Foeticide: More ‘Missing’ Girls Among Hindus Than Muslims in Last Two Decades, Official Data Shows

New Delhi: Hindus have the highest number of missing girls attributable to female foeticide in India, a new research report prepared by the Pew Research Centre has revealed.

The researchers got their data from the last three rounds of the National Family Health Survey (NFHS), including the fifth and latest one (2019-2020). The NFHS is conducted by the Union government.

The Pew Research Centre is a Washington-based nonprofit that undertakes research on, among other topics, demographics. The team that put together the current report comprised five people: primary researcher Yungping Tong; associate director and demographer Conrad Hackett; researchers Stephanie Kramer and Anne Fengyan Shi; and director of religion research Alan Cooperman.

According to their analysis, at least 9 million girls are ‘missing’ in India as a result of female infanticide from 2000 to 2019. To compare, this is slightly lower than the entire population of Uttarakhand.

India banned prenatal sex-determination testing in the Pre-Conception and Pre-Natal Diagnostic Techniques Act 1994. The clandestine use of ultrasound facilities for this purpose continues, however, as do sex-selective abortions.

“As India’s largest religious group, Hindus make up 79.8% of India’s total population and account for a disproportionate share, 86.7%, of the missing female births,” the report reads.

These girls are said to be ‘missing’ because they should have been born and become part of the population – but didn’t because the foetuses were aborted. According to the centre’s report, Hindus alone ‘lost’ some 7.8 million girls.

The Sikhs follow the Hindus: they comprise only 1.7% of the country’s population but contributed 4.9% of ‘missing’ female births (around 4.4 lakh girls).

Muslims account for 14% of the population and were responsible for 6.6% (5.9 lakh) ‘missing’ female births. Finally, Christians make up 2.3% of the population and accounted for 0.6% (50,000) ‘missing’ births.

A simple equation

When girls go ‘missing’ in this manner, the male-female ratio in the general population becomes skewed in favour of the males.

The NFHS data doesn’t provide a direct estimate of the number of girls missing as a result of sex-selective abortions. The researchers determined these numbers by comparing the numbers of observed and expected female births.

In addition, they figured the expected number of female births by applying India’s sex ratio in the given period – 105.3 boys per 100 girls. Put differently, if X boys are born, then X multiplied by 0.95 girls will have been born in the same period.

Information on the actual number of births over the three decades came from the Sample Registration System (SRS), which is also undertaken by the Union government.

The difference between the number of female births that should have happened and the actual number of births is equal to the number of ‘missing’ births.

Ultrasound of a growing foetus in the mother’s womb at 23 weeks old. (Photo by: Flickr/hose902 CC BY-NC-ND 2.0)


The concept of ‘missing’ girls flows from a skewed sex ratio at birth, which has slightly stabilised over the last few years in India. (In this article, ‘sex ratio’ refers to ‘sex ratio at birth’ per se.) However, the sex ratio has been the topic of public discussions more often than the reality of ‘missing’ girls.

As it happens, a 2010 discussion paper prepared by researchers at the Institute for Labour Economics (IZA), a German nonprofit research institute, offers a useful comparison. The paper’s authors examined such ‘missing’ girls in India from 1995 to 2005 – which is roughly the decade before the ones to which the Pew group paid attention.

The paper stated that “as many as 0.48 million girls per annum were selectively aborted during 1995-2005”. Further:

“For a given sex history of births, substantially more sex selection was conducted post-ultrasound by families with wealth (top 20%) and relatively educated women (attaining at least secondary education) and, conditional on wealth and education, by Hindus as compared with Muslims.”

The paper called the practice of sex-selective abortions among educated women “striking” and the concentration of girls in more prosperous households a “challenge” to “the popular notion that the exercise of son preference is a marker of economic backwardness and ignorance”.

Killing female foetuses also resulted in what the Pew Research Centre report’s authors call a “marriage squeeze”, especially among Sikhs: a shortage of marriageable women. The ‘squeeze’ is also fuelled by the fact that few Indian men marry outside their religion.

It can eventually be responsible, in part, for increases in sex-related violence and crimes and trafficking of women.

The usual suspects

The report’s authors write that there is a correlation between three parameters and an individual’s decision to conduct a sex-selective abortion (but not necessarily causation). These are caste, wealth and education.


As an example, according to the Pew Research Centre report, among ‘general category’ Sikh women, the sex ratio (121 boys per 100 girls) is more imbalanced than it is among ‘Scheduled Caste’ Sikhs (102). The authors attribute this to land ownership and therefore wealth.

Specifically, taking a cue from NFHS data, the authors write that upper-caste Sikh households are more unlikely to be underprivileged than ‘Scheduled Caste’ Sikh households to own a piece of land (59% vs. 8%). Therefore:

“… land-owning caste groups may be more motivated than others to avoid having daughters, especially when their regional norms exclude daughters from inheriting family property.”


Next, the report’s authors didn’t find a significant sex-ratio gap between general- and reserved-category Hindus and Muslims. But when they combined the use of ultrasound facilities with caste patterns, the picture changed.

The authors found that the sex-ratio tended towards more males than females among the so-called ‘upper-caste’ women compared to women of ‘Scheduled Caste’ groups.

“For instance,” they write, “the birth ratio among general category women was 9 points wider than among women of ‘Scheduled Castes’ in the 2005-2006 NFHS, when the rate of ultrasound use among upper-caste women was twice as high.”

Access to ultrasound facilities and a skewed sex ratio were in combination taken to be suggestive of sex-selective abortions.


The Pew Research Centre authors inferred from the NFHS data that women who were wealthier and more educated were less likely to favour sons over daughters. Families that lived in cities were also less likely than their rural counterparts, according to the data, to favour having sons, likely because they were more educated.

However, the authors also warned that education, wealth and urbanicity could together improve a woman’s or a family’s access to an amenable ultrasound facility or other forms of prenatal sex-screening.

They use the example of South Korea, where studies have found that the most educated groups were the first to report a widening gap in the sex ratio – in the 1980s – before the gap spread to other demographic groups. Similarly, later, the country’s most educated groups were the first to report a narrowing of the gap towards the natural 1:1 ratio.

While the study periods of the IZA discussion paper and the Pew Research Centre report are a decade apart, and both NFHS and SRS data indicate a declining trend in the preference for sons, both analyses indicate that sex-selection abortions are continuing in the country. They might be becoming less common as well, but we don’t yet know if that is in proportion to the rate of growth of India’s economy and literacy.

All we know for now is that in the 20 years until 2019, India lost almost ten million girls.