Conflict management and the misapplication of moral hazard theory

Abstract

This piece develops a formal model of moral hazard and evaluates its utility for the study of conflict management. Although the concept of moral hazard is perhaps heuristically useful, the analysis shows that it generally does not accurately describe the incentive structure produced by the conflict management activities of third parties. Moral hazards only occur when there is an opportunity for agents to take ‘hidden actions’, and this is seldom the case where humanitarian intervention is concerned. Thus, the application of moral hazard theory to conflict management may lead analysts and policy makers astray.

Acknowledgements

For helpful comments, I thank Timothy Crawford, Benjamin Cohen, Peter Digeser, Alan Kuperman, Rose McDermott, Harrison Wagner and members of the Political Science Faculty Colloquium at UCSB. Kirk Lesh provided valuable research assistance.

Notes

1. An under-specified or ambiguous contract can cause problems if the insuring and insured parties have different views about the nature of the contract. Indeed, this may be one of the causes of unwanted outcomes in humanitarian intervention. Nevertheless, the security guarantee, by logical necessity, must be similar to a contract, otherwise the insured party's behaviour would not change. In international politics there is no external authority to enforce contracts; therefore contracts must be self-enforcing. In the case of humanitarian intervention costly signals, reputational effects and other costs might serve as a commitment mechanism that allows an insured party to realize that the insuring party is not merely engaging in ‘cheap talk’ that should be completely discounted.

2. Kreps defines moral hazards as situations “where one party to a transaction may undertake certain actions that (a) affect the other party's valuation of the transaction but that (b) the second party cannot monitor/enforce perfectly” (1990, p. 577). Using an example of a shop owner attempting to contract with a potential manager, Mas-Colell et al. (1995, p. 477) define moral hazard as “the inability to observe how hard [a] manager is working”.

3. See, for example, Mas-Colell et al. (1995, p. 477) who discuss “the hidden action case, also known as moral hazard”. Kreps (1990, p. 578) also uses the terms interchangeably, but notes that some scholars have attempted to make nuanced efforts to differentiate the two.

4. See Mas-Colell et al. (1995, pp. 487–488) for a discussion of these conditions.

5. Friedman (1966) and Waltz (1979) provide good explanations of why models and theories should be judged on their ability to explain or predict, not on whether their assumptions perfectly depict reality and can be deemed to be ‘true’ or ‘false’.

6. On the problems associated with uncertainty over preferences before issuing a contract, see Rauchhaus (2005a) for a discussion of adverse selection.

7. For an analytic treatment of the three-player issue, see Crawford (2003). For a formal model of this issue, see Rauchhaus (2005b). Cetinyan (2002) also develops a three-player game where a third party intervenes, but the purpose of intervention is not conflict management. Instead, the third party is essentially a co-belligerent who supports the ethnic group.