Sex-Selective Abortion in India: The Impact on Child Mortality B ...

The most significant predictor was a composite of birth order and the number of older living sons. The estimated sex ratio at birth was significantly higher among second-order births with no surviving older brothers (110) and among third-order births with no surviving older brothers (113); however, this was only the case for births reported in NFHS-2 (1984–1998). Also in NFHS-2, births to women who had completed middle school or higher were more likely to be male (112), as were births to mothers of religions other than Hindu or Muslim (115) and births to women with greater media exposure (112). Among births reported in NFHS-1 (1978–1992), urban residents had a significantly elevated sex ratio at birth (110), and Muslims had a significantly lower sex ratio at birth (102). These authors also found limited evidence of the selective abortion of boys in particular states among third-order births to women with two living sons: In NFHS-1, sex ratios at birth were 99 in Punjab, 73 in Delhi, and 93 in Maharashtra, and in NFHS-2, they were 90, 87, and 87, respectively.

A recent study by Jha et al. (2006), utilizing birth history data from the 1998 Special Fertility and Mortality Survey of 1.1 million households nationwide, found that among 133,738 births in 1997, the sex ratio for second births when the preceding child was a girl was 132, compared to 90.7 if the preceding child was a boy. Similarly, for third births, when the previous two children were girls the sex ratio was 139 and when they were boys—85.0. Mothers with a grade 10 or higher level of education had a sex ratio of 146 for the second birth when the first child was a girl; however, no differences were found by religion. By geographic location, the highest sex ratios were found in the North, in the state of Punjab, the city of Delhi, and the states of Bihar, Haryana, Gujarat, Rajasthan, and Uttar Pradesh.