Double Counting in Hedonic and Travel Cost Models
探讨了当环境改善的收益分别通过住房市场（享乐模型）和休闲需求（旅行成本模型）估算时，是否应将两者相加，以避免重复计算污染损害。
agency is responsible for reducing pollution in a lake. The pollution causes economic damage because the lake is less attractive for fishing and swimming. Evidence of the influence of pollution can be found through judicious sifting of the prices of houses near the lake and from reduced recreational use of the lake. The administrator of the environmental agency responsible for the lake wants to reduce the pollution in the lake but, cognizant of scarce resources, wants some evidence that the expensive investment in publicly owned treatment works will yield commensurate benefits. Consequently, she contracts with two economists for research on the benefits of environmental improvements. One will study the damages from the perspective of the housing market. The other will study the damages through the demand for recreation. When the environmental administrator receives the estimates of damages from the two economists, should she add them together? This is the basic question explored in this paper. Both travel cost models of recreation demand and hedonic models of housing can produce measures of pollution damages. When applied to the same polluted resources, do travel cost models and hedonic models measure the same damage? This analysis is similar in spirit but much simpler than that of Rosen (1979) and Roback (1982). The latter work examines an equilibrium when the wage and the land price are affected by amenity levels. The use of hedonic models to value access is not new. The spatial equilibrium model of a city can be viewed as a predecessor of the hedonic price model. In this model, formalized by Alonso (1964), the land rent gradient represents a tradeoff between commuting time and space. In the hedonic context, Nelson (1977) shows how the demand for commuting time and the price of housing interact to influence housing prices. Accessibility in some form, as a proxy for travel time, is a popular attribute of houses in hedonic models. The issue addressed in this paper has relevance for litigation and for policy. It arises in cases of CERCLA (Comprehensive Environmental Response, Compensation and Liability Act) concerned with the recovery of natural resource damages. Freeman (1987) describes a case where a hazardous substance released into New Bedford harbor apparently had an impact on the housing market near the harbor and on the recreational use of the harbor. This impact was studied with both a hedonic model and a travel cost model. Double counting is also a relevant issue when access to resources which are valued for their recreational returns are addressed in a hedonic framework.