Agriculture, Income and Conflicts

Abstract

Since the 1950s, armed conflicts have become more and more recurrent. Most conflicts occur in countries where incomes are heavily dependent on the agricultural sector. This dissertation aims to systematically investigate the interconnection between agriculture, income and conflicts. The first chapter is the introduction of this dissertation. Background information about conflicts is provided in this chapter. In particular, we offer statistical evidence about the quantity of conflicts, distribution of conflicts in terms of time and location and the number of deaths as the result of these conflicts. This background information is important for understanding the severity of conflicts and the significance of reducing them. The second chapter is the overview chapter. The purpose of the overview chapter is to provide a literature review about the interrelationship between agriculture, income and conflicts. In this overview chapter, we start our discussion about how conflicts can hinder economic development emphasizing the importance of studying conflicts. We also discuss the development of conflict-related studies in the literature, estimation methods and potential issues when researchers attempt to investigate the effect of income variations on conflicts. We then end this chapter by analyzing the role of agriculture, especially the impact of agricultural productivity on conflicts. The third chapter is the main study in this dissertation and is about estimating the effect of rainfall shocks on conflicts. We thoroughly examine the results on the negative relationship between rainfall shocks and conflicts in African countries from Miguel, Satyanath and Sergenti (2004). We consider the role of data revision and cross-sectional dependence in their estimation. We find that the negative relationship between rainfall shocks and conflicts in Miguel, Satyanath and Sergenti (2004) is not valid when the revised rainfall and conflicts datasets are used in their estimation. However, we propose a new estimator that is able to take cross-sectional dependence arising from spatially-dependent weather patterns and cross-border conflict spill-overs into account to examine the link between rainfall shocks and conflicts. Using this new estimator, we find that rainfall variations are indeed a determinant of conflicts. The fourth chapter is another main study and examines the effects of productivity-enhancing technology in agriculture on conflicts. We consider the commercial legalization of Genetically-Modified (GM) soybean cultivation in Brazil in 2003 and investigate the effects of GM soybean cultivation on land conflicts in Brazil. In this chapter, we provide a theoretical model to show that the enhancement of agricultural productivity induced by GM soybean cultivation can reduce land value and then mitigate land conflicts. To assess the validity of this theoretical prediction, we employ the Difference-in-Differences estimation and find that states that have more land that is suitable for cultivating GM soybeans after the legalization in 2003 are negatively associated with land conflicts. The empirical results on the mitigating effect of GM soybean cultivation on land conflicts are reinforced by a series of robustness checks. The fifth chapter is the conclusion of this dissertation. Specifically, we summarize the contents and achievement of this dissertation.