Author contributions: R.J.D. wrote the paper.

The author declares no conflict of interest.

This article is a PNAS Direct Submission.

aThe number for PhDs has increased from 36 to 42 y during the same period. See Average Age of Principal Investigators with MD, MD-PhD, or PhD at the time of first R01 Equivalent Award from NIH, fiscal years 1980 to 2011 (2).

bPercentage of NIH R01 Principal Investigators Age 36 and Younger and Age 66 and Older (Fiscal Years 1980 to 2010) (3).

cThe data are somewhat different for those with doctoral degrees, who are subject to their own unique pressures and opportunities. For these individuals, the time to first R01 has increased 6.4 y over the last three decades, from 35.7 in 1980 to 42.1 in 2013. The data as to age to appointment are inconclusive: according to one set of data, the average age at first medical school appointment for PhDs has increased by roughly the same amount (12). Conversely, a separate set of data shows that the median age to first tenure track job for all US trained doctorates in the biomedical sciences has held roughly steady over time (5).

dA comparison of the success rates yields P < 0.001.

eThis analysis assumes an increase in appropriations to allow the NIH to lift the success rate of young scientists.

fThis is at some odds with Alberts et al. (18), who see the treatment of indirect costs of federally funded research as a source of subsidy for other university activities.

gIn the NIH lexicon, “new” investigators are first-time investigators. Early-stage investigators are new investigators who are within 10 y of completing their terminal research degree or within 10 y of completing their medical residency at the time they apply for R01 grants. Only about 55% of investigators who receive their first NIH grants are at an early stage of their career (35).

hR03s are relatively small, short-term grants offered by some of the institutes that can be used to run pilot projects. The R03 program faced criticism for being underfunded, nonrenewable, about as burdensome to apply for as an R01, and with only a slightly higher success rate. The program was so flawed as a gateway to preliminary data that the NIH was compelled to issue a notice that “strongly encourage[d]” young scientists not to apply for the R03s, for fear that the program was dissuading them from applying for R01s (36).

iEconomic tournament theory tells us that competitions (such as that for a move from a postdoctoral position to a tenure track position) are a useful and effective mode of allocating a scarce resource when reward structures are based on relative rank rather than absolute levels of output. However, the theory also tells us that to achieve optimal results it is critical that the competition be transparent, open, and fair (41).

jIt would be advisable to extend the 5-y period during periods of family or medical leave.