I never quite understood why you need to do a postdoc after completing your PhD. Is it like getting an unofficial second PhD?

  • Comments are not for extended discussion; this conversation has been moved to chat.
    – eykanal
    Jan 8 '18 at 17:44

No, it's not like a second PhD, because there's little to no documented requirements, no thesis, no defense, no coursework, etc.

In the 'old days' (prior to the squeeze on academic funding and exponential increase in PhD candidates), one took a post doc primarily to:

  1. Learn a new skill, technique, or method
  2. Dedicate a few years time conducting research to strengthen one's CV without the added stress of teaching and other tenure-track requirements.

Ideally, the post doc came out a very desirable candidate armed with a great research program and toolbox.

Today, a post doc is quickly becoming mandatory for academic-track scientists, as the pool of jobs shrink and the pool of candidates expand. It is not uncommon to find people doing multiple post docs, or people finding themselves in a permanent post doc (or research scientist) position.

While this is somewhat discouraging, the simple fact that funding is scarce and science can be expensive suggests that the ideal candidate has a proven track record and is likely to secure additional grants. Post docs provide one the time and resources to prove that tangibly.

  • 2
    I would upvote this except that I'm skeptical about the historical picture being painted. You make it sound like there was a golden age of research, and then came horrible funding cuts. There are probably multiple trends that have happened, and it may be different in different countries (e.g., Thatcher-era UK might fit your model). But I suspect the main thing that's happened is just that more people have been going to college and grad school, and although the supply of research jobs expanded exponentially, the expansion in demand was a faster exponential.
    – user1482
    Jan 6 '18 at 19:31
  • 4
    @BenCrowell it's also possible that the growth in both jobs and demand was linear, or at most polynomial, but that the misuse of the word exponential has grown exponentially. Jan 7 '18 at 14:22
  • @Bob way to call me out on the use of 'exponential' as a literary device. I can't argue with you...a polynomial trend is far more likely. I'm guessing not linear, but would love to see the data. Touché.
    – HEITZ
    Jan 9 '18 at 1:56
  • @HEITZ this Nature article is a bit out of date but has some very interesting graphs. Japan and the UK are linear at most, Germany is flat(!), but China is definitely nonlinear. Nearly all the growth in PhD numbers in the US is in biomedical sciences. Jan 9 '18 at 15:17
  • @Bob That's a great reference (though somewhat distressing). Thanks!
    – HEITZ
    Jan 11 '18 at 5:41

Why do you need to do a postdoc after getting your PhD?

In practice? Because very likely you won't be able to get any better (academic) job than that.

These days "postdoc" simply means a research position that is not permanent or indefinite (e.g. tenure track). It is far, far easier to land (yet another) postdoc than a permanent position.

This applies to the fields I'm familiar with (mostly sciences). The humanities may be different.


It's not a second PhD, because the length of the postdoc is typically much shorter than that of a PhD.

The function of a postdoc is sort of an "advanced apprenticeship," where you may choose your advisor (who should be different from your graduate advisor!) according to several possible goals, including:

  • Learning a new technique or field
  • Gaining further experience in a slightly different area from your graduate research
  • Personal needs (such as the "two-body" problem)

Such positions usually last anywhere from one to four years, depending on funding. Your duties will also likely involve helping to supervise some of the graduate or undergraduate students in the group. You may or may not also have some teaching duties. But you typically don't have the obligation to write grants and "bring in" money.

  • 32
    “The function of a postdoc is sort of an ‘advanced apprenticeship,’” — I strongly object to that depiction, though it’s fashionable to treat postdocs as students nowadays (many institutes explicitly hire for the post of “postdoctoral student”). It’s a racket to justify decreased salaries and benefits. A postdoc is no more an apprenticeship than any other normal job is. You always learn on a job. Accepting such a description from employers erodes the employees’ negotiation power and should therefore be opposed at every turn. Jan 5 '18 at 12:22
  • 4
    I think it is the other way around: Postdocs have no bargaining power and therefore they can be treated as students. This is not to say that that is any better, but just that you cannot solve that problem by avoiding terms like apprenticeship or student. Jan 5 '18 at 13:35
  • 6
    You seem to suggest that universities are evil. Maybe, but lack of bargaining power of post docs is not the reason. That is just a consequence of the fact that there are many more desirable candidates as there are positions. What makes universities more or less evil is how they choose to use that power. Jan 5 '18 at 14:54
  • 3
    @KonradRudolph: Assistant professor is also sort of an advanced apprenticeship. The contract is fixed-term, because the university is still not sure if the person is qualified for the job. Jan 5 '18 at 15:01
  • 4
    It's not a second PhD, because the length of the postdoc is typically much shorter than that of a PhD. The difference in length is not the key difference. Also, I'm told that in biology, for example, usually people spend longer as postdocs than as PhD students.
    – Thomas
    Jan 6 '18 at 2:07

