Risk preferences, uncertainty, and war

In the game-theoretic literature on international conflict, risk-acceptance leads to war if it is severe enough to eliminate the bargaining range, in which case war occurs even under complete information. I analyze the effect of varying risk-propensities in incomplete-information crisis bargaining. In this setting, an additional lottery (beyond just the war lottery) is involved due to the proposer's uncertainty about whether the proposal will be accepted. This has the effect that even small increases in either side's willingness to take risks increase the probability of war, as opposed to the all-or-nothing, eliminating-the-bargaining-range mechanism. This result holds regardless of whether either side is risk-averse or risk-acceptant on the whole, and even if no bargaining range is being eliminated. I also show that uncertainty about the opponent's risk-propensity can lead to war via a standard risk-return tradeoff, which is another way that risk-preferences can affect war and peace.