'Scenario Adjustment' in Stated Preference Research

To assess demand for non-market goods, researchers must sometimes resort to direct elicitation of consumer tradeoffs with the use of surveys. Stated preference (SP) methods typically involve surveys of consumers wherein choice scenarios are posed to respondents and individuals are asked to indicate their preferred alternatives. As SP research has matured, much progress has been made to address a variety of well-known biases that can afflict demand estimates produced by these methods, but some concerns still remain. We use an existing survey designed to ascertain willingness to pay for private health-risk reduction programs to illustrate yet another potential source of bias. This bias is caused when not all respondents answer exactly the choice question they are asked and that the researcher intended for respondents to answer. SP researchers are familiar with the problem of outright “scenario rejection,” where respondents may choose the status quo alternative because they reject the viability of the proposed alternatives. In contrast, we address the more subtle problem of “scenario adjustment,” where respondents impute that the substantive alternative(s) in a choice set, in their own particular case, will be different than the survey instrument suggests. We demonstrate a strategy to control and potentially correct for scenario adjustment in the estimation of willingness to pay.