Explaining the causes and consequences of internationally monitored elections

Until 1962, there were no recorded cases of international election observation in sovereign states. By 2004, more than 80 percent of elections held in non-consolidated democracies were internationally monitored, and all internationally legitimate leaders outside of the developed democratic world were expected to invite international observers. This dissertation is motivated by the empirical puzzle that a substantial portion of leaders invite international election observers and then orchestrate massive electoral fraud. My argument is that election observation began as a signal to the international community that the incumbent leader was committed to democratizing. During the Cold War, leaders committed to democratization invited international observers in order to distinguish themselves from other leaders and curry favor with Western-oriented states. As international benefits for democratizing states increased, the act of inviting observers was imitated by leaders who wanted the benefits of looking like a committed democrat without actually becoming one. I provide a formal model of the decision by incumbent leaders to invite international election observers. I also provide a general model of international benefit allocation, and show how the goal of democracy promotion is related to other goals of foreign policy. The formal model yields propositions that I test with several original large-N datasets. I show that election observation grew as the benefits of being perceived as a legitimate democracy increased, and as leaders grew more skilled at manipulating the election without being detected. The growth of international election observation also carried domestic consequences. The empirical evidence presented in the later part of this dissertation demonstrates how this change in international politics influenced domestic politics in states that invite observers. First, I demonstrate that observers can reduce election day fraud. This causal claim is supported empirically with natural experimental evidence. A second effect of the spread of election observation is that improvements in certain leaders' ability to manipulate the election led to observable changes in opposition party behavior. Because leaders were more likely to invite observers as they improved their ability to manipulate the election undetected, opposition parties responded by boycotting elections more frequently when international observers were presen