Abstract
Abstract The authors use US data on media coverage of politics and individual survey data to document that citizens exposed to more politicized newspapers have more extreme political preferences. This polarization effect of media is mainly driven by individuals who harbor liberal opinions reading more newspapers, as opposed to individuals endorsing rather conservative positions. More politicized media also reinforce other aspects of citizens’ political sophistication such as political knowledge. This enhanced political sophistication materializes in observable involvement in politics, measured by campaign contributions.
Highlights
Most democracies have never been as politically polarized as they are today.1 Holding polar political opinions bears important socio-economic consequences, since in such contexts political decisions are likely to deeply divide the population and to bring along important economic costs
We explore the effect of newspaper coverage combined with individual survey data from the 2008 wave of the American National Election Studies (ANES) on whether surveyed individuals have an interest in politics and read news
To analyze the effect of individual characteristics conditional on a variable aggregated at a higher level, we follow Facchini and Mayda (2009) and estimate the following probit model: Prob(Yi = 1 | Xi) = Φ(β1Interesti + β2Interesti × Medias + βKZi + ψs) where Φ(.) represents the cumulative distribution function of a standard normal, Yi captures the individual’s ideological position, monetary contribution or political knowledge, Interesti the individual’s interest in following campaign or media exposure, Medias the state media coverage of politics in the individual’s state of residence, Xi is a vector of individual controls, ψs state dummies to control for additive state-specific unobserved effects, and standard errors adjusted for clustering at the state level
Summary
Most democracies have never been as politically polarized as they are today. Holding polar political opinions bears important socio-economic consequences, since in such contexts political decisions are likely to deeply divide the population and to bring along important economic costs. Kennedy and Prat (2017) show in a study on 36 countries that 40% of the surveyed subjects get informed by reading newspapers Combining this finding with the strong political content of newspapers (Campante and Do 2014), one realizes the importance of exploring the potential causal effect of newspapers on political polarization. We explore the effect of newspaper coverage combined with individual survey data from the 2008 wave of the American National Election Studies (ANES) on whether surveyed individuals have an interest in politics and read news. Our central result suggests that politicized information polarizes political opinions This could be interpreted as a confirmation of previous findings that news spread quickly in networks of the same political side (Halberstam and Knight 2016), eventually producing more polarized political attitudes. We find that more politicized media improve individuals’ political sophistication as measured as political knowledge regarding the U.S political life. Lastly, we demonstrate that an additional positive implication of a high media coverage of politics is a stronger involvement of citizens in politics, as measured as campaign contributions
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
