The validity of the contingent valuation method: Perfect and regular embedding

Abstract

This article deals with one aspect of the validity claim of the contingent valuation method, namely: to what extent does the method produce different values in situations for which economic theory claims different values. Two aspects of this validity test — perfect and regular embedding — were tested in a field experiment comprising six samples of the Dutch population. Perfect embedding occurs when the value for a specific good is similar to the value for a more inclusive good. Perfect embedding did not manifest itself in the experiment. Respondents considered a package of six goods as well as a package of two goods more valuable than one of these goods. It is argued that this supportive evidence of the method's validity claim occurred because the goods involved were well-defined. Regular embedding occurs when the same good receives a lower value if the value for it is inferred from the value for a more inclusive good rather than if that good is valued on its own. Regular embedding was only found when respondents were given the opportunity to value the inclusive good before valuing the specific good. Respondents who were given information on the inclusive good without valuing it did not state different values than respondents who were not given that information. It seems that respondents perceive an inclusive good as being relevant to their valuation decision only when they are asked to value it. Further research is necessary to shed more light on the underlying processes that may account for this.