Testing the reliability and construct validity of a simple and inexpensive procedure to measure the use value of recreational fishing

There is currently much interest in measuring the economic benefits of recreational fisheries: (a) to defend the sector per se, (b) to estimate environmental damage and (c) to improve recreational fisheries management decisions (Arlinghaus, Mehner & Cowx 2002). Because of the inherent difficulty in transferring benefit estimates from one particular location to another, there is an urgent need to value as many recreational fisheries as possible (Arlinghaus et al. 2002). In this respect, the survey-based contingent valuation (CV) method has been recommended (e.g. Hudgins & Malvestuto 1996) and used to measure total economic value (TEV; i.e. the sum of use and non-use values) of recreational fishing or fisheries resources (e.g. Navrud 2001). However, there is an ongoing controversy about the reliability of the CV method (Carson 2000). The lack of consensus on standard CV procedures and the potential biases associated with it demand that every CV survey must be designed, implemented, interpreted and reported carefully (Mitchell & Carson 1989). The most pressing need for widespread application is, how to reduce the costs of CV surveys while still maintaining a high degree of reliability (Carson 2000). Therefore, extensive application of the CV method will only be possible in recreational fisheries, if the CV procedures are simple, inexpensive and reliable. This applies because public bodies charged with the management of recreational fisheries often lack personnel with sophisticated economic and econometric skills, and funding to finance large scale valuation projects (Arlinghaus et al. 2002). The most straightforward way of estimating both the benefits generated by angling in local, regional and national economies (economic impact), and also the benefits of the current resource use experienced by an individual angler [net economic value (NEV) or consumer surplus (CS)], is to first ascertain total expenditures (economic impact) and then ask respondents in an open-ended question to estimate the maximum amount over and above those expenditures that they would be willing to pay before they chose to stop angling (e.g. Pollock, Jones & Brown 1994). The latter is a relatively simple approach to estimate willingness-to-pay (WTP – an estimate of NEV or CS) and seems suitable in recreational fishing studies because, in contrast to users of public goods such as clean air or biodiversity, anglers (1) know what their recreational fishing experiences are about, i.e. the good to be valued is well defined and hypothetical, and information bias and scope effects should be minimal, (2) are used to paying for part of the quasi public good angling and thus there is no need for a rather abstract payment vehicle and (3) the series of questions on expenditure make anglers recall and think about how much more money he or she would be willing to pay, which should enhance the accuracy of the elicited maximum WTP values (Hudgins & Malvestuto 1996). Combining open-ended question formats with off-site angler contact methods such as mail or telephone surveys results in a simple and inexpensive CV study. However, environmental economists have sometimes criticized open-ended question formats and non-personal contact methods as being unreliable (e.g. Mitchell & Carson 1989; Arrow, Solow, Leamer, Portney, Radner & Schuman 1993). Irrespectively, the higher costs of in-person contact methods and the sophisticated econometric skills needed to analyse other question formats such as binary discrete choice formats (e.g. Mitchell & Carson 1989) limit their applicability to recreational fisheries. Nevertheless, every published CV study should contain a reliability