The problem in scientific funding is stability, not overall size

Michael S. Teitelbaum has an editorial in Science about scientific funding that echoes a point that I have been making for a while: the issue with scientific funding is as much about volatility (bigs ups and downs) as it is total funding.

For NIH, more research funding does produce increased research output, as intended. Yet, because the system as currently structured employs graduate and postdoctoral research assistants to do much of the laboratory work, increased research funding also produces (after a multiyear lag) additional Ph.D.-level applicants for NIH grants. No effective mechanisms are in place to align these increased numbers with expanding career opportunities.

In theory, the resulting chilly job markets for recent biomedical Ph.D.'s should generate negative feedback that would tend toward more stable equilibria. In a closed system, and one with full information available to prospective graduate students, some fraction of undergraduates who might otherwise consider becoming Ph.D. students and postdocs would correctly perceive the difficult career paths and would pursue other options.

In practice, however, the system is not closed. Given increased research funding, additional graduate students and postdocs can be readily recruited from large potential pools in countries with fewer such opportunities--precisely what took place as the NIH budget was rapidly doubled. Nor is there anything even approximating full information about career prospects available to prospective entrants, whether domestic or foreign.

The employment and career implications of such a positive-feedback structure were modeled during the 1990s. One simulation (for mathematics) assumed a 2% real increase in annual research funding for 5 years, followed by stabilization. The result was short-term improvement in employment prospects for recent mathematics Ph.D.'s, followed by deteriorating career prospects that ended up worse than before the funding increases began, even with the assumption of a permanent 10% increase in real research funding. (Emphasis mine. Citations removed.)

Basically what happened when NIH funding doubled between 1998 and 2003 was that academic institutions went hog wild to gobble up as much of the new funding as possible. They expanded their graduate programs, hired new professors, and invested in new buildings. That is all fine and good, but when funding stopped growing as fast it resulted in something like a market bubble in academic scientists. New professors were much less likely to get new grants -- putting themselves in funding trouble, and newly minted graduates students were unable to get new postdocs. The problem is especially acute because the training scientists is such a long process. Individuals starting at the beginning with a reasonable expectation of academic jobs are hurt when much less of those jobs are present than were at the start of their training. Considering the length of the process, it may be unreasonable to expect new students to incorporate this uncertainty into their decision about entering graduate school. No one can predict 20 years in the future. (The article also indicates that much of the money for capital projects was borrowed, making the academic institutions themselves much more vulnerable to any downturns.)

Now there are two general strategies to solve this problem.

One, we could pray for the immediate resumption of continual 10% increases in the NIH budget. This is not in my opinion a realistic solution. We just cannot expect Congress to keep giving us more money at that rate, even with all the good will in the world.

Two, we could indulge in a variety of fiscal methods to "smooth" changes in the NIH budget. Teitelbaum mentions several. A simple one would be to allow researchers or NIH-funded institutions to save more of the money they get each year from grants to generate a sort of slush fund to sooth funding gaps. (Right now...I am not sure of the percentage...but there is a limit on how much grant money an individual can save from year to year. What you don't spend goes back to the Treasury. This also results in wasteful investment in capital at the end of the year to make sure you don't have any money left over.)

My dream solution -- because I doubt it would happen -- is that the NIH budget could be made non-discretionary and indexed to inflation. You could increase it at say 1% plus inflation every year as part of the HHS budget. This result in smaller but much more predictable increases in scientific funding. This stability would prevent future market bubbles in research science.

Anyway, I am just happy that people are thinking about this issue in terms of stability. No one benefits from a business cycle in science -- not graduate students, not professors, and not the taxpayer who expects good science to be produced. If a policy solution that provides stability can be found, I would much rather have that than any transient increase in NIH funding.