PREVENTING_CONFLICTSProject reference: 203576
Funded under :
Understanding and preventing conflicts: on the causes of social conflicts, and alternative institutional designs for their prevention
Total cost:EUR 1 330 000
EU contribution:EUR 1 330 000
Call for proposal:ERC-2007-StGSee other projects for this call
Funding scheme:ERC-SG - ERC Starting Grant
The research project will use theoretical models and empirical techniques to explore the causes, consequences, and prevention mechanisms of conflicts. The aim is to determine the basic elements that make countries more prone to social conflicts and then identify a set of feasible policies to prevent future episodes of violence. The project considers the causes and the propagation mechanisms of social conflicts of different intensity. The main objective of the project is to the study the institutional designs that may prevent, or mitigate, such social conflicts. Therefore, the analysis of economic institutions (such as property rights, etc.), political institutions and structure (democracy, decentralization, political systems, etc.), and the type of political leaders, that can help to prevent, conflict in potentially conflictive societies. From a methodological perspective, the project proposes to overcome some statistical pitfalls present in most of the previous literature on the determinants of civil wars and conflicts. The use of simple linear regressions, or a probit/logit specification, imposes very strong identification conditions that are likely to be violated. The current consensus, which emerges from those analyses, is that poverty is the single, most important determinant of civil wars. This result could be an artifact of simultaneity problems: the incidence of civil wars and poverty may be driven by the same determinants, some of which are probably missing. We propose to check the robustness of this consensus idea, and the importance of the institutional design, using other econometric procedures (instrumental variables and matching methods) which are subject to weaker identification conditions than the traditional regressions. Finally, we plan to investigate methods to deal with the missing data problem that plague the study of the determinants of civil wars.