Lake Water Quality Valuation —Benefit Transfer Approach vs. Empirical Evidence Miko³aj Czajkowski, Agnieszka Markowska, Olimpia Markiewicz, Anna Bartczak, Warsaw University, Department of Economics, Warsaw Ecological Economics Center, Poland Milan Scasny, Jan Melichar, Hana Skopkova, Charles University Prague, Environment Center, Czech Republic 1. Introduction If decision-making follows economic optimality criterion, a public author- ity may rely on benefit-costs analysis in which involved costs and benefits are compared. Acquiring monetary values particularly for non-market goods and services by primary studies may however be time-consuming and costly. One of the possible solutions is Benefit Transfer (BT). This method has been in- creasingly popular in the literature, since it is much faster and cheaper (Champ et al. 2003). Benefit Transfer basically presents a technique of adopting values being derived by original study to a new (policy) site. The method has its origins in publications by U.S. Water Resources Council of daily estimates for recre- ation activities, which were used in evaluating water-related projects (Ro senberger and Loomis 2001). The method was first formally described by Freeman (1984). The first approaches explored value transfer techniques based on transferring original results from a ‘study’ site directly to a policy site with some adjustments where needed. Such adjustments of the existing values could be caused by inconsistency in time (CPI corrections), currency (PPP) or income (income elasticities). The basis for function transfer were laid down by Loomis (1992). The idea behind this approach is transferring an entire estimated demand (or WTP) function, which obviously depends on a study site context. This approach is believed to be a more reliable method since statistical analysis allows for preparing a function in a way that would fit the policy site characteristics best. Implicit assumption however also results in a drawback of this ap- proach; i.e. the same predictors and regression coefficients for original and policy site are expected. This assumption doesn’t have to be satisfied, espe- cially for regions differing in many characteristics (e.g. Loomis 1992, Loomis et al. 1995, Downing and Ozuna 1996, van den Berg et al. 2001). There are many empirical studies testing validity of benefit transfer be- tween countries (e.g. Krupnick et al. 1996, Ready et al. 2004 or Rozan 2004). 156 ekonomia 19