Sex-Selective Abortion in India: Inference from matching trends in conditional sex ratios with trends in availability of ultrasound

Sonia Bhalotra, Tom Cochrane
University of Bristol, UK

The unprecedented rise in the sex ratio between the 1991 and 2001 census has prompted a new wave of research on son preference in India. We conduct a difference in difference analysis, using information on the timing of the introduction and diffusion of prenatal sex determination technology to identify structural breaks in the sex ratio of higher order births conditional upon the sex of the first child. The premise, supported by field-based research, is that families seldom attempt sex selection for first births. We allow heterogeneity in the conditional sex ratio by family level indicators of heterogeneity in son preference and access to ultrasound. Using non-parametric and parametric techniques, we find evidence consistent with sex selective abortion. Sex selective abortion appears to have started earlier in and remains more prevalent amongst families with greater wealth, more educated women, urban location, residence in the north-west of the country and high caste Hindu status. To back our inference that sex selective abortion is at play, we consider a range of alternative hypotheses, arguing that no one of these can, by itself, explain the patterns in the data. Our estimates suggest that the ideal composition of births in India is 2 boys and 1 girl.

View full paper