Further tests of entreaties to avoid hypothetical bias in referendum contingent valuation

Thomas C Brown, Icek Ajzen, Daniel Hrubes

2003-09-01

Abstract:Over-estimation of willingness to pay in contingent markets has been attributed largely to hypothetical bias. One promising approach for avoiding hypothetical bias is to tell respondents enough about such bias that they self-correct for it. A script designed for this purpose by Cummings and Taylor was used in hypothetical referenda that differed in payment amount. In comparisons with behavior observed in otherwise identical real payment referenda, the script worked remarkably well at higher payment levels (dropping the likelihood of a yes vote to the level obtained in separate sample real referenda), but less well at a lower payment level. The repeated success of the script at higher payment amounts is a clear sign that additional research is warranted—research that will bring us a more complete understanding of how the entreaty works. This research may need to explore its workings at the individual respondent level.