37

Multiple users on this site have stated that only around 1 in 10 (or maybe an even smaller ratio) PhD students are able to successfully continue in academia, beginning with an Assistant Professor position and working their way up towards tenure. (I am unable to find links for this, but I have read it in at least 5 different answers/comments on this site.)

For the purposes of this question, I am ignoring those who actively want to pursue a career in industry, and am focusing on those PhD students who wish to go into an academic career and fail to successfully do so. I wish to know why students fail to make the transition, and what exactly is the hardest part of this.

For instance, is it:

  1. The transition from a PhD position to a postdoctoral position: A postdoc has to work much more independently than a PhD student, and maybe people struggle with this.
  2. The next transition, from a postdoctoral position to an assistant professor position, or,
  3. Simply making a bigger impact on one's research field of interest, after securing a tenure-track position, and hence moving away from academia into industry.

In which part do newly graduated PhD students, who actually want to go into academia, "fall by the wayside"? Or, which of these transitions is the hardest to make, resulting in students not making the transition to a full academic career?

astronat supports the strike
  • 20,380
  • 6
  • 67
  • 91
  • 22
    The title is unclear to me, PhD students are in academia – Azor Ahai -him- Jun 08 '21 at 03:03
  • 3
    When I say academia, I mean a full-time career in academia, affiliated to a University, conducting research, handling classes, basically a professor position. –  Jun 08 '21 at 03:12
  • 2
    Are you saying that you didn't work full time while earning the PhD? Shame on you :-) – Jyrki Lahtonen Jun 08 '21 at 12:42
  • @JyrkiLahtonen Haha, I've not yet earned a PhD, I have only just finished my undergraduate degree :) –  Jun 08 '21 at 13:16
  • 6
    Highly suggest you read Lantsoght, The A-Z of the PhD Trajectory (available online), esp., Sec. 13.5.1: "Transitioning from PhD Student to Faculty Member". – Daniel R. Collins Jun 08 '21 at 13:17
  • 1
    @DanielR.Collins Thanks for the suggestion! Will definitely check it out. –  Jun 08 '21 at 13:24
  • 2
    "Hardest" is ill-defined. Maybe one could answer a question about transition rates. – user2705196 Jun 08 '21 at 13:41
  • 1
    This could do with an answer that looks at actual data: How many PhD students are hired per year in a given field, how many postdocs (including re-hires), and how many faculty. Preferably world wide. I expect that would confirm my experience that it's relatively easy to go from PhD to postdoc if willing to move world wide, but that going from postdoc to faculty is hard. – gerrit Jun 08 '21 at 15:52
  • @gerrit I think this thread provides a few statistics, but I am yet to read through it completely. It is also only for the transition from PhD's to tenured professors. –  Jun 08 '21 at 15:57
  • PhD students are (almost always) full-time academics. – Konrad Rudolph Jun 08 '21 at 17:49
  • Please note that the remaining 9 over 10 PhDs (in the STEM area) that did not find their way in Academia usually got a good position in
    • research institution
    • public employee and similar etc
    • companies
     – EarlGrey Jun 08 '21 at 18:48
  • @EarlGrey Absolutely, I do not mean that these people have "failed" in any manner, I just mean that they have not made the transition, and that among them, there will be people who originally intended to go into a full-time career in academia. –  Jun 09 '21 at 00:51
  • 2
    The hardest part? More candidates than available academic positions. And departments have politics of who to hire. It is not always the best candidate. – stackoverblown Jun 09 '21 at 13:52

7 Answers7

63

Most PhD students will not have academic careers. This is not because the duties of academic careers are "hard." Academic careers do involve hard duties. But so do non-academic careers.

Most PhD students will not have academic careers for economic reasons. There is simply not enough demand to employ more academics. The economic factors have the biggest impact on those who seek so-called permanent positions. Permanent jobs require a long-term financial commitment from the employer, and there is very little demand for making that sort of commitment.

