Summary
There is increasing concern about political violence in the US, with incidents like the 2021 Capitol Attack highlighting these worries. Similar instances of political violence in democracies around the world are observed, although the actual prevalence of political violence differs. For instance, expert assessments for Protect Democracy indicate a severe threat of political violence in India, a significant one in the US and Poland, and a low one in Germany.
We find that the prevalence of political violence in democracies correlates with public support for such violence. Comparative data consistently show high levels of both political violence in India, for which we also observe the highest level of public support for political violence. Respondents from the United States, Germany, and Poland show similar support levels, with Brazilian respondents showing slightly higher support. The Israeli sample is strikingly opposed to political violence, which may be explained by the societal implications of the October 7 terrorist attack. Second, while differing partisan identifications show little correlation with support for political violence, a general distrust in the political system is consistently linked to such support across democracies.
The survey
This report summarizes the results of a cross-national survey conducted in collaboration with YouGov during January and February 2024. It specifically addresses survey items that inquire about support for political violence. Respondents were asked whether they supported various forms of violence directed at individuals from opposing political parties, including:
- Protesting without a permit
- Vandalism
- Assault
- Arson
- Assault with a deadly weapon
- Murder
Respondents who reported ‘Support’ or ‘Strongly support’ for the first item (i.e., protesting without a permit) were then asked about the second item (i.e., vandalism), and so on. Respondents were not presented with subsequent items if they selected ‘Neither support nor oppose,’ ‘Oppose,’ or ‘Strongly oppose’ to any of the preceding violent actions.
In this report, we limit our analysis to the more severe forms of political violence: assault, arson, assault with a deadly weapon, and murder. We present the percentage of respondents who selected ‘Support’ or ‘Strongly support’ for each type of violent action. All other respondents are coded as not supportive of the violent action. This includes those respondents who were not asked about a particular item because they did not support a less severe form of violence beforehand.
Support for political violence across democracies
This graph reports respondents’ support for political violence by country. We find the strongest support for political violence for Indian respondents and the lowest for Israeli respondents. We find similar levels of support for political violence in the US, Germany, and Poland, with Brazilian respondents showing slightly higher support.
Support for political violence by partisan groups
Below we show support for political violence for the main partisan group in each country.
Poland
In Poland, supporters of Confederation are the most supportive of political violence, while supporters of the Civic Coalition show the lowest support. Supporters of the Law and Justice (PiS) party and Poland 2050 are located between these two poles, but overall support for political violence is relatively low.
Germany
In Germany, while low overall, the supporters of the main parties CDU/CSU, FDP, SPD, The Left show similar levels of support for political violence. In comparison, the Greens and AfD supporters were more reluctant to endorse political violence toward out-partisans.
Brazil
In Brazil, PT supporters (Worker’s Party) show slightly more support for political violence than PL supporters (Liberal Party).
Israel
In Israel, support for political violence is very rare and only occurs among a very small number of Likud supporters.
India
In India, while comparatively high across both party camps, BJP supporters are considerably more supportive of National Congress supporters.
United States
In the United States, support for political violence is similarly low among both partisan camps (Democrats and Republicans).
Affective polarization is not consistently related to support for political violence
The following figures display the relationship between affective polarization (feeling thermometer of in-party minus feeling thermometer of most disliked out-party) and support for acts of political violence. Across countries, affective polarization is not consistently linked to support for political violence. In Poland, Brazil, and India, more affectively polarized respondents were more supportive of acts of political violence. In Germany and the United States, moderately affectively polarized respondents were most likely to support political violence.
Poland
In Poland, more affectively polarized respondents were marginally less likely to support political violence.
Germany
In Germany, low and highly affectively polarized respondents are more likely to support political violence. Moderately polarized respondents are less likely to support political violence.
Brazil
In Brazil, more affectively polarized respondents were marginally less likely to support political violence.
Israel
In Israel, affective polarization and support for political violence are not linked to each other. Note that only a few respondents reported to support any support of political violence.
India
In India, more affectively polarized respondents were less likely to support political violence.
United States
In the United States, moderately affectively polarized respondents were more likely to support political violence.
Younger respondents are more supportive of political violence
The following figures display the relationship between age and support for acts of political violence. Across most countries, older respondents were less supportive of political violence. Conversely, younger respondents were more likely to support acts of political violence. An exception is Germany, where age and support for political violence are unrelated.
Poland
In Poland, the older respondents were, the less they supported political violence.
Germany
In Germany, age and support for political violence are not related to each other.
Brazil
In Brazil, except for assault, the older respondents were, the less they supported political violence.
Israel
In Israel, except for murder, the older respondents were, the less they supported political violence.
India
In India, younger respondents were more likely to support political violence than older respondents.
United States
In the United States, the older respondents were, the less they supported political violence.
