No abstract

This paper focuses on organizational issues of allocating authority between an uninformed principal and an informed expert. We show that the standard result that delegating decisions to a perfectly informed expert is better than communication is reversed if the principal can restrict the precision of the expert's information (without learning its content). We demonstrate that these organizational forms-informational control and delegation-can be either complements or substitutes, depending on the principal's ability to a¤ect the expert's discretion about the set of allowed policies.JEL classi…cation: C72, D81, D82, D83

No abstract

No abstract

This paper studies a multi-stage version of Crawford and Sobel’s communication game. In every period the receiver determines a test about the unknown state whose result is privately observed by the sender. After the sender sends a costless message about an outcome of the test, the receiver selects a test in the next period. After a finite number of periods of interaction, the receiver makes a decision. The paper offers a sequence of tests that refine sender’s information step-by-step and preserve truthtelling in every period. This sequence allows the receiver to learn the state in a subinterval of the state space with an arbitrary precision and has appealing theoretical properties. It consists of simple binary tests which reveal whether the state is above a certain cutoff, where the cutoffs are monotonic across periods and independent from results of the previous tests. Finally, we show that the relative payoff efficiency of multi-stage interaction compared to a single-stage game increases without a bound as the bias in preferences tends to zero.