The present article reviews the effects of wartime and post‐war policies on regional agglomeration of landuse and transportation activities in Japan. For this purpose the authors have formulated a long‐term model over the period 1920–85, and conducted its final test. After confirming the accuracy of the entire model, it is used to evaluate the wartime policies which caused excessive employment agglomeration in Tokyo, and post‐war policies concerning the high speed Shinkansen rail network. We conclude that the choices of alternative policies would have led to a more balanced regional structure than there is now, with less agglomeration in both Tokyo and the Kanto region.

No abstract

Urban agglomerations in Tokyo and Osaka in post‐war period are largely attributable to the policies underlying the four National Development Plans implemented in that period. These policies, which were originally intended to achieve an effect on the opposite of agglomeration, relate primarily to the high‐speed transit network known as the Tokaido shinkasen and the expressway segments linking the Kanto and Kansai regions.For the purpose of capturing the historical scenario of regional agglomeration, we have formulated a long‐term macro‐econometric model from 1950–85, with 5‐year simulation cycles. This constitutes an employment distribution block, an input‐output block, and an inter‐prefectural activities block, incorporating all major macro‐variables.In an earlier stage of this research the model was calibrated and the total tests upon all formulations were completed. Ascertaining that the accuracy of the model is well within permissible limits, an evaluation of policy scenarios due to the delay in the introduction of the shinkansen on the Tokaido line, has been conducted. Results indicate that Tokyo grows much less rapidly than in reality, Osaka maintains the same rate of growth, and the other areas show a substantial growth by 1985.In the present version, we improve on the model with respect to its qualitative performance by relaxing the exogeneity of national employment and GNP to allow for the evaluation of these policy variables. In completing these final tests on the model, we are able to confirm the validity of the model by demonstrating its accuracy to be in a satisfactory range. In a case study, the policy areas that have contributed to the present distribution of development have been explored in detail. Comparative policy transformations, necessary to achieve the opposite effects to agglomeration, are discussed.