The hazards of thinking about moral hazard

Abstract

Some scholars have argued that expectations of humanitarian military intervention might actually increase the incidence of intrastate violence, and have invoked the concept of moral hazard from the economics literature on insurance and government regulation to explain why. However, the term ‘moral hazard’ is no substitute for a careful analysis of how expectations of intervention might influence the decisions made by participants in intrastate conflicts. I argue that an analysis of those decisions implies that the prospect of intervention might indeed make violence more likely, but for reasons that have little to do with what is commonly called moral hazard.

Acknowledgements

I would like to thank Alan Kessler, Pat McDonald and Robert Rauchhaus for their comments on earlier versions of this paper, and Alan Kuperman and Tim Crawford for helpful editorial suggestions.

Notes

1. These questions are also raised in Grigorian (2005) and Rauchhaus (2005).

2. A preliminary version of some of these ideas can be found in Wagner (2004). A more extended version is presented in Wagner (2005).

3. It is important to bear in mind, however, that the fact that possible compromise settlements are available does not alone guarantee that there will be some compromise that everyone would prefer to a winner-take-all contest (Fearon, 1995, pp. 386–388). If no feasible agreement is jointly preferred to a contest, then a contest would not be inefficient.

4. Fearon acknowledges that such inconsistent expectations may exist, but attributes them to “bounded rationality”, which seems to imply that fully rational decision makers with the same information would always have consistent expectations (Fearon, 1995, pp. 392–393). But since personal or subjective probabilities just reflect people's preferences between lotteries with known probabilities and uncertain outcomes with unknown probabilities, it is not clear in what way such incompatible expectations are inconsistent with rational behaviour, or how they could be avoided by rational decision makers.

5. This is the central theme of much of the recent literature on the relation between war and bargaining. The idea can be found in Clausewitz (1976), and is the main theme of Blainey (1988). Note, however, that, while Blainey popularized the idea that states may have conflicting beliefs about their relative military capabilities, conflicting beliefs about the distribution of the costs of war may be at least as important. For a fuller discussion and further references, see Wagner (2000). See also Muthoo (2000).

6. For an extended discussion, see Wagner (2005).

7. For a more extended discussion of the familiar contrast between ‘idealism’ and ‘realism’ in writings about politics, see ibid.

8. This is perhaps what Kuperman has in mind when he says that a norm of humanitarian intervention would undermine the ability of rulers to deter rebellion by threatening massive retaliation.

9. My thanks to Pat McDonald for helping me see this point more clearly. For a discussion of how other forms of humanitarian involvement in intrastate military conflicts can also have unexpected adverse consequences, see Terry (2002).

10. See the discussion by Marc Trachtenberg (1993) of exceptions to the norm of non-intervention in the 19th and early 20th centuries.