GradPay: Crowdsourcing Graduate Student Working Conditions

Over the past few years, a slew of articles have warned would-be academics away from graduate school, arguing that graduate students are overworked, underpaid, and underemployed once they complete their studies. Other commentators have passionately defended graduate education, dismissing such concerns as overblown. This debate highlights the dearth of information about graduate student well-being. [pullquote]Altogether, the typical graduate student is paid less than half of the average starting salary of a college graduate.[/pullquote]

Certainly, some graduate students struggle to make ends meet, and others live relatively comfortable lives. But, at the moment, it’s hard to know how much money a typical graduate student is paid, or how much work she does. There’s also little information about differences in working conditions across departments: my limited personal experience suggests that students in the natural sciences earn more than students in the humanities, but there is little hard evidence to back up intuitions like this.

To address these questions, I recently launched GradPay, a survey of graduate student working conditions. The project is inspired by the Adjunct Project, but focuses on graduate students rather than adjunct faculty.

The survey asks Master’s and Ph.D. students about the work they do, the stipends they earn, and the health benefits they receive (if any). Some results are available on the site now; many new analyses will be added in the coming months. The results are available in real time on the project site, but I’ve highlighted some of the most interesting findings here.

GradPay Graduate Student Stipends By State

Median Stipend by State: Reds indicate high stipends, blues low stipends; black indicates missing data.

Although the survey has only been available since early in the year, the results have already started to shed light on graduate working conditions. Out of a total of 1,670 doctoral student respondents, the median annual stipend was $21,000. Respondents also reported working as teaching assistants for a median of 50% of their terms in school.

Nearly all doctoral students received a stipend of some kind: less than 3% of respondents reported a stipend of less than $10,000, and just 1.3% reported that they received no stipend at all. Over 15% of respondents reported that they worked as a teaching assistant every semester; 6.3% of respondents indicated that they were never required to teach. Over one-fourth of respondents reported taking out loans to support their graduate education.

Altogether, the typical graduate student is paid less than half of the average starting salary of a college graduate. Because the typical Ph.D. program takes at least five years to complete, choosing graduate school may entail taking a big financial penalty in the short term.

Factors Affecting Graduate Student Stipends

Stipends also varied widely across states, institutions, and departments. Doctoral students in Kansas reported the lowest stipends, at $12,875; Connecticutensians earned the highest median stipend of $27,500 (for a visual take on the geographic diversity of graduate stipends, check out the map visualization at the bottom of the post). At the institutional level, median stipends ranged from $11,580 at the University of Houston to $32,600 at Harvard. Last, at the disciplinary level, stipends varied from $12,000 in urban and regional planning to $30,000 in nuclear engineering.

Overall, the results of the GradPay survey paint a more complicated picture of graduate working conditions than one might expect. As warned by some commentators, some graduate students are in dire financial straits. Others, though, earn relatively respectable stipends and health benefits, and are rarely required to teach. The survey also reveals staggering differences across institutions and departments: [pq align=right]Students at the highest-paying university earned stipends over 180% more than stipends at the lowest-paying university.[/pq]

There’s still a lot of work to be done on the project. The results will become more accurate as more students complete the survey, and the website will become more informative as more analytics are added. If you want to get involved, you can forward the survey to graduate students, student advocacy groups, and university administrators: the address is Or, if you have suggestions for new directions for the project or want to start a collaboration, get in touch with me at


Written by Joshua Carp, a doctoral candidate at the University of Michigan and the creator of GradPay. GradPay was inspired by the Adjunct Project


  1. Srijan Sengupta June 24, 2013 at 9:57 PM

    Nice job! One feedback, the results might be more meaningful if they were adjusted with respect to the cost of living, which can vary quite a bit across campuses.


    1. Thanks! Great point about cost of living–I’ve been thinking about this recently and will add it as soon as I have time.


      1. Absolutely – the map above, while interesting, is fairly misleading without cost of loving adjustments. When I was comparing offers between a number of schools, one of my top choices offered $27k per year, but another, one that I thought was a better fit, offered $16k. I was worried about taking that much lower stipend, but once I did cost of living calculations I found that functionally the two stipends were actually fairly equivalent, and in the end went for the better fit with the “lower” stipend.

        Another factor to consider is how long assistantships last. Most that I’m familiar with (being in the social sciences) are for 9 months, and don’t include the summer. I’ve been lucky to have an adviser that’s always had work for me to do over the summer, but many folks in the program do not and are often scrambling fill in this funding gap, often through taking jobs unrelated to their field of study.


  2. Do these salaries include the tuition waiver. I know that the whole never sees that money but the school at least considers it past of the package, like health insurance. Incidentally, when I was a graduate student I taught every semester and was paid $8500 a year. Needless to say, I was on food stamps and HUD. So I am not defending the schools, just saying that the spend may not be the whole picture.


    1. Good question! No, the stipends don’t include tuition waivers at the moment. In my experience, students whose tuition is covered by their programs don’t always know how much it would cost if they had to cover it themselves. This kind of information might be easier to get from universities or department coordinators directly.


  3. […] the survey are available on the site in real time, in anonymized form, and Joshua has put together a brief summary of the results so […]


  4. I wanted to fill this out, but my school (Drew University) is not available as one of the possible institutions. My experience as a PhD student does not fit at all with your data (FAR less than the median funding), and so I wanted to contribute to help flesh out the data. I hope the data isn’t being skewed in this way, as I think this is a great project.


  5. Can you add a button to download data as a CSV?


Leave a Reply

Your email address will not be published. Required fields are marked *