The world is currently living through an economic recession caused in part by an overbuilding of housing in the U.S. and several other countries. People built too many houses because they made unrealistic forecasts about future growth in house prices. When housing prices fell due to oversupply, people owning houses (by definition) lost considerable wealth. People who built or sold houses lost their jobs. At the heart of this problem, many have argued, were perverse incentives, such as reduced requirements for down payments, that encouraged people to invest too much in housing. People could enter the housing game without put up much of their own money. The problem is that they couldn't stay there.
In a recent issue of Science, Bruce Alberts argues that federal support for scientific research is similarly threatened by perverse incentives that lead to too much investment in research.
One such perverse incentive encourages U.S. universities, medical centers, and other research institutions to expand their research capacities indefinitely through funds derived from National Institutes of Health (NIH) research grants. A reliance on the NIH to pay not only the salaries of scientists but also the overhead (or indirect) costs of building construction and maintenance has become a way of life at many U.S. research institutions, with potential painful consequences. The current trajectory is unsustainable, threatening to produce a glut of laboratory facilities reminiscent of the real estate bust of 2008 and, worse, a host of exhausted scientists with no means of support.
This is an argument I don't want to make. Like Aaron Carroll, even with academic tenure, I make my living by winning competitions for NIH and AHRQ research grants. Moreover, like Kevin Murphy and Robert Topel (can't find a good download) I think that in the long run, medical research is one of the best investments our society makes, and therefore we should increase the resources that we spend on it.
But it needs to be done gradually and rationally. Otherwise with universities and research institutes building science towers on spec,
The results are not pretty. With success rates for acquiring an NIH grant below 10% in some cases, achieving a stable research career now has elements of a lottery, with one's future depending on a chance ranking assigned through a peer-review process that is unable to discriminate adequately among a sea of research proposals. Biomedical scientists are spending far too much effort writing grant applications and reviewing those of others, leaving precious little time to do what they should be doing: reading the scientific literature and thinking deeply about their research and teaching. And I worry about the future of universities, given the incentives to overinvest in research at the expense of needed facilities and faculty for teaching.
Alberts argues that current federal funding policies require institutions to commit less of their own cash up front than they should. His remedy is that
A new NIH policy must make it unambiguously clear that expansion through laboratory building construction requires a substantial, nonreimbursable, long-term commitment of resources, including "hard-money" faculty support, by any institution that wants to increase its facilities and research staff.
This is, by the way, how the Canadian research grant system works. The Canadian Institutes of Health Research do not allow you to put your salary on a grant, and hence your institution has to be committed to supporting you with hard money.