Does highlighting COVID-19 disparities reduce or increase vaccine intentions? evidence from a survey experiment in a diverse sample in New York State prior to vaccine roll-out.

Racial identity and political partisanship have emerged as two important social correlates of hesitancy towards COVID-19 vaccines in the United States. To examine the relationship of these factors with respondents' intention to vaccinate before the vaccine was available (November/December, 2020), we employed a multi-method approach: a survey experiment that randomized a vaccine-promotion message focused on racial equity in vaccine targeting, stepwise regression to identify predictors of hesitancy, and qualitative analysis of open-ended survey questions that capture how respondents reason about vaccination intentions. Experimental manipulation of a racial equity vaccine promotion message via an online survey experiment had no effect on intention-to-vaccinate in the full sample or in racial, ethnic and partisan subsamples. Descriptively, we find heightened hesitancy among non-Hispanic Black respondents (OR = 1.82, p<0.01), Hispanics (OR = 1.37, p<0.05), Trump voters (OR = 1.74, p<0.01) and non-Voters/vote Other (OR = 1.50, p<0.01) compared with non-Hispanic White respondents and Biden voters. Lower trust in institutions, individualism and alternative media use accounted for heightened hesitancy in Trump voters, but not non-Hispanic Blacks and Hispanics. Older age and female gender identity also persistently predicted lower vaccine intentions. Qualitatively, we find that most hesitant responders wanted to 'wait-and-see,' driven by generalized concerns about the speed of vaccine development, and potential vaccine side-effects, but little mention of conspiracy theories. Identity appears to be an important driver of vaccinate hesitancy that is not fully explained by underlying socioeconomic or attitudinal factors; furthermore, hesitancy was not significantly affected by racial equity messages in this setting.