Are Hypothetical Referenda Incentive Compatible? A Comment

In this comment we argue that the experimental evidence provided by Cummings et al. is not inconsistent with hypothetical referenda being an incentive compatible mechanism for eliciting economic values. Differences in response proportions between hypothetical and real referenda can be expected if the underlying distributions have equal means but differ in scale. Once rescaled, Cummings et al.’s experiments provide evidence that responses to hypothetical referenda in experimental markets are not incapable of accurately estimating willingness to pay relative to responses to real referenda in experimental markets. By relaxing the assumptions of equal covariate effects and homoskedastic responses, we show that the Cummings et al. experiments are indistinguishable from an experiment with random draws from a binomial distribution with equal probabilities of success and failure. Corresponding Author Timothy C. Haab Department of Economics East Carolina University Greenville, NC 27858 (919) 328-6820 [email protected]