Abstract: In times of political unrest, some leaders and regimes resort to the use of force in order to maintain societal equilibrium. This practice has short and long term effects, quelling current disorder as well as generating a reputation which will inform the decision of potential dissidents in the future. In this paper, we explore whether patterns of repressive response to protests events are consistent with reputation building efforts of individual leaders. Using duration analysis, we find strong support for hypotheses linking reputation concerns to repression propensity. Leaders who represent a break from prior regimes repress with significantly higher probability when they are new in office. For those who are heirs to a repressive regime, however, time in office has no relationship with repressive tendencies.
Abstract: Human rights abuses vary widely between and within states, yet a large amount of this variation remains unaccounted for in the literature. Recent theories that seek to explain this variation have focused on domestic factors that explain why states choose to violate the human rights of their citizens. Adding to this literature, we seek to explain variation in human rights variations by looking at two distinct domestic features that influence the likelihood of human rights violations. First, we explore the potential threat of domestic opposition. The stronger domestic opposition is, we argue, the more likely the leader will resort to repressive methods, including human rights violations. However, we consider further internal variation via individual leaders' tenure: we argue that the longer the leader has been in power, the more secure she is in her position and the less likely she is to resort to repressive methods. This theoretical approach of investigating the incentives leaders have to resort to repressive action is particularly useful to explain human rights violations in non-democratic states. Initial quantitative investigations of these propositions have produced confirming evidence.
Current status: This paper was presented at APSA 2011 as a poster. The slides are available below. Though the paper is currently on hiatus, we would appreciate any comments about further refinements in the empirical test and theoretical expectations.