论文部分内容阅读
The spatial econometrics enable us to decompose the marginal effect of the concerning variables into direct and indirect effects. LeSage and Pace (2009) proposed the summary measures for them and showed that the Bayesian approach makes us implement the statistical inference for the measures. This study examines the productivity of roads in Japan from a viewpoint of the spatial econometrics using this technique and argues the direct and indirect effects from a Bayesian point of view. From the empirical results, we can find that the marginal effect is underestimated if the indirect effect is ignored.