The impact of news shock of the openings or expansions of large-scale semiconductor plants on local labour market in Japan

Abstract

This study investigates the causal effect of news shock of openings or expansions of large-scale semiconductor plants in Japan on local labor market, especially measured by new job opening ratio, using DiD (Difference-in-Difference) applied to prefecture-level data. The results shows news shock of large-scale openings or expansion of semiconductor plants increase new job opening ratio by 0.05 - 0.08pt significantly. To address the endogeneity problems arising from firms’ site selection, I matched control group consist of prefectures where news about the openings or expansion of large semiconductor plants were announced but the amounts of investment were relatively small compared to the treatment group and run DiD analysis, and finally obtained the similar robust results. Furthermore, it has been recently pointed out in econometrics that the heterogeneity of treatment effects in a TWFE (Two Way Fixed Effect or Staggered DiD) analysis in which the timings of treatments differ, can lead to biased estimator. The similar robust results were also obtained in the analysis using CS (Callaway and Sant’Anna) DiD, a type of DiD method that overcomes this bias problem. This paper is in the literature which analyzes the impact of large-scale plant openings (million dollar plant openings) on the local labor market, and contributes to the accumulation of knowledge in this field, which is surprisingly scarce in Japan.