Abstract
In this paper, we verified whether the Marshall-Dupuit measure (MD) used in practice can be a sufficient approximation to Equivalent Variation (EV) in the benefit evaluation of transport improvement. We used the transportation behavior model in the context of classical consumer behavior theory proposed by Morisugi, Ueda and Le. We estimated model parameters by inter-regional trip data in Japan and then measured user's benefit by each of MD and EV. We have proved that MD is fully effective as an approximation of EV.