EVALUATION OF INFRASTRUCTURE STOCK EFFECT BY CAUSAL DISCOVERY

Abstract

In regional economic activities, the stock effect brought by road infrastructure may not always follow the fundamental economic theory since the technological innovation would surely occur after the investment. Therefore, finding on causal relationship between the actual regional characteristics is the big issue to understand how infrastructure stock effects appear. In the past studies, the causal structure of the variables was assumed on the basis of economic insights and validity of economic theory. In this study, the causal relationship between improvement of accessibility by road improvement and economic growth is found by the statistical characteristics of the data, called causal discovery. Causal discovery requires no assumption on causal structure, but purely investigate the independence among the variables. The proposed procedure in causal discovery is applied to confirm the validity.