Universities Are Profiting Off Federal Research Grants at Taxpayers’ Expense—Trump Reforms Could Curb the Abuse

The Trump Administration has proposed a much-needed reform of how we reimburse so-called “overhead expenses” associated with federal research grants made by the National Institutes of Health (NIH). It is reasoned that research activity financed by the Feds involves not only paying researchers, buying lab equipment, and some travel but also increases needed administrative oversight, depreciation from wear and tear on buildings, and other “indirect” costs. Universities have administrators who negotiate these costs with the government. Some schools, such as Harvard, get a much larger percentage of overhead reimbursement than others. Presumably, a school that pays its Vice President of Research $400,000 may get a larger overhead percentage than another institution paying only $250,000. If Harvard had marble floors that were expensive to maintain in its laboratories, but nearby, Bridgewater State had cheap linoleum flooring with low-cost upkeep, Harvard would likely get bigger reimbursement—an invitation for what economists would call massive rent-seeking behavior.

The new rule, held up for now by wrangling in the courts, is that every college will get 15 percent of direct research costs for overhead expenses. There is no need for armies of expensive administrators on both sides negotiating overhead.

In reality, colleges currently make money off overhead reimbursement. A little basic economics: The appropriate measure of institutional financial burden is the marginal (extra) overhead costs resulting from a research project, not some broader average cost measure. At present, at the margin, universities profit from research overhead. Many universities use some of their overhead funds to give additional financial incentives to researchers to submit proposals, hoping to obtain more grants. The grants are a win-win for the school, both financially profitable and enhancing institutional academic reputation. Indeed, they are probably lucrative enough for both researchers and universities to contribute importantly to a huge problem, namely the growth in fraudulent research.

But research funding reforms have their limits. Those receiving funding are monopolists: they are hired to be the only research provider on a very specific topic. Why not advertise topics needing research and fund the lowest qualified bidder, as we do with highway and building projects? Since the prospects for research success vary with many things, including the past performance of the researchers, a hybrid system may be optimal, where both the quality of the research design and the past record of the researcher count a good deal, but so does the cost of the project.

I discuss this in more detail in my forthcoming book, Let Colleges Fail.

One approach: assign a score from zero to 100 for each research proposal, funding those with the highest scores. A large portion—say, 70 points—should be determined by traditional academic standards, such as the research’s potential value and the qualifications of the research team. The remaining 30 points should be based on cost considerations, with more points awarded to proposals where investigators request lower overhead reimbursement.

If you want to demand high overhead reimbursement—say, 50 percent of direct costs—you may do so, but it will significantly reduce your chances of receiving funding.

There are various alternative approaches. For example, in the request for proposals, it could be made clear that requests for less than $1 million will have a higher probability of acceptance than large grants. However, it is also possible that scientific knowledge and human progress would be better advanced by funding a small number of large grants—$10 million or more—rather than many smaller ones.

Given the magnitudes involved, we probably need more objective research on the effectiveness of federally funded research done by scholars without “skin in the game,” maybe even foreigners ineligible for funding. Are there economies of scale in academic research? Shouldn’t we again revoke the policy of using earmarks in federal appropriations to fund research projects at politically favored schools with inordinate amounts of lobbying clout? If projects have commercial applicability, how do we distribute royalty income that accrues? Should the 1980 Bayh-Dole Act governing this be reviewed and modified? How do you divide financial gains between researchers and the institutions they work for? Should funded institutions be required to affirm commitment to free expression and even institutional neutrality in political matters as a condition of grant eligibility, as seemingly suggested by Trump NIH director nominee Jay Bhattacharya? (I say yes). Can we protect scientific dissenters whose views are generally disdained but sometimes, like Galileo, are right?

Lots of questions needing answers. Stay tuned.