Choice experiments addressing outcome uncertainty (OU) typically reframe continuous probability densities for each risky outcome into two discrete categories, each with a single probability of occurrence. The implications of this simplification for welfare estimation are unknown. This article evaluates the convergent validity of willingness-to-pay (WTP) estimates from a more accurate multiple-outcome treatment of OU, compared to the two-outcome approach. Results for a case study of coastal flood adaptation in Connecticut, United States, suggest that higher-resolution OU treatments increase choice complexity but can provide additional information on risk preferences and WTP. This tradeoff highlights challenges facing the valuation of uncertain outcomes.
