Diffusion process of fertility transition in Japan: regional analysis using spatial panel econometric model

Kenji Kamata, National Institute of Population and Social Security Research, Japan

This study aims to verify the diffusion process of fertility transition in Japan. I analyzed the Japanese regional fertility using spatial panel econometric models (Spatial Lag and Spatial Durbin model) to verify diffusion process between regions. The working hypothesis is which effects of adaptation or diffusion are influence the fertility decline, or the effects of both are detected? The sample is 47 prefectures, and the observed periods are from 1920 to 2010 (every 5 years, it is excluded in 1945). The dependent variable is the standardized marital fertility ratio (MFR). The standard population is 1930. Independent variables are (a) female mean age at first marriage, (b) infant mortality, and (c) proportion of primary industries. Covariate variables are population density (per km2), the sex ratio of total population, and the train time to nearest mega city. As a result of the analyses, for hypothesis, the effect of both the adaptation process and the diffusion process was confirmed. It hopes to continue model estimation and that separated the time in the future, such as the possibility of different covariates in each period, the study centered on deepening of the model.