Essays on Conflict and Terrorism

Abstract

This dissertation explores how aid affects the probability of a conflict onset and how the phenomenon of political terrorism can be studied using network analysis. In the first chapter I exploited new geo-localized data available through the geographic information system (GIS). Given that the literature on conflicts demonstrates that civil conflicts tend to be highly localized within a country and project-based aid are becoming one of the main channels for donors, I examined the impact of foreign aid at a local level on the probability of conflict onset. I will show that there is a positive link between aid and the probability of a conflict onset, in particular for aid directed to sectors classified as fungible, while there is no effect on the persistence of the conflict. In the second chapter of my thesis, I proposed the first original micro dataset on Italian terrorists’ socio-economic characteristics. I reconstructed the network of the Italian terrorist groups of all political colours and I proposed an analysis using the concept of centrality in order to detect characteristics that help an individual to reach a central position in this peculiar setting. Results highlight the role of women and the importance of the trade-off which characterises those connections. In the final chapter, I analysed the dataset created in the previous chapter in order to study recruitment choices made by the Italian terrorist groups. By using the concept of homophily, the goal of the work is to understand whether the recruitment policy of a terrorist organization follows a strategic path. Results confirm a different behaviour with respect to what the standard literature on social network suggests, which tends to enlarge through time, reinforcing the idea that recruitment choices are not merely based on political ideology but they are made considering strategic aspects.