ABSTRACT
Using an original vignette survey experiment, we explore how white Americans’ evaluations of policy arguments in the local context change depending on the rhetoric used to criticize the policy. We compare the effects of three rhetorical frames used by critics of a policy: inefficiency, unfairness, and racism. We find that respondents use political rhetoric as a strong ideological cue: criticisms that invoke both unfairness and racism make respondents perceive policy critics as more liberal and policy supporters as more conservative. The effects of rhetoric on criticism evaluations are moderated by partisanship: Democrats find criticism about unfairness and racism more persuasive, while for Republicans the effect is opposite. We are unable to establish reliable effects of rhetorical framing on the respondents’ support for the policy itself or their readiness to engage in local political action.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Author Note
Earlier versions of this project were presented at the Annual Meeting of the International Society for Political Psychology and the Interdisciplinary Workshop in American Politics at the University of Michigan. We are grateful to Lauren Van De Hey for excellent research assistance. We would also like to thank to Sara Morell and others for their helpful comments.
Notes
3 We did not pre-register the experiment because of its exploratory character.
4 Given our interest in respondents’ evaluations of policy arguments, we did not include a condition in which no argument from policy critics would be present.
5 Respondents’ demographics are generally balanced across the free conditions, although respondents who got efficiency vignettes are somewhat more educated (see Table S1 in Supplementary Material). Attrition is also balanced across conditions with the completion rate of about 98% (F2, 700 < 0.01, p = .998).
6 Figure S1 in Supplementary Material presents the results for 311 nonwhite respondents.
7 We also estimate treatment effects controlling for demographic covariates. Estimates presented in Figure S2 in Supplementary Material show the same results as simple mean comparisons.
8 For instance, our vignettes described the locality as a city while only one third of Americans live in urban cores.
9 We are grateful to an anonymous reviewer for making this point.
10 The design of our survey experiment does not allow us to assess potential institutional consequences of racial rhetoric such as building coalitions in the council and setting the policy agenda.