Are sex-selective abortions skewing population: Myth Versus Truth
Background on missing women
In March 1992, economist Professor Amartya Sen published a paper titled the “Missing women” (Sen, Missing women, 1992) in which he confidently stated that social factors “must…explain the low female:male ratios in Asian and north African countries. These countries would have millions more women if they showed the female: male ratios of Europe and the United States.” However, realizing that comparison with the United States and Europe may not be appropriate, Sen uses the sub-Saharan ratio and calculated that India had 37 million missing women.
Eleven years later, Sen, in another paper titled “Missing women – revisited”  went on to speculate whether the rise of ‘Hindu right wing parties’ had a link with hidden cultural values and the emergence of female feticide. (see box: 1). That such absurd guesswork was even published by the British Medical Journal reflects how certain papers are published without any editorial intervention, the author’s reputation eliminating the need for an important inquiry on the absence of substance in the claims being made. Ironically, in 1999, Amartya Sen had been awarded India’s highest honor, the Bharat-Ratna, by the Bharatiya Janata Party.
Since colonial times, much of the European and American media have peddled the narrative of Indian culture being evil and discriminatory enough to the extent of indulging in extreme violence. After the horrible Nirbhaya rape episode, the western media went to town sermonizing about India’s “rape culture”. In a general narrative where India is painted as being an evil society, the charge of Indians also practicing massive number of abortions fits into the western storyline about India being a nation that needs to be civilized.
Under these circumstances, exploring the sources and validity of such accusations becomes important.
Sources for estimating the “Gender gap”
When it comes to the issue of sex-selection and subsequent abortion of female fetuses, the allegations of Amartya Sen, a 2006 paper published in Lancet titled “Low male-to-female sex ratio of children born in India” (Lancet-2006) and another 2011 paper in Lancet titled “Trends in selective abortions of girls in India“ (Lancet-2011)  are important sources for propounding that Indians (and by implication “Hindu culture” ) have a massive socially evil behavior and indulge in massive elimination of female fetuses by using the medium of pre-natal sex-selection.
The Lancet-2006 states: “Although further research is needed, in our opinion, the most plausible explanation for the low female-to-male sex ratios reported at birth is prenatal sex determination followed by selective abortion. Other explanations, including infections, smoking, maternal nutrition, and hormonal factors during pregnancy, might play a part in reducing the overall sex ratios, but they are unlikely to explain the discrepancies noted for second-order and higher-order births. The results of a US study of 6000 children born indicate that sex of subsequent births is independent of sex of earlier births.
….Results of studies done in European, North American, Asian, and African countries suggest a natural sex ratio (that is, in the absence of selective medical or social pressures for fewer females) of about 950–980.” (The article had referred to “Missing women: revisiting the debate” by Stephen Klasen and Claudia Wink). The paper authors then go on to estimate that in 1997out of a total of 28 million births, 13.6–13.8 million girls should have been born, but that the actual number was 13·1 million—a deficit of 0·59–0·74 million female births.”
In short, the Lancet-2006 paper assumes a) that the sex-ratio of Indians is 1:05 and b) the sex of subsequent births is independent of previous birth. Then, based on a survey that 13% of pregnant women have access to pre-natal sex-determination, they assume that a third of the female fetuses are aborted. The Lancet-2006 paper goes on to state: “If we assume that selective abortion explains half the missing girls born first or girls born at third or higher birth order, that is a number of 0·14–0·20 million; a total of 0·45–0·54 million selective abortions yearly.”
Referring to a survey conducted by the International Institute for Population Science, the Lancet-2006 paper continues, “A much smaller national survey, done at the same time as the SFMS, noted that 13% of women (8% of those living in rural areas and 31% of those living in urban areas) self-reported prenatal sex determination during their last pregnancy. Thus, at least 3·6 million (28 million x 0.13) women have access to prenatal sex determination, of which about half—1·8 million— would be expected to be carrying female fetuses. If only one-third of these fetuses are aborted, then our calculations are not overestimates.” (emphasis added)
The Lancet-2006 paper makes some assumptions about Indian sex-ratios – such as, Indian sex-ratios must be the same as that of western and developed nations which are racially and nutritionally different and may also have different age for parenthood. The Lancet-2006 paper also assumes that sex of a baby is not dependent on the previous birth when there is literature and surveys that suggest otherwise. William James, in his article “The Human Sex Ratio,” writes, “Two large sets of data have been published on this point. Data on the sex ratios of four million French births by the numbers of prior brothers and sisters were given by Malinvaud (1955) and reprinted in James (1975a). Similar data in respect of 150,000 U.S. births were given by Ben-Porath and Welch (1976)…….. In “both sets of data the sex ratio rose significantly and substantially with the number of prior boys; and it fell significantly and substantially with the number of prior girls”.
In the subsequent Lancet-2011 paper, the authors go on to substantiate their findings as follows: “… Declines were much greater in mothers with 10 or more years of education than in mothers with no education, and in wealthier households compared with poorer households. By contrast, we did not detect any significant declines in the sex ratio for second-order births if the firstborn was a boy, or for firstborns..…After adjusting for excess mortality rates in girls, our estimates of number of selective abortions of girls rose from 0–2·0 million in the 1980s, to 1·2–4·1 million in the 1990s, and to 3.1–6.0 million in the 2000s. Each 1% decline in child sex ratio at ages 0–6 years implied 1.2–3.6 million more selective abortions of girls. Selective abortions of girls totalled about 4·2–12·1 million from 1980–2010, with a greater rate of increase in the 1990s than in the 2000s”.
After such a damning indictment of Indian societal situation, one would have to really be concerned about Indian society. However, as we shall show, the reality may be somewhat different than that is painted by the authors of the Lancet papers and the likes of Amartya Sen.
Issues that affect gender gap
Understanding the factors that influence human sex ratio is a complex issue and is still not very well understood. From various studies, it appears that sex ratio is dependent on multiple factors such as race, age of parents, nutrition level of parents, order of birth, time of conception, hormonal influences etc. In the United States, the sex ratio for blacks is estimated to be 1.03, whites have a sex ratio of 1.054, Asian Indians have 1.066 and Chinese have 1.074 (see Table 1). In addition to these racial differences, it is known that there are differences in sex ratios even within a relatively racially similar Europe itself, and not much is known as to why these differences exist.
The Lancet-2011 paper makes an important assumption that Indian sex ratio is between 1.05 and 1.02. The authors utilize that number based on the “ranges reported in most high-income countries where social pressures for fewer girls do not exist.” Amartya Sen, on the other hand, relies on the work of Ansley Johnson Coale to compute a gigantic “Missing women” number. Such guesswork and estimates may be fundamentally flawed as discussed by the author Sivan Anderson and Debraj Ray in their work. Anderson and Ray observe, “Coale (1991) used a reference sex ratio at birth of 1.059 for all groups. This is problematic: there is substantial variation in the sex ratio at birth across race and ethnicity. The average sex ratio at birth for developed countries is in the range of 1.05 to 1.07, with a median equal to 1.059, but this range is non-trivial. Even within Europe, the average in Northern Europe is around 1.05, whereas for the Mediteranean it is in the range of 1.06 to 1.07.”
This in effect, casts doubt on whether the standard gender ratio of 1.05 for India, assumed by the authors of the Lancet papers is valid.
Gender Ratio at birth among different races
Table 1 shows the sex ratio for various races in the United States and the Asian Indian sex-ratio is shown to be 1.066, materially different enough from the 1.05 ratio assumed by authors of the Lancet papers to come to their conclusions. However, even this number may not be applicable for Indians in India because there is a difference between Asian-Indians in USA and Indians in the age of marriage as well as nutritional factors. Low nutrition as well as low parental age has been shown to increase male births .
Anderson and Ray further state, “Biological determinants of the sex ratio at birth include the timing of conception and hormonal variations (James, 1987). However, these factors have proved difficult to measure and most research has relied on variables which are more easily observable at a large scale such as parental age and birth order. In general, the proportion of male births increases with the number of prior births and shorter birth intervals and it decreases with parental age and the proportion of multiple births”.
