Are International Election Observation Missions Able to Deter Election-day Fraud?

International actors play a prominent role during processes of democratization also through the engagement in elections all over the developing word. While scholars and policy makers admit how complex and multi-faceted the problem of democratization is, in practice, policy measures were strongly focused on elections. Implicit in this emphasis on multi-party elections is the assumption that elections are pivotal for democracy and democratization. Western countries and international organizations as the European Union pressure governments to hold democratic elections by the mechanism of political conditionality, by providing founds and technical assistance and by sending thousands of international electoral observers. The European Union, for example, in its external action, uses election observation as one of the instruments of EU external policy to promote democracy, the rule of law and human rights. Actually, international observation of elections is one of the more diffused forms of democracy promotion. Surprisingly, very little attention has been paid to the effectiveness of electoral observation missions in detecting fraud and few empirical works have tried to discover whether electoral observation missions can bring more “free and fair” elections, as proponents of this democracy-promotion instrument claim. Following the valuable work of Susan Hyde, one of the few scholars dealing with the assessment of election observation, this article will evaluate the effectiveness of international election observation in deterring election-day fraud during the 2004 presidential elections in Ukraine. I propose a framework for evaluating the effectiveness of the observation work relying on a subnational level of analysis: actually, I examine the effects of international election observation at the micro-level, meaning by this at polling station level. I use a quasi-experimental research strategy to verify if international election observation has some effects on domestic political actors'' behavior: actually, by comparing election results of polling stations visited by observers with results of polling stations not visited by them, it is possible to notice if the presence of observers causes a reduction in election-day fraud, therefore evaluating the “observers'' effect” at the sub-national level (obviously, controlling for specific polling-station characteristics such as dimensions, ethnic affiliation, and whether it is rural or urban). In other words, if international election observation reduces election-day fraud directly, all else held equal, the party that is cheating should gain, in mean, less votes in polling stations visited by observers compared to polling stations that were not visited.