Anonymous Physicist
  • 98,828
  • 24
  • 203
  • 351
  • 38
    Yes, exactly this. The hardest part of transitioning from PhD to academia is simply finding a job in the first place. – mathkb8 Jun 08 '21 at 02:13
  • Isn't there demand for new professors who can make an impact on the research output of a department/University? Is it really that hard to find a department with open, tenure-track positions? –  Jun 08 '21 at 02:20
  • 20
    @RamPadmanabhan It is easy to find one department with an open tenure-track position. You might find tens of open positions. But in most fields of research there will be hundreds to thousands of qualified applicants, and tens of thousands of qualified people who don't bother applying because there is so much competition. – Anonymous Physicist Jun 08 '21 at 02:41
  • 3
    The money has to come from somewhere. – Anonymous Physicist Jun 08 '21 at 02:42
  • 6
    @AnonymousPhysicist So, in a sense, it is the transition to a tenure-track position that is hardest, because there are so many applicants, all of whom are well-qualified? –  Jun 08 '21 at 02:46
  • 10
    I would not say that; there are plenty of other careers where there are many applicants. It is the low demand that stands out. Pick any professional sports league in the US. The number of athletes one league hires in a year is greater than the number of tenure track physicists hired in one year. – Anonymous Physicist Jun 08 '21 at 02:53
  • 4
    Also, keep in mind that many advertised "open positions" are never filled; an open position is not necessarily a real job opportunity. – Anonymous Physicist Jun 08 '21 at 02:54
  • 4
    The competition is simply brutal. And it’s not clear, at least to me, that society would be better served otherwise. – A rural reader Jun 08 '21 at 02:54
  • 8
    It's easy to see what @AnonymousPhysicist is saying. Look at your department. How many tenured or tenure track faculty are there. How many PhD students, and how many graduate each year. That ratio (faculty to students) represents a slice of the market realities in your domain. – Flydog57 Jun 08 '21 at 18:45
  • 4
    Here's a particular datapoint: when I was applying, ~6 years ago, I got a polite and appreciated rejection email that stated, in part, "it was very difficult to pick among the 700+ qualified applicants". Mid-tier comprehensive (USA), not particularly specific (that I recall) subarea mathematics, tenure-track of course. For that one seat. From some sort of ID on their system, their numbers aren't fake (+ they have no reason to). – obscurans Jun 08 '21 at 22:19
16

In addition to the point that @AnonymousPhysicist makes that there are far fewer postdocs positions available than there are PhD students, and far few faculty positions available that there are postdocs, there is also the up-or-out mentality: after a certain number of years as a postdoc, many people will start to think you are past it - that if you were any good, you'd already have a faculty position by now. This means there is really a limited amount of time you can spend as a postdoc looking for a faculty position, even if you were willing to stay as a postdoc.

