The Causal Effect of the Shinkansen on Population Change: An Application of Propensity Score Method

Shinkansen, a high-speed rail in Japan, has been in use for approximately 60 years since it started service in 1964. The Shinkansen has promoted the movement of people and has had a significant impact on the demographic changes of the areas along the Shinkansen lines. This study aimed to analyze the population change around the Shinkansen station using tertiary grid data from the National Population Census. Specifically, we estimated the causal effects of Shinkansen stations on population changes in the surrounding areas between 1995 and 2015 using the propensity score method. In the analysis, we calculated the average treatment effect in the treatment group (ATT) using the weighting, stratification, and matching methods of the propensity score method. In particular, the propensity score matching method calculates ATT using subclassification matching, nearest neighbor matching, and caliper matching, and the results are compared and investigated with the results of the genetic matching method. Finally, the inverse probability weighting (IPW) method was used to calculate the average treatment effect (ATE), which was compared with the results obtained by multiple regression analysis. The findings show that the population grew by an average of approximately 300 people per block within a 15 km radius of the Shinkansen stations.