De Figueiredo and Weingast: The rationality of fear

De Figueiredo and Weingast. 1999. The rationality of fear: Political opportunism and ethnic conflict. In Civil Wars, Insecurity, and Intervention, eds. Walter and Snyder. New York: Columbia University Press, pp. 261-302.

In Brief

The authors attempt to answer to puzzling question:

1. Economic puzzle: Why would individuals and groups accept the high costs of ethnic conflict rather than the high gains of cooperation?
2. Political puzzle: Why does conflict occur at the moment it does?

In the authors' words:

"Understanding the ethnification of politics requires addressing two fundamental puzzles, one economic and one political (Weingast 1998). The fundamental economic puzzle of ethnification concerns its huge costs. Individuals and groups locked in these struggles forgo the enormous benefits of economic and social cooperation in favor of bitter violence and hardship. Why do citizens take actions leading to this negative-sum outcome? The fundamental political puzzle of ethnification concerns its timing. How do we explain the often sudden eruption of ethnic violence, especially when it follows a long period of peace? Addressing this puzzle requires three cornponents: an explanation of the period of quiescence, the timing of violence, and why violence erupts so suddenly, often in full force in a very short period."

Main Argument

"Three factors interact to produce ethnic violence: leaders with a tenuous hold on power, fear among the citizenry, and uncertainty about the true intentions of propagators of violence." The third factor is essential to understand "the link between a leader's claims and the rising fear among citizens."

The variables, then, are as follows:

• Y: Ethnic violence
• X1: Leaders with a tenuous hold will make a very risky gamble (conflict) to "gamble for [political] resurrection."
• X2: Citizens who fear for their (group, thus individual) survival will support such claims--if they believe that they are in danger (and that it isn't their leader's fault--see X3).
• X3: Uncertainty about the nationalistic politician's true intentions--voters must believe that the nation faces genuine danger, not danger resulting only from X1.

More on the importance of X3:

"The critical factor beyond Milosevic's direct control was that the Croatians' actions 'confirmed' (in the Bayesian sense) Milosevic's claims about them. Given the causal ambiguity about growing tensions [i.e. Serbs weren't sure who was causing the pot to boil], Croatia's actions increased the 'average' Serbian citizen's subjective assessment that the Croatians were bent on violence against Serbs. Had the average citizen known for sure that Milosevic sought to incite the Croatians [i.e. that it was Milosevic, not the Croatians, causing tension], the average citizen would not have reacted with the fear of bad consequences that moved them to support Milosevic." (See example on 264-265.)

Two Further Results

The authors' model also leads to these conclusions:

1. If beliefs about the probability of victimization (by another group) reach a certain point, the average citizen will support the nationalist leader. As the expected costs of victimization rise (relative to costs of war), then the average citizen will support the nationalist leader even if the probability of victimization is much closer to 0 than 1.
2. Conflict in Yugoslavia wasn't inevitable. If Milosevic wasn't unsure about his political future, or if his reform-minded opponents were in power, then nobody would have wanted to pursue an ethnic strategy.