The other thing that drew my eye in the recent piece in the NYT by Bhatias, Fan, Smith and Hwang was the table which reports the number of grants and the funding amount by Institute and Center. Again, it compares 2025 with the average of 2015-2024 (in constant dollars for the funding amount) and reports a percentage change.
Unsurprisingly, the NIMHD took the sharpest reductions in the number of grants (61%) and funding (38%). It is perhaps obvious why an institute dedicated to minority health and health disparities across all of the IC health domains took a big hit under the current regime. NINR funded 51% fewer awards. [The Table, I swear, said they spent 37% less money but I blinked and now it reads as a 0.5% increase. We shall see if it reverses because half as many awards for the same money seems weird.] NINR was identified in the Lauer et al 2021 report as being of disproportionate interest to Black PIs so…yeah, fewer awards tracks. This is of course not driven by “government efficiency” since these two institutes get a tiny fraction of the overall NIH appropriation. In Lauer’s paper, 0.99% for NIMHD and 0.52% for NINR of the subset of 20 ICs. In FY2015 NIMHD got 0.89% of the overall NIH budget and NINR got 0.47%. In FY2020, 0.81% and 0.41%.
NIAAA experienced a 47% reduction in competitive grants from a 12% reduction in funding. NIDA funded 13% fewer grants and actually had a small increase (0.9%) in FY2025 funding amount. It’s always a little weird to see such seemingly similar ICs, which support similar and overlapping research communities, have such different funding outcomes. My browsing of awards in the waning weeks of the FY suggested that NIDA had gotten away with very few multi-year awards. Given my potentially erroneous recollection in the last post, I checked. NIDA funded what looks like 1 of 129 new R01 in FY2025 as multi-year. They did not fund any RF1 conversions. The NYT data incorporate all competitive mechanisms, perhaps they chose to prioritize R01s. Still, I’d really like to know how and why NIDA was able to essentially refuse the multi-year mandate strong suggestion. And whether we might see some variety of punishment (reduced budget) or corrective action (an unusually high multi-year rate for 2026 perhaps?).
In contrast NIAAA funded 13 of 60 new R01 as multi-year. Since NIAA funded 64 new R01 in FY2024, 110 in FY2023, 93 in FY2022, 71 in 2021 and 71 in 2020, I’m not sure what to make of their seemingly large FY2025 reduction in all competitive grants versus a decade average. Maybe they were already starting a planned reduction of R01 support in FY2024. Maybe the relatively higher FY22-23 allocation was receiving a correction. Or, maybe, the multi-year out-years reduced the intended count back towards FY22-23?
I next tend to look at NINDS (26% fewer awards, 26% less funding) and NIMH (43% fewer awards, 31% less funding) as adjacent ‘brain institute’ comparisons. NIA tends to be a bit strange because Congress periodically lavishes “omg, Alzheimer’s!” money on it. They had 31% fewer awards with 17% less money.
FWIW, Lauer et al 2021 identified the National Eye Institute as having the highest success rates and here they funded 21% fewer grants and spent 10% less.
Anyway, all of this is a good reminder that each IC does things a little differently and freaking out / feeling relieved about our prospects based on NIH wide averages or on some other IC’s behavior is unwise.
There’s a new article on the NIH funding results for FY2025 in the NYT by Bhatias, Fan, Smith and Hwang. It has some data indicating that total competitive grant funding ($) ended up 13% down from the average for FY2015-2024, all represented in inflation-adjusted dollars. The piece shows a 22% reduction in the number of competitive grants awarded, again compared with the decade average*. The article also shows that the NSF funding amount was down by 18% versus the 20150-2024 average, corresponding to a 25% reduction in the number of NSF awards.
One part of this article I don’t fully understand is the description of the impact of multi-year funding. This is, of course, a major reason for any reduction in numbers of grants that exceeds the reduction in overall funding. Simplistically, ever year past the first one that is funded in FY2025 via multi-year funding prevents the funding of one additional new award under traditional budgeting. The infographic they included was addressed with the following text comments:
- In the past, the N.I.H. typically awarded grants in five annual installments.
- Researchers could request two more years to spend this money, at no cost.
