The US federal government gives professors and universities tens of billions of dollars per year to conduct research. There are problems with how we allocate these funds—several federal research agencies are risk averse, and funding decisions have at times been captured by insiders. New Science’s report on the NIH gives a detailed and nuanced overview. Nevertheless, our research system is the envy of the world, and few believe we should cut research spending.
To get the most out of our research system, we need to introduce competing funding streams. ARPA-H is promising, for example, not just because ARPAs are effective, but also because the agency introduces a new mode of health research funding that is independent of NIH study groups, although sadly, not of the NIH itself.
Yet it is a mistake to place the entire burden of research funding diversity on the federal government. Civil society can and should fund basic research. Non-government research funding is valuable both for the discoveries that it directly enables and also for its salutary effect on the overall research ecosystem through competition and funding diversity. For example, the private Fast Grants project was able to quickly deploy money for Covid-related research at a time when federal agencies were simply too slow to respond, yielding many discoveries that would not have happened if the researchers had to apply for federal funds.
Fortunately, we have non-federal institutions that conduct research, are swimming in money, and have a non-profit mission theoretically consistent with basic research funding: our elite universities. Unfortunately, several highly-endowed universities spend relatively small proportions of their own funds on research. It’s worth taking a close look at how highly-endowed universities could contribute to diversifying research funding if they wanted to—or they were required to.
Data on endowments and research
The National Center for Education Statistics maintains a list of the top 120 university endowments in the country. The numbers are staggering. There are a dozen endowments worth more than $10 billion. The biggest endowment belongs to Harvard University, with over $41 billion at the end of the 2020 fiscal year. These enormous numbers have led some to joke that Ivy League schools are really hedge funds with universities attached. And because these universities are non-profit entities, the funds’ profits are reinvested year after year without paying any taxes.
Data on research funding is available from the National Center for Science and Engineering Statistics, part of the National Science Foundation. The NCSES’s Higher Education Research and Development Survey records how much each university spends on research as well as the source of funds.
Combining data from these two sources lets us see how much highly-endowed universities spend on research as a fraction of their total endowments. For example, in fiscal year 2020, Harvard University spent $1.24 billion on research, of which $606 million came from the federal government and $391 million from its own coffers. In other words, Harvard spends about 0.93% of its endowment on research. (These numbers were not significantly affected by Covid). To get research spending of its own funds to be equal to a mere 2 percent of its endowment would require an additional $447 million in annual spending.
Many institutions spend much more than 2 percent of their endowments on research annually. UC-San Francisco and Indiana University-Bloomington each have endowments over a billion dollars and spend an amount equal to more than 20 percent of their endowment on research. The University of Utah, Baylor College of Medicine, Virginia Tech, University of Georgia, University of Florida, UCLA, Iowa State University, University of Missouri, Rutgers, University of Texas Southwestern Medical Center, University of Illinois at Urbana-Champaign, and University of Arkansas all spend funds greater than 10 percent of their endowments. Across the 82 universities that have endowments greater than $1 billion and receive federal grants greater than $5 million, the average percentage is 6.16 and the median is 3.63.
The elephant in the room is that all of the deficient institutions are private, while most of the universities that spend more relative to their means are public. These data show that elite, well-endowed private universities differ in their mission from public universities in at least this one important way.
If the private schools in the chart above actually spent 2 percent of their endowments on research every year, it would add up to more than $2.4 billion in additional research funding per year. That’s basically like adding a new ARPA-H.
Could we make elite private universities spend more of their own funds on research?
We could strongly induce elite private universities to spend their own money on research by making it a condition of receiving federal grants. The federal government already imposes numerous requirements on federal grant recipients, most of which are to do with administration, compliance, and auditing. These rules are burdensome but surely necessary in some form, if not the exact shape they exist in today.
Adding a requirement that institutions with an endowment over a billion dollars spend at least 2 percent of their endowment value on direct research, in contrast, could be simple. It would require minimal additional paperwork. The burden on the institution would be substantive, not performative.
Alternatively, we could impose some additional requirements that might get us more bang for the buck. For example, we could require that the 2 percent cover direct costs of research, not indirect costs like research buildings, utilities, janitors, and so on.
Another important consideration is that the additional institutional funding should be uncorrelated with federal funding decisions. Suppose a university decided to get to its 2 percent requirement by simply topping up every professor who received a major federal grant—in that case, we wouldn’t be introducing any additional institutional competition into the system. Alternatively, if the university used its 2 percent requirement to fund only grant applicants who were denied federal funding, it would likewise be parasitic on federal agencies to make funding decisions. Neither option would add any dynamism to US research funding.
Instead, a university could run their own internal grantmaking system, or follow a system of just dividing research funds equally and unconditionally between eligible researchers. Either would introduce an element of competition into the research ecosystem. Researchers could decide to apply for the internal grant instead of a federal one or try to make do with unconditional basic research funding, cutting time spent on grant applications. In a Fast Grants survey, the median respondent said she spends 25-50 percent of her time applying for grants. 82 percent of respondents said that if they had unrestricted funding, their research program would become more ambitious.
One thing is for certain: multi-billion-dollar nonprofits should not just accumulate funds without providing public benefits. Tax-exempt status does not exist for the purpose of empire building. A good way for the Ivy League to start pulling its weight would be to increase its own research spending.