Ian Sudbery
  • 38,074
  • 2
  • 86
  • 132
  • 7
    "after a certain number of years as a postdoc" That number of years varies a lot by field. – Anonymous Physicist Jun 08 '21 at 09:12
  • 2
    @AnonymousPhysicist but is absolutely true (rightly or not) -- especially if you're a postdoc at the same institution for a number of years. When I started postdocing, I remember a mate advising me not to stay longer than 2 years at any position, as "after 2 years, you will benefit as much as you can from them. They can still benefit from you working there, but the benefits of you start going down". – penelope Jun 08 '21 at 11:53
  • 1
    @AnonymousPhysicist Yes. This will vary by field (and by supervisor) - I have one colleague who is very much against hiring anyone who has already done a postdoc somewhere. In my part of biology that is somewhere between 2 and 3 postdocs. – Ian Sudbery Jun 08 '21 at 11:55
  • 3
    Here, in Germany, there is a hard limit on the number of years (12) that you can work on temporary contracts in academia, so you literally cannot stay as a postdoc. I'm not sure whether similar arrangements exist in other countries? – Jack Aidley Jun 08 '21 at 12:17
  • 3
    @penelope That's not good advice if you work in a laboratory field where each position uses different, complex instruments. When you move you must retrain. It is even worse if you spent two years building an instrument and then move, leaving it behind. – Anonymous Physicist Jun 08 '21 at 12:21
  • 1
    @JackAidley I think the US DOE has a 3 year rule. But that is another question. – Anonymous Physicist Jun 08 '21 at 12:21
  • 1
    "there are far fewer postdocs positions available than there are PhD students", citation needed. My experience is the opposite, that there are more postdocs than PhD students. – gerrit Jun 08 '21 at 15:57
  • 1
    @JackAidley In US, Canada, and UK, I've known people who spent more than 30 years on postdocs. I've known postdocs to retire due to old age. So such rules certainly do not exist everywhere. – gerrit Jun 08 '21 at 15:57
  • @gerrit Numbers of postdocs is hard to find, but this: https://www.statista.com/statistics/185167/number-of-doctoral-degrees-by-gender-since-1950 says around 188,000 graduate every year in the US, so assuming it takes 5 years to complete a PhD, that around 940k PhD students. This link: https://www.postdocacademy.cam.ac.uk/aboutus/postdocs-in-the-uk-higher-education-landscape suggests 80-100k postdocs in the US. – Ian Sudbery Jun 08 '21 at 16:49
  • This tallies with my (mol biol) department - we have 84 PhD students and 25 Postdocs – Ian Sudbery Jun 08 '21 at 16:54
  • @gerrit: Yeah, I'm from the UK so I knew it wasn't universal, but I'm not sure whether it's a quirk of Germany or quite common around the world. Sorry for being unclear. – Jack Aidley Jun 08 '21 at 17:13
  • @gerrit How could that be mathematically possible, unless the average time each PhD entrant spends in postdocs is longer than the time spent in PhD (wildly untrue, since many PhD students do not undertake any postdocs) or the number of PhD students overall has sharply decreased since the time the current population of postdocs was in school (not true, at least in fields I'm familiar with)? – Kevin Carlson Jun 09 '21 at 03:13
  • 1
    @JackAidley on paper, yes, 12 yrs is the limit. In fact, however, there are a few exceptions to that rule, and I know people who's been on temporary contracts for longer. – sleepy Jun 09 '21 at 08:58
  • 1
    @KevinArlin In my experience, the average time people spend in postdocs is much larger than the time they spend in PhDs. Like I've said, I've seen people being in postdocs for more than 30 years. Here I use "postdoc" loosely to include any time-limited research position, even if the job title is not "postdoc", because the role and job situation is essentially the same (just possibly a bit more senior and a bit better paid). – gerrit Jun 09 '21 at 10:16
  • 1
    @gerrit - this is incredibly discipline and country specific. In mol biol in the UK, people are PhD students for 3.5-4 years, and postdocs for 6-10 years (after 10 years you are often deemed "past it", and the only way to stay in academia is to stick with one supervisor and hope they don't have a gap in funding before they retire), in the US thats different because PhDs take 5-6 years. But even then we have 3 times more PhD students than postdocs. In other subjects people do 2-3 years of postdocs, and in others most go direct from PhD to faculty. – Ian Sudbery Jun 09 '21 at 11:07
  • @IanSudbery - If PhD students stay 5 years, that is ~16 students per year. If postdocs stay 2 years, that is ~12 postdocs per year - not that far off in reasonable probability of a PhD student getting an academic postdoc. Then there are postdocs in other non-university settings. – Jon Custer Jun 10 '21 at 17:08
  • @JonCuster Our PhD students stay 3.5 years, and our postdocs 3. – Ian Sudbery Jun 10 '21 at 17:35
12

