Democracy in the Crossfire: The Electoral Fallout of Extremist Political Violence

Abstract

What are the effects of political violence on election outcomes in established democracies? We argue that exposure to political violence raises citizens' demands for security. Existing accounts disagree on whether this demand rewards incumbents, conservative parties, or the radical right, but they agree that it should cost extremist parties perceived as contributing to instability. Using a novel police dataset of 148 incidents of political violence, almost all from the far left or the far right, during Greece's snap parliamentary elections of May and June 2012, we employ a difference-in-differences design at the voting center level in Athens and Thessaloniki. We find no evidence that political violence affects voter turnout. Further, we find only tenuous evidence of an electoral boost from violent incidents for populist radical right parties. Political violence does, however, reduce the combined vote share of extremist parties in exposed areas. Separating far-left from far-right incidents suggests that each extremist party loses support after violence from its own ideological side, though the evidence is stronger for the far right and we treat the pattern as descriptive. We argue that the backlash against extremist parties as a whole speaks to research on how voters punish parties associated with violence, as well as to wider debates about democratic resilience in times of crisis.