A panel data analysis of the probability of childbirth in a Japanese sample: New evidence of the two-child norm.

Abstract

To reveal the conditions that could facilitate childbirth in modern humans, it is necessary to analyze not only cross-sectional surveys but also panel data that track the same person for a long period. In this study, we analyzed factors that would influence the probability of childbirth. We analyzed Japanese panel data with a Cox proportional hazard model. Subjects of our analysis were married women and their childbirth records from 2004 to 2009. Contrary to the predictions based on the theory of behavioral ecology, we found no positive relationship between good parental conditions for childcare, such as high income, increase in income, or coresidence with parents (i.e., grandparents of children), and the occurrence of childbirth. We found that the number of existing children had a significant impact on the probability of childbirth. The likelihood of further childbirth by couples with one child was nearly equal to that of childless ones. However, the corresponding likelihood of couples with two children was about five times lower than that of childless ones. The total fertility rates in modern developed societies are quite low and couples prefer having two children. This trend is known as the two-child norm, but it is a paradoxical phenomenon in terms of fitness maximization. Our result provided new quantitative evidence of this norm. This study revealed that the number of existing children being less than two was one of the factors associated with further childbearing in our Japanese sample.