Who is it Okay to Punch? An Experimental Investigation of Support for Intolerance in the Form of Physical Violence

By Frank J. Gonzalez and Alexandra McCoy

The following blog post is a summary of the research that co-won the Midwest Political Science Association’s Best Paper in American Politics Award (presented at the 2021 MPSA Annual Conference).

The variety of ways in which we have seen political violence coupled with the (relative) rarity with which it seems to reach its most extreme forms (e.g., violent attempts to overthrow government) raises important questions. When is support for political violence most likely? What factors are associated with individuals being willing to climb the ladder of intolerance from negative affect, to support for denial of rights, to willingness to shed blood for a cause?

A great deal of scholarship has been focused on political intolerance and, more recently, affective polarization. This research focuses on Americans’ negative feelings toward other groups and the factors that make one more or less politically intolerance and supportive of political violence. However, we turn our attention to the other side of the equation: the characteristics of individuals and groups that make them most likely to be the targets of political violence.

More specifically, what makes some groups perceived as more “punch-able” than others? Which perceived characteristics of a potential target matter most? And finally, do the same characteristics matter across types of intolerance, such as support for violence versus support for limiting a group’s rights?

In this paper, we used two experimental tasks – one “real-world rating task” (RWRT) involving evaluations of real people and groups and one conjoint task manipulating attributes of hypothetical groups –  to test expectations derived from prominent theories in political science as well as social psychology. In the RWRT, participants rated a random selection of 32 well-known individuals and groups from a list of 64 (some political and some not, some real and some fictitious) along several dimensions, including positive feelings toward the target and how upset they would be if the target were physically harmed. For the conjoint experiment, participants engaged in a 30-trial conjoint task where they chose between two groups (“Group A” and “Group B”) based on the following question: “Which group would you be more likely to support violence against in the form of a member of that group being punched in the face while in public?”

In our analyses, we were mainly interested in the roles played by target race, political alignment (either via partisanship or expressed ideological values), threat, disgust, warmth and competence (as per some social psychological theories), and expressed support for violence by the target.

We find consistent support for effects of political alignment, including independent effects of both partisan alignment and ideological values, with perceived pro-dominance values of the target group (e.g., “The most important thing is winning, or else all is lost.”) being robustly associated with greater intolerance across participants on the political left and the right. We also find some factors with mixed results, including race, where we interpret findings as reflecting a mix of prejudice and deference to egalitarian norms. However, the most robust predictor is the target’s perceived tendency toward violence. Across types of intolerance, groups and individuals perceived as supporting violence, themselves, are the most likely to be targets of not only negative affect, but also support for physical violence.

The version of this paper presented at MPSA used a sample of just 90 undergraduates, but we have since replicated the study (with some design adjustments) on a sample of nearly 1,000 Lucid respondents, and our main findings – especially the robust effect of perceived violence – remains.

There are some potentially important implications of these findings regarding how we might not only predict, but also prevent groups from being the target of violence. Our results suggest that voicing support for violence against a group may, in a cyclical way (and somewhat ironic if the violence is being called for against a group perceived as being violent), actually increase the probability of violence expressed by the target group against one’s own group. If violence is seen as a justification tool – a means to an end – then “fighting fire with fire” will lead to further acts of retaliatory violence, a phenomenon not at all novel to scholars of international and intrastate conflict (e.g., Liberman, 2006).