非集計行動モデルを用いた工業立地分析

概要

論文の詳細を見る
Industrial location causes a wide range of changes in the surrounding region. Therefore it is necessary to forecast the future location of manufacturing industries in the regional planning. Industrial location has been widely analyzed by the location theory and the urban economics. But few operational or policy-oriented models, which can be applied in actual planning process, have been developed. One of the reasons is that most popular techniques in land use modelling, so-called aggregate type models, cannot describe well the changes in industrial location. The locational behavior of each industrial firm differs from that of the other. And the number of location is so small that it is not adequate to analyze it using zonal data.On the other hand, disaggregate behavioral models have sound bases of consumer theory and are expected to describe locational behaviors more explicitly with a much broader range of explanatory variables. Moreover, they do not require so many samples. In applying disaggregate behavioral models to location choice analyses, however, there are some problems which are caused mainly by a number of potential alternative sites.In this paper, after some discussions on the applicability of the disaggregate behavioral models to land use analyses, industrial location behavior is analyzed by using nested logit models. Each nested logit model used in the analysis has enough variables to explain its location behavior, and parameters are estimated with good results. In the stage of building the forecasting model, this study discusses the ways not only to introduce capacity constraint caused by lack of land supply, but also to estimate the numbers of different types of industrial locators which are indispensable in forecasting total demand for location in the area.The models are built in both metropolitan areas of Tokyo and Sapporo, Japan, and some comparison between them are also discussed.
日本地域学会の論文

非集計行動モデルを用いた工業立地分析

スポンサーリンク

概要

著者

関連論文

スポンサーリンク