A salary with strings attached

Usually, if you’ve done a job fine and been paid for your work, the money is yours no matter what you do after you’ve left the job. However, if you are a postdoc, you may find a big caveat.

Many postdocs’ salaries are paid with grants won by their lab heads, the principal investigators. A PI gets a grant, such as an R01, and part of it goes to paying the salaries of those in their lab working on the project. Another common way postdocs are funded is by Kirschstein–NRSA (National Research Service Award) awards — either individual fellowships (e.g. F32s) or training grants (e.g. T32s). In this case, the award (not the PI’s grant) pays the postdoc’s stipend. If you’re on one of these awards as a postdoc (not a predoc), you probably signed a form about a payback obligation.

The payback obligation is based on the fact that NRSA funding is to pay for your training. This funding doesn't support a specific research project like an R01 does; rather, it funds your development (though good research helps!). This is also why if you are paid from your PI’s grant it’s a salary, but money from an NRSA is technically a stipend, not a salary. With a stipend, you have to figure out how to pay your own taxes and you aren’t typically eligible for your institution’s retirement policy. Also, the stipend usually doesn't count as earned income for Social Security purposes. On top of all that, an NRSA award was issued on the premise that you are going to use your training to do great things for health-related research. So, if you take the funding and then do something else, the federal government basically says, “Fine, but give it back then.”

If you receive NRSA funding as a postdoc, during that first year you are funded, you accrue an obligation to work in a health-related field for another year, specifically, “by engaging in health-related research, health-related research training or health-related teaching.” If you don't, you must pay your stipend from the first year back to the government.

The debt accrues as you go — up to a year. So, if you leave after one month, you owe one month of work or stipend back. If you leave after a year, you owe a year. The details of this payback obligation can be found starting on page 327 of this National Institutes of Health policy document.

The payback obligation is incurred only during the first year of your postdoc. If you just continue to work at your postdoc for another year, even if you continue to be funded by an NRSA, you’re all good. By the end of your second year in your postdoc, you are generally no longer obligated to pay anything back should you leave then.

The NIH details this common way to fulfill the obligation: “For individuals receiving postdoctoral support under individual fellowships or institutional research training grants, a payback obligation is incurred for the first 12 months of Kirschstein–NRSA support. However, the 13th and subsequent months of postdoctoral NRSA-supported research training serves to pay back this obligation month by month.”

I’m in my second year of an NRSA-funded postdoc and I’m assuming I’ll be able to just continue working here and fulfill my payback this way. Hopefully no biggie!

However, for a not insignificant minority of postdocs whose plans or lives change, the obligation to pay back a year’s salary becomes a big deal.

If you want to or have to leave your postdoc with a payback agreement, it can be unclear what you are allowed to do for work. As noted above, the NIH document states that payback can be done by “engaging in health-related research, health-related research training or health-related teaching”. They define those terms this way (we've edited their descriptions only to match our style, but they're otherwise verbatim):

• Research. Research is defined as an activity that involves designing experiments, developing protocols, and collecting and interpreting data. In addition, review of original research or administration of original research that includes providing scientific direction and guidance to research may be acceptable if a doctoral degree and relevant research experience is required for individuals filling such positions. Such research can be conducted in an academic, government, commercial or other environment in either a foreign or domestic setting. In addition, when consistent with the cumulative amount, type and frequency of research or research training experiences, functions that involve analytic or other technical activities conducted in direct support of research, as defined above, will also satisfy the service payback obligation. 
• Teaching. Teaching is an instructional activity that takes place in an organized educational or other instructional environment. Activities classified as teaching are generally carried out in a formal didactic setting, but other activities will be considered if they are consistent with the certifying institution’s policy on the definition of teaching responsibilities. Such teaching can be conducted at universities, professional schools, research institutes, teaching hospitals, primary schools, secondary schools or colleges. When calculating hours of teaching per week, it is permissible to include three hours of preparation time for each hour of direct instruction. Acceptable teaching activities must have a biomedical or health-related relevance. 
• Health-related. Health-related means related to the description, diagnosis, prevention or treatment of disease. Fields other than those usually considered to be directly related to human disease, such as agriculture, environmental sciences, biotechnology and bioengineering, also will be considered health-related.

Clearly, there is a range of allowed work. To better understand what is an acceptable job that won’t require a postdoc to pay back their salary, I reached out to Beverly Venable, an NRSA payback specialist at the NIH. I asked her about common jobs that people might seek after their postdocs. I was happy to hear from her that all the jobs I asked about would satisfy the requirement and not require payback. The jobs I asked about were:

• Teaching science in a college (no research component)
• Teaching science in K–12
• Research position in industry/biotech
• Managerial or administrative position in industry/biotech
• Medical writing
• Science writing (for magazines, etc)
• Medical consulting
• Medical science liaison

Things that wouldn’t fulfill the waiver include leaving science completely. Nonmedical-related consulting, for example, is a common one. Since I’m not an NIH payback specialist, I’d recommend contacting the office yourself if you are wondering about whether the specifics of your new job fit or not. Contact information of payback specialists (who I have found to be very knowledgeable and responsive) is here.

If you have a payback obligation, you must start paying it within two years. However, there are situations when you can wait longer to start and in the most extreme cases you can apply for a waiver to not pay at all. If you have a temporary disability, for example, you can file to defer payment. If you have a permanent and total disability, you can file for a total waiver.

Criteria for waivers and deferrals are described in the same NIH document linked above. They include extreme hardship and situations when not granting an extension “would be against equity and good conscience.”

NRSA awards can be great sources of funding and training opportunities. Hopefully this column can help people be aware of the details and make sure it’s something they can take on.