Economic and political agencies use information to manage their relationships with oversight committees, other agencies, and the general public. Information releases are timed and crafted to manage expectations about current and future developments within the agency. This proposal studies how information management policies can be and are used to shape incentives of counterparties to further the interests of the agency. The key consideration is in how the timing and content of releases trades off good and bad news. As a general theme it is in the interest of the agency to release minimally good news whenever possible and maximally bad news as infrequently as possible.
This is modeled formally as a principal agent relationship in which a principal is continuously monitoring a stochastic process and strategically releasing information about the process to an agent who takes actions that affect the principal. The agent knows the law of the process but relies on the principal for specific information. Absent any information the agent will eventually evolve beliefs inducing actions against the interest of the principal. The proposal studies how information is managed over time in order to minimize the frequency of these events and maximize the duration of cooperative behavior by the agent.
