Coercive Institutions and State Violence Under Authoritarianism

Abstract

Why do we observe such widely differing patterns of repression and state violence under authoritarian rule? Despite a wave of recent interest in authoritarian politics, the origins, design and behavior of the coercive institutions that embody the state's monopoly on violence remain relatively unexamined. This project draws on new statistical and geographic data, elite interviews, and archival evidence from the U.S. and Asia to chronicle the origins and operation of the internal security apparatus in three Cold War anti-communist authoritarian regimes – Taiwan, the Philippines, and South Korea – and compares them to similar processes in Communist authoritarian regimes in North Korea and China. Its findings challenge dominant narratives about contentious politics and state-society conflict in Asia; offer an unprecedented view inside 'secret police' use of surveillance, coercion, and violence; and provide a new understanding of the institutional and social foundations of authoritarian power.I argue that autocrats face a fundamental tradeoff between designing their internal security apparatus to deal with a popular threat, or coup-proofing it to defend against elite rivals. Coup-proofing requires an internally fragmented security force drawn from narrow segments of society; managing popular unrest requires a unitary apparatus with broadly embedded, socially inclusive intelligence networks. Autocrats construct coercive institutions based on the dominant perceived threat when they come to power, but these organizational tradeoffs, exacerbated by institutional stickiness, blunt their ability to adapt as new threats arise. Organizational characteristics thus give rise to predictable patterns of state violence. A more fragmented, exclusive security apparatus – associated with a high initial threat from fellow elites – is likely to be more violent, both because it has stronger incentives to engage in violence and because it lacks the intelligence capacity to engage in discriminate, pre-emptive repression. In contrast to existing threat-based explanations of repression, I demonstrate that autocrats who are deeply concerned about popular threats use less violence rather than more, and do so because they mobilize organizations expressly designed for that purpose. In these organizations, intelligence becomes a substitute for violence, and citizens relinquish their privacy, but less often their lives.