Replication Data for: \"Violence on Many Sides\" Framing Effects on Protest and Support for Repression

The success of protests depends on whether they favorably affect public opinion: nonviolent resistance can win public support for a movement, but regimes counter by framing protest as violent and instigated by outsiders. We argue public perception of whether a protest is violent shifts based on framing the types of action and the identities of participants in those actions. We distinguish dimensions of (1) threat of harm, (2) bearing of arms, and (3) identity of protesters. Using survey experiments in Israel and the United States we find support for framing effects. Threat of harm has the largest positive effect on perceptions of violence and support for repression. Surprisingly, social out-groups are not perceived as more violent, but respondents favor repressing them anyway. Support for repressing a nonthreatening out-group is at least as large as support for repressing a threatening in-group. The findings link contentious action and public opinion, and demonstrate the susceptibility of this link to framing.