A Multi-Attribute Extension of Discrete-Choice Contingent Valuation for Valuation of Angling Site Characteristics

(ProQuest-CSA LLC: ... denotes formulae omitted.)The Contingent Valuation Method (CVM) (e.g. Bateman, et al. 2002; Venkatachalam, 2004) is an established method for the valuation of recreational activities such as angling. A popular version of the method is the discrete choice CVM (Bishop & Heberlein, 1979) whereby the respondent accepts or rejects bids. Different approaches for modeling such data have been proposed by Hanemann (1984) and Cameron (1988). Also, Kristrom (1997) presented a "spike" model accounting for potential individuals with zero willingness to pay.The CVM has mainly been used for measuring values of the gross effects (use and/or non-use values) of environmental projects or damages (e.g. Carson, et al. 1992). However, it is a frequent interest to managers of wildlife areas to obtain information about angler's valuations of different angling site attributes. The CVM could be used for this purpose but several alternative stated preference methods are available. The use of choice experiments was proposed by Boxall, Adamowicz, Swait, Williams, and Louviere (1996) and Adamowicz, Boxall, Williams, and Louviere (1998). Another alternative is contingent ranking (e.g. Garrod & Willis, 1999). Also, Haefele and Loomis (2001) proposed a stated preference method based on ratings of alternatives. Among the alternatives to CVM, the choice experiment method has received most attention and is increasingly applied in research on environmental valuation.This paper focuses on a version of the CVM method in which the attributes in the scenario, in addition to the bid, is varied over respondents. CVM designs of this kind have earlier been employed (e.g. Boyle, Welsh, and Bishop, 1993). The name MACVM (Multiple Attribute Contingent Valuation Method) is here suggested and the method is proposed for valuation of angling site attributes. An alternative would be to consider the method as a special case of the choice experiment method. In a choice experiment the respondent considers a choice set with a number of hypothetical alternatives (usually two) and a base case alternative. The MACVM corresponds to a choice set containing only one hypothetical alternative apart from the base case scenario. However, while a choice experiment can for some purposes be designed without the base case alternative, the MACVM is built on the comparison between the base case and the hypothetical alternative. This is also the idea behind the CVM wherefore the interpretation of the method as an extension of the CVM is preferred.When choosing to accept or reject the suggested bid of a new angling site, the respondent needs to evaluate the new site in relation to the existing set of angling sites. The traditional approach to this problem is to collect data on the sites available and develop a model for site choice (e.g. McFadden, 1978). We suggest a new and simpler approach. A model of respondents' choice behavior is developed by treating the maximum utility over the set of available angling site as a random variable. In this way estimation of the utility function can be done without data on the existing set of alternatives. This is a considerable simplification allowing cost-effective analysis of preference.The MACVM is presented in the next section with a derivation of a statistical model within the random utility maximization framework. section 3 reports the results of an application of the MACVM and the CVM to the valuation of angling site attributes in the county of Jamtland in Sweden. A discussion of results and potential advantages of the MACVM is saved for the final section.MethodologyIn a discrete choice CVM, a hypothetical scenario is offered to the respondent at a cost (c) and the task of the respondent is to accept or reject this cost. By offering the scenario at different costs to different respondents, the population mean willingness-to-pay can be estimated using statistical analyses of responses (Hanemann, 1984; Kristrom, 1990). …