A committee of experts must decide on the value of a numerical variable (the interest rate for instance). Experts have different views on the value that should be adopted so that usual methods face a number of shortcomings, such as allowing for manipulation or failing to give incentives for coordination. We design the class of Approval mechanisms which allow the experts to report ranges of values. We prove that these mechanisms admit a unique equilibrium outcome and hence are able to moderate the final decision. Moreover, we test experimentally these mechanisms and give evidence that the experimental subjects prefer reporting ranges of values to the more conventional single-outcome announcements.
Approval Mechanisms: Theory and Experiments
Jan 31st, 2019