Legislating during war: Conflict and politics in Colombia

Abstract

This paper studies how politicians and their constituents respond to political violence by investigating the Colombian civil conflict. I use data on rebel attacks, legislators’ tweets, and roll-call votes, and I employ event-study and difference-in-differences empirical methods. After rebel attacks, Twitter engagement (as a proxy for popular support) increased both for incumbent-party legislators’ tweets and for legislators’ tweets using hard-line language. Legislators increased their support for the incumbent party after attacks but only when the government had a hard-line policy position as inferred both from the recent historical context and from text analysis of the president’s tweets. Though the effects were initially large, they lasted less than two weeks. The empirical results are consistent with a political economy model of legislative behaviour in which violent events, that produce rally-‘round-the-flag effects and shift voters’ views, elicit different responses from politicians depending on the policy position of the incumbent party. Finally, I identify a set of potentially affected congressional votes, revealing that these conflict-induced swings in incumbent support can have persistent policy consequences.