Conflict, interest and strategy: A risk limit approach to conflict

The paper criticizes Axelrod's (1970) measure of conflict, which is basically a measure of the conflict of interest and does not account for the potential for overt conflictful behavior based on how the actors are strategically interpositioned. Differences between the two dimensions of conflict are illustrated, and it is shown how both dimensions are necessary for a theory of conflict. Conflict is conceptualized as a struggle for preferred equilibria in 2x2 games with at least two equilibria. A new measure of conflict for this class of games based on the concept of risk limit is developed in both a static and a repeated game setting. The risk limit approach measures the actors' inclination to behave in a conciliatory or challenging way. This is different from Axelrod's approach, which measures the incompatibility - and hence the conflict - of the actors' preferences. We show that higher incompatibility of preferences between the actors does not necessarily imply that they are more inclined to behave challengingly. Furthermore, we show that the more actors value future payoffs in terms of present payoffs, the greater their inclination will be to behave challengingly.