Federal Research Fund and STEM Education

Policy Research Working Paper Federal Research Funding and STEM Education Emily E. CookDevaki GhoseEkaterina Khmelnitskaya Policy Research Working Paper11411 Abstract This paper examines how federal science and engineeringresearch funding—although intended to advance research—affects degree production and programs offered in science,technology, engineering, and mathematics (STEM). Usingdata from 1971–2016, the study implements a triple-dif-ference design that exploits variation across colleges, time,and fields of study. The findings show that federal grants and 3.7 percent of undergraduate programs in science, tech-nology, engineering, and mathematics annually across 200U.S. research universities. The impacts are concentratedin biology and engineering, aligning with the prioritiesof major funders such as the Department of Health andHuman Services, the National Science Foundation, and The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about developmentissues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry thenames of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely thoseof the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and Federal Research Funding and STEM Education Ekaterina KhmelnitskayaUBC Sauder School of Business Keywords: federal research funding, higher education, STEM, major choice, innovation JEL Codes: H52, I23, I28, O31, O38 1Introduction Government research grants constitute a major source of support for academic researchworldwide.1In the United States, federal agencies spend over$60 billion annually on universityresearch and development, largely motivated by the goal of advancing scientific discovery.2Yetuniversities jointly produce research and education, raising a central question: do policies designed Concerns about the supply of STEM workers have long shaped policy debates over howgovernment policies can help expand the supply of skilled labor (Atkinson 1990; PCAST 2012;Bloom et al. 2020). While public research funding promotes innovation (Babina et al. 2023),research support and educational access are often viewed as competing priorities for scarce public This paper studies whether federal science and engineering (S&E) research funding—the mostprominent form of government research support—affects the production of human capital in STEMfields, measured through changes in the number of STEM graduates and degree programs at the We address this question by linking degree completion data from the National Center for Education Statistics (NCES) with agency-university-year funding data from the NSF’s NCSES,creating a panel that extends back to 1971. Our empirical strategy uses a triple-difference design,comparing STEM and non-STEM (control) outcomes within the same university over time as A potential concern with the triple-difference method is spillovers across fields—either positive,if S&E funding indirectly benefits other fields through universities reallocating resources internally,or negative, if students switch fields. The availability of multiple non-STEM fields as controlsallows identification to draw on alternative comparison groups, mitigating concerns that spilloversaffecting any single field drive the results. We also address potential concerns about spillovers by We document two main findings. First, increases in federal research funding substantially expandSTEM degree production at all levels of university education. Funding increases raise doctoraland master’s degree completion and generate sizeable spillovers to undergraduate education—even though undergraduate students are rarely directly supported by research grants. On average, the share of STEM human capital formation in the United States. Second, universities respond along the extensive margin by expanding capacity and broadeningthe range of educational offerings in funded fields. Aggregate annual S&E funding to universitiesleads to the creation of roughly 161 doctoral programs across research universities within sevenyears of funding—about 6.3 percent of the average number of doctoral programs offered. Similar Disaggregating the overall STEM effect by subfield— Engineering including Computer Science,Biology, Physics, Mathematics, and Economics — shows that the effects are primarily driven byEngineering and Biology. This reflects the funding composition: throughout most of our panel, over Two considerations help interpret these results. First, federal research funding is primarilyintended to support scientific research—not student education—with most spending directedtoward salaries, subcontractors, and equipment (NCSES 2024). Thus, our estimates reflect indirect triggers such complemen