Voters from both parties largely agree on how to punish acts of political violence

Recent research published in the journal Political Behavior shows that Americans largely prioritize the severity of a violent act over the perpetrator’s political identity when determining punishment. While partisanship does influence how individuals judge political violence, the study suggests that the specific nature of the crime plays a much larger role in public opinion.

Political violence has become a prominent concern in the United States following high-profile events such as the Unite the Right rally in Charlottesville and the January 6 Capitol attack.

Social scientists have debated whether the American public is becoming more tolerant of violence as a means to achieve political goals. Previous scholarship has often focused on whether individuals support violence in the abstract or in relation to specific, highly charged events.

The authors of the current study sought to understand how citizens apply standards of accountability when faced with concrete scenarios. They aimed to determine if Americans are willing to punish individuals from their own political group who engage in violence.

“We saw a good deal of unresolved scholarly debate on the extent to which Americans support political violence in surveys,” said study author Joseph Phillips, a lecturer in politics at Cardiff University.

Free daily newsletter

However, the key scholars involved with that debate all agreed the answer likely depends on context. Therefore, we wanted to know, if we supplied people with different kinds of perpetrators who performed different acts of political violence, how would that affect their reactions?”

The study investigates whether the “fairness standard”—the democratic expectation that laws should be applied equally regardless of identity—holds up when political tribalism is involved. In a highly polarized environment, there is a risk that citizens might excuse violence committed by ideological allies while demanding harsh punishment for opponents.

The researchers designed the study to disentangle the effects of the perpetrator’s identity from the severity of their actions. This approach allows for a precise measurement of which factors weigh most heavily in the minds of American voters.

To test these questions, the research team recruited participants through Dynata to create a sample that approximated the demographic composition of the United States. The researchers conducted two separate pre-registered conjoint experiments.

Conjoint experiments are designed to present participants with hypothetical scenarios where multiple factors vary simultaneously. This method helps prevent participants from giving socially desirable answers by forcing them to weigh different attributes against one another.

The first experiment included 3,402 participants who evaluated a total of 17,010 different perpetrator profiles. Participants read a scenario describing a protest in a state capital that turned violent. They were presented with a single perpetrator profile that included randomized details about the person’s age, occupation, gender, race, party affiliation, and the specific act they committed.

The acts varied in severity and target, ranging from spraying graffiti to injuring people with weapons. Participants were asked to recommend a jail sentence for the perpetrator and to evaluate the likelihood that the person was acting for a “good cause.”

The second experiment involved 3,065 participants who evaluated 38,682 profiles in a “forced choice” format. In this version, participants were presented with two different perpetrators side-by-side. They were asked to decide which of the two should be allowed to walk free without charges. This design forced respondents to make difficult trade-offs and revealed which factors they considered most unforgivable.

The data from the first experiment revealed that the nature of the act was the single most powerful predictor of the recommended punishment. Participants recommended longer sentences for severe violence compared to minor property damage.

For example, acts involving severe physical injury resulted in sentence recommendations that were years longer than sentences for vandalism. The target of the violence also mattered considerably. Acts of violence directed at police officers attracted the harshest penalties. Attacks on government buildings were punished more severely than attacks on small businesses or counter-protesters.

The researchers found that partisanship did play a role, but its effect was secondary to the crime itself. Participants were slightly more lenient toward perpetrators from their own political party.

On average, an individual belonging to the opposing political party received a sentence recommendation that was approximately four months longer than a co-partisan who committed the same act. This indicates that while partisan bias exists, it does not completely override the public’s assessment of the crime’s severity.

A key finding from the study is the level of agreement between Democrats and Republicans regarding the hierarchy of offenses. Despite the common narrative of two distinct political realities in America, voters from both parties tended to rank the severity of different acts in the same order.

Both groups viewed violence against police as particularly egregious. Republicans were slightly more punitive overall, particularly regarding property damage and crimes against police. However, the general pattern of condemnation was consistent across the political spectrum.

The researchers also examined the role of affective polarization, which refers to the emotional dislike of the opposing political party. Participants who scored high on measures of affective polarization were more likely to display double standards.

These individuals were more inclined to recommend harsher punishments for political opponents and lighter sentences for allies. This suggests that as emotional polarization increases in the general public, the impartial application of democratic norms may face greater strain.

“Americans, when they get the option, want to punish political violence, especially when the act is severe,” Phillips told PsyPost. “The public do let political allies off the hook to an extent, but the extent is rather small. And despite a lot of discourse, Democratic and Republican voters largely agree on which acts are worse.”

Perceptions of the perpetrator’s motive showed a different pattern than the sanctions themselves. Participants were often willing to attribute “good cause” motivations to perpetrators who shared their political identity or demographic characteristics.

However, this sympathy did not necessarily translate into leniency in sentencing. The analysis showed only a weak connection between viewing a perpetrator’s motives favorably and recommending a lighter punishment. This implies that Americans may sympathize with a cause but still believe that violent actions require legal consequences.

“Going from threatening a police officer to injuring one with a deadly weapon, according to our respondents, nets a perpetrator four extra years in jail,” Phillips explained. “By comparison, the perpetrator coming from the opposite party, as opposed to the same party, only nets four extra months.”

As with all research, there are some limitations. “We didn’t show people visual cues of perpetrators – we told people their characteristics through text,” Phillips noted. “It’s possible people respond to the race, class, and gender of political perpetrators more when they are simply shown, rather than explicitly stated.”

Additionally, the study left the specific cause of the protest vague to avoid biasing the results based on the issue being protested. This lack of context might have dampened the effects of identity, as real-world protests are often deeply intertwined with specific ideological causes.

Future research could address these limitations by incorporating visual media or varying the political context of the protests. The authors also suggest that future studies should investigate the communication dynamics that shape how people process news about political violence.

The current study focused on how people evaluate the violence itself, but it did not explore what makes people perceive an act as “political” or “violent” in the first place.

The study, “When Push Comes to Shove: How Americans Excuse and Condemn Political Violence,” was authored by Joseph B. Phillips, B. Kal Munis, Nicole Huffman, Arif Memovic, and Jacob Ford.