Soldiers' Funerals Increase Nationalism: Evidence from a Natural Experiment in Turkey

Abstract

Soldiers’ funerals are a potent symbol of sacrifice. We investigate how these funerals influence public opinion and political behavior. On the one hand, soldiers’ funerals could undermine support for the conflict and government (the ‘casualties hypothesis’); on the other hand, this exposure to violence could increase antipathy toward out-groups and galvanize support for the campaign. This empirical question is difficult to answer: in many settings, more hawkish constituencies send more soldiers to the front lines, generating selection bias. We provide new causal evidence from Turkey, where military service is compulsory for men, and soldiers are randomly assigned to posts across the country. Some military posts are peaceful, while others are on the front lines of conflicts with Kurdish insurgents or the Islamic State, providing as-if random variation in which districts lose soldiers and host state-organized funerals. Following soldiers’ funerals, we show that attitudes become more hawkish — support for coercive ‘solutions’ to the insurgent conflict increases by 28% — and hostile to the political integration and equality of Kurds. Public displays of nationalism, including street protests and attacks against the pro-Kurdish party, also double after funerals. Despite these attitudinal and behavioral shifts, the funerals decrease the vote share of the nationalist party by an additional point, with a corresponding increase for the incumbent party. We suggest this entails both strategic campaigning from the incumbent and strategic voting from the public: the AKP emphasize the importance of stability during conflict, encouraging voters to prioritize electing a majority government to aggressively prosecute the conflict, rather than risk unpredictable coalition bargaining during a period of escalating violence.