The Impact of News Shock of the Openings/Expansions of Large-scale Semiconductor Plants on Local Labor Market in Japan

This paper examines the impact of news shocks of the openings/expansions of large-scale semiconductor plants in Japan on the local labor market, especially on the jobs-to-applicants ratio, using monthly prefectural panel data in a causal inference framework based on the Difference-in-Differences (DiD) method. The results show that news shocks created by the openings/expansions of large-scale semiconductor plants significantly increase the local job-to-applicants ratio by about 0.05pt-0.08pt. In order to overcome the endogeneity problems arising from the site selection of the firms, I conducted a matching DiD analysis using other prefectures whose investment amounts were relatively small compared to the treatment group and where the news of the same openings/expansions of semiconductor plants was announced as a control group and obtained similar significant results. Furthermore, regarding TWFE (Two-Way Fixed Effects, or Staggered DiD) analysis, where the timing of treatments differs, it has been pointed out that bias can occur when there is heterogeneity in the treatment effects. Nevertheless, similar significant results were obtained in the analysis by CS-DiD, a DiD method that overcomes these bias problems arising from TWFE of staggered DiD. This analysis serves as an agile policy formation and evaluation tool in the government's EBPM projects and provides positive evidence for the project in progress.