HYPOTHETICAL BIAS AND AN EXPLICIT CONTRACT

A concern about the contingent valuation method (CVM) is the finding that hypothetical and actual payments often differ. In particular, hypothetical payments are generally higher than actual payments resulting in “hypothetical bias”. Many methods have been proposed for reducing this bias, such as certainty adjustment and cheap talk, but none of these appear to work well in all situations. In this paper we test for the effect on hypothetical bias of a social contract. Specifically, we test whether hypothetical bias is decreased by asking participants to sign an oath of honesty. An experimental approach was used which asks for actual and hypothetical group payment to Heifer International using a modified BDM framework. One treatment received an oath while the other did not. Two different univariate and multivariate tests for the effect of the oath on hypothetical bias were conducted. In both cases, the oath reduced hypothetical bias. We concluded that an oath of honesty reduces hypothetical bias significantly and that this may offer a new and better way of controlling for this bias in CVM studies.