said “Where are they? I don’t see any here.” At first, I thought it was a joke, then I realized that he was serious. Of all people, an astronomer should know how to take selection effects into account. (Yes, there are a few people who still show up at conferences when they are no longer employed in academia, but it is a very small number.)

    Another disadvantage is that people are trained for years then leave, then new ones are trained, partially to do the same things. Very inefficient.

    Why won’t this happen? One reason is that short-term funding means short-term commitment. Not necessarily good for science, but perceived to be good by the funding agencies. Another reason is that soft-money people are cheap and undemanding labour. Another is that it means that someone will have to make a decision as to who stays and who goes. With the survival-of-the-fittest scheme, this responsibility can be avoided to some extent.

    • What do you mean by “soft money”?

      • Phillip Helbig Says:

        It’s a common term (at least, I didn’t invent it and I have often heard it) to mean salary from temporary employmen, e.g. the typical 2-year postdoc appointment or whatever.

    • i would normally ask why are “more” permanent jobs the answer?

      like peter – i would think the better way to match supply and demand would be to reduce supply at an earlier stage.

      however – the ~3% per annum growth in the numbers of UK astronomy academics over the past 10+ years suggests that UK universities do agree with you (at least to some extent). although this leads to problems upstream with a lack of equivalent growth in research funding (which might undercut the university’s model of how they expected to pay for these posts).

      as it is, i don’t think our new graduate students are so naive that they think they should expect a permanent position, even if that is what they are after from a PhD. these are skilled and resourceful people, and the simple maths of the situation (as laid out by peter) isn’t beyond their grasp.

      • Phillip Helbig Says:

        “i would normally ask why are “more” permanent jobs the answer? “

        I am assuming that, for a given amount of money, the goal is to maximize the amount of “astronomy done”. Any scheme will need some permanent jobs and some non-permanent jobs, mainly to determine whom to hire for the permanent jobs. I just think that the scales are tipped too far in favour of temporary jobs.

  2. Joern Wilms Says:

    I disagree. Your statement is correct if you assume that the only aim of a PhD is to pursue a career in academia afterwards. I do not think this is true, to pursue a PhD means to get a high academic qualification that is usable also outside of academia. Despite all hearsay to the contrary, the low fraction of unemployed PhDs means that this is seen similarly in those sectors of industry that hire (STEM) PhDs.

    I’m based in Germany, where this is especially true: The “overproduction” of PhDs is even higher than in the US or the UK (typical ratios of PhD students to advisors are more 5:1 to 10:1, so as a professor one can expect to graduate 50-100 PhDs during one’s tenure), but industry still asks us to graduate more PhDs and none of the (more than 10) PhDs I’ve advised in the past decade are out of a job. Only three of them are in academia… Granted, this is still hearsay, but given that my colleagues make the same experience.

    I think there is a problem in the sense that many academics think that if one doesn’t get a position in industry, one has essentially failed, but if you look at the bigger picture that many industrialized countries need a very highly educated workforce, I think there is no such thing as an overproduction of PhDs. Granted, this means that academics have to ensure a breadth of education and soft skill training that goes beyond pure research during the PhD and also to manage the expectations of people starting to do a PhD, but in my experience that is easily implemented and even helps the research. What is necessary, though, is a change in academic culture that accepts that doing a PhD is not the same as starting an academic career, but that sees doing a PhD as a continued increase in qualifications that are broadly applicable inside and outside of academia.

    • Phillip Helbig Says:

      Maybe a PhD is better than a lesser qualification, but perhaps it is still not ideal for “industry”. However, that is beside the point. The question is, if you have a fixed amount of money for astronomy, how do you best spend it? Sure, people with knowledge required by industry are needed, but why should university budgets have to pay for them getting this knowledge? This reminds me of the “recruit the best heads” approach, i.e. let some other country pay for education and then we’ll hire them (perhaps at a lower salary than a native would demand).

    • Phillip Helbig Says:

      One could argue that if someone doesn’t understand the calculatons Peter mentioned, he gets what he deserves anyway. Still, academia relies to a large extent on the hope of young people to stay in academia.