In my field (cond-mat physics), there are really two main reasons. First, a lot of people want a job in the industry, and they leave after a PhD. Then, if you want to find a postdoc job, chances are that you'll get it rather soon: there are many open postdoc positions (in Europe, at least) and it is usually a problem to find a candidate to fill it, not vice versa. Second, the main bottleneck is between the postdoc and the adjunct or Jun.-Prof. level. Those jobs are scarce.

sleepy
  • 1,795
  • 8
  • 16
  • 2
    Yes. This might be field dependent of course, but I would say that the majority of PhD students do not stay in academia mainly because they do not want to. They have seen academic research from the inside for long enough to decide that it is not the right fit for them. I've seen this realization with many people roughly half way through. Some of them stop right there, while others finish the PhD to improve their C.V., but none of them even try to apply to post-doc positions. – mlk Jun 08 '21 at 09:01
  • 1
    Also, there are quite a few research institutes in Europe. People have a third option besides academia (as in being employed by a university) and industry. –  Jun 08 '21 at 09:05
  • 1
    You have confused "adjunct" with "assistant." – Anonymous Physicist Jun 08 '21 at 09:11
  • 1
    No I haven't. But you could also include Asst. Prof. to that list. – sleepy Jun 08 '21 at 11:37
  • +1 for mentioning that most are choosing to leave. – Jack Aidley Jun 08 '21 at 15:29
  • 6
    @mlk The fact that many do not want to stay in academia is not entirely unrelated to the fact that it's very hard to find a permanent job in academia. – gerrit Jun 08 '21 at 15:46
  • 6
    @gerrit Certainly, and the two are hard to separate. But I know more than one person who was good enough to likely have made it in academia, but still decided against it. One classic recurring theme seems to be that many people begin to think about starting a family around the age where they finish their PhD. And suddenly a career that will likely involve several long distance moves, frequent travel and long hours doesn't sound so appealing anymore. – mlk Jun 08 '21 at 16:40
11

I doubt there is a common answer across all disciplines. I'd highlight a few challenging inflection points, whose relative painfulness will vary.

1a. Being genuinely self-driven (work ethic). During a Ph.D., one is meeting frequently with an advisor, who may themselves have deadlines. There is at least annual reporting of progress to stay enrolled (varies a bit). So while a lot of Ph.D. students struggle a bit to buckle down and get stuff done, there's still a fair amount of regular external pressure. A post-Ph.D. but pre-tenure academic faces that less frequently, so may not make enough progress to make the cut to the next stage.

1b. Burnout, the flip side of 1a. Many junior academics, rightly or wrongly, always feel time is breathing down their necks and just find the number of hours they feel they need to put in is not worth it. This is doubly so if their personal lives deserve time too!

  1. Continuing to find interesting but answerable research questions. This is what - in many disciplines - the post-doc stage is supposed to help bridge, but regardless can be challenging. Aim too high and you don't get publishable results. Aim too low and your career becomes a yawn.

  2. Building teams and getting funding. More relevant for high-capital disciplines like experimental science, but also others. Even conference travel takes $, and you need to learn how to ask for it and get it!

And then there are the more pragmatic ones:

  1. More applicants than jobs. In many disciplines, the supply of Ph.D.s far exceeds demand, in terms of junior academic jobs. That can translate directly into no job, but also insidiously into getting tracked into not-so-good jobs, cut off from your community, with loads of service responsibilities, very temporary, etc. -- all of which impact your ability to do great research to land the next job.

  2. Mobility issues - the best or only job available may be somewhere you can't go, for personal or family reasons.

Oversimplifying greatly, in my experience, 4 and 5 are problems universally, felt particularly keenly where 1 or more postdocs is the norm (and so more opportunities to be zinged by these factors). 1a tends to be the biggest problem where people go from postdoc to independent researcher, where that's applicable, or Ph.D. to junior faculty member, where there are no postdocs. 2 and 3 at the more senior pre-tenure stages.

