America’s college-admissions debate has been shaped by a small number of universities with many more qualified applicants than seats. Most colleges — and many programs inside selective universities — operate under very different conditions. As international graduate enrollment falls, the consequences can reach well beyond campus.
MADISON, Wisconsin, September 21, 2026 — Much of America’s debate over college admissions begins with Harvard. There are good reasons for that. When a university has far more qualified applicants than seats, the rules used to decide who gets those seats matter enormously. Evidence of differential treatment deserves scrutiny, and that principle should not depend on which group is helped or hurt.
Before the Supreme Court’s 2023 decision in Students for Fair Admissions, Harvard, like many selective universities, explicitly considered race in admissions. At an institution where many qualified applicants compete for very few seats, preferences are consequential because helping one applicant changes the chances of another. But institutions like Harvard are not representative of most American higher education.
A new report from the Center for Research on the Wisconsin Economy (CROWE), “Our College System Would Go to Hell Very Quickly”: America’s Two College Admissions Markets, by Junjie Guo and Ananth Seshadri, examines what changes when the focus moves beyond this small group of unusually selective universities. 88% of four-year colleges are open admission or admit at least half of their applicants. Only 31 institutions (1.4% of the total) admit fewer than one applicant in ten. In fall 2023, Harvard admitted 3.5% of applicants, compared with 43% at UW–Madison and 77% at Minnesota–Twin Cities.
“When seats are truly scarce, admissions decisions have real consequences,” said Seshadri, co-director of CROWE and professor of economics at UW–Madison. “But most of higher education is not Harvard. For many colleges and programs, the difficult question is not simply whom to admit. It is whether enough of the students they admit will actually show up.”
That difference is especially clear in graduate education. UW–Madison’s Graduate School admitted 23.3% of applicants for fall 2025, but only about 39% of admitted students enrolled. International students accounted for 58% of applicants, 43% of admits, and 33% of new enrollees. Similar patterns appear at the University of Minnesota and in University of California academic master’s programs. A program can reject applicants and still compete hard for the students it admits.
Different parts of the same university can also face very different conditions. An undergraduate college may have many more qualified applicants than seats. A PhD program may be constrained by the number of funded positions it can support. A tuition-dependent master’s program may instead be trying to persuade enough admitted students to enroll. Among 28 highly selective private research universities examined in the report, first-year undergraduate enrollment grew only 2.9% from 2019 to 2023, while engineering and computing master’s degree production grew 49.2%.
That distinction changes how to think about a common question surrounding international students: does admitting an international student mean that an American loses a seat? If capacity is fixed, it can. But universities do not always operate with fixed capacity. Evidence reviewed in the report shows that stronger international master’s demand can lead universities to expand programs and increase domestic graduate enrollment. Displacement can occur, but the answer depends on whether capacity responds.
International demand is now moving sharply in the other direction. International enrollment in science and engineering master’s programs fell from roughly 211,000 students in fall 2024 to 161,000 in fall 2025 — a decline of 24%.
Institutions entered that decline from very different positions. International students accounted for 4.5% of graduate enrollment at the median institution, 19% at Georgetown, 29% at UW–Madison, 55% at North Texas, and 75% at Clark University. At the extreme, international students accounted for 98% of graduate enrollment in Trine University’s separately reported regional/nontraditional unit.
What happens when those students do not come depends on the institution. A university with a deep applicant pool may enroll somebody else. A program closer to the enrollment margin may have to recruit more aggressively, accept a smaller class, or eventually shrink. A university can replace a missing student with someone who otherwise would have enrolled elsewhere, shifting rather than eliminating the loss.
“If an international student does not come, it does not follow that an American simply takes the seat,” Seshadri said. “At some universities that may happen. Elsewhere, the class may simply get smaller. And when a college or program contracts, the effects need not stop at the campus boundary.”
Colleges are also employers, customers, and anchors of local economies. Students pay tuition, rent apartments, buy food, and spend money where they live. When enrollment falls far enough, the effects can reach university budgets, programs, employment, and businesses in the surrounding community.
Highly selective universities have played an outsized role in shaping public perceptions of college admissions. Gallup polls show that confidence in higher education has fallen sharply over the past decade. Americans are often asked to understand a large and varied sector through the admissions practices of institutions that are least representative of how most colleges operate.
The report’s title comes from remarks President Donald Trump made during an August 2025 Cabinet meeting about foreign students. Trump warned that without them, “our college system would go to hell very quickly,” but the more specific observation came next: the pressure, he said, would fall not on the top colleges but on colleges “that struggle on the bottom.” On that economic point, the evidence in this report supports the distinction Trump drew. Institutions with deep applicant pools have more ways to replace students who do not come. Colleges and programs operating closer to the enrollment margin have fewer options, and the consequences of lost enrollment can extend well beyond the campus.
Admissions power depends on which side of the market is scarce.
Read the full report: “Our College System Would Go to Hell Very Quickly”: America’s Two College Admissions Markets.
About CROWE
The Center for Research on the Wisconsin Economy (CROWE) was established in 2017 in the Department of Economics at the University of Wisconsin–Madison. CROWE supports and disseminates economic policy research from a market perspective, with a particular focus on Wisconsin and state-level policy. Its work combines original research, student engagement, and public outreach.
For more information, visit crowe.wisc.edu.
For questions about the report, contact Ananth Seshadri at ananth.seshadri@wisc.edu.
