A one dollar increase in federal funding leads to a 33-cent increase in non-federal funding at U.S. universities.
Federal spending on R and D spurs funding from the private sector and from state and local governments, according to researchers Margaret Blume-Kohout, Krishna Kumar, and Neeraj Sood. In Federal Life Sciences Funding and University R and D (NBER Working Paper No. 15146), the researchers estimate that a one dollar increase in federal funding leads to a 33-cent increase in non-federal funding at U.S. universities. This may be because the non-federal funders interpret successful applications for federal funding as a signal of quality. There is a greater increase in non-federal funding per federal dollar received among those universities that do not grant Ph.D.s, among lower ranked universities, and among universities that have historically received less funding institutions for which the signaling value of receiving federal funding may be greater.
Federal agencies spend billions of taxpayer dollars funding academic research each year. Especially in the life sciences, funding for academic research has risen rapidly over the last decade. In 2007, total federal obligations for R and D at academic institutions totaled over $25 billion, of which $15.5 billion was provided by the National Institutes of Health (NIH), primarily to support basic and applied research in the biomedical sciences.
The American Recovery and Reinvestment Act of 2009 (ARRA) provided an additional $8.2 billion to NIH to fund extramural life sciences R and D. This spending was justified as a means to speed economic recovery and on the grounds that it would promote long-term scientific progress, but some critics suggested that it might crowd out funding from other sources.
The authors use a panel of 272 U.S. universities, with data on federal and non-federal life sciences R and D funding spanning nearly a decade, to explore this issue. In contrast to the claims of crowd-out, their findings suggest that federal funding "crowds in" other spending. These findings differ in scale from previous studies, which find only an additional eight cents in academic spending, compared with the 33 cents found here. The use of the university as a unit of analysis, and a reliance on cross-sectional as well as time-series variation (as opposed to aggregate time-series variation alone) might partly account for this difference.
-- Lester Picker