how long should you stay in graduate school?

How  many years should a PhD program take? Let’s assume that you are well prepared and that you finish exams and courses in a brisk pace. First, you have to realize that some disciplines (like engineering) have short clocks. You have no choice. You will graduate in about four or five years.

Now, for the rest of academia, you choose how long to stay in. Previously, I talked about how you can nudge your committee into letting you finish. Roughly speaking, once you publish, you are done. But when should you do this? To figure this out, consider the following:

  • There are many strong arguments for *not* staying an extra year. Every year you are in graduate school, you loose a lot of money. Your potential salary – graduate student stipend >> 0. Also, there’s an argument for “growing up” and being a professional with a real job and office.
  • There are a few arguments for staying in. If you have no publications, spending a year to get a paper accepted can have a huge impact. There are some specialties that simply require a long time. If you happen to be in a horrible job market, it may help to wait a little.

My argument is that there is a presumption for a shorter period of time. Most people should aim for “short” of 4-6 years unless you have a really, really good excuse.

Impress your friends: From Black Power/Grad Skool Rulz


Written by fabiorojas

December 11, 2012 at 12:01 am

Posted in fabio, grad school rulz

13 Responses

Subscribe to comments with RSS.

  1. My department (top 15 political science) made a policy decision to stop funding students past 5 years except in special circumstances. This decision was made before the financial crisis and ensuing budget cuts. The idea is that 5 years is plenty to complete the degree, and if it isn’t (e.g. because of time spent doing field work) then getting external funding should be the goal anyway.

    I think this is generally good policy. Under the previous regime too many students stuck around for 7-10 years and either never finished, or if they did they were clearly not cut out to be good academics (or they would have finished sooner).

    It can make the job market a bit difficult, since the peer review process takes so long that you’ve got to submit stuff by the beginning of the 4th year (at the latest) to have a chance a getting a publication before going on the market. For us, that’s only a few months after defending our dissertation prospectus.

    In my case it worked out. My MA thesis was R&R at a fairly big journal, I had a co-authored a forthcoming article, was co-editor of a book under contract, and author/co-author of several book chapters. But a lot of that was because my advisor was kind enough to work with me on project relatively early on. I haven’t submitted any of my dissertation chapters to journals yet, but I got a good job anyway because of the other stuff (at Indiana-Bloomington, actually). I had other offers too.

    But I planned all this relatively early and got a bit lucky. For others, coming out with publications in 5 years is tough, and in this market no ABDs are making long shortlists for R1 TT without publications, much less getting interviews. So pushing students out early(ish) is basically asking them to take a post doc or VAP while they revise/refine/resubmit their dissertation. That’s not necessarily a terrible thing either, but there are likely downsides.


    W. K. Winecoff

    December 11, 2012 at 12:40 am

  2. One downside of pressuring students to get out quickly is that it dissuades them from taking on more ambitious projects. I know that had the department put more pressure on me to finish within five years I could have done it with a quantitative dissertation based off my M.A. project, but as one of my advisors once said, it would have been a series of “punts” to get shorter, less ambitious projects out and under review. I’m incredibly grateful our program supports us for longer because I felt comfortable taking on the risk of a much more ambitious multi-method research project with a full year of data collection in the field, which will ultimately contribute much more to sociology. It was a wonderful opportunity to be able to do that without the pressures of teaching multiple courses or a tenure clock ticking. I’d be curious to see the average quality of research that grads produce who are pressured to finish quicker, and where they are placed in jobs.
    It often takes time to learn the craft of how to publish in sociology where good work makes theoretical, methodological AND substantive breakthroughs. This is different than articles in some other fields (which get students out faster) but publish shorter empirical articles.



    December 11, 2012 at 1:24 am

  3. I was talking to a professor once who actually said he regretted ending his Ph.d studies too early. He said he took 5 years to finish, but wished that he’d stayed an extra year or two because of the “experience” of grad school is just so enjoyable. He also said that he thinks he could’ve learned more by staying an extra year or two.

    That being said, he’s a tenured professor fairly early into his career so perhaps he can afford to have these kinds of “regrets.”


    Andrew B. Lee

    December 11, 2012 at 1:33 am

  4. 1. Jaime: Get back to work!

    2. WK and Andrew: In thinking about PhD completion time, we have to distinguish between those who take long because they are getting better (e.g., more articles) vs. those who take long because they simply don’t have the capacity to complete their degree. The current system does not make this distinction.

    WK raised the issue of how hard it is to get jobs w/o pubs. Agreed. But remember, that’s true because we are all letting our students stay in school forever! In disciplines with short times, people are able to get jobs. Thus, if could solve the collective action problem and get all programs to systematically shorten times, you would have a labor market that is less publication oriented.



    December 11, 2012 at 4:21 am

  5. I think it’s hard to compare “short clock” fields like engineering or physics to the social sciences. It’s my understanding that while the PhD completion time is very short in these fields it is customary to have one (or often more than one) post-doc position before taking a tenure track faculty position precisely because the grads from these programs are not ready to be faculty members after only 4 or 5 years. Is it possible to think of the longer programs as sort of mashing the PhD and post-doc together?



