Do United Nations peace operations lead to a decrease in post-civil war state repression? If so, how? While the literature is clear that UN peacekeeping decreases the likelihood of civil war recidivism, much less is known about whether the UN is able to promote liberal post-conflict human rights practices. This dissertation argues that UN peace operations, in general, produce significantly greater improvements in post-civil war state repression compared to cases where the UN does not intervene. I argue that the UN is able to accomplish its ameliorative effect on human rights practices by increasing the probability that belligerent parties will comply with peace agreement provisions, thus reducing supply- and demand-side factors of repression. On the supply-side, i.e., factors that influence the ability and/or willingness of government principals and agents to repress, peace agreement compliance typically leads to the disarmament and demobilization of government forces, the restructuring of the security sector, the monitoring of state security forces, and political power-sharing. On the demand-side, i.e., factors that influence the ability and/or willingness of opposition groups to violently dissent from the state, peace agreement compliance typically leads to the disarmament and demobilization of opposition militias, the transition of rebel groups into political parties, and the integration of opposition groups into political, military, and economic institutions. Hypotheses derived from this theoretical argument are quantitatively tested using both the universe of post-Cold War civil wars that experienced a period of peace prior to 2005, as well as a matched subset of these conflicts in which cases are matched along important confounding variables in an attempt to mimic random assignment to treatment and control groups. The analysis finds that cases in which peacekeeping forces are deployed experience a significantly greater improvement in human rights practices than cases in which belligerent groups are left to fend for themselves. Importantly, these estimates are robust to possible selection effects, meaning the UN did not tend to intervene in easier cases, nor are post-civil war improvements among UN cases simply regression toward the mean. Moreover, this improvement is even greater in the matched subsample, which gives a more accurate estimate of the UN's effect in places where UN peace operations typically go. The improvement is also greater when the UN is given a multidimensional mandate. These hypotheses are further tested using causal mediation analysis, which tests whether the UN effect is mediated through the hypothesized causal mechanisms. This quantitative analysis is supplemented with a qualitative analysis that illustrates the causal mechanisms theorized above. In these chapters I analyze two very different cases where UN peace operations were generally considered successful: ONUSAL in El Salvador and ONUMOZ in Mozambique. If the UN is indeed able to exert a causal impact in reducing state repression, it should be seen in these two cases. I therefore closely examine primary and secondary accounts of the years following the end the civil wars in El Salvador and Mozambique to determine the role the UN played in peace agreement compliance and reductions in supply- and demand-side factors of repression. I find compelling evidence that both cases experienced declines in state repression that are directly attributable to UN efforts. Quantitative synthetic control testing provides further evidence that UN involvement was at least partly responsible for the decreases in post-conflict state repression in these two cases.
|In Administrative Set: