Jonathan Birge

In the stimulus package passed by the wise souls of the US Senate, the NIH budgets will get a boost of around ten billion, along with many more to health services in general, with little to nothing going to the NSF. I won’t venture to judge the relative merits of the NSF versus the NIH in the grand scheme of things, but since this is supposed to be a stimulus bill, let’s consider the two agencies for a minute from the perspective of economic returns.

NIH funding generally falls into two categories: on the one hand you have very long term research on grand problems, such as understanding cognition or the ability to regenerate nerves via stem cells, and on the other, short term research which is essentially medical in nature. The former has essentially zero short term benefits, other than “ah, that’s nice” articles in Nature, in which we find out something vague about the way the human memory works (”something to do with this here chemical, plus a few thousand others we don’t know yet!”). These large scale efforts will eventually, we hope, yield true revolutions in the way we perform medicine, and might be have huge economic returns. But they are decades away. The latter category, on the other hand, provide immediate health benefits, but generally in the form of linear improvements in care which come at exponential increases in cost. For example, new cancer drugs may extend the lives of a fraction of patients, but they cost $60,000 or more.

Given that any advances born of NIH funding turn into private profits funded by health care consumers, I’d argue NIH funding generally has a negative short term return on investment; instead of adding to our productivity as a country, it results in products which we are forced to buy (via givernment regulated insurance) no matter how marginal their improvement. If a pill works 5% better than the previous one, but costs 10 times more, no politician has the guts to say it’s not worth it. In this environment of completely uncontrolled cost management, for every dollar we spend on the NIH, the “return” is that we have to spend far more on the marginally effective technologies produced. This is completely different from most engineering and science research, where the products are, at worst, pointless amusements (e.g. most Media Lab research), and at best make society more efficient (e.g. networking, industrial robotics, digital imaging).

I don’t know what the Senate is thinking, but if their idea of stimulus spending is throwing money at healthcare research, we’re in trouble. You don’t grow a productive nation by spending a quarter of your GDP on visits to the doctor, but that’s where we’re headed in the next decade. I’m not a blind cheerleader for the NSF, either, but at least the research it funds has a net positive return on investment, even if it’s close to zero. Things like better computers, artificially intelligent robots, telepresence, etc. would all be much more beneficial to our economy than another $1000 a month pill we’ll be forced to buy for people who will die anyway. And if you’re apalled at my mercenary separation of morality from economics, ask yourself this: Is it in our long term moral interests to bankrupt the country? Healthcare quality tends to go down in collapsing economies…