    December 11, 2012 at 4:49 am

  6. JD: Get back to work!!!!!

    But yes, I’ve often thought that the modern PhD combines the degree + post-doc. Once again, though,the question is whether this is a good arrangement. For universities, it clearly is – you get cheap labor. For individuals, not so clear. You are essentially functioning as a prof, but at a fraction of the pay.

    Yet, still short clock field come in different flavors. Biology is clearly a mash up of short PhD and post-docs, economics is short and has few post-docs. What I would hate is the biology system of short PhD and endless post-docs. My preference is for something almost like economics. Short is the norm, but we’d grant exceptions for well argued cases.



    December 11, 2012 at 5:04 am

  7. If what you need is an extra year to get the paper accepted, you should be looking for post-docs or VAP positions. Some VAP positions are high workload, but if you’ve got the paper mostly done this is doable. If we require three years of coursework, I think the expectation should be five years for those using existing data, six years for those of us who are doing fieldwork or constructing our own datasets.



    December 11, 2012 at 5:35 am

  8. Mikalia: Graduate program reforms requires many things. First is ditching unneeded course work. For example, IU requires 4 extra courses in a minor. I used to be in a physical science program that had two language requirements. Etc. That would help. In my view, 4-6 years is fine. But once you get past that, I start questioning the value of the time.



    December 11, 2012 at 5:52 am

  9. Some potential unanticipated consequences of shortening grad school:
    1) As above, long times to publication means that students have to start solo publishing in year 2 or 3, when many simply aren’t capable of pulling together a good independent project. This creates an incentive for departments to admit students who have masters degrees. At best this offloads some of the cost of training to other universities, at worst it offloads it to the students.

    Yes, we could collectively agree to shorten review times. I’d also like a pony.

    2) The post-doc model isn’t perfect, either. Moving around creates enormous inefficiencies in the form of transition costs. It’s also really hard on families and employed partners. Which brings me to 2a:

    2a) To the extent that women are more likely than men to be married to /partnered with other professionals than men, less likely to be married to a stay-at-home spouse, and more likely to prioritize their spouses’ careers (see Clayman report on dual career couples), the post-doc/VAP model will disproportionately discourage women from the PhD/academic route. We see this in the sciences.

    2b) Where is all this post-doc funding supposed to come from? Do we really want to encourage universities to hire more VAPs, usually at lousy wages and high teaching loads, because we want to get people out of grad school quicker?

    3) More PhDs produced in the aggregate. If you get students through quicker, all else being equal, there are strong incentives to simply admit more students. Faculty need grad students to be RAs, and departments need grad students to TA or teach courses. Besides, it’s a rare department that will say to an administrator, “oh, no, we don’t need all this money any more, please give it to some other department.”

    I completely agree that we need some better way of easing out weak or even just middle-of-the-road students, and much earlier in the program. Yes, there are some late bloomers and some late disappointments, but I’d wager that predictive power of assessed performance in year 2, if not sooner, is extremely high. (It’s also partly self-fulfilling, but that’s a different issue.)



    December 11, 2012 at 11:35 am

  10. In Denmark (and some other European countries) a PhD is set for 3 years – that’s the funding you will get (a salary almost equal to an assistant professor). These 3 years include one semester of teaching. Not saying it’s good, just to compare. It certainly limits where one can publish.


    Danish system

    December 11, 2012 at 12:12 pm

  11. Having spent an excessive amount of time in grad school due to lengthy fieldwork, I agree that it makes sense to get students through more quickly. However, I’ve encountered quite a few programs that are really pushing the 6 year model, and I have serious concerns about that. It makes it very difficult to do qualitative international research, which I think is absolutely essential for the discipline in an increasingly interconnected world. Foreign research often requires language skills. But such a timetable basically means only those who are native speakers of a language or have previous experience in a foreign country will be able to do that kind of work. I also am concerned more generally about the short timetable encouraging people to start their sociological career as undergrads, then move quickly through a doctoral program. I spent two years working in the “real world” after college, and I think that kind of experience was actually quite crucial for my intellectual development. My undergrad major was not sociology, and so grad school was a very different experience from college for me, and took me in completely new directions. I’m increasingly seeing people continuing their undergrad work into doctoral programs, graduating quickly and becoming professionalized in a very particular way, and I think this results in narrowly trained sociologists and is not good for the discipline.



    December 11, 2012 at 2:13 pm

  12. I think that 6 years is a good goal for any sociology program, but that it shouldn’t be a hard and fast cutoff.

    There are many things that get in the way of completion, including the fact that some weaker students will never finish. But a major one is funding. If the student has to work a random job/s, or teach an extra class, or become an RA on a project that will never lead to publication credit or continued interest in the substance, then that lengthens the time in the program.

    If we want to get students through, we’ll have to start funding them for the entire time though. Including summers. (This has been the move with 5 year programs–but as many others have pointed out, 5 years may not be enough.)



    December 11, 2012 at 6:36 pm

  13. IMHO time-to-degree is a fetish that obscures the real question which is progress and ambition. Fabio’s point above that there’s a difference between taking a long time because you’re doing good work and taking a long time because you can’t hack it is absolutely correct. (As others have pointed out there’s also taking a long time because of health/personal/etc situations.) The solution isn’t a target number of years but that we ought to become less scared of exercising judgment in order to distingush ambitious performance from procrastination.



    December 17, 2012 at 2:02 pm

Comments are closed.

%d bloggers like this: