Many impacts involve changes in the direct and/or indirect flows of valued services to society. These services can offer a wide range of valuable attributes, but they frequently go unpriced in the economic sense. Markets simply do not exist for some attributes and some services; contemplating markets for some others (e.g., health services) has been questioned even given extensive competiton for services and products. For others, markets that do exist fall short of being comprehensive or complete in the presence of externalities of production or consumption. In either case (and others), researchers have recognized the need to develop alternative means with which to assess value. More precisely, they have tried to extend the scope of the economic paradigm so that implicit and explicit tradeoffs between development and conservation of unpriced resources can be explored within the structures of standard decision analytic tools such as cost-benefit analysis, cost-effectiveness analysis, and so on. Parikh and Parikh (1997, 1998) provide a primer on valuation with case studies.
To be more specific, economists have built a theory of choice on the basis of the notions of consumer sovereignty and rationality. Economists assume, therefore, that individuals are able to value changes in nonmarket goods and services as easily as they can value changes in marketed goods and services. The only difference between the two cases is that markets provide the researcher with some indirect data with which to assess individuals' values of marketed products. Nevertheless, individuals should be able to tell researchers what they would be willing to pay for changes in nonmarket conditions or willing to accept as compensation for those changes. In fact, willingness to accept (WTA) payment for foregoing a good and willingness to pay (WTP) for a good are the two general yardsticks against which values are judged.
It should be noted that WTA and WTP are seldom the same for most nonmarket goods or services. In fact, WTA and WTP can give wildly different estimates of the value of these services if there are no perfect substitutes (i.e., if it is impossible to fully compensate individuals unit by unit for their loss). When such a substitute does not exist, WTA > WTP. By how much? Cummings et al. (1986) report that it is not uncommon for estimated WTA to be more than 10 times larger than estimated WTP. These differences might be derivative of the method of estimation, but they also reflect the fact that WTA and WTP are two different concepts that need not match.
It also should be noted that WTA and WTP have analogs in the market context. Compensated variation (CV) is the extra income that individuals would require to accept an increase in the price of some marketed good; CV is the analog of WTA. Equivalent variation (EV) is the income that individuals would be willing to forego to see the price of some marketed good fall; EV is the analog of WTP. These measures sometimes are used in market-based analysis. It should be no surprise that EV < CV unless the good in question has a perfect substitute.
Valuation methods usually are divided into two distinct approaches. Direct methods try to judge individuals' value for nonmarketed goods by asking them directly. Contingent valuation methods (CVMs), for example, ask people for their maximum WTP to effect a positive change in their environments or their minimum WTA to endure a negative change. Davis (1963) authored the first paper to report CVM results for environmental goods. Comprehensive accounts of these methods appear in Mitchell and Carson (1989), Hanley and Spash (1993), and Bateman and Willis (1995). This is a controversial method, and current environmental and resource literature continues to contain paper after paper confronting or uncovering problems of consistency, bias, truth-revelation, embedding, and the like. Hanley et al. (1997) offer a quick overview of these discussions and a thorough bibliography.
Indirect methods of valuation try to judge individuals' value for nonmarketed goods by observing their behavior in related markets. Hedonic pricing methods, for example, assume that a person buys goods for their various attributes. Thus, for example, a house has attributes such as floor area; number of bathrooms; the view it provides; access to schools, hospitals, entertainment, and jobs; and air quality . By estimating the demand for houses with different sets of attributes, we can estimate how much people value air quality. One can thus estimate "pseudo-demand curves" for nonmarketed goods such as air quality. Travel costs are another area in which valuation estimates of the multiple criteria on which utility depends can be finessed out of observable behavior. The hedonic method was first proposed by Lancaster (1966) and Rosen (1974). Tiwari and Parikh (1997) have estimated such a hedonic demand function for housing in Bombay. Mendelsohn et al. (2000) brought the hedonic approach to the fore in the global change impacts arena. Braden and Kolstad (1991) and Hanley and Spash (1993) offer thorough reviews of both approaches. Is there a scientific consensus on the state of the science for these methods? Not really. There is, instead, a growing literature that warns of caveats in their application and interpretation (e.g., health services) and/or improves their ability to cope with these caveats. Smith (2000) provides a careful overview of this literature and an assessment of progress over the past 25 years.
Other reports in this collection