Son Preference, Gender Composition, and Parental Time Allocation: Evidence from Rural China

Yi Fan 2, Junjian Yi1, Junsen Zhang1
1The Chinese University of Hong Kong, Hong Kong, 2The London School of Economics and Political Science, UK

This paper empirically estimates the effect of child gender and gender composition on parental joint time allocation between the labor market and the household using micro data from rural China. To control for the endogeneity of child gender, we exploit a new instrumental variable (IV) that is the number of paternal brothers. We find a significant time reduction in maternal involvement in household chores with a presence of sons. Mothers with sons but no daughters have a significant time decrease in chores by 48.8%, compared to those with daughters but no sons. Having at least one son and a first-born son decrease maternal time spent on household work by 65.9% and 41.0%, respectively. This pattern is consistent with the story that giving birth to sons increases mothers' intrahousehold bargaining power. We compare the IV estimates with the ordinary least squares (OLS) estimates, and suggest that the upward bias in OLS estimates may result from the unobserved heterogeneity in son preference across households. Furthermore, we employ a different empirical strategy under fixed effects (FE) estimation, and find a consistent pattern as the one under IV estimation.

View full paper