Restrictions and agent selection

초록

In an agency model, we compare the use of an insider versus an outsider agent when the principal can restrict their choice sets. The insider is familiar with the old project but not with the new project, whereas the outsider is unfamiliar with both. The principal chooses between the insider and the outsider, and the selected agent manages the project—deciding between the old and the new project, as well as choosing the project's attribute. Gathering information about a project is costly to an agent. We show the following. Restricting the selected agent's choice set allows the principal to induce information gathering more effectively. Under sequential learning, the principal (weakly) prefers the insider. Under simultaneous learning, however, the outsider may be preferred. Moreover, restricting the agent's choice set enlarges the parameter range where the outsider is preferred. Sequential learning is dominated by simultaneous learning from the principal's perspective.

키워드