Search

Both in their quantity and their quality, informal political conversations can provide an important bellwether for democratic health. However, not everyone is willing to participate in political conversations in all settings, and systematic imbalances in who chooses not to share political attitudes can distort perceptions of public opinion. Using data from three original surveys, including both observational and experimental analysis, we examine people's decisions to initiate political discussions using a psychological framework of self-threat and self-affirmation. We find that political conversations pose a higher level of self-threat when disagreement is probable and the relationship with the potential discussion partner is weaker. High levels of self-threat, measured via self-reported anxiety, are associated with a lower willingness to initiate a political conversation. However, self-threat can be counteracted. While it does not reduce the anxiety associated with a threatening situation, self-affirmation increases people's willingness to initiate a political conversation in higher threat circumstances. This suggests that efforts to find common ground or boost confidence by reflecting on non-political values could increase the pool of people willing to bring up and share their political views. [ABSTRACT FROM AUTHOR]

Prominent accounts of public opinion argue that citizens’ preferences are unstable, with stated desires on policies varying wildly from survey to survey, and ideologically incoherent, with preferences on multiple policies evidencing little or no structure. In the aggregate, these findings suggest that many voters are not capable of fulfilling their normative role in the democratic system. In this article, we challenge this conventional view and argue that the apparent instability and incoherence among the public are both overstated and outdated. Using panel surveys from the 1970s, 1990s, and 2010s, we conduct a multi-trait multi-method (MTMM) confirmatory factor analysis of citizen preferences in multiple issue areas. Our results reveal a surprising degree of preference stability in all three time periods across many policy domains. Furthermore, our results reveal increasing levels of ideological thinking over time and that these patterns of stability and coherence hold across subpopulations defined by levels of sophistication. [ABSTRACT FROM AUTHOR]