Improving the Performance of Contingent Valuation Studies in Developing Countries

Abstract

This paper discusses three main reasons why so many of the contingentvaluation studies conducted in developing countries are so bad. First,the contingent valuation surveys themselves are often poorly administeredand executed. Second, contingent valuation scenarios are often very poorlycrafted. Third, few CV studies conducted in developing countries aredesigned to test whether some of the key assumptions that the researchermade were the right ones, and whether the results are robust with respectto simple variations in research design and survey method. The paper concludesthat research on stated preference methods in developing countries iscritically important to the successful implementation of these methodsbecause (1) there is no empirical evidence to suggest that rapid,”streamlined” CV surveys yield reliable, accurate results, and (2)there is a significant risk that the current push for cheaper, simplerCV studies could discredit the methodology itself. Moreover, the policydebates to which CV researchers are asked to contribute are often oftremendous importance to the well-being of households in developingcountries. Because the costs of policy mistakes can prove tragic, itis critical that VC researchers push for excellence in this researchenterprise and that funding agencies think more carefully about thevalue of policy-relevant information in the fields in which thecontingent valuation method is being used to study household preferencesand behavior (e.g., water and sanitation services, urban air pollution,soil erosion, deforestation, biodiversity, watershed management,ecosystem valuation, vaccines for the poor).