![Sample our Humanities journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5939/TOC-Subject-Sample-2021-Humanities.jpg"})
Abstract
Political observers of all types often express concerns that Americans are dangerously polarized on political issues and are, in part due to the availability of opinionated niche news programming (e.g., ideological cable, radio, and Internet news sources), developing more entrenched political positions. However, these accounts often overlook the fact that the rise of niche news has been accompanied by the expansion of entertainment options and the ability to screen out political news altogether. We examine the polarizing effects of opinionated political talk shows by integrating the Elaboration Likelihood Model of attitude development into our own theoretical model of selective media exposure. We employ a novel experimental design that gives participants agency to choose among news and entertainment programming by including treatments that allow participants to select the programming they view. The results from two studies show that ideological shows do indeed have the power to polarize political attitudes, especially among individuals who possess strong motivations to craft counterarguments. However, the polarizing force of cable news is diminished considerably when individuals are given the option to tune out.
[Supplementary material is available for this article. Go to the publisher's online edition of Political Communication for the following free supplemental resource: Appendix: Arguments Presented to Participants in Studies 1 and 2.]
Acknowledgments
This research was supported by National Science Foundation Awards SES-0752546 and SES- 0752354. Thanks to Jackie Filla, Byran Martin, Chad Murphy, Andrew Flores, Justin Nelson, Carrie Skulley, Arash Aalem, Caleb Cavazos, Jennifer Dubé, Rachelle Jung, Phillip Lee, Lauren Menor, Kyla Persons, Joel Ruvalcaba, Roxanna Sanchez, and Rebecca Tekeian for their research assistance, as well as Kevin Esterling and Indridi Indridason for help developing a subject pool for these experiments. All errors are our own.
Notes
1. For instance, in mid-January 2012, the evening news shows on the three broadcast networks drew 26 million viewers a day combined, while the most popular conservative cable news talk show (The O'Reilly Factor on Fox News) and liberal talk show (The Rachel Maddow Show on MSNBC) drew 4.5 million viewers combined (CitationTV by the Numbers, 2012).
2. We measured subjects’ ideology prior to the treatments. The results are unaffected if moderates are included in the analysis. Results are available from the authors upon request.
3. Age, race, partisan identification, and ideology do not jointly predict treatment assignment, χ2(36) = 35.16, p = .51.
4. The composite segments for The O'Reilly Factor were drawn from shows airing on September 30, October 6, and October 8, 2009. Each segment featured criticism of health care legislation in Congress. The composite segments for The Rachel Maddow Show were from shows airing on October 6 and 7, 2009, and focused on benefits of health care reform.
5. The argument rating items scale together (α = 0.65).
6. We also observe that participants found the forced news conditions to be more partisan (M = 2.63 on a folded partisan assessment scale where 0 = no perceived partisanship and 4 = strong Republican/Democrat orientation in the clip) than participants exposed to only entertainment shows (M = 1.12), t(61) = 4.30, p < .001 (one-tailed).
7. Although it would also be appropriate to analyze these data with ANOVA, we chose OLS regression because its more general framework allows us to compare more easily the effects of the experimental conditions. Because our hypotheses imply a series of difference of means tests, regression is more intuitive than ANOVA, which is geared toward portioning explained variance. Nevertheless, it is important to remember that ANOVA is analogous to OLS and produces the same results with a different presentational emphasis.
8. Unfortunately, we do not have sufficient observations to estimate reliably the effects of the news shows separately for liberals and conservatives. Our experimental design averages the effects of the liberal and conservative shows. We leave it to future research to address this limitation.
9. For subjects high in need for cognition, the choice condition had a smaller effect on resistance to opposing arguments vis-à-vis the forced counterattitudinal condition (difference = −2.35, p = .004, one-tailed).
10. We measured subjects’ ideology prior to the treatments. The results are unaffected if moderates are included in the analysis. Results are available from the authors upon request.
11. Age, race, partisan identification, and ideology do not jointly predict treatment assignment, χ2(33) = 32.37, p = .4981.
12. The Last Word stimulus originally aired August 15, 2011, and featured guest host Chris Hayes, editor at large of The Nation, arguing for higher tax rates for wealthy Americans. The O'Reilly Factor stimulus was from an August 16, 2011, show featuring guest host Laura Ingraham interviewing Heritage Foundation tax policy analyst Curtis Dubay, both of whom advanced conservative arguments opposing tax increases and government spending.
13. The argument rating items scale together (α = 0.68).
14. We also observe that participants found the forced news conditions to be more partisan (M = 2.1 on a folded partisan assessment scale where 0 = no perceived partisanship and 4 = strong Republican/Democrat orientation in the clip) than participants exposed to only entertainment shows (M = 0.8), t(78) = 3.72, p < .001 (one-tailed).
15. We do not have a sufficient number of observations to estimate reliably the effects of the news shows separately for liberals and conservatives.
16. These findings may be limited to situations, such as television watching, where participants have limited opportunity to engage discursively people who hold views different from their own. Authentic deliberation may produce opportunities for finding common ground (e.g., CitationNeblo, Esterling, Kennedy, Lazer, & Sokhey, 2010).
17. While we do possess data on what shows subjects watched in the choice condition (and how long they watched these shows), it would be inappropriate to use this information in an attempt to estimate the causal effects of these shows. What subjects chose to watch in the choice condition is endogenous to the choices available, and any differences we observe between these groups could have more to do with selection than the effect of these shows. In other work, we propose a different experimental design to investigate the causal mechanisms underlying attitude formation given self-selection into partisan media (Arceneaux & Johnson, 2013).