The paper aims at evaluating the existent transport statistics system and searching for the new direction of it under the recognition with "The reform of it is an important policy subject and argent issue". Therefore it reviews some previous researches briefly and looks back at history of transport statistics after Meiji period. And next, it examines about the existent statistical data sources and the various statistics system. Stood on the above, policy and technical change, which occurred around the statistics at present, are considered and three proposals are arranged for.

In this paper, we explain three basic models of benefit-estimation for transport projects, which are the general model,the perfect-substitute model, and the logit model, and show their merits and demerits. The main contribution of this paper is to clarify the relationship among these three models and show that the general model includes the perfect-substitute model and the logit model as special cases. This result suggests that we can basically rely on the general model, regardless of the estimation method of transport demand. After deriving theoretical results, we use examples to illustrate how to estimate the benefits of transport projects practically.

The trucking industry in Japan was deregulated in 1990. This study develops a model of evaluating welfare effects by the deregulation. We could say the deregulation increased shippers' surplus through promoting the market competition, while the surplus gain was not so large comparing with whole economy since the freight transport market had already been competitive before the deregulation.