      One might argue that a PhD (or othe doctoral degree) is useful outside of academia, but the same argument probably doesn’t work for a string of postdocs.

    • telescoper Says:

      I have heard this argument often, and disagree with (most of) it. Not all research is in academia, and there is some demand for people with PhDs in industrial research. In my own field of astronomy, however, I think the number is fairly small. As for non-research jobs I think properly trained Masters graduates would do just as well in such jobs. In other words we produce my PhDs who end up in jobs in which the very specific skills they need. The fact that few have difficulty finding employment does not provide evidence against my statement as employers will always go for the employees with the highest qualifications whether they are necesary or not.

      I think the PhD should be a training for potential researchers (either inside or outside academia). The vast majority of PhD students in the UK system enrol because they want to have a career in academic research despite the fact that the odds are heavily against this being possible. The result is a large pool of disgruntled PhDs and very low morale. At the very least we should be more honest with our graduates about what the PhD means, and more importantly what it does not mean, for their future careers.

      Of course in the UK the entire system of postgraduate education is a complete mess, so we don’t have sufficient highly-trained Masters students.

      • Phillip Helbig Says:

        “As for non-research jobs I think properly trained Masters graduates would do just as well in such jobs.”

        Objectively, this is true. They might even be objectively better. Subjectively, some employers are willing to pay more for someone with a doctoral degree, because it looks good on nameplates and so on.

      • One important difference between the UK and Germany here is the amount of R&D done in industry. There is way more of this done in Germany (just look at the %age of GDP spent on R&D in industry in the two countries) so it can be argued that Germany needs to produce more PhDs to fill those jobs than the UK does.

        Of course UK industry should be doing more R&D, but that isn’t a problem that will be solved by continuing the overproduction of PhDs.

      • Phillip Helbig Says:

        Yes, there is more industrial research in Germany, but one can still ask the question whether a doctoral degree (there is no “PhD” in Germany) is necessary and, even if it is, whether the degrees of such people should be funded by “blue-sky” research funds. Of course, industry wants this: they get something and don’t have to pay.

      • Joern Wilms Says:

        I did not make this clear in my reply, I think, but I am an X-ray astronomer, so what I said above about my students specifically refers to people who did their PhD in astronomy (let’s not go into the different versions of a doctoral degree, in practice, there’s no discernible difference between astronomy/science degrees in the UK and, e.g., Germany).

        It turns out that the skill set taught in astronomy (analytical thinking, handling data sets, modeling of data, numerical modeling and simulation etc. etc. etc.) is widely usable. The specifics of the physics that’s done are not transferable, but that is not the point.

        From the old times of the diploma system, Germany has a long tradition of producing master’s level physicists (virtually all of our BScs continue to a Master’s), but there a master’s doesn’t show the level of ability of doing independent research that the PhD does. And this turns out to be a sought after skill.

        I agree with others that yes, the expectations of students starting the PhD differ from what realistically will happen to them, but that I think is mainly a cultural problem of managing expectations. The point I was trying to make is that the first thing that we as academics need to change is the expectation of people pursuing a PhD that this will automatically lead to an academic career. This is not the case, and this is also not why one should pursue a PhD in the first place. The reason why one should pursue a PhD is for the luxury of being able to do independent research, combined with the knowledge that the skills obtained during the PhD are very widely usable and the guarantee that a PhD will allow you to lead a much more interesting professional life in industry or academia than a BSc or a MSc.

      • “The vast majority of PhD students in the UK system enrol because they want to have a career in academic research”

        This is just plain wrong. It may have been right in your time or what you’re always hearing from interviewing candidates or from a supervisor’s position but I can tell you from a PhD who speaks to other PhDs that perhaps 1 in 10 actually want to pursue a career in academia, regardless of difficulty, and this is probably being generous. This isn’t to say they aren’t interested in science and doing a PhD, but come on most students burn out or realise they just don’t have the enthusiasm to really carry on in academia and just move on to other things and it has nothing to do with the job situation.

      • telescoper Says:

        Well, I base my statement on polls of new PhD students. I stick to my view that most start out (i.e. enrol) because they want to have a career in academic research. I won’t argue that fewer feel that way at the end of their PhD.

