Partisan Public Responsiveness to the News

• In Event: Responsiveness and Representation of Public Opinion

Abstract

Research demonstrates thermostatic public responsiveness to policy that is mediated by news coverage of government actions (Neuner, et al 2019; Soroka and Wlezien 2022). It is however not clear whether and how groups respond differently to the news, and there is special reason to expect responsiveness to vary across partisan groups (see, e.g., Druckman et al., 2013; Mullinix 2015), who not only may have different information about policy change (Benkler et al. 2018) but may also react differently to the same information (e.g., Branham 2018). To consider the possibility, we analyze the relationships between policy, the news, and partisan public preferences in the United States. For the analysis, we focus on three spending domains where measures of the different variables are available and thermostatic responsiveness is evident: defense, welfare, and health. Our analysis reveals substantial differences in levels of policy support for Democrats, Independents, and Republicans. Partisan preferences do move in parallel over time – though they do so imperfectly, with a mean correlation of 0.80, and only 0.67 between Democrats and Republicans. This heterogeneity appears to reflect differences in responsiveness to the news, where Republicans are more responsive to the media spending signal than Independents and especially Democrats. Indeed, preferences for the latter do not reliably reflect media coverage. Thus, party may not fundamentally condition thermostatic feedback, but it does appears to matter for the information citizens have. Put differently, all partisan groups respond to information about spending; but Independents and Republicans (particularly the latter) are responsive to the “other” media coverage that is not correlated with spending. In the paper, we unpack the how this produces divergence in preferences over time and explore how different media sources drive the results. We also demonstrate that the patterns are not a recent development associated with the rise of social media. Although it is well known that media coverage can both inform and misinform, this has been true for some time.

Authors

• Christopher Wlezien, University of Texas at Austin

• Stuart N. Soroka, UCLA