‘Speaking Freely’: Partisanship and ideology

This is the second installment in a series of posts delving into the results of FIRE’s ‘Speaking Freely’ report on college students’ attitudes toward expression on American campuses. In this post, we will talk about the partisan and ideological breakdown of our sample, and how similar our respondents are to college students that have been surveyed elsewhere. Next week we will explore the second section of the ‘Speaking Freely’ report, which focuses on how students react to the expression of their peers.

Political partisanship and ideology

Political partisanship is a demographic characteristic that is used by survey analysts and students of political behavior to understand people’s attitudes toward politics. Before presidential elections, many polling companies and news organizations track the partisanship of registered voters carefully, because partisanship is a factor that closely influences who individuals vote for on election day. Although Americans may consider themselves to be members of any number of political parties — from the Democratic and Republican parties, to the Libertarian, Communist, Working Families, or Green Parties — American surveys usually follow the American National Election Study (ANES) lead and limit responses to Democrat, Republican, or Independent.

When we surveyed college students for our ‘Speaking Freely’ report, we followed the ANES tradition and worked with YouGov to ask our sample of 1,250 college students which political party they felt closest to, so that we could use both three- and seven-point partisanship scales to better understand students’ attitudes toward issues of free expression on college campuses. Here’s a quick explanation of what that means: A person’s self-identification with the Democratic or Republican party, or as an Independent, is their partisanship. When these are the only three answer options, this self-identification is known as a “three-point” partisanship scale. There is an additional “seven-point” partisanship scale that asks how closely an individual feels to the Democratic or Republican party. This scale includes the following options: strong Democrat, lean Democrat, weak Democrat, Independent, weak Republican, lean Republican, and strong Republican.

Political ideology is another demographic characteristic used to understand people’s attitudes toward politics. Ideology is a more complicated concept to define than partisanship because a person’s ideology is based on his/her views toward a number of social, economic, and political issues. Because of this, many polling companies and scholars create ideology scales based on a respondent’s answers to multiple questions about current social and political issues. For example, Pew updates its political typology questions every few years and allows members of the public to take a quiz to determine their closest political typology. Although it is true that political ideology is multi-faceted, for our ‘Speaking Freely’ report, we approached the question in a way that was simpler for respondents, and worked with YouGov to ask respondents whether they self-identify as one of the following: very liberal, liberal, moderate, conservative, or very conservative. This creates a five-point ideology scale similar to the seven-point ideology scale used by the ANES.

The partisan and ideological breakdown of college students

Although organizations like Pew and the ANES have collected excellent data about the political partisanship and ideology of American adults over time, no data source has consistently tracked the political partisanship of American college students. The Higher Education Research Institute (HERI) has collected data on college freshmen’s ideology since 1966, but has only started collecting similar data on college seniors in 2007. This limits scholars’ ability to look at how students’ partisanship and ideology change through their college experience, and prevents those who study public opinion from having strong baseline statistics to use when analyzing college students’ attitudes toward political issues.

Although we are not completing a comprehensive literature review on the partisanship and ideology of college students, it is important to note that there are excellent academic works on the subject. For example, Abramowitz analyzed college student partisanship, Jennings et al. looked at how family socialization affects the partisanship of young adults, and other pieces (found here and here) have analyzed how the college experience changes the ideology of young people.

Despite the shortcomings with current data, it is important to compare the partisan and ideological breakdown of our survey’s respondents to the partisan and ideological breakdowns found in similar surveys of college students. In Tables 1 and 2, we refer to the following surveys: HERI 2016, Villasenor 2017, Bucknell 2017, and FIRE 2017.

Table 1: Political ideology of college students

Table 2: Political partisanship of college students

Looking at the information presented in Table 1, we are encouraged to see that the ideological breakdown of our sample is similar to that reported by HERI earlier this year. HERI reports that 35% of students self-identify as liberal or very liberal and 22% of students self-identify as conservative or very conservative. In our sample, 34% of students self-identify as liberal or very liberal, and 23% self-identify as conservative or very conservative. There is an 8 percentage point difference between the proportion of students in our sample that self-identify as moderate, compared to the proportion of students that self-identify as moderate in the 2016 HERI survey. This is likely due to the additional “I don’t know” answer option that we allowed students to choose when completing the ‘Speaking Freely’ survey.

Additionally, our partisanship data closely mirrors that analyzed by Rothman et al. in The Still Divided Academy. They report that 32% of college students self-identified as Democrats in 1999. Our respondents, surveyed in 2017, are similar in this regard, as 33% of them self-identified as Democrats. Additionally, in both surveys, 7% of students report not knowing their political partisanship. Our samples included different proportions of Republicans, however. In 1999, Rothman et al. found that 26% of students self-identify as Republican, whereas 20% of our sample self-identified as Republican. As you can see in Table 2, these students appear to have shifted to self-identifying as Independent. Although we don’t have enough information to present a decisive answer as to what caused this partisan change in the eighteen years between when our respective samples were surveyed, we do know that other organizations have found similar statistics. Pew reports that, compared to the generation that came before them, Millennials are more likely to identify as Moderate than Republican. The difference in partisanship between our data and Rothman et al.’s data likely reflect this generational turnover.

Thank you for your continued interest in our ‘Speaking Freely’ survey. If you would like to find out more about our survey, check out our post on college students’ self-expression, or check back next week to read about students’ reactions to the expression of their peers.