A principal-agent perspective on counterinsurgency situations

Abstract

The successful conclusion to the war in Afghanistan necessitates an array of solutions that includes non-kinetic actions. In particular, there is growing interest in having the local population take control of the regional security, providing protection from external threats as well as internal sources of violence. In this thesis, the principal-agent paradigm is used to model the relationship between coalition forces (principal) and local tribes (agents).These relationships are affected by the lack of alignment of interests of the principal and the agents. To achieve the desired alignment, the principal must provide incentives to the agents. Two scenarios are considered: The simplest where the agents do not have private information and their actions can be verified. This scenario is used to determine a baseline for the transfers offered by the principal. The second- and more realistic- scenario captures the principal's lack of knowledge about the actions taken by the agents. The last model provides qualitative insights about the cost to the principal due to unknown information.