Using Panel Data to Analyze Partisan Changes in Vaccine Opinions

During the first half of the 2010s, vaccine skepticism appeared either unrelated to partisanship or somewhat more associated with the cultural left. By the early 2020s, however, it had become substantially more prevalent among Republicans than Democrats. Standard theories of partisan elite cue-taking (e.g., Druckman, Peterson, and Slothuus 2013) do not fully explain this shift, given the inconsistency of Republican elite messaging on vaccines. This transformation raises several questions: Did partisan divergence emerge only with COVID-19, or begin earlier? To what extent does it reflect individuals changing their views versus changes in party composition? And when opinions changed, were those shifts driven by demographic and attitudinal factors, or does partisanship remain independently important? Existing studies cannot adequately address these questions because they rely on cross-sectional data or panels beginning after COVID-19. We leverage the 2016–2020 ANES Panel, supplemented by Gallup and Pew time-series data, to investigate these dynamics.