A Hicksian Approach to Cost-Benefit Analysis with Discrete-Choice Models

Abstract

Although Hicksian consumer's surplus are considered the theoretically correct underpinnings of applied welfare economics, the Hicksian approach has not been fully integrated into the discrete-choice literature. This paper presents a new specification of utility functions for discrete-choice models that satisfy both the homogeneity property of demand and Roy's Identity. By solving the path dependency problem inherent in Marshallian measures, this specification permits the derivation of Hicksian measures. The paper presents a novel technique of identifying the entire indirect utility function for transport analysis without data on the complete budget set and with data only on mode choice. The cost-benefit framework advanced here has been used to derive exact theoretical of the welfare impact of changes in prices and quality at both an individual and aggregate level using numerical methods (Hau, 1981, 1983, 1984). In these works, the theoretical framework was applied to a demand-supply corridor simulation model of the San Francisco Bay area using micro-data, thus showing the theoretical model to be empirically viable. Note that the traditional definition and estimate of the valuation of time emerged as a byproduct, further supporting the empirical validity of our theoretical framework. The first attempts to derive an index of user benefit within the travel demand context were those of Neuberger (1971) and Harris and Tanner (1974). These authors defined consumer's surplus as the area to the left of the probabilistic demand functions. Satisfaction of the Hotelling integrability conditions for the equality of cross-partials guarantees path independence, but only the additional condition that the Hessian matrix be negative semidefinite ensures that the demand functions are derived from utility maximization. Several authors derived measures of that are consistent with random utility maximization and established some of their properties (Williams, 1977; Ben-Akiva and Lerman, 1979; and Daly and Zachary, 1978). Domencich and McFadden (1975), Bruzelius (1979), McFadden (1981) and Small and Rosen (1981) demonstrated the connection between neoclassical consumer theory and an accessibility measure. All of these authors employed the constancy of the marginal utility of income or money (numeraire) to justify the use of Marshallian consumer's surplus measures, known to be plagued by the path dependency problem. The main difference between Small and Rosen (SR) and my approach here can be seen in the respective implicit compensation objectives involved.' SR derive an expression for the amount required to compensate an individual for a change in price or quality, given that the individual chooses to consume the