STATED VS IMPLICIT WILLINGNESS-TO-PAY UNDER RISK

Individuals' willingness-to-pay (WTP) for risky alternatives was assessed using four different procedures in order to test hypotheses about psychological processes underlying WTP. Key variables manipulated were (1) costs described as losses vs payments (framing), (2) integration vs segregation of costs and risky outcomes, and (3) judgment vs choice-based response mode. The largest and most striking effect was that, contrary to a hypothesis suggested by loss aversion, WTP was actually higher when the cost of an alternative was described as a ''loss'' than when it was described as a ''payment.'' This difference was replicated in a second experiment. The Transaction Analysis framework of Fischhoff and Furby (1988) was used to generate a list of contextual factors that might influence subjects' WTP and perhaps explain this puzzling effect. In Study 2, subjects rated the importance of statements corresponding to these and other factors. These ratings suggested that the operative difference between the ''loss'' and ''payment'' conditions was not framing. Instead, implicit norms associated with transactions were apparently evoked less strongly by the loss condition. These results are evidence that experimental manipulations of theoretical interest in behavioral economics may induce unintended disparities between the problem structures and assumptions of researchers and subjects. Beyond the clear methodological implications, these findings point to the need for behavioral decision theorists to pay more attention to the ''gap-fillers'' and ready-made scripts that individuals bring to every task. (C) 1995 Academic Press, Inc.