This website uses cookies.

By clicking the "Accept" button or continuing to browse our site, you agree to first-party and session-only cookies being stored on your device to enhance site navigation and analyze site performance and traffic. For more information on our use of cookies, please see our Privacy Policy.

  1. Home
  2. ASSA Annual Meeting
  3. 2026
  4. Preliminary Program
  5. Higher Education and Political Polarization
« Back to Results

Higher Education and Political Polarization

Paper Session

Saturday, Jan. 3, 2026 8:00 AM - 10:00 AM (EST)

Philadelphia Convention Center
Hosted By: American Economic Association
  • Chair: Joshua Goodman, Boston University

Do Faculty Affect Their Students’ Political Beliefs?

Micah Baum
,
University of Michigan
José Joaquín Endara Cevallos
,
University of Michigan
Annaliese Paulson
,
University of Michigan

Abstract

Funding and instruction at American universities has become highly politicized, in part due to claims by Republican politicians that Democratic college professors “indoctrinate” their students to become liberal. In this paper, we study whether there is scope for professors to do so. We first use voter rolls and voter history files linked to publicly available payroll records from 34 state flagship universities to document the partisanship of college professors and course instructors. We then use natural language processing tools to describe the content of individual course materials and study whether Democratic instructors assign different materials than their Republican or non-partisan counterparts. Finally, we use student transcript data from one university to link students to instructors and study the causal effect of instructors on student self-reported ideology and party registration. College faculty are predominantly Democrats, especially in the humanities and social sciences. Although Democratic instructors are more likely to assign course materials that conservative policymakers would deem controversial, this is entirely explained by the types of courses Democrats teach; we find little difference in course content when comparing sections of the same course taught by Republican and Democratic instructors. College students are mostly liberals when they enter college, and many become more liberal during college. However, faculty do not cause this leftward shift. This is because students sort across academic fields on their political beliefs, leaving little room for indoctrination. Exploiting plausibly random variation in when instructors teach a given course, we estimate precise null effects of faculty partisanship on student beliefs or party registration. Ideological diversity among college students is limited, but claims that this is because of what college faculty teach are not justified.

Political Polarization in Higher Education

Jacob Light
,
Hoover Institution
Gideon Moore
,
Stanford University
Sam Thau
,
Stanford University

Abstract

Public trust in higher education has declined sharply in recent years, amid growing concerns about perceived ideological bias in college classrooms. This project investigates the politicization and partisan slant of U.S. college course content over the last 25 years using a novel embedding-based method that contextualizes ideological language in course descriptions. Drawing on a unique, large-scale dataset of course offerings from over 750 colleges and universities, we quantify two key dimensions: (1) polarization—the degree to which course content aligns with liberal or conservative ideological positions, and (2) politicization—the presence of explicitly political themes. On both dimensions, we document measurable shifts in course content over our study period. We further explore whether course content is primarily shaped by student demand, instructor ideology (proxied by political donations), or institutional priorities, employing a series of quasi-experimental designs including a "movers" design and a structural student demand model. Our findings provide the first large-scale, data-driven assessment of ideological trends in college instruction.

The Long-Run Effects of Colleges on Civic and Political Life

Lauren C. Russell
,
University of Pennsylvania
William Marble
,
University of Pennsylvania
Michael J. Andrews
,
University of Maryland-Baltimore County

Abstract

Colleges and universities are among the most important local institutions. In this paper, we examine the long-term effects of college and university establishment on social and political outcomes in the United States. We use a quasi-experimental design that compares counties in which colleges are located to "runner up'' locations that were considered but ultimately not selected. We show that the presence of higher education institutions has significant impacts on civic engagement, social trust, and political preferences. Counties with universities exhibit higher rates of volunteering, have more civic organizations, and populations with higher levels of social trust. We find muted effects on other measures of social capital. We also find aggregate political effects: places with colleges cast more votes in elections --- implying greater political power --- even as their turnout rates are not significantly different from runner-up locations. Furthermore, while there are no significant partisan differences prior to 2000, there is a growing and significant divergence in presidential vote shares in favor of the Democratic Party after 2000. Finally, we find that places with colleges elect members of Congress that are increasingly liberal on both economic and racial issues. These findings suggest that universities play a causal role in shaping the civic and political culture of the communities in which they are located.

Changing Minds: How Academic Fields Shape Political Attitudes

Yoav Goldstein
,
Tel Aviv University
Matan Kolerman-Shemer
,
Hebrew University

Abstract

This study investigates how academic fields influence political attitudes, analyzing data from 310,000 students entering 477 American colleges between 1990 and 2015. The primary analysis employs a conditional-on-observables approach, leveraging extensive pre-college controls. Results indicate that choosing a major in business or economics, rather than humanities or social sciences, decreases liberal self-placement by 0.22 standard deviations (approximately 10 percentage points). To reinforce the causal interpretation, a complementary quasi-experimental analysis uses expansions of specific academic programs at the college level. This approach confirms that increased availability of business and economics programs reduces students' liberal self-placement, partly by influencing their choice of major.
Accounting for students' self-selection into majors, we find that these choices amplify political polarization, as students often select majors that align with and reinforce their existing political views. Major choices also account for approximately one-sixth of the gender gap in political ideology observed among graduating college students, reflecting women's greater propensity to choose liberal-leaning majors.
Furthermore, the study identifies significant effects of majors on policy attitudes. For example, students majoring in business or economics adopt more right-wing views on taxation and national healthcare policies but maintain relatively left-wing positions on social issues such as abortion, compared to peers in other majors.
Finally, the study explores peer influence as a channel. By examining variations in students' baseline ideology within the same major and college over time, we find that while students’ political ideologies strongly correlate with those of their peers, this relationship arises primarily through selection rather than influence. Our precise estimates indicate that peer effects account for less than 5% of the total influence of academic majors. The evidence suggests that alternative mechanisms, such as academic content, may play a more substantial role.

Discussant(s)
Riley Acton
,
Miami University
Ro'ee Levy
,
Tel Aviv University
Lois Miller
,
University of South Carolina
Joshua Goodman
,
Boston University
JEL Classifications
  • I2 - Education and Research Institutions
  • P0 - General