Can stated measures of willingness-to-accept be valid? Evidence from laboratory experiments

Abstract

Willingness-to-accept (WTA) questions have been largely abandoned in stated preference empirical work in favor of eliciting willingness-to-pay (WTP) responses, mainly due to perceived unreliability of questions that ask respondents for compensation amounts. This paper reassesses whether stated WTA welfare measures can be valid in public and private good contexts. We conduct two sets of laboratory experiments to analyze whether elicitation format, survey design and framing, and follow-up questions can generate truthful responses. For public goods, we adapt the existing WTP incentive compatibility theoretical framework to the WTA context and test the theory using an experiment involving voting. Results are consistent with the WTP literature and suggest that WTA values can be valid as long as responses have consequences for respondents. For the private good experiment, we focus on whether respondents are motivated to affect the price or the provision of the good. We find that strategic behavior is present and in the direction expected by theory. Survey framing and the use of follow-up questions can provide bounds on the value estimates. These findings raise potential concerns with the use of non-incentive compatible elicitation mechanisms in WTA contexts.