Delegation to a possibly ignorant agent
Aggey Semenov
Abstract
Aggey Semenov
Abstract
Abstract In a delegation problem, an uninformed principal delegates decision‐making powers to a biased and possibly ignorant agent. The principal cannot use monetary payments but can restrict the set of the agent's choices. I show that in the general case, the principal may offer a disconnected set of choices. In a setting with arbitrary bias the uncertainty principle holds—the principal benefits as the likelihood that the agent is informed increases. When the bias is constant, I show that the ally principle holds—the principal benefits as the bias becomes smaller. Finally, when the likelihood of the agent being informed is determined by the agent's effort, then for small biases the principal benefits from limiting the agent's choice. When the agent is sufficiently biased, the principal gives more choice to the agent, so as to to improve incentives for information acquisition.
OpenAlex reports 2 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Abstract In a delegation problem, an uninformed principal delegates decision‐making powers to a biased and possibly ignorant agent. The principal cannot use monetary payments but can restrict the set of the agent's choices. I show that in the general case, the principal may offer a disconnected set of choices. In a setting with arbitrary bias the uncertainty principle holds—the principal benefits as the likelihood that the agent is informed increases. When the bias is constant, I show that the ally principle holds—the principal benefits as the bias becomes smaller. Finally, when the likelihood of the agent being informed is determined by the agent's effort, then for small biases the principal benefits from limiting the agent's choice. When the agent is sufficiently biased, the principal gives more choice to the agent, so as to to improve incentives for information acquisition.
Key concepts: Delegation, Principal (computer security), Principal–agent problem, Incentive, Limiting, Set (abstract data type), Payment, Microeconomics