In India, the age of marriage and age of conception is considerably less than in western societies, the nutritional level of both the parents is very low and there is difference in birth intervals as well. The implication then would be that the sex ratio for Indians in India could easily touch 1.07 or more, which would mean that the ratios deduced by Sen used as well as the estimates about the supposed gender-gap in the Lancet studies are significantly wrong.
It is also very important to note that, in India, the supposed “Missing Women” are evenly distributed and the pre-natal factors account only for 11%. To quote Siwan Anderson, “First, although the overall sex ratios in India and China are similar—both around 1.06—the two countries have distinct age profiles of missing women. A large percentage of the missing women in China are located before birth and in infancy. We estimate that around 37–45% of the missing women in China are due to prenatal factors alone. But the numbers for India are more evenly distributed across the different age groups. Prenatal factors account for around 11%, and if we add up all the female deficit up to age 15, we do not get to a third of the total”. (emphasis added)
It is noteworthy that Anderson and Ray, in their paper, find it necessary to state that Sen’s “terrible story of inequality and neglect” is possibly true in large part, but other stories may need to be told as well.” (emphasis added) It is the ‘other stories’ that we want to mention in this article.
Assumptions and statistics
The Lancet-2011 paper at the end of the paper, goes on to accept “our annual estimates of selective abortions of girls relying on the census are, by necessity, crude.“ (pg 1927) and they also accept that the sex-ratio ratio that they have assumed is “based on findings on findings in Europe and North America, and might not apply to Asian populations for unknown biological reasons” (emphasis added). The authors also mention the caveat “unmeasured biological factors, such as infections, might reduce or increase overall sex ratios at birth” and another issue they state is that “unpublished data on birth histories in Indian diasporas (data not shown), a small and presently inexplicable excess of third girls after the birth of two earlier boys.” (emphasis added). In short, all the caveats stated in the Lancet-2011 paper, makes the claims in the whole paper useless and unreliable.
The Lancet-2011 paper also makes another point, “With respect to second-order births after a female birth, the more years of education the mother had the less likely they were to give birth to a girl; mothers with grade 10 or higher education had a significantly lower adjusted sex ratio “. The paper does not even consider that the more years of education of the parent may also imply different nutritional status, different age of parents, birth-order etc, all of which could influence the sex-ratio. Merely assuming that the parents practice sex-selective abortion would be rash.
In short, the Lancet-2011 paper has numerous huge statistical holes that raise grave doubts on the popular narrative that selective-abortion of the girl child is rampant in India. What is even more puzzling is that these kinds of tendentious conclusions get past the review process of Lancet. One must wonder if an assertion that the English were aborting their children in massive numbers could get past the reviewers of the Lancet, as easily compared to the assertion that Indian parents are prone to sex-selective abortions.
The situation in scientific publications seems to be so bad that Richard Horton the editor in chief of Lancet has written a scathing note  stating: “A lot of what is published is incorrect….The case against science is straightforward: much of the scientific literature, perhaps half, may simply be untrue. Afflicted by studies with small sample sizes, tiny effects, invalid exploratory analyses, and flagrant conflicts of interest, together with an obsession for pursuing fashionable trends of dubious importance, science has taken a turn towards darkness. As one participant put it, “poor methods get results” “. The incessant drive for publishing any kind of research may also be a reason for these kinds of papers. In short, the Lancet-2011 paper has numerous huge statistical issues that raise grave doubts about the popular narrative that selective abortion of the girl child is rampant in India. What is even more puzzling is that these kinds of tendentious conclusions get past the review process of Lancet. One wonders if an assertion that the British indulge in massive sex-selection abortions could get past the reviewers of the Lancet, as easily when compared to the assertion that Indian parents are prone to sex-selective abortions
|“A lot of what is published is incorrect …. The case against science is straightforward: much of the scientific literature, perhaps half, may simply be untrue. Afflicted by studies with small sample sizes, tiny effects, invalid exploratory analyses, and flagrant conflicts of interest, together with an obsession for pursuing fashionable trends of dubious importance, science has taken a turn towards darkness. As one participant put it, “poor methods get results” – Richard Horton, Editor-in-Chief of Lancet|
Developing Narrative of an Abusive Indian culture
The accusation that Indians practice rampant sex-selective abortions has also gained currency because it is of a generic nature, accusing Indian ‘culture’ rather than an individual. So, the authors can get their papers published and get their glory, while not having to be scrutinized. Painting an eastern-culture as bad is easier for publications like the Lancet as it fits the popular western narrative that Asians in general and Indians in particular have to be civilized.
