top of page

Be Notified of New Research Summaries -

It's Free!

Does State Funding for Public Universities Produce More STEM Graduates?

Writer: Greg Thorson
Greg Thorson
6 minutes ago
7 min read


El Fatmaoui (2026) finds that a 10% increase in state appropriations for public four-year universities raises the number of STEM bachelor's degrees they award by about 3.4% four years later, while non-STEM degrees show little consistent response. The study examined whether state funding affects degree production differently across fields. It analyzed federal data on degrees, spending, and programs at more than 600 public universities from 2003 to 2019, using each institution's historical reliance on state funding to isolate the effect of budget changes. The effect is concentrated in the biological, mathematical, and physical sciences, among male graduates, and at less selective institutions. Additional funding also increased spending on academic support and the number of STEM programs offered.


Why This Article Was Selected for The Policy Scientist

State appropriations remain a central source of support for public four-year universities, which educate a large share of the nation's undergraduates, including many students from lower-income families. These appropriations have fluctuated considerably over the past two decades, even as policymakers have placed growing emphasis on expanding the science and technology workforce. Whether changes in state budgets affect some fields more than others is a timely question as legislatures weigh higher education against other priorities. Published in Education Finance and Policy, a respected journal in education policy research, this study extends a substantial literature on state funding by showing how its effects differ across academic fields. Its national panel of more than 600 public universities over 17 years provides a broad empirical foundation, and its instrumental variables design, supported by placebo tests and alternative specifications, strengthens causal inference.


Full Citation and Link to Article

El Fatmaoui, A. (2026). Dollars and degrees: The asymmetric impact of state appropriations on STEM and non-STEM fields. Education Finance and Policy. Advance online publication.


Central Research Question

Do changes in state appropriations affect the number of bachelor's degrees that public four-year universities award in science, technology, engineering, and mathematics (STEM) differently from the number they award in other fields? The study also asks whether any such effects operate through spending, program offerings, or tuition, and whether they differ by selectivity, discipline, gender, and race.


Between 2001 and 2019, state appropriations for public higher education fell 2.2% in constant 2019 dollars, from $82.6 billion to $80.8 billion, while full-time-equivalent enrollment grew 25.3%. Meanwhile, the President's Council of Advisors on Science and Technology recommended in 2012 a 34% increase in STEM graduates. Because many STEM programs are more expensive to deliver than other programs, reductions in state support could fall disproportionately on the fields policymakers have sought to expand.


Previous Literature

A substantial body of research documents how public universities and their students respond to changes in state funding. Deming and Walters (2017) and Bound et al. (2019) find that universities reduce spending when appropriations fall, and Webber (2017) shows that state divestment raises in-state tuition. Bound et al. (2020) find that a 10% reduction in state funding leads to a 16% increase in foreign undergraduate enrollment. Chakrabarti et al. (2020) find that funding reductions increase student debt, and both they and Bound et al. (2019) find larger effects for lower-income students and less selective institutions. Goodman and Volz (2020) show that lower statewide appropriations reduce public enrollment and increase for-profit enrollment.


A second body of research establishes that the cost of instruction varies considerably across fields. Studying Florida's public universities, Altonji and Zimmerman (2019) find per-credit costs of about $450 in engineering and the health sciences, compared with $200–$250 in the social sciences, mathematics, business, and psychology. Hemelt et al. (2021) similarly find that engineering programs cost more than English programs, largely because of higher faculty wages and lower teaching loads. Although the importance of STEM education for economic growth and social mobility is well documented (Wolniak et al., 2008; Carnevale et al., 2011), how state funding affects STEM and non-STEM degree production differently has received little attention. The present study extends this literature by estimating these field-specific effects and examining the institutional channels through which they operate.


Data

The study uses the Integrated Postsecondary Education Data System (IPEDS), the federal data collection covering U.S. colleges and universities, to construct a panel of public four-year institutions from 2003 to 2019. IPEDS records the number of bachelor's degrees awarded in each detailed major at each institution, along with institutional spending and tuition. State appropriations come from the State Higher Education Executive Officers Association's Grapevine database, which provides more complete state figures than IPEDS. Programs are classified as STEM using the Department of Homeland Security's STEM Designated Degree Program list. The models also control for state and county economic and demographic conditions.


