Conflict and Leadership: When is There a Hawkish Drift in Politics?

We analyze an agency model of political competition to examine whether conflict encourages hawkish behavior, and if such behavior can itself aggravate conflict. We consider situations of conflict between a state and an insurgent group, such as conflict over a piece of land. Negotiations are carried out on behalf of the state by a democratically elected leader whose ability and ideology are imperfectly observed by the electorate. A more capable leader can take a hardline position in the negotiations (i.e. cede less land to the insurgents) at a lower expected cost (modeled as the cost of continued insurgency) than a less capable one. Similarly, an ideologically hawkish leader enjoys greater intrinsic utility from retaining land than a less hawkish leader. Two main results that emerge are: certain types of politicians may be excessively hawkish as compared to their first best policy choices, which itself increases the probability of conflict; and for any credible voting strategy, the re-election probability of a hawk is greater than that of a dove. Finally, we show that the voting equilibrium of this game does not always achieve a constrained Pareto optimum suggesting that third party mediation may improve welfare.