An Endowment Effect for Risk: Experimental Tests of Stochastic Reference Points

Recent models of reference-dependent preferences indicate that expectations may play a prominent role in the presence of behavioral anomalies. A subset of such expectations-based models predicts an "endowment effect for risk": that risk attitudes differ when reference points change from certain to stochastic. In two purposefully simple risk preference experiments, eliminating often-discussed confounds, I demonstrate both between and within subjects such an endowment effect for risk. These results provide needed separation between expectations-based reference-dependent models, allow for evaluation of recent theoretical extensions, and may help to close a long-standing debate in decision science on inconsistency between utility elicitation methodologies.