Son Preference and Early Childhood Investments in China⇤

Where the fraction of male births is abnormally high, heterogeneity in son preference would suggest that parents of sons may have a stronger son preference than parents of daughters. Child sex may have become a stronger signal of parental sex preferences over time as the cost of sex selection has declined and sex ratios at birth have increased. In this paper, we build on Chen, Li and Meng’s 2013 analysis of ultrasound di↵usion across counties in China, which was found strongly predictive of increased sex ratios at birth. Here, we consider whether ultrasound di↵usion changed the pattern of early childhood investments in girls versus boys. If parental investments (like sex ratios) respond to parental sex preferences, postnatal investments in girls should increase with the di↵usion of ultrasound and increased prenatal sex selection. In contrast, the prediction for investments prior to birth is ambiguous. For pregnancies carried to term, ultrasound revealed sex as much as six months prior to delivery, enabling gender discrimination in in utero investments. In contrast, sex selective abortions would tend to increase in utero investments in girls through preference sorting. We evaluate these competing predictions using microdata on investments in children using the 1992 UNICEF Chinese Children Survey, conducted by the National Bureau of Statistics. We find that prenatal discrimination against girls seems to be confined to sex-selective abortion. There are some small positive compositional e↵ects on postnatal investment measures for girls. JEL Classification: J13, J16, O33 ⇤We wish to thank Mark Duggan, John Ham, Jeanne Lafortune, Soohyung Lee, Ren Mu and seminar participants at the 2010 AEA session “Circumstance and Selection in Early Childhood Health” and the University of Maryland for helpful comments. All errors are our own.