https://orcid.org/0000-0002-9535-3866
![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
ABSTRACT
In recent decades, the discussion of terrorist threat in the United States has focused almost exclusively on radical Islam. Now, progressive politicians increasingly talk about white supremacy as a form of terrorism. We explore reactions to this new rhetoric using two survey experiments, in which non-Hispanic white respondents are exposed to real tweets of Democratic and Republican politicians discussing white supremacy and radical Islam as security threats. We formulate two alternative expectations: after seeing messages from Democratic politicians linking white supremacy to terrorism, whites may either diverge in their assessments of Republicans’ anti-Muslim appeals depending on their own partisanship or they may uniformly rate anti-Muslim rhetoric as more acceptable. Both experiments show significant increases in the perceived acceptability of anti-Muslim appeals when they are presented after messages about white supremacy, and these results are driven by white Democrats. Our results highlight the potential side effects of political communication about terrorism.
Acknowledgements
This research was supported in part by Gerald R. Ford Fellowship at the University of Michigan and the Democracy Initiative at the University of Virginia. An earlier version of the project was presented at the Annual Conference of the European Political Science Association. The authors are grateful to Leah Christiani for access to the student data and to Olga Gasparyan for helpful comments on the paper draft.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
3 While our design is based on the similarity between messages, we do not imply comparability of white supremacy and radical Islam as ideologies (or of whites and Muslims as the implicated groups).
4 See Supplementary Material for sample demographics.
5 In the control condition, respondents saw a Democratic tweet after a Republican tweet. We use the corresponding comparison (perceived appropriateness of a Democratic tweet on white supremacy before vs. after seeing a Republican tweet on radical Islam) to test whether the exposure effect is one-sided (only exposure to a tweet about white supremacy increases acceptability of anti-Muslim rhetoric) or two-sided (exposure to a tweet about radical Islam also increases acceptability of rhetoric linking white supremacy to terrorism).
6 When presented with tweets, respondents were informed of representatives’ partisanship to account for potential differences in political knowledge.
7 See Supplementary Material for the full list of tweets.
8 “Do you agree with the general point that the politician is trying to make in this tweet—regardless of the language that they used to express it?” Answers coded from 1 = Strongly disagree to 7 = Strongly agree.
9 Presented results are only for non-Hispanic white respondents. See Table S1 in Supplementary Material for results among non-white respondents (including Hispanics).
10 When we interact the treatment with racial attitudes, we find significant positive treatment effects for racial liberals but not for racial conservatives (see Table S2 in Supplementary Material).
11 Only 13.1% of Republicans in the treatment condition have the maximum acceptability score of seven, thus suggesting enough space for a non-trivial treatment effect.
12 See Supplementary Material for sample demographics.
13 Similar to Study 1, respondents saw a Democratic tweet after a Republican tweet in the control condition.
14 See Supplementary Material for the full list of tweets.
15 Presented results are only for non-Hispanic white respondents. See Table S1 in Supplementary Material for results among non-white respondents (including Hispanics).
16 Like in Study 1, we find significant positive treatment effects for racial liberals but not for racial conservatives (see Table S5 in Supplementary Material).
17 These results may be more conservative compared with Study 1 due to timing. Study 2 was fielded in July 2020—two months after the murder of George Floyd and amid national protests for racial justice. The salience of racial issues at that time might have further decreased white Democrats’ tendency to rate explicit group-based appeals as “appropriate” and suppressed the average treatment effect. We are grateful to an anonymous reviewer for making this point.
18 A good example is Media Matters for America posting videos containing controversial statements made on air by prominent conservative media personalities like Tucker Carlson and Matt Walsh.