Validation of Empirical Measures of Welfare Change: A Comparison of Nonmarket Techniques

Since the passing of the Flood Control Act of 1936, the attention of many resource economists has been directed towards measuring the benefits and costs of alternative uses of natural resources. Increasingly, it has been recognized that the value of some of these uses is not explicitly determined through market transactions. This has led to the focus of much attention upon the development of nonmarket methods to estimate the value of these uses. In the 1940s the groundwork for the travel cost method was established by Hotelling (Prewitt 1949). Following this, in the 1960s Davis initiated the basic foundation for bidding methods (Davis 1963), which were later to be subsumed under the heading of contingent valuation methods (Brookshire, Randall, and Stoll 1980; Schulze, d'Arge, and Brookshire 1981; Thayer 1981). Both of these methods will be further discussed below. In addition to these two methods, a third approach, the unit day value method, has been recognized as acceptable but not preferred for the valuation of recreational uses of water and related land resources (Dwyer, Kelley, and Bowes 1977; U.S. Water Resources Council 1979). Given the existence of alternative methods of estimating the value of resource uses, several obvious questions arise. One of the first of these is, under what circumstances is each method most appropriate? Existing literature has focussed sufficiently on this question and pointed out limitations of each, especially the travel cost method and unit day value methods (Brookshire, Ives, and Schulze 1976; Dwyer, Kelly, and Bowes 1977; Freeman 1979; McConnell 1975; Randall, Ives, and Eastman 1977; Schulze, d'Arge, and Brook-