Low Fertility and Female Labor Supply in Japan—Time Series Analysis Using Bayesian VAR Approach

Abstract

Japanese society faces a low fertility rate and population decline, and the declining population brings about a shortage of labor forces. Therefore, female labor participation is attracting attention. However, cross-sectional data show a positive relationship between the total fertility rate (TFR) and female labor participation in recent years, while the time series data suggest a negative relationship. This may be a puzzle waiting to be solved. This chapter empirically interprets the dynamic relationship between the TFR, female wage, and labor participation from the time series perspective, using Japan’s experience. Generally, the VAR model is used to confirm the dynamic relationship among relevant variables; however, it is not easy to use the VAR model because the variables have a unit root. This study verifies the dynamic relationship using the error correction model (ECM) and Bayesian vector autoregression (BVAR) models. Therefore, both ECM and BVAR models show a similar result. They show that female wages negatively impact the fertility rate, and a similar result was obtained from both ECM and BVAR models. Women’s wages are an opportunity cost of having children, and we reconfirmed this fact in Japan’s experience. Additionally, this means that the compatible situation for women between labor provision and rearing children has been insufficient in Japan. Furthermore, adding the government benefits to young families to the BVAR model shows that this positively affects the TFR by calculating impulse response function. This can be interpreted as the government benefits reducing the opportunity cost of children. Hence, the intervention by the government to improve the TFR is appropriate.