Do we say what we do? Testing cheap talk to reduce hypothetical bias in moral choice experiments

OBJECTIVE: Discrete choice experiments (DCEs) are widely used in healthcare to estimate willingness-to-pay (WTP) but may be affected by hypothetical bias (HB), especially in morally sensitive contexts. While cheap-talk is proposed as a mitigation strategy, its effectiveness in health-related DCEs involving moral trade-offs remains unclear. This study examines how cheap-talk influences WTP in such settings. METHODS: A split-sample DCE on organ transplantation policies was conducted, involving trade-offs between cost and morally salient outcomes: saving lives ("being alive") and improving quality of life ("having a life"). Respondents (N = 651) were randomly assigned to one of three survey arms: control (no manipulation), cheap-talk, or cheap-talk with follow-up question. Multinomial logit model in WTP space with a Taboo Trade-off Aversion (TTOA) specification was used to estimate treatment effects and interactions with religiosity. RESULTS: Exposure to the cheap-talk script reduced WTP for saving lives, indicating increased attention to financial considerations. WTP for quality-of-life improvements and avoiding taboo trade-offs remained unchanged. Religious respondents reported higher WTP to avoid taboo trade-offs in the control arm, but this gap disappeared under cheap-talk, showing that deliberation moderates monetary expressions among religious individuals without altering underlying convictions. CONCLUSIONS: Cheap-talk promotes more reflective decision-making in morally sensitive health-related choices, particularly among individuals with strong moral or religious convictions. It reduces elevated WTP for taboo trade-offs, while its effect on other respondents is limited. Future research should combine stated and revealed preference data and explore models that account for non-compensatory moral decision rules to better capture complex moral preferences DCEs.