NIH-funded researchers can't catch a break

Changes appear to be underfoot in how the NIH pays out awards. It may seem a minor technicality, but it will likely have far-reaching consequences.

The NIH appears to be changing its awards payout model, from incrementally to all at once. Historically, the NIH has always paid out awards one year at a time. If the NIH awarded you a 4-year grant, they didn’t give you all the money upfront, but instead initially gave you only the first year. Then, towards the end of that year, you had to submit a progress report where you described what you did and how funds were spent, and then the NIH gave you the next year of funding.¹ This process meant the NIH provided oversight during the entire period of performance, and in principle it could modify the award along the way or even terminate it early. In practice, this has rarely happened, but the process kept researchers on their toes and ensured they dotted all the i’s and crossed the t’s. The alternative payout model is full funding, where the entire award amount is made available at the beginning of the funding period. While the NIH has not typically done this, other funding agencies do use this model, for example the NSF as far as I’m aware.

Photo by National Cancer Institute on Unsplash

What appears to be happening is that the NIH is now transitioning from incremental funding to full funding of awards. On first glance, this may sound like a non-issue. At steady state, either model works just fine, and there are only minor differences for the project PIs.² In fact, PIs likely prefer full funding upfront, as it removes some friction that otherwise exists with waiting for the next installment.³ And, the fact that other funding agencies are successfully using the model of fully funding grants upfront demonstrates that there is nothing wrong with this model in principle. This doesn’t mean, however, that the switchover is going to be a non-event.

The problem is the transition period. Which can be brutal. Let’s consider a simple example. For simplicity, let’s focus only on 4-year awards, which in the past have been funded in annual installments but are now being transitioned to full funding in the first year. Under the past model, assuming a steady state of no major budget increases or decreases from year to year, each year 25% of available funds can go towards making new awards while the remaining 75% are needed to fund the out years of the previously awarded grants. If the NIH takes all the available funds that can go towards making new awards and uses them to fully fund those awards upfront, then it can only make one quarter of the awards it otherwise could have made. That is a reduction by 75% in the first year of the transition! In practice, it seems that the current plan is to fully fund half of the newly made awards, which means the number of awards that can be made assuming no changes to the budget is reduced by 3/8. This is still almost a 40% reduction in new awards. And if there are any budget cuts they will have to be added on top and come primarily out of the funds available for new awards. As the NIH still has to fund the out years of existing grants, and 75% of the budget is already committed, only 25% is available to be cut. If the total budget were cut by 25% then to first approximation no funds would be left over to make new awards.⁴

I’ve seen comments along the lines of “so what, it’s going to be a few years of a difficult transition, but then we’ll be back to a new steady state where things are fine.” These comments betray a misunderstanding of how precarious the situation of many NIH-funded labs is at any point in time. For many NIH-funded labs, an interruption of funding leads to termination of researchers, up to and including the PI. A PI who needs to cover the majority of their salary from federal funding will simply not receive a paycheck when funding runs out, and so may have to look for alternative employment. And even if universities agree to cover PI salaries for a while (some may do so for some subset of PIs), all the other researchers in the lab—PhD students, postdocs, technicians—will face termination, and this may lead to eventual lab shutdown anyways (see next paragraph). Importantly, even under the current status quo of NIH funding, many labs routinely find themselves at the brink of shutting down. They have already spent their last grant money, and they have also used up any institutional discretionary funds they may have had, and maybe the institution has provided some bridge funding that is now also running low. Such labs need a grant ASAP or they will have to shut down. They definitely won’t survive having to wait another year or more for their next grant. You can be certain there are numerous labs in this situation right now.

