Risky Sure Things

Decision research often takes the variability of potential outcomes as a measure of risk. It thus characterizes sure things, which, by definition, guarantee a specific outcome, as safe. But is this characterization always empirically valid? We show that, when the prevailing reference point is an uncertain option or position, sure things can be perceived as risky rather than safe. Furthermore, preferences may hinge on such perceptions: when construed as risky, sure things can be less appealing. Our findings suggest a perception-based explanation for why the classic tendency to favor sure things over uncertain options is often attenuated given an uncertain reference point. More broadly, they tap an unresolved debate about the determinants of decisions. Much research focuses on taste-based determinants, such as attitudes toward perceived risks, and thereby underplays the critical role of the underlying perceptions. Appreciating the impact of perceptions on decision making leads to novel prescriptive recommendations. This paper was accepted by David Simchi-Levi, decision analysis. Supplemental Material: Data and the online supplement are available at https://doi.org/10.1287/mnsc.2022.4590 .