Usual Suspects or Scapegoats? Role of (Social) Media in Political Polarization

• In Event: Polarization: Causes, Consequences and Remedies

Abstract

Evidence based on expert assessments, voting data, as well as ideological positions of the two main parties on the left-right political spectrum highlights how political elites in the United States have polarized over the last decades. Does the increasing polarization among the legislators, as well as more broadly, among politicians, spill over the public? If one were to accept this spill-over logic, the public would have to get information from politicians or about politics, in general, including about the degree of polarization. While the public does rely on the cues provided by the politicians, these cues first pass through the lens of media before they reach ordinary voters. In the last two decades, the frequency of media discussions of political polarization has increased. News media mostly cover ideologically polarized politicians – they are given a disproportionately large airtime.
Based on the American National Election Studies 2020 survey data, this paper uses a binary approach to political polarization, operationalizing it in terms of affect and ideology. The measure for affective polarization is measured by taking an absolute difference between the feeling thermometer scores of Republicans and Democrats, whereas ideological polarization is measured as the distance from the center of political spectrum, either to the left or to the right. Keeping in mind the distinction between the two dimensions of polarization and using robustness-tested multiple linear regressions, this paper argues that partisan identity, exacerbated by partisan media (online, TV, radio) use, activity on social media, and political discussions, are tied to more extreme political views. Media partisanship measurement is based on Ad Fontes Media and Allsides.com media bias rankings. A total of 15 online news websites, 42 TV programs, and 14 radio programs are included in the analysis.
The findings show that, in line with the theoretical expectations, receiving political information from partisan media is correlated with affective polarization, but the effect is weak. Out of three media sources – news websites, TV, and radio – the latter is the most polarizing, which is also manifested by the highest average bias as compared to the other two sources and many strongly conservative commentators. Against this background, use of news websites has a de-polarizing effect, whereas getting news from TV has an ambiguous effect in that it correlates with higher affective and lower ideological polarization, hinting at sensationalism of TV programs.
For social media, platform-specific differences are found. Facebook use (operationalized as visiting the platform) is correlated with lower polarization – something that does not apply to Twitter use. Conservative social media users display higher levels of polarization than liberals. Finally, both offline and online political discussions are correlated with higher polarization.
While providing insights that future research could build upon, the findings of the given study need to be taken with scrutiny, as most of the variables in the analysis are only based on survey data or expert assessments of media bias. The outlined theoretical model could benefit from an experimental research design. This is especially important for (social) media exposure, which may suffer from social desirability bias.

Author

• Shota Gelovani, Technical University of Munich