Environmental and Resource Economics 19: 173–210, 2001.© 2001 Kluwer Academic Publishers. Printed in the Netherlands.173Contingent Valuation: Controversies and EvidenceRICHARD T. CARSON1, NICHOLAS E. FLORES2and NORMAN F. MEADE31Department of Economics, University of California, San Diego;2University of Colorado, Boulder;3National Oceanic and Atmospheric Administration, U.S. Department of CommerceAccepted 14 August 2000Abstract. Contingent valuation (CV) has become one of the most widely used non-market valuationtechniques. CV’s prominence is due to its flexibility and ability to estimate total value, includingpassive use value. Its use and the inclusion of passive use value in benefit-cost analyses and environ-mental litigation are the subject of a contentious debate. This paper discusses key areas of the debateover CV and the validity of passive use value. We conclude that many of the alleged problems withCV can be resolved by careful study design and implementation. We further conclude that claimsthat empirical CV findings are theoretically inconsistent are not generally supported by the literature.The debate over CV, however, has clarified several key issues related to nonmarket valuation and canprovide useful guidance both to CV practitioners and the users of CV results.JEL classification: Q26, D611. IntroductionTo fully assess the economic desirability of environmental policies, analystsmust estimate the value of non-market commodities. Overlooking or ignoring theservices provided by non-market commodities in cost-benefit analyses and otherempirical economic studies severely undermine the accuracy and relevance ofthe results. Since the 1960s, several non-market valuation techniques have beendeveloped in recognition of the importance of these services. Of these techniques,the most commonly used is contingent valuation (CV).1Its flexibility facilitatesvaluation of a wide variety of non-market goods, including those not currentlyprovided. Perhaps more importantly, CV enables researchers to assess total valuethat includes passive use value.2In many instances, the magnitude of passive use value may be substantial.However, it has often gone unmeasured. Inclusion of passive use value potentiallyincreases the stakes in natural resource damage assessments and may tip the scalesin favor of preserving natural resources over development in individual projectanalyses.3Although CV is the most frequently used non-market valuation tech-nique for environmental goods, debate persists over the reliability of CV and theoverall suitability of passive use values in economic policy analysis. Consequently,several industry groups have voiced their opposition by lobbying against the use174 RICHARD T. CARSON ET AL.of CV and by sponsoring research aimed at investigating CV’s reliability and thenature of passive use value. In the course of this debate, many theoretical andempirical issues have been raised.4As a consequence, there has been a funda-mental rethinking of many of the key issues related to assessing the benefits ofenvironmental amenities. What emerges is a much richer theoretical frameworkfor non-market valuation and a variety of approaches for assessing the quality ofCV results. Within this context, the purpose of this paper is to provide a conciseoverview of some of the most commonly alleged weaknesses of CV and passiveuse value and to offer counter-arguments.The paper is organized into seven sections. Section 2 addresses welfareeconomic issues associated with CV and passive use. Section 3 identifies keyelements in the design, implementation, and analysis of CV surveys and theirrelevance to CV criticisms. Section 4 discusses empirical results from CV studiesand their consistency with economic theory. Section 5 discusses strategic behaviorand its relevance to CV elicitation formats. Section 6 discusses issues related to thevalidity and reliability of CV estimates. Section 7 offers concluding remarks.2. Welfare Economic IssuesThe link between welfare economics and CV is quite direct: CV offers the poten-tial to trace out the willingness to pay distribution for a population of economicagents for a proposed change in a good. If properly executed, CV is a useful toolfor benefit-cost analysis. CV can also be used for other purposes where knowl-edge of the willingness-to-pay distribution and its relationship to other variables,e.g., income and geographic location, is of interest. Policy makers often considerdistributional and political criteria in addition to welfare economic criteria.Welfare economics, through benefit-cost analysis, seeks to reveal whether thepotential change in utility resulting from a change in an economic variable, such asa change in a commodity’s price or the level of provision, is positive (Just, Huethand Schmidt 1982). The welfare implications are often expressed in terms of achange in an index, usually the monetary amount which would need to be takenfrom or given to the agent to keep the agent’s overall level of utility constant. Atthe level of an individual economic agent, these monetary measures take a partic-ularly simple form: for a desired increase in the good, the maximum amount theagent would be willing to pay to obtain the improvement, and for a decrease, theminimum amount the agent would be voluntarily willing to accept in compensationin exchange for accepting the decrease.5Whether willingness to pay (WTP) orwillingness to accept (WTA) is the appropriate measure depends upon the relevantproperty right to the good. A number of different proposals for how to aggregatethe monetized measures obtained from agents have been advanced (Mueller 1989).CONTINGENT VALUATION 1752.1. TOTAL VALUEThe term total value, synonymous with true WTP or WTA, arose in environmentaleconomics with the awareness that sometimes substantial portions of WTP or WTAwere not accounted for in the measure of economic value obtained using marketprices or revealed preference techniques. In this sense, WTP and WTA estimatesderived using those approaches are defective since welfare gains or losses may beoverlooked if passive use value is decisive.While a variety of distinctions have been proposed (Carson, Flores and Mitchell1999), for our purposes here, it is useful to decompose total value into direct usevalue and passive use value. Direct use can be most easily thought of as requiringthe agent to physically experience the commodity in some fashion. Since an agentneed not directly use a