e obtain the following result: when there are negative income shocks the opportunity cost of conflict diminishes. This is exacerbated by the presence of imperfect information, as the contenders second guess each other’s intentions.
The economic costs of civil war are devastating (Collier, 1999). It remains a major cause of underdevelopment and poverty, particularly in Sub-Saharan Africa. Indeed, the indirect effects due to increased incidence of disease and famine are huge (Sambanis, 2002; Montalvo and Reynal-Querol, 2007). This issue has come to the forefront of Economics research. Indeed, there is a number of recent empirical studies on the causes of civil war (Collier and Hoeffler, 1998, 2004b; Fearon and Laitin, 2003; Sambanis, 2002; Hegre and Sambanis, 2006; Montalvo and Reynal-Querol, 2005; Miguel, Satyanath and Sergenti, 2004). Two robust empirical patterns emerge from this literature. First, civil wars overwhelmingly take place in poor countries. Second, a civil war tends to occur right after an unfavourable economic shock.
Existing economic models of conflict have difficulty dealing with these stylized facts. These models follow one of two traditions. On the one hand, there are static models in which the contenders can devote resources to a technology of expropriation (Skaperdas, 1992; Grossman, 1991, 1995, 1999; Hirshleifer, 1995, 2001). These models are income-neutral in the sense that the amount of resources diverted to expropriation is typically not diminishing in income. The reason is that while the opportunity cost of these resources increases with income, the potential returns to expropriation also increase. On the other hand, other scholars have focused on models in which conflict occurs as a consequence of bargaining breakdown (Fearon, 1995; Powell, 1996a, 1996b, 1999). In these models, the bigger is the prize to be divided, the bigger are the incentives to fight. Hence neither of these literatures can accommodate the fact that civil wars occur in bad times in countries with low GDP.
The immediate objective of this research project is to fill this important void in the literature. The aim is thus to formulate a theoretical framework of civil war that generates predictions consistent with the stylized facts unveiled in the empirical literature.