      • telescoper Says:

        PS. I don’t accept usually accept anonymous comments either. If you really believed your statements you wouldn’t hide behind a pseudonym and give a phony email address.

    • I decided to avoid European PhD programs completely because I heard about the EU-wide funding cuts, and professors in German-speaking countries seem to have too many doctoral students for these students’ good. I think that, to ensure that a given doctoral student will get the training they deserve, a good amount of attention must be given at both ends of a PhD by the supervisor (but for completely different reasons)

      On the other hand, about the suggestion of funding councils going back to funding masters education, one has to ask how Canada benefited from SSHRC/NSERC (in humanities/social sciences and science/engineering respectively) funding masters education…

  3. Could it not be argued that before they have completed a PhD, it is very hard to judge which students have the necessary skills to be successful in research? And that therefore from the point of view of getting the best researchers out at the end, it is worthwhile starting from as large a pool of talent as possible in the beginning?

    I’m not sure this argument is completely correct, of course. But I do have a suspicion that even after completion of the PhD, it is still hard to judge the worth of a researcher, because so many of the standard ‘signals’ we use to do this are actually rather dependent on the advisor (and luck) rather than the students themselves. (In fact I’m pretty convinced that getting one’s first postdoc is largely a matter of random chance …)

    The point being that making such judgements before a PhD might be even harder.

    • Phillip Helbig Says:

      If the decision is made earlier, it has to be made more objectively.

      In a programme where a master’s degree includes a full year of thesis work, I think this is enough to decide.

      It might be true that one can’t be sure after, say, 1 year of research work if someone really has what it takes, but one can probably say if someone doesn’t have what it takes. It would be better for such people to leave before continuing in academia, and it would be better for the better people since it will free up a position for someone who can do good, original work and not just be a robotic assistant.

  4. Adrian Burd Says:

    I agree with Peter completely on this. The same situation occurs in other scientific disciplines, such as the one I’m in at the moment. I have been involved in a couple of faculty searches over the past few years, and in both cases one could have staffed several complete good departments from the applicant pools. In one case, one could have had 2-3 world class departments and 4-5 very good ones. That’s just from one search!!!

    Perhaps I could offer another possible solution; combined degrees. There are many careers where a good knowledge of science is, at the very least, a desirable thing. One could think of law, journalism, etc. Combined Masters degrees could be a way of handling that, and at the same time, provide a solid career path for those who students who want to be involved in science but not necessarily follow an academic career in it. Such a structure would require careful thought, and a willingness on folks in multiple disciplines to talk to each other about the training that would be required to pull it off.

    In Marine Sciences in the US there seems to be a trend (possibly a successful one, it’s too early to tell yet) towards these non-traditional degrees.

    Just a thought……

  5. Phillip Helbig Says:

    Whether or not it is the best use of research money is another question, but I think that most people who leave academia are smart enough to realize (at least by the time they leave) that Peter’s calculation is correct that not all can stay and would have no big emotional problem with leaving, as long as they felt that they were leaving because those who stay are really significantly better. But when the average age for getting a permanent job is 40 or whatever, many other factors play a role in determining who stays and who goes, especially for those with children and/or not independently wealthy. I think this is the bigger problem, the “I could have been a contender” feeling. (This is independent of the much larger (morally) but smaller (in numbers) problem of people getting permanent jobs, even at prestigious institutes, due to reasons other than academic qualifications.)

    Again, it might not be the best use of research money, and if I were distributing my own funds I would do so in a matter which maximizes the scientific output, but most people who leave academia probably look back on it with some degree of fondness. (In my own case, Dickens said it best: It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness, it was the epoch of belief, it was the epoch of incredulity, it was the season of Light, it was the season of Darkness, it was the spring of hope, it was the winter of despair, we had everything before us, we had nothing before us, we were all going direct to Heaven, we were all going direct the other way….)

