ベイズ型統計モデルによる日本経済の地域分析
野田 英雄, 姜 興起
pp. 19-28
DOI:
10.5687/iscie.23.19抄録
This paper aims to develop an alternative framework for empirical analysis of regional economic growth. We apply our proposed method to analyzing the Total Factor Productivity (TFP) changes in regional economies and structural characteristics among the regions in Japan. Conventional approaches to empirical analysis of regional economic growth are insufficient for an appropriate description of the complicated behavior of TFP. Accordingly, we assume that TFP is a time-varying parameter and estimate it using a Bayesian method that employs a smoothness priors approach. Compared with conventional approaches, this application of the Bayesian method facilitates a more detailed and sophisticated analysis of TFP. A major finding of the analysis is a significant change in the TFP trends of the Japanese economy, beginning around the time of the first oil shock. Furthermore, our analytical results suggest a meaningful direction for government policies to promote economic growth in Japan.