This is somewhat confused, unsurprising because these technical details of NIH funding are indeed confusing to newcomers (and old hands). The annual installments of the usual grant award typically permit no more than 25% of the annual budget to be retained and rolled forward into the next year. If one wishes to do this for the final year originally awarded (which could be five, or could be fewer), one is allowed to request a No Cost Extension (NCE). The first year NCE request is approved by default** in most cases. A second year requires an additional hurdle of justification and permission- although I do not know if these were rejected with any frequency before 2025. The “no-cost” means no additional funds past those originally committed will be awarded. The phrasing by NYT “, at no cost” confuses this.
The phrasing “two more years to spend this money” is also a bit confusing / misdirecting. The 25% limit on budget “carry forward” is not cumulative, so the NCE will generally be for up to 25% of one year’s worth of the budget.
Of course, the NIH has many areas of operation in which there are exceptions to what is, or appears to be, standard operating procedure. It can permit the carry-forward of more than 25% of the unused budget. It can, therefore, permit what amounts to a cumulative carry-forward. It can permit what amounts to full on skipping of any expenditure in one year to be pushed forward into the next. There are always exceptions.
With that cleared up, the multi-year funding description is likewise confused. First, this seems to imply NCEs past one year will not be approved. They are rare, so meh whatever.
- Under the new system, the N.I.H. pays up front for four years of work.
- And researchers can get one more year to spend this money.
- Which means that they get less money on average, and less time to spend it.
The records on RePORTER show that NIH has awarded new, five-year R01 proposals in FY2025 with anywhere from one (the normal funding system) to four (multi-year) of the budget years awarded from the FY2025 budget. This is where “pays up front” thing comes from. The infographic suggests the “can get one more year” sentence means the researchers obtain this interval via a NCE.
[record scratch]
Uh-oh. My memory from looking at the multi-year funding issue through the summer says to me “This is not what the RePORTER record suggests“. The record for each new R01 lists Project Start and End Dates and Budget Start and End Dates. For a new R01 awarded in FY2025 the Start Dates will all be in 2025, save for a few Dec 2024 awards. Under traditional award, the Budget End Date will be one year after the start, mostly in 2026. A multi-year funded award will have a Budget End Date more than one year away, i.e., in 2027, 2028 or 2029. Thus, says my memory, one can easily confirm five year awards funded partially via multi-year that appear to anticipate the return to good old, Type 5 noncompeting award for Years 3, 4 or 5, depending on the length of the multi-year.
However. In attempting to put some numbers to this today, the most recent set of records varies substantially from my recollection. Now the multi-year awards have both Budget and Project End dates that are the same. Yes, there are something like 45 new R01s awarded for only two years. Another 56 awarded for only three years. And about 260 awarded for four years. These latter span the range of ICs, not just the ones (hi NIGMS) that traditionally only award a maximum of four years for new R01s, regardless of what was proposed. This seems a bit unusual IF the multi-year grants were selected at random with regard to requested interval of funding.
There are only 2 (out of 3,443 new R01) that appear to have been funded up to 2030, i.e., the full five year proposed. The balance of five year awards, about 1,831 of them, were budgeted as traditional single-year funding as indicated with 2026 Budget End Dates.
We can’t tell from the RePORTER record what was originally proposed and what went through Council decisions. We cannot validate that these end up with “less money” compared with what was proposed and what would likely have been awarded, in total, under normal operations. As a final clue, it is possible to divide the awarded direct costs with the number of years awarded as multi-year grants and see a Gestalt impression that these are not five year award budgets stuffed into four, three or two years. So the current RePORTER record does suggest that multi-year funded awards were indeed cut by years and probably be total award amounts.
I would be very fascinated to hear from some of these multi-year PIs whether their Notices of Award have been changed from the original.
*I am really not sure the point is best made this way. The most important depiction would be compared with FY2024 and FY2023 in my view. Not against a decade long average in which budget levels varied.
**This became an issue in the first half of 2025 when the pro forma approval of first year NCE requests was halted and all of them were supposed to go through the more involved approval process. This led to a lot of concern that NCEs were going to be clipped as part of the induced chaos. There was a related set of rumors that maybe the carry-forward default of 25% for any budget year was going to be denied. Which was very concerning for many PIs, myself most certainly included.
DrugMonkey 