Back in July, Erin Shaw and I wrote about results from a survey showing that 29% of prospective first-year college students reported removing a college from their “might apply” lists based on political reasons. We noted the obvious consequences of political self-sorting for individual intellectual growth, but also the general learning climate on campus. Since then, a new working paper suggests that not only has this sorting been in motion for decades, but students would pay a pretty penny to avoid political opposition on campus.

Using data from The Freshman Survey administered by UCLA since 1966, the researchers assess the extent to which the proportion of students identifying as liberal, middle of the road, and conservative has changed over time at various types of institutions. Due to variation in data availability, the main analysis is restricted to 1982 - 2019, including first-time students, who responded to the political views question, and for which at least 100 students from that institution responded. The final data analysis is based on 7+ million students from over 1,000 colleges and universities.

A few notable trends emerge. First, students are shifting from “middle-of-the road” to relatively more liberal. This increased liberalism among students is growing disproportionately at liberal arts colleges, HBCUs, and research universities, with the biggest shifts occurring at the most selective or “elite” universities. Religious colleges, on the other hand, are becoming more conservative.

Using regional public colleges as the baseline for comparison, the researchers report that “the share of students identifying as liberal or far left has increased by 1.4pp per decade at public research universities, 2pp per decade at private research universities, 2.2pp per decade at HBCUs, and 2.5pp per decade at liberal arts colleges,” and “the most selective colleges (5th quintile) have seen a 3.4pp increase in liberal students per decade and a corresponding 3.3pp decrease in the share of conservative students.”

So, our most elite universities and our major research universities are becoming quite a bit more liberal, while religious colleges are becoming more conservative—trends that have been in motion since at least the 1980s.

But most interestingly, this paper also shows that students would literally pay thousands of dollars more for tuition if they could spare themselves the discomfort of being around students who are politically different from them.

To find out just how much students would pay to be around their political allies, the group used a survey to ask more than 1,000 undergraduate students in June 2025 to choose between various comparisons of hypothetical colleges that varied on key attributes (cost of attendance, student body size, institution type (private-religious, private-non-religious, public), quality (measured by average SAT/ACT scores), metropolitan area size, distance from home, state political leaning, and student body political leaning) to isolate the effect of politics on students decision making.

Across these scenarios, liberal students preferred campuses with more liberal students and fewer conservative students, whereas conservative students preferred campuses with fewer liberal students but without an explicit preference for more conservative students.

The authors also calculated a dollar price reflecting what students were willing to pay for these preferences:

In the full sample, [liberal or far left] students are willing to pay $1,162 more to attend a college with a 10pp greater share of students identifying as liberal, $2,617 more to attend a college with a 10pp lower share of students identifying as conservative, and $3,064 more to attend a college in a state with a 10pp higher share of Democratic voters… [Conservative or far right] students are willing to pay $2,201 more to attend a college with 10pp lower share of liberal students and $2,720 to attend a college in a state with a 10pp lower Democratic vote share. [These] students are not, however, willing to pay more to attend a college with a higher share of conservative (vs. moderate) students.

The bottom line is that both liberal and conservative students put a high value on seeking out campuses with fewer peers who politically disagree with them, and liberal students are also willing to pay to find campuses with more political agreement. When students avoid the discomfort of political adversaries, they miss the point of a college education, especially those who chose to study social disciplines.

Liberal students, perhaps, have the most to lose in this self-sorting game.

Student populations, especially at our nation’s research universities and elite institutions, are more liberal than conservative, with faculty populations that also skew liberal, located within mostly democratic states or regions. In such environments, liberal students simply don’t get the pushback or exposure to genuine ideological difference that the conservative students on most campuses do.

As Lauren A. Wright wrote last year in The Atlantic, “These [intellectual and social] challenges impart educational advantages by forcing conservatives to defend their points of view. Liberal students, surrounded by like-minded peers and mentors, have less opportunity to grow in this way.” Sure enough, in her interviews with Princeton undergraduates, liberal students were less familiar with and less prepared to refute conservative views than conservative students were with liberal views. Liberal students at Yale also agree: “While liberal students are cushioned by a sense of majority, conservatives must grapple constantly with difficult questions. They are forced to interrogate their own beliefs in ways their liberal peers need not.”

Students seem to prefer avoiding their ideological opponents, but research tells us that contact with political opponents actually helps reduce polarization. If we want our universities to be places where viewpoint diversity can benefit all on campus, we must show students that ideological bubbles are the problem, not the solution.