If you intend to work in industry, you don't need and should not do a postdoc.

If you want a job in academic, it often requires a strong profile. You need to do a postdoc to continue doing research, and building up your profile.

  • My post-doc has been immensely useful for my future work in industry. It's not all about what you put on your resume - the experience can be invaluable. Not always, but often.
    – einpoklum
    Jan 7 '18 at 17:54

1. There isn't really such a thing as a "post-doc"

A "post-doc" is just a non-tenured fixed-term research position which requires the maturity and experience which a Ph.D. degree is a recognition of. That's it.

Now, it's true that it's customary to spend time in such a position, and in many academic institutions tenured position candidates are considered to be inadequate if they have not first had a stint (or two) as non-tenured "post-docs" somewhere - but there's no fundamental principle at work here.

2. Weakness of the academic staff (in terms of collective ethos, organization and action)

You could rephrase your question this way: "Why not just hire people with Ph.D.s into tenure track positions?" And indeed, there is no good reason not to (generally). Tenure track does not mean you actually have tenure immediately on employment; and even tenure is not some carte blanche to not do your job, or a guarantee to not be terminated. It should really be referred to as the "normal track" for employment of academics.

The thing is, non-tenure-track researchers are cheaper, and require less commitment. Additionally, universities and high-and-mighty PIs can cast this employment as something inherently temporary, increasing the chance that these employees not expect much to begin with - they'll just shuffle along. Thus in many countries there's a glut of post-doc positions and a dearth of proper academic staff positions.

Had there been a resistance from the tenured academic, or support by them to their non-tenured, more junior colleagues, in a demand to be moved into the tenure track / promoted to associate professors based on good performance, plus for permanent positions for research assistants where relevant - the balance of forces would change, and fewer people would spend many years moving from one post-doc to another despite being worthy of a proper position.

Of course, this requires academics to have a strong sense of what really benefits research; a memory of the history of their field going back several decades (if not a century); a sense of community or solidarity with their junior colleagues; and organs for collective action independent of university management, namely - academic staff unions. As the latter are often missing or weak, the former tend to be missing or weak as well.


First, you don't need to have a postdoc. The standard for whether or not a postdoc is necessary (and indeed how much "postdoc" is enough) varies wildly by field. This, incidentally, makes interdisciplinary hiring committees an interesting experience.

But it's not a "2nd PhD". What it's meant to accomplish is to be the transition period where you start to view yourself (and the field starts to view you) as an independent researcher. While working on your PhD, it's quite likely that one's advisor had a significant hand in formulating the question you were addressing, etc. The idea of a postdoc is that you can begin to mature as an independent scientist, potentially mentoring others, seeking funding, determining the direction of a project, etc. while still in a protected, mentored environment.

By analogy, if a PhD is learning to ride a bike with training wheels, a postdoc is taking the training wheels off, but still having a parent jogging by your side.

  • My experience has not been quite as you describe. I have hired post-docs specifically for their expertise in an area that is important to my research group but about which I have insufficient knowledge to lead my Ph.D. students and myself. Aside from that leadership, they have not been expected to set group directions, seek funding, and the usual other PI responsibilities. I have had post-docs who previously were faculty or industrial research lab employees, and so did not need mentoring. Jan 9 '18 at 20:15
  • @E.DouglasJensen Hence "potentially" for many of those activities. Though I'd suggest "We hired an expert in X who is how we determined how we did Y" is a form of setting group directions.
    – Fomite
    Jan 9 '18 at 20:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.