The Authoritarian Wager: Political Action and the Sudden Collapse of Repression

Abstract

Authoritarian rulers tend to prevent political action, but sometimes allow it even if it leads to social conflict. The collapse of preventive repression is especially puzzling when rulers have reliable security forces capable of preventing protests. We develop a game-theoretic model that explores the incentives of authoritarians to repress or permit political contestation. We show that rulers with the capacity to fully repress political action create despotic regimes, but rulers with more moderate capacity might opt to allow open contestation. The status quo bias that favors regime supporters weakens their incentive to defend it. Rulers take the authoritarian wager by abandoning preventive repression and allowing opposition that threatens the status quo. The resulting risk gives incentives to the supporters to defend the regime, increasing the rulers’ chances of political survival. Even moderate changes in the structural capacity to repress might result in drastic policy reversals involving repression.

State Repression and Democratic Dispensation in Uganda 1996–2016

“…Specifically, Rozenas (2020) argues that state repression encourages citizens to engage in activities that oppose the regime. Other literature contends that when the opposition is increasingly becoming stronger to threaten the democratic incumbent government, the incumbent uses actions that prevent activities opposition from effectively challenging it to sustain its life span (Slantchev & Matush, 2020).…”

A Theory of Demographically Targeted Repression

“…Kalyvas 2006;Gregory, Schröder, and Sonin 2011). The model analyzes repression against imperfectly observable, clandestine opposition, which complements the existing literature on public forms of opposition like protest (Ritter 2014;Slantchev and Matush 2019;Shadmehr and Bernhardt 2011). Finally, the model incorporates the idea that repression can motivate defensive opposition, which has been explored widely in the empirical literature (Kalyvas 2006;Finkel 2015), but not so much in the theoretical one.…”