College roommates influence each other’s political ideology

Universities have become an important front in America's "culture war." Today, many conservatives are deeply skeptical of–or even hostile to–institutions of higher learning. Pundits and politicians on the political right routinely accuse colleges and universities of "indoctrinating" students into "leftist" political ideologies. And while prior research has found that college graduates tend to be more liberal and more likely to vote for Democrats, there has been no scientific study indicating that trends result from some causal force in universities.

So, does college cause students to become more liberal? And if so, why? The stakes of answering these questions are quite high, as universities face increasing pressure from Republican-controlled state legislatures around the country.

We fielded an original panel survey that allows us to address these questions. Our survey was fielded on the full freshman classes at two major universities in the US. We asked questions directly linked to their ideology, including political views, far left or right, liberal, conservative etc. It allowed us to collect data on over 1600 students who completed both waves of our survey (one before and one at the end of their freshman year), and whose roommate completed at least one wave of the survey. These high-quality data give us unprecedented leverage to address questions of ideological change at universities.

We find that, on balance, student ideology did not change much over the course of the freshman year. Only 24% of the students changed their ideological self-identification over their first year, and their change in ideology was minor. Very importantly, given claims of indoctrination, among those who changed most became more moderate. Indeed, the students overall became slightly more conservative after the freshman year.

We then turn to the question of what explains the change we do see. Since we suspect that peer socialization is a key pathway of influence, we take advantage of the quasi-random roommate assignments to see whether students are influenced by the ideology of their roommate(s). We find that when students are randomly assigned to a roommate who is more conservative than they are, the student tends to become a little more conservative over the course of the freshman year. Conversely, when students are randomly assigned to a roommate who is more liberal than they are, they tend to become a little more liberal themselves over the course of the year. This finding demonstrates that roommates influence each others' political views.

Ours is the first to rigorously examine changes in student ideology over the course of the freshman year of college. For this reason, our research design allows us to weigh in on this important topic in contemporary politics. Despite the widespread indoctrination claim, our results show that there is no evidence of broad left-ward ideological movement among students–most students' ideology is unchanged over the course of the year. Even so, students whose ideology does change seems to be a function of socialization, with roommates reporting more similar ideologies after a year together.