By Gus Hoffmann

In Adam Weinberg’s welcome-back email, he encouraged students to “Develop a wide set of friendships, especially with people who are different from you.” However, during these politically-charged times that can be a difficult task to accomplish. With increasing political fissures regarding Covid-19, a wide variety of social justice issues, and economic policy it is often easier to converse about these issues with people who share similar political beliefs. Back in 2016, Oliver Gladfelter wrote a post titled “How Segregated is Denison?” Oliver found that in 2016 the student population demonstrated homophily, the idea that people develop friendships based on shared ideologies and political preferences. We at 127 were curious about whether the Denison student body in 2021 demonstrated less homophilous behavior.

We are drawing on data from February of 2020 from around 500 student participants, in which 19% identified as Republican. Though there are other measures of it, this does not show a lot of political diversity at Denison. The survey also posed the question, “Out of the 5 people you are closest to on campus, how many are Republicans?” Paired with the demographics of survey participants this study led to interesting findings regarding bipartisan friendship at Denison.

If friendships are randomly mixed on campus, then the response should be that 1 out of 5 (~20%) are Republican. And that is exactly what we found. For individuals answering the survey, one of their five closest friends at Denison is a Republican on average. However, that average is not true of everyone. As we would expect if politics is the basis of friendships in a polarized age, the strength of partisan identity skews the average of their Republican friends. “Strong Republicans” who answered the survey indicated 3 of their closest 5 friends are Republican, while “strong Democrats” show on average that just 1 of every 13 of their friends are Republican. Unsurprisingly, on Denison’s campus Republicans have more Republican friends (see the figure below) and only independents come close to the expected value of 1 out of 5.

I was curious if partisan bias was reinforced by any of the demographic and organizational divisions on campus. Or if US politics, which has been deeply divided by and over race for a long time, shows any replicated divisions on campus. In the figure below, there is almost a persistent gap where white students have more Republican friends. That is especially true toward the middle of the partisan scale and among weaker Republicans. There is no gap among those who identify as strong Republicans, for whom ⅗ of their closest friends are Republican, and somehow the few non-white strong Democrats report a lot (⅖) of Republican friends.

Other elements of Denison civil society may affect bipartisan friendship, especially greek involvement. Republicans are more likely to be in sororities and fraternities (not so much among women, but among men) and, overall, those in the Greek system report more Republican friends (about .25 more). Moreover, non-whites are not that involved in greek life, which may compound the social skew, but among those who are, they have more Republican friends – in two cases non-white Greeks have more Republican friends than white Greeks. Outside of the Greek system, the racial gap in Republican friendship reappears. Non-white individuals seem to meet their Republican friends in the Greek system.

But the lessons of Greek life are not shared for other forms of involvement. In terms of group involvement and political diversity, the campus looks well mixed. Republicans are just as involved as Democrats when it comes to participating in clubs and other extracurriculars so diverse friendships follow. This conclusion aligns with the data – all partisan identities are involved in about 2.5-3 groups on average. The small faction of Independents are involved in a bit less, but not significantly so. Denison’s clubs and other organizations seem to exhibit bipartisanship, composed of individuals with different perspectives and political affiliations (which is good!).

Splitting the data by gender and race shows more extreme findings between the five closest friends of men and women. It is clear that there is a gender gap – men have more Republican friends than women. The gap in Republican friendship by race all but disappears among men – for example, both non-white and white men who identify as Republican have half of their five closest friends being Republican. Among women, however, sizable racial gaps emerge among independents and Republicans. Across the board, non-white women have almost zero Republican friends (interestingly a bit more among non-white Democratic women). Due to this gender gap, the more women friends we have, the fewer Republicans are in their network – that’s true especially for men.

Denison urges its students to mix with those who are different from you. Adam Weinberg’s welcome-back email reiterated this point, the administration actively encourages this advice, and Denison organizations help make it happen. However, existing Denison demographics (race, gender, involvement in greek life) constrict people to hang out with people similar to them. Whether students prioritize these demographics about themselves or others do, I find homophilous behavior for a variety of reasons. There’s also no getting over how few Republicans there are on campus and that they tend to stick together. But because they are such a minority, Republicans have more Democratic friends than Democrats have Republican friends. No matter how much Denison likes to preach that the student body is in harmony, this data shows that there are deep political divides in friendships that follow long-salient demographic divisions.

Gus Hoffmann is a senior at Denison University where he majors in Global Commerce and one day hopes to win Lampson Late Night Trivia.