The 1-in-5 odds that were counted per application

Be the first to star this exhibit

A postdoc newsletter we invented charts NIH’s research project grant success rate from 1998 to 2024 and tells readers that fewer than 1 in 5 researchers who apply get funded. The 18.5% is real, but it counts applications. Count the 45,555 people who applied that year and 27.8% of them were funded; set the people holding any award in 2024 against everyone who held one or applied without success since 2020, and it is 40.8%.

NIH Data Book (NIH Office of Extramural Research, Division of Statistical Analysis and Reporting), research project grants, fiscal years 1998–2024: report 20, competing applications, awards and success rates; report 164, competing investigators and funding rates; report 166, new R01-equivalent funding rates by career stage; report 303, cumulative investigator rate. Rates to one decimal are computed by us from NIH’s published counts. The newsletter and both drawings are ours., 2026Added October 24, 2026

01The claim

Fewer than 1 in 5 researchers who apply to NIH get funded, so a postdoc planning a first grant should expect to come away with nothing four times out of five.

02The trick

Every point is NIH’s own success rate and the axis starts at zero. The slip is one word: the success rate counts applications, and the headline says researchers. In fiscal 2024 NIH funded 10,265 of 55,418 research project grant applications, 18.5% (resubmissions within the same year count once). But a researcher is not an application. One person can send several in a year and needs only one to succeed, so a rate per application usually runs below a rate per person: in NIH’s own counts, by four to ten points every year since 1998. Turning 18.5% of applications into “1 in 5 researchers” moves the number to a unit it was never measured in.

03The fix

NIH publishes the same year at the grain of people. Of the 45,555 investigators who competed for a research project grant in 2024, 12,671 were funded: 27.8%. And of the 93,638 people who either held an award in 2024 or applied without success at some point from 2020 to 2024, 38,200, or 40.8%, held one. The redraw puts all three rates on one zero-based axis and labels each by what it counts. The grains even disagree on the trend: from 2013 to 2023 the per-application rate rose 4.5 points while the per-investigator rate rose 8.3. None of the three is any one person’s odds; each pools renewals with first tries and established labs with new ones. For the newsletter’s actual reader the closer grain is narrower still: on new R01-equivalent applications in 2024, NIH funded 20% of first-time applicants and 26% of established investigators. A question about people needs a rate whose denominator is people, and the closer that denominator gets to the reader, the better the answer.