Abstract This paper examines how different measures of son preference—stated son preference, sex ratios at birth, and implied son preference—capture varying degrees of fertility control and gender discrimination. Using data from India, I show that states with high sex ratios at birth and implied son preference (i.e., continuing childbearing until a son is born) engage in sex selection, while states with high stated son preference do not. However, female infants in high stated son preference states experience higher infant mortality rates, a pattern not observed in states with high sex ratios or implied son preference. States with high sex ratios and high implied preference also had excess female mortality before the widespread use of ultrasound in India, suggesting that the rise in sex selection has contributed to closing these gaps. To formalize this relationship, I define the unwanted girls ratio as the difference between desired and actual sex ratios at birth, which increases with higher stated son preference but decreases with sex selection. States with higher unwanted girls ratios have greater levels of excess female mortality. The paper highlights that linking specific measures of son preference to sex selection helps explain why some measures correlate with gender health gaps while others do not.
Priti Kalsi (Tue,) studied this question.