We provide controlled laboratory evidence that open-ended hypothetical surveys do not always accurately elicit real economic commitments from individuals. We argue that they can provide biased measures of true values, where the latter are elicited using incentive-compatible institutions. We also...