The STEM sample covers 617 institutions and 339 STEM majors, and the non-STEM sample covers 663 institutions and 783 majors. Between 2003–2007 and 2015–2019, average state appropriations per institution fell by $15.4 million, or 15.3%, from $99.9 million to $84.5 million. Over the same period, average STEM degrees per institution rose 67%, from 319 to 533, while non-STEM degrees rose 11.5%. Because IPEDS records counts of degrees and programs rather than individual students, the data cannot follow students from enrollment to graduation or distinguish newly created programs from reclassified ones.


Methods

The study employs a Bartik-style (shift-share) instrumental variables design to estimate the causal effect of state appropriations on degree completion. A simple comparison could be misleading, because funding may respond to institutional performance or need. The instrument combines each institution's historical reliance on state funding, measured as the share of its revenue from state appropriations averaged over 1987–1991, with current statewide appropriations per college-age resident. Institutions that historically depended more on the state are more exposed to changes in its budget, and measuring that reliance more than a decade before the study period limits the influence of current circumstances.


The models estimate elasticities, the percentage change in degrees associated with a 1% change in appropriations, for degrees awarded three to six years after a funding change, and include institution, major, and year fixed effects. The instrument strongly predicts actual appropriations, with first-stage F-statistics between about 100 and 680. Robustness checks include more demanding fixed effects, the exclusion of the most influential institutions, an alternative instrument based on aggregate state appropriations, and a placebo test that randomly assigns STEM labels to non-STEM majors 10,000 times.


The design assumes that historical reliance on state funding affects degree production only through its interaction with later appropriations. The instrument is unrelated to most baseline characteristics and to pre-existing trends in STEM programs, although it is correlated with baseline state per capita income. The analyses linking spending and program offerings to degrees rely on ordinary least squares with sensitivity tests for omitted variables, and the author interprets them as descriptive associations rather than causal effects.


Findings/Size Effects

The principal finding is that a 10% increase in state appropriations raises the number of STEM bachelor's degrees awarded in each major by about 3.4% four years later, the standard time to graduation (significant at the 5% level). Across students graduating in four, five, or six years, the STEM elasticities range from 0.24 to 0.34. Non-STEM estimates are statistically insignificant at four years and significant only at the 10% level at five years, where the STEM estimate is about 1.6 times as large. At average values, roughly $10 million in additional funding corresponds to about one additional four-year graduate in each STEM major. Appropriations also increase STEM master's degrees, but not doctoral degrees.


The effects differ substantially across institutions and fields. They are statistically significant only at institutions outside the most selective, which can more readily replace lost state funds with gifts, endowments, and nonresident enrollment. Within STEM, the effect is concentrated in the biological and biomedical sciences, mathematics and statistics, and the physical sciences, with a negligible effect for engineering overall. The response is concentrated among male graduates, with elasticities of 0.31 at four years and 0.30 at five years (significant at the 10% and 5% levels), while estimates for female graduates are small and statistically insignificant. There are no significant effects for Hispanic graduates in any major.


The mechanism analysis indicates that state funding operates mainly through institutional resources rather than prices. Higher appropriations increase total, instructional, and academic support spending (such as tutoring and advising), with larger effects on academic support than on instruction. Appropriations also increase the number of distinct programs offered, with an effect on STEM programs about 1.6 times that on non-STEM programs. These resources are more closely associated with STEM degrees: the association with total spending is about 64% larger for STEM, the association with academic support spending more than three times as large for four-year graduates, and more programs are associated with more degrees only in STEM. A 10% funding increase reduces sticker prices by 1.04% and net prices by 1.8%, effects the author considers too small to account for much of the result. The STEM results remain robust to the placebo test and to excluding the most influential institutions.


Conclusion

The study provides credible evidence that state appropriations affect STEM degree production at public four-year universities more than degree production in other fields. A 10% increase in funding raises STEM bachelor's degrees by about 3.4% four years later, with the effect concentrated in the sciences and mathematics, among male graduates, and at less selective institutions. The evidence on mechanisms is consistent with funding operating through academic support spending and the expansion of STEM programs, although these links are descriptive.


Because the data record degrees rather than students, the study cannot show whether funding changes alter who enrolls, who persists, or who switches majors. The author suggests that reductions in state support may have disproportionately large effects on STEM education and that less selective institutions may warrant particular attention in funding decisions. The author also notes that university assets per student have grown even as operating support declined, and recommends research on how such long-term investments affect student outcomes and on why effects are absent for Hispanic students.

Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
Screenshot of Greg Thorson
  • Facebook
  • Twitter
  • LinkedIn


The Policy Scientist

Offering Concise Summaries*
of the
Most Recent, Impactful 
Public Policy Research

*Summaries Powered by ChatGPT

bottom of page