Houska
  • 7,170
  • 1
  • 18
  • 32
7

The academic career after PhD starts with at least one postdoc position. In order to get a postdoc offer you need to show strong ability and talent to do research (in math it means good letters of recommendation, publications, good PhD granting department, well known and good advisor, etc.). To transition from a postdoc to a tenure track position you need more of the same. Usually only a few people with PhD can become postdocs and even fewer will get tenure track positions.

markvs
  • 3,008
  • 6
  • 15
0

If we leave aside financial reasons and difficulty of finding a job, it's simply difficult because most people aren't that good at doing research.

I have actually thought about this a lot: I'm curious as to why my friend who was a stellar student in his Mathematics degree and got a high First (obviously seems to know a fair amount of mathematics) has really failed to make it in academia.

One thing which I think people do wrong is that they don't seem to read many research papers. This really seems like a big mistake. You will have to read a lot to be able to zoom in on a section where you are able to make a contribution. Similarly, it's hard to know how to write papers properly when you aren't reading them, then you become dependent on other people to help you, and if you are dependent on others, you will probably fail at postdoc level.

Another of my friends seemed to do well on his Engineering PhD and published two good papers (with several co-authors) but has struggled on his postdoc and not published anything (possibly because he now has to direct his research a bit more independently, as opposed to being told what to do). It's definitely worth trying to reflect on this a bit, try to identify what others do wrong that they fail, and so on.

Tom
  • 3,733
  • 1
  • 11
  • 29
  • True, but I am more interested in the case of graduated PhD students who could not transition, though they were excellent researchers, published quality papers in highly ranked journals, etc. etc. I think the "demand" answer addresses this pretty well. –  Jun 09 '21 at 01:01
  • This is a bit surprising to me that this would happen: almost all PhD students are not cut out for research and there are enough academic positions to go round amongst the few people that remain. I think probably independence of the researcher plays a role ie. they publish quality papers but they are essentially being told what to do by a supervisor. – Tom Jun 09 '21 at 07:13
  • almost all PhD students are not cut out for research I'm not so sure about this; particularly at top-20 programs in the US, I would say that most PhD students will be cut out for research. I do agree with the independence point, but this might vary from field to field. –  Jun 09 '21 at 07:38
  • @Tom True enough but the "alickers" often get the job over the claims of the self-directing researchers. – Trunk Jun 09 '21 at 11:04
  • Yes I see your point. For example, it took Einstein a surprisingly long time to land the job in academia which he wanted because he didn't get on with the relevant people or didn't hide that he thought they were stupid (hence he had to work at an office for quite a few years). Is this the sort of thing you mean? – Tom Jun 09 '21 at 12:46
  • What I mean is more middle-of-the-road that Einstein. Typically, it would be some "Rob Jones (24) PhD" with his well-organized, elegantly phrased but specious research together with adroit diplomacy on Prof Honeywell's intellectual blindspots and mid-life apathy ahead of some "Tamer Karadeniz (28) MSc (METU), PhD" whose largely independent, unstinting experimental work and study has effectively made him an unacknowledged world authority. In hindsight, Einstein was always going to come through and his "exile" to the Swiss Patent Office gave him time to theorise. "Karadeniz" may end up unknown. – Trunk Jun 09 '21 at 13:38
0

My two cents : I don't agree with your first premise:

The transition from a PhD position to a postdoctoral position: A postdoc has to work much more independently than a PhD student, and maybe people struggle with this.

My experience of supervision as a PhD student :

  • First 2 month : 30min / 1h talk with supervisor ~ once every two days during lunch.
  • Following 6 months : The same, once a week
  • The second year : A 2 hour talk, circa every month
  • The 3rd year: My supervisor left and was replaced with another one. A 1h talk circa every two weeks or so.

I would call this very independant work, as I was basically making all the calls, I just needed to justify them adequately with my supervisors. I did not go for a postdoc afterwards (I transitioned to industry), but I doubt a postdoc position would have left me "less independant".

G. Fougeron
  • 121
  • 6