Although determining the sex of a fetus—and terminating the pregnancy based on the test results—is illegal in India, about half a million female fetuses in the country are aborted each year because of a cultural preference for sons, according to estimates from a national survey.¹ As a result, in 1997, only 899 girls were born in India for every 1,000 boys, and the sex ratio was even more skewed among families who had had only girls. Moreover, the proportion of births that were female was particularly low among educated women, who may be best able to afford ultrasound to identify the sex of their fetus.

The findings come from an analysis of data from the Special Fertility and Mortality Survey, a 1998 survey in which the Indian government asked a nationally representative sample of ever-married women about their fertility history. Respondents also provided demographic information. The survey did not ask about respondents' use of prenatal sex determination or whether they preferred children of a particular sex. The current analysis focused on 133,738 births that occurred in 1997, including 71,666 boys and 62,072 girls.

After adjustment for consistency with India's annual nationwide demographic survey, the birth data yielded a sex ratio (defined as the number of female births per 1,000 male births) of 899. For first-born children, the sex ratio was slightly lower (871). However, among families that already had one child, the sex ratio for the second birth differed dramatically, depending on whether the last child born had been a girl or a boy (759 vs. 1,102, respectively), a pattern that was apparent in all but one of the 17 Indian states included in the study. The bias toward male births was even greater when a family already had two female children: The sex ratio for births to families with two girls (719) was far lower than that for births to families with two boys (1,176) or with one boy and one girl (908). Overall, about 5,000 fewer girls were born to respondents in 1997 than would have been expected if the births had been evenly distributed between the sexes.

Of the various demographic variables examined in the study, maternal education proved to be a strong predictor of a newborn's sex. Among women who had previously given birth to a girl, the likelihood that the second child would also be a girl was significantly lower for women with at least a 10th grade education than for women who were illiterate (sex ratios, 683 and 869, respectively). Sex ratios for second births did not differ by the mother's religion, the family's land holdings or whether the family lived in an urban area.

According to the researchers, infanticide is unlikely to account for the skewed sex ratios. Infanticide usually occurs on the day of birth, and such cases would presumably be reported as stillbirths or neonatal deaths; however, respondents reported more stillbirths among boys than among girls, and the number of stillbirths (1,203) was substantially smaller than the number of missing females (although stillbirths may be underreported). Another 4,173 births resulted in early neonatal deaths, but these, too, were more prevalent among boys than girls, and the sex ratio for these deaths—unlike that for infants who survived—did not vary by maternal education.

The authors argue that "the most plausible explanation" for the preponderance of boys born in India "is prenatal sex determination followed by selective abortion," although hormonal, nutritional and other factors may also affect sex ratios. The low sex ratio observed in the study, they say, reflects a cultural preference for male children, who not only carry on the family name but generally earn more money than females and are better able to financially support their parents during the latter's final years (India does not have a social security program).² Sex-selective abortion may be particularly common among more educated mothers, the researchers suggest, because these women have more access to, and can better afford, prenatal ultrasound.

Data from other countries suggest that the natural sex ratio should be between 950 and 980; if that is the case, then the number of missing female births in 1997 for all of India would be between 590,000 and 740,000, of which the researchers estimate that 450,000 to 540,000 can be attributed to selective abortion. These figures suggest that selective abortion has resulted in about 10 million missing female births in India during the 20 years (1985– 2005) since ultrasound use became common. Given that some evidence suggests that India's sex ratio has fallen further since 1997, the authors argue that "the absolute numbers of missing females [are] likely to grow."—P. Doskoch

References

1. Jha P et al., Low male-to-female sex ratio of children born in India: national survey of 1.1 million households, Lancet, 2006, 367(9506):211–218.

2. Sheth SS, Missing female births in India, Lancet, 2006, 367(9506):185–186.