Do numerical probabilities promote informed stated preference responses under inherent uncertainty? Insight from a coastal adaptation choice experiment

Inherent outcome uncertainty within stated preference surveys is invariant across valuation scenarios. It has received relatively little attention in the environmental stated preference literature. Specifically, it is unknown whether percentage probabilities—a ubiquitous means of communicating uncertainty in questionnaires—are an effective risk communication tool. This article systematically evaluates two treatments in a discrete choice experiment survey related to coastal climate change adaptation in Connecticut, USA: one provides only raw frequencies as a risk communication tool, while the other provides implied numerical probabilities in addition to the same raw frequencies. Results from a mixed logit model and from a latent class model that controls for sociodemographic influences show that the use of percentage probabilities to communicate inherent uncertainty has no additional effect on average welfare estimates or the choice behavior of respondents. Our findings suggest that percentage probabilities may not be an impactful way to communicate inherent uncertainty in environmental stated preference questionnaires.