  6. I agree completely, and in fact was saying that the UK should move to a proper Bologna scheme with more funded Masters courses and fewer PhDs a decade ago – see the appendix I wrote as a ‘minority report’ to this document:

    Click to access careers_v12.pdf

    The other authors did not agree with my conclusion and I suspect an aspect of that is the self interest of overworked academics who want cheap labour to work with them on research. Now I am one I can see the problem, but I remain convinced that the system does not serve the career aspirations of our PhD students as well as it might given some restructuring. I suspect the same is true for the interests of industry.

    The response of PPARC to this report is also interesting. See discussion in A&G and links to the response itself here:;jsessionid=DA7A98F69DED8FE4D0D9DBBAE2E7A189.f01t04?v=1&t=i6c9wn8u&s=198083b173b01195ba426ac2c6d00cd9dde39363

    I suspect my position on this is why my attempts to get onto the STFC Education & Training committee have always been unsuccessful 🙂

  7. Reblogged this on Disturbing the Universe and commented:
    I agree with Peter on this, and have done for some time. See appendix 3, which I wrote, here:

    Click to access careers_v12.pdf

  8. Chris Chaloner Says:

    Lots of good discussion which will clearly go on…
    As someone who did a D.Phil and then moved into a public sector lab to do a job clased as “experimental science” but really systems engineering; and then into industry proper as a systems engineer, I can make several comments:
    1) going into industry is not a failure!
    2) in considering economic activity,don’t just thing about Research, but R&D as a whole – and the D needs typically at least an order of magnitude more funding than the R.
    3) the quality of development engineers in industry is universally high, whether they have an MSC or a PhD or neither – there is no “dead wood” – simply from commercial pressures, such people are removed very rapidly.
    4) the management structures in industrial R&D mean that in general the rate of progress is much higher in an industrial development than in a university.
    5) what we have lost is a consensus on what university researchers should do, and what industrial researchers and developers should do – so we have academics trying to hide any commercially valuable developments until the patents and spinout companies are in place – and because of the lack of understanding of the costs of development and therefore an inflated idea of the value of low-TRL developments, these new technologies often lose out to overseas competition.
    6) we definitely need a new start to postgraduate education – the Bologna criteria are very interesting. My sister ran an academic department in the US and always said that when it came to appointing faculty members, the quality of applicants from the UK was very poor compared to the US, the rest of Europe, and even the rest of the world.

  9. Iain Steele Says:

    I am pretty sure I made this comment last time you talked about this, but anyway…

    With your focus on UK academic appointments you forget about the very large number of technical and scientific staff working at observatories all over the world, building satellites at places like RAL etc. for whom a PhD is vital training.

    Also you seem to think you know better than industry what industry needs and wants – do you have any evidence they would prefer Masters level students, or is it just a prejudice from your ivory tower that they can’t possibly need a PhD to do “industry”?

    • Phillip Helbig Says:

      “With your focus on UK academic appointments you forget about the very large number of technical and scientific staff working at observatories all over the world, building satellites at places like RAL etc. for whom a PhD is vital training.”

      OK, each academic can produce 2 PhDs who stay on in academia (in the broader sense of the term), not 1. There is still over-production since most who don’t become professors or whatever don’t end up at the likes of RAL.

  10. Yes, of course, there is an overproduction of people with PhDs.

    The whole issue is complex and it is difficult to address many of the issues without going into depth, so I’m not going to try now.

    A central problem is that a core argument made for state funding of basic science is that universities train considerable numbers of people to PhD standard who then are in great demand by industry. That in turn creates a careers crisis in academic science. It creates a need to fund facilities to produce data to fund PhD students. It also leaves the funding of science dependent on politicians and civil servants continuing to believe that the flow of PhDs into industry justifies spending large sums of money. It’s a mess.

  11. […] side, but it is by no means a complete solution to the problem that, at least in the UK, we produce many more people with PhDs than are needed to sustain academic research and we need to think much more carefully about whether this route provides the correct career […]

  12. […] answer is a bit complex, but chances are that universities and industry prefer #2, while students (and/or recent grads without a permanent position) prefer […]

  13. […] chorus of bloggers and journalists have come to the conclusion that science produces far too many PhDs. One even […]

  14. […] be welcomed, but I do wonder about the wisdom of increasing PhD student numbers still further. As I have stated before, I think we already produce far too many PhDs. I think this money might be better spent increasing […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: