What would others do? Revisiting inferred valuation to mitigate social desirability bias in stated preferences. / Job Market Paper

Abstract

Stated preference practitioners have focused on designing methods to reduce errors in willingness to pay (WTP) and willingness to accept (WTA) valuation caused by well-known biases. Among these methods, inferred valuation (IFV) offers an approach to identify social desirability bias (SDB) by asking respondents to infer the behavior of a reference group. To test the validity of this approach, we construct a conceptual model of the relationship between inferred values and the similarity of the respondent to the reference groups. We examine the respondent’s perceived income similarities to their neighbors and to the average citizen when inferring their preferences regarding the reclamation of Oil Sands Processed Water (OSPW) in Northern Alberta. We find evidence of SDB in our sample and our results show that perceived similarities influence the inferred WTP and WTA. However, the direction of the effect of these perceived similarities does not always conform with our theoretical expectations, calling into question the validity of IFV as a method for capturing SDB. Future research would benefit by being cautious when implementing IFV for both welfare measures and by providing detailed information about the reference group