This paper addresses the lack of empirical evaluation of the use of web based surveys in the context of contingent valuation surveys. We compare, using a case study, in-person interviews with web based surveys regarding response rate, information additivity effects and respondents' attitudes towards paying. The web based survey had a much lower response (5.1%) than the in-person interviewing (84%). We find the web based contingent valuation surveys to be neither more susceptible to information additivity effects nor more prone to zero protest responses. We conclude in favor of the use of web based surveys,
IntroductionThe contingent valuation method (CVM) is a stated preference method 1 for valuing changes in the provision of nonmarketable goods (e.g. environmental amenities). The method was conceptually proposed by CiriacyWantrup in 1947, but the first empirical application only took place in 1961 with Robert Davis, in the US (Jackobsson and Dragun, 1996; Venkatachalam, 2004). The contingent valuation method rose to high prominence in the 1980s and in the last two decades has been routinely used in the appraisal of projects with impact on the environment. Its potential to capture both use and non-use values of environmental assets and the flexibility of its application largely contributed to its massive use (Carson, 2000).Basically, CVM is a survey-based methodology which relies on obtaining monetary estimates for the economic value of a specified change in the provision of the environmental good of interest, which typically is not traded in the conventional markets. Hence CVM application consists in the construction of hypothetical markets (also called elicitation scenarios) where the individuals' valuation of the good is elicited. Monetary estimates are obtained either as individuals' Willingness to Pay (WTP) or Willingness to Accept (WTA). WTP measures the amount of income a person is willing to forego in exchange for an improved state of an environmental good. E-mail address: cristina.marta@ist.utl.pt (C. Marta-Pedroso). 1 Stated preference, or direct, methods employ constructed markets in an experimental setting to elicit individuals' preferences. Other stated preference valuation approaches include contingent ranking, contingent choice and conjoint analysis (see Boxall et al., 1996, for a comprehensive comparison of the various stated preference approaches).