From coeducation to integration, institutions of higher education in the United States of America have been the regular target of politically driven criticisms. Chief among these criticisms in recent years has been the idea that institutions of higher education are lacking in ideological diversity. These critiques have had significant impacts on institutions of higher education, as colleges and universities have faced persistent disinvestment and restructuring as a result of partisan policy making. These criticisms are not, however, new and have been a hallmark of higher education since its inception in the United States of America. In response to these criticisms, a substantial amount of research has been conducted on student politics, student political behavior, and the politics of college campuses. This research, however, has been consistently driven by survey data and is dated, leaving room for a contemporary exploration of student politics. This study was designed with the limitations of previous research in mind, and with a focus on the modern information landscape and social media’s inextricable connection to contemporary expressions of political ideology, and utilizes digital trace data to investigate student politics, ideological diversity, and ideological skew on college campuses.
Using digital trace data collected from Twitter and a latent attribute analysis of that data, I constructed a novel dataset of 8,554 students representing 43 states and 139 unique institutions of higher education. The average estimated political ideology of the students in the dataset was -0.337, which represented a left of center political position. With respect to students’ information networks, the students in the dataset followed 43,958 unique information sources on Twitter, which had an average estimated political ideology of 0.4234, which represented a right of center political position.
The findings of the study indicate that, while the average college is moderate but leans left, there is no lack of ideological diversity on college campuses in the United States of America and previous survey-based research on the topic may over- and under-represent certain political populations. Institutional variables such as cost of attendance and institutional selectivity were not significantly predictive of student politics and campuses in general are exceedingly moderate, ideologically diverse, and not as politically extreme, specifically with respect to liberal skew, as they are accused of being. Similarly, the study finds that students consume a diverse swathe of information online, but that information is likely to be significantly more moderate than the political positions of most students. Similarly, students’ information networks were ideologically diverse, but that diversity was less prevalent in the information networks of more conservative students. Given a lack of alignment between student ideologies and the available information online, a theory of constrained choice online was proposed and substantiated. Finally, most students appear to prefer to associate with peers who share their political views and subsequently consume information that is aligned with those views, but extreme homophily, siloing, and selective exposure to ideologically consonant information is most prevalent among conservatives.
Implications from this study for research, practice, and policy largely centered on the utility of digital trace data as an alternative and novel data source and the reality that colleges and universities are not as politically extreme as they are perceived. In sum, student politics largely reflect the politics of the country as a whole and there is no significant liberal skew.