But even if the PI can survive a year or two with little to no funding, their fate may be sealed. With every month that passes, their lab gets less competitive for future grants. They are not generating preliminary results. They are not publishing at the rate their competitors are. At the same time, the fully funded labs are forging ahead, and they’re not putting a pause on writing grant applications either.⁵ So, six months to a year later, when the PIs who are out of funding try to apply for their next grant, their applications will simply not be competitive compared to those from labs that were funded in the interim. Therefore, rather than finally receiving a grant after waiting a year or two, these PIs will continue to be unsuccessful. Instead, it will be the labs that got the first round of fully funded awards that will also receive the awards in the later years. I don’t see how this kind of a transition in payout model, unless implemented with extreme care, won’t lead to numerous labs shutting down and disappearing, even if in the end there is as much grant funding available as there was before the change. (And of course that’s not a given either.)

Jeremy Berg, former director of the National Institute of General Medical Sciences at the NIH, has put together a computer simulation that predicts the change in total number of awards made over time, taking into account more factors than I did here in my simple back-of-the-envelope calculation. We can see how the total number of awards made drops in the first and second year of the transition before eventually recovering. (Click through to the thread to see results under various different assumptions, including with budget cuts.)

Let me bring up one final pernicious effect of transitioning to fully funding grants that I haven’t seen anybody mention so far: It will reduce indirects received by universities, at least temporarily. This is because universities only receive indirects for dollars actually spent. Universities don’t receive indirect funds in a lump sum when a grant is awarded. Instead, they charge in proportion to exactly the direct costs spent by each research project each month. If a PI receives a million-dollar grant but only buys $5000 worth of reagents the university only receives $2800 (or thereabouts) in indirects. To put another way: Assume in the past a university had four researchers that all had one year of funding they were spending. In this case the university was receiving 100% of the overhead associated with the funding provided in that year. Instead, now, a single researcher will have four years of funding. The university has received the same total amount of funds, but initially 75% of the awarded funds or more will sit in an account unused, and so will the associated indirect costs.

Between labs running out of money and having to close and reduced indirects, I expect medical schools and biomedical research institutes to be hit especially hard by this one policy change alone. Their entire business model depends on researchers who fund themselves by bringing in federal grants. With paylines in the single percentage points, possibly as low as 4%, the self-funded model is simply not sustainable. Four-year colleges may weather the storm a little better, and may not have to fire faculty members outright, though labs will close there also and the universities may stop hiring faculty members to reduce overall headcount.

Even if the transition period from one payout model to the other is temporary, we have to expect that the consequences will be long lasting and not easily reversed. Every closed lab is lost expertise. Once a lab is closed, it generally doesn’t reopen. And if an entire medical school or research institute is forced to close, it for sure won’t reopen. Additionally, I wouldn’t be surprised if the more junior PIs (in particular, Assistant Professors) will be the hardest hit. At 4% paylines, it’s going to be difficult for them to compete against the biggest names in their field, who will still get funded. And if universities experience budgetary constraints and see a need to reduce their faculty numbers, the easiest choice for them is to not give Assistant Professors tenure, in particular if it can be justified by the candidates’ poor track record of obtaining external funding. So, in 5–10 years, when the senior people start retiring, there won’t be many junior people to take their place, and the US may have permanently lost the ability to compete at the highest levels of international science.

1. This is called a non-competitive renewal. “Non-competitive” because the expectation is you will get the next funding increment. You don’t have to compete for it against other investigators.

2. The PI (Principal Investigator) is the lead scientist on an NIH award. This is typically a professor at a medical school, but it can also be a professor at a four-year college or a senior scientist at a university or research institution.

3. This can be a real issue for subawards. Ask me how I know.

4. In practice, at that point the NIH would likely give substantial haircuts to the existing grants, but even so there wouldn’t be much money left to pay for new awards.

5. It’s normal for an NIH-funded PI to submit at least 5–10 applications per year, to ensure there is a steady stream of new projects coming in. You never want to be in the situation where you’re down to a single grant and once it runs out you have to close down shop. You also want projects to be staggered so they don’t run out at the same time. Ideally, you would receive at least one new grant every two years, but since you can’t be certain to be successful in any given year you just try all the time.