Experimental results on expressed certainty and hypothetical bias in contingent valuation.

• Author: Blumenschein, Karen

1. Introduction

   In order to determine the optimal level of nonmarket goods, such as environmental quality, information about the willingness to pay for the good is needed. The contingent valuation method has been developed to measure the willingness to pay for environmental changes (Mitchell and Carson 1989). The method means that individuals are asked about their hypothetical willingness to pay for a defined good. The use of the contingent valuation method is controversial among economists (Diamond and Hausman 1994; Hanemann 1994). The controversy is centered around the extent to which the hypothetical choices in the contingent valuation method correspond to real economic choices. Despite this controversy, relatively little experimental work has been carried out directly comparing hypothetical and real willingness to pay.

   A number of different techniques exist to elicit the willingness to pay in a contingent valuation study. Currently, the most commonly used elicitation format is the dichotomous choice approach (Hanemann 1994). In a dichotomous choice contingent valuation question, a subject responds yes or no to a hypothetical question about paying a specified price for a defined good. By varying the price in different subsamples of respondents, the mean willingness to pay for the good can be estimated (Hanemann 1994).

   Cummings, Harrison, and Rutstrom (1995) recently presented data from an experiment comparing the dichotomous choice contingent valuation method for measuring willingness to pay with real decisions. Three experiments were carried out on three different consumer goods (an electric juicemaker, a calculator, and a box of chocolates). For each good, the proportion of hypothetical yes responses exceeded the proportion of real yes responses. In early work by Bishop and Heberlein (1979) and in more recent work by Nape et al. (1995), dichotomous choice contingent valuation questions to measure willingness to accept also have been shown to lead to similar problems of overestimation. Overestimation in hypothetical willingness to pay questions has also been shown for hypothetical referenda (Bjornstad, Cummings, and Osborne 1997; Cummings et al. 1997) and open-ended willingness to pay questions (Neill et al. 1994; Loomis et al. 1996).

   The overestimation of the dichotomous choice contingent valuation approach noted by Cummings et al. (1997) was also confirmed in another experiment on private goods by Johannesson, Liljas, and Johansson (1998). In that experiment, the hypothesis that a more conservative interpretation of the approach, where only definitely sure yes responses are counted as yes responses, correctly predicts real purchase decisions was also tested. Definitely sure yes responses were found to significantly underestimate the real yes responses and thus provided a lower bound for the real willingness to pay. A related approach was also used in a recent study by Champ et al. (1997), which compared hypothetical dichotomous choice questions about donating a specified amount to a public good with actual donations to the public good. They also assessed the certainty of the hypothetical donation responses on a 1-10 scale from very uncertain to very certain. They found that hypothetical donations significantly exceeded real donations but that there was no significant difference if only subjects that were very certain of their yes responses were counted as real yes responses.

   In this study, we carried out another experiment to compare the dichotomous choice contingent valuation method with real decisions for a consumer good. We confirm the results of Cummings, Harrison, and Rutstrom (1995) that hypothetical yes responses overestimate real purchase decisions, but we cannot reject the null hypothesis that definitely sure yes responses correspond to real purchase decisions. The design of the experiment is outlined below along with the results and some concluding remarks.

2. Design of the Experiment

   Our experiment involved a pair of sunglasses made by UVEX. These sunglasses are used as protective eyewear in professional laboratories in the U.S. and are not typically available in ordinary retail shops. The sunglasses are similar in style to sunglasses used for sporting activities such as bicycling and playing volleyball. We ordered the sunglasses directly from the wholesaler at a bulk price of $9 per pair. Undergraduate students taking an introductory economics class at the University of Kentucky College of Business and Economics were included in the experiment. Participation in the experiment was voluntary and each student received $5 for participating.(1) In total, 133 subjects participated in the experiment, which was carried out in August 1996.

   The subjects were divided into two groups. The first group received a hypothetical dichotomous choice question followed by a real purchase question, and the second group received only the real purchase question. The hypothetical question was worded in the following way:(2)

   We are interested in how much you are willing to pay for the sunglasses displayed at the front of the room. These sunglasses are made by UVEX and block 99% of ultraviolet light (both UVA and UVB). They have an antifog and antiscratch coating as well as sideshields, which make them suitable for driving, hiking, and other outdoor activities. These sunglasses have moveable temples so that they can be adjusted for individual fit. The sunglasses will be brought around for you to examine before you tell us your answer.

   Assume that you were offered the chance to buy these sunglasses here and now with your own money. Observe that you are not actually being offered the chance to buy the sunglasses but that you are asked to think about what you would do if you were offered the chance to purchase the sunglasses here and now. Would you buy the sunglasses here and now at a price of $5.00? Note that you are to answer if you would buy the sunglasses here and now at a price of $5.00 and not if you would buy the sunglasses on any other occasion at this price. Please circle yes or no below when you have made up your mind.

   This question was...