📝 SummarXiv

Abstract

I study dynamic contracting where a principal (Sender) privately observes a Markovian state and seeks to motivate an agent (Receiver) who takes actions. Sender can both use payments to augment ex-post payoffs or persuasion to alter the informational environment as ways to provide incentives. For any stage-game payoffs, cost of transfers, rate of future discounting, and Markov transition rule, optimal transfers are backloaded-payments occur only when Sender commits to reveal the state at all continuation histories. In a rideshare example, the optimal contract is a loyalty program: drivers receive the static optimal information structure until a random promotion time, after which the state is fully revealed and only payments are used to motivate the driver.