| In this dissertation, I bring together three papers in which I apply recent advances in political methodology to problems in comparative politics.;In Chapter 1, I extend previous work on the effects of economic growth on civil wars in sub-Saharan Africa (Miguel, Satyanath, and Sergenti, MSS, 2004). Whereas MSS employ a linear model in their analysis, I use a probit model. In addition, I control for the likely possibility that the occurrences of civil wars in a country are serially correlated, and I use a transition model to investigate war onset and war termination separately. I find with the transition model that the effect of growth on civil war onset is even stronger than what MSS report and highly non-linear for negative values of lagged growth.;In Chapter 2, I present joint work with Michael Gilligan that examines the effectiveness of United Nations (UN) peacekeeping missions into post-conflict and in-conflict countries. Previous statistical studies of the effects of UN peacekeeping have generally suggested that UN interventions have a positive effect. Recent methodological developments, however, have questioned this result, because the cases in which the UN intervened were quite different from those in which they did not. We correct for the effects of nonrandom assignment with matching techniques on a sample of UN interventions in post-Cold-War conflicts and find that UN interventions are indeed effective in the sample of post-civil-conflict interventions, but that UN interventions while civil wars are still ongoing have no causal effect.;Last, in recent years, computational models have become more and more common in political science. Thus far, however, little guidance has been provided on how to analyze and characterize the outputs of such models in a way that enables rigorous and precise inferences. In Chapter 3, I present a comprehensive methodology that can be used with any computational model that tends towards one or more steady-states. I demonstrate how to calculate mean estimates of output variables and to determine how long or how often to execute a model to ensure that mean estimates are representative and that the precision of the estimates meet certain minimum criteria. |
