The Political Consequences of Criminal Violence: Applying MRP to Mexican Survey Data

Online Seminar

How does exposure to violence shape political behavior? Does criminal violence push citizens to participate more or less in politics? These questions are becoming more and more important as crime rates rise in many parts of the developing world, but especially in Latin America. We argue that exposure to criminal violence, rather than generating anger or fear, perpetuates political cynicism. As crime increases, citizens update their perceptions of the state’s capacity and willingness to protect them — reducing the stakes of political competition. Studying the effects of exposure to crime is made difficult by the fact that many crimes go unreported, especially in the developing world. Instead of relying on official crime rates, we adapt and apply the methods of multilevel regression post-stratification (MRP) to self-reported victimization in national surveys from Mexico. We first validate the applicability of MRP to this context using census data. Then, we use MRP estimates by municipality to study the effects of crime on turnout. Finally, we use survey-based measures of public attitudes to adjudicate among alternative mechanisms.