According to academicpositions.be, which quotes the New York Times, PhD students earn 15000 and 30000 dollars per year. But according to Educationdata:

In 2012, the average cost of a full year in a Ph.D. program was $21,400. Of the $21,400, the net price for the student was $8,480. Of the $21,400, roughly $12,920 was covered by grants. 39.2% of surveyed Ph.D. students offset the cost of their degree with teaching assistantships.

Why do these sources seem to contradict each other?

Does this also differ by country?

  • 1
    This varies around the world. Where do you propose study?
    – Buffy
    Commented Nov 30, 2022 at 22:43
  • Costs whom? The professor pays stipend, tuition, and research costs.
    – Jon Custer
    Commented Nov 30, 2022 at 22:48
  • I wonder how Glassdoor defines "PhD student". That definition might not match the definition in Academia. Also, my understanding is that "PhD student" is not considered a job in the US. Thus, I'm confused how there could be a salary.
    – user9482
    Commented Dec 1, 2022 at 6:31
  • I replaced the Glassdoor reference
    – Riemann
    Commented Dec 1, 2022 at 8:31

3 Answers 3


It's kind of both. When you study, you (or someone) has to pay the university the tuition fees. However, in many countries, the professor/department/scholarship will have funding to pay the university and pay you a stipend.

  • If you don't have funding, then the net result is you pay the university.
  • If you do have funding, then the net result depends on how much funding is provided. Quite often, the net result is you are paid. If the funding is only partial, you still might have to pay the university.

Don't confuse cost with price.

The following applies in the US. There is also some difference by field and by university. Generally, though,...

I think $57K sounds a bit high, though it probably includes such things as health insurance. The stipend is set to cover modest living expenses of a student in that locality, recognizing that many are married. Enough for frugal living, but little enough that you want to finish ASAP.

And for almost all doctoral students, a TA or an RA will come with forgiveness of almost all costs. Those costs are borne by the institution or the government that supports it. But a privately funded student will have high cost of tuition. Again, I'd guess $99K is at the top end of the scale, and those are total costs over the multi-year program, not annual costs. And, most students in doctoral programs have a TA or (fewer) an RA. Both provide valuable service to the institution, especially in making the faculty more productive.

But yes, the costs are high, since in some fields the labs are costly and the professors are well paid. Even building maintenance isn't free.

You aren't normally paid for your research. A TA means you assist in teaching (normally) undergraduates. An RA probably means that you assist in someone else's research, though your project might grow out of that. You are paid a stipend, not normally a "salary" (US). The tax implications of that difference might be important, but most likely taxes will be due.

  • A bit high? The NSF graduate research fellowship pays $36,000 and represents the largest pay a lot of fields can hope to get, by like $10,000 ... Commented Nov 30, 2022 at 22:59
  • @AzorAhai-him-, yes, 36K sounds about right, but does it include health insurance - a big cost in the US. And the 57K was listed as a max figure. But 10K sounds too low.
    – Buffy
    Commented Nov 30, 2022 at 23:06
  • 4
    One valuable thing a PhD in an empirical field teaches is a reminder to always concern yourself with where data come from. From the page OP links: i.sstatic.net/5nhwO.png ...a couple at $25k, a couple around $35k, one at $150k... and GlassDoor happily reports "average 57,029". These are just reported salaries by users, and the average may not be particularly meaningful if it includes outliers that are data entry errors or unusual circumstances.
    – Bryan Krause
    Commented Nov 30, 2022 at 23:14
  • @Buffy It does, yes. But that is not usually included when reporting salaries. And no, no one said $57k was a max. If you click through, it is clearly implied to be some sort of mean. Commented Nov 30, 2022 at 23:26
  • And "largest pay ... by $10,000," i.e. $25,000 not $10,000 for the year. Commented Nov 30, 2022 at 23:28

I am now living in a country where generally PhDs are not paid unless you find a grant and where a lot of people do a PhD for free. In my country this can also happen, however it's now getting more and more common to require actually having a grant or some kind of compensation before being able to start a PhD, thus most people over there earn a salary. I'm paid a salary for my PhD as I'm employed outside of university.

Generally around here doing a PhD does not cost money, outside of registration fees and stuff like that, which is generally quite low and definitely below 1000€ for the whole PhD. If you get a grant these fees are generally covered by the grant. Of course, if you're not paid you'd have to account for how much it costs you not to have another jobs that pays you.

I do not know about you, but I'd never accept to work for free. Doing a PhD is a job in my mind, since you're doing something that takes your time and generating useful research for other people to use (which is quite different from what a student does for example during his master), thus I'd never accept to do a PhD for free. On a side note, I personally disapprove of people who accept to do unpaid PhDs or to work overtime; this because you're basically lowering the salary of other people doing a PhD by accepting to do the same job for free.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .