## Hidden Repression Under Dictatorships

##### Author

##### Type

**Trabajo de grado - Maestría**

##### Document language

**Español**

##### Publication Date

**2014**

##### Metadata

Show full item record##### Summary

This paper presents a model of the interactions between a dictator and a mass of citizens in which the dictator might have an incentive to use hidden repression. Most papers on political economy of dictatorships assume that the role of repression might work as a signal about the strength of the regime. Here, under a global games framework, we endogenize this decision in a situation where there exists a possible threat of an uprising that might topple the regime. Citizens interact in such a way that the collective action problem of a revolution is not solved beforehand, and so each one takes the decision to participate or not in the revolution independently. These decisions are such that there are strategic complementarities but each citizen is unsure about the actions of her fellow citizens. They receive two signals about the kind of regime they are facing: one, informing about the strength of the dictator to withstand a revolution; the second, informing how repressive is the regime. Given this information, using Bayesian updating, they decide to participate or not. We show that as long as citizens have perfect information about at least one parameter of the regime, there exist a unique equilibrium in which regimes which are strong enough have an incentive to increase the noise informing their repression profile, i.e. to use hidden repression. We also analyse the robustness of these results by relaxing the quality of information agents receive. We extend the model to the case where citizens have imperfect information about both parameters and reach a solution coherent with the previous one.##### Keywords

##### Collections

This work is licensed under a Creative Commons Reconocimiento-NoComercial 4.0.This document has been deposited by the author (s) under the following

**certificate of deposit**