It is very evident that the Lancet-2011 paper has numerous huge statistical issues that raise grave doubts on the popular narrative that selective abortion of the girl child is rampant in India. One must wonder if an assertion that the English indulge in massive sex-selection abortions could get past the reviewers of the Lancet, as easily when compared to the assertion that Indian parents are prone to indulge in sex-selective abortions
Many educated Indians (as Amartya Sen and the Lancet paper authors) are very willing to promote demonization of India, and attribute the worst possible motives Interestingly, even the authors of the Lancet papers, while admitting that they cannot estimate the degree to which prenatal sex determination affects sex ratio, are very sure that they cannot exclude selective abortion. When extreme claims are made about India in western sources, the Indian media also mindlessly propagates it as the gospel truth.
Could Indias sex-ratio improve ?
To quote from Anderson and Ray, “The implications of the data on improving healthcare could be even more interesting. There is enough literature documenting that the sex ratio in utero is considerably higher than at birth, and that the sex ratio of miscarriages, spontaneous abortions, and stillbirths is much larger than the sex ratio at birth. As a result, improving health conditions that reduce the incidence of miscarriages and stillbirths will increase the sex ratio at birth, which is consistent with the secular trends in the sex ratio at birth.” (emphasis added). So, improved healthcare and nutrition may in effect increase the sex-ratio gap. However, this gap cannot continually increase and there has to a level at which the sex-ratio begins to stabilize.
The United Nations has published a report by Mary E. John titled “Sex Ratios and Gender Biased Sex Selection: History, Debates & Future Directions”. This document catalogs several views, opinions and is a good compendium of the sociological demonization of Indian society that is peddled in the highly Eurocentric academic circles.
The report points out that in the 2011 census “it would appear that there has been a peaking (or plateauing) of the practice of sex selection in states like Gujarat, Haryana, Delhi and Himachal Pradesh, with actual small improvements from very low levels in Punjab and Chandigarh.” However, the report goes on to express concern that “CSRs are falling in large parts of western, central and eastern India…. In other words, the state wise figures demonstrate a widening of the circle, well beyond the so-called prosperity belt of north-west India, to the poorer states.”.
As has been previously quoted from the paper by Anderson & Ray, improving healthcare may initially, negatively impact sex-ratio and later the sex-ratio at birth tends to plateau out. This is exactly what is happening in India. Considering the very bad state that India was in at the time of independence, it has taken a few decades for India to start lifting the hundreds of millions out of poverty. The highly biased Eurocentric atmosphere that exists in sociological departments prevents academics from seeing the larger picture of Indian society that is generally positive and admirable even as it may have problems.
The overall improvement and stabilization in sex ratio will happen only when the marriage age and nutrition levels of the parents increase. Only a growing economy, with increased education and prosperity, which would increase the age for marriage (and thus the parents’ age at the time of conception of a baby) and improve the nutrition levels (of the parents at conception), may potentially bring about the change that we wish to see.
Effects of bad science and biased media reports
The New York Times has quoted a survey to claim that India is the worst country for women amongst the G20, because of a “combination of infanticide, child marriage and slavery”. This kind of scare reports could have effects on the tourism to India and affect the earnings of those dependent on the sector. After the horrific Nirbhaya episode in Delhi and ensuing media frenzy, a biased German lady professor even rejected considering an Indian male student for an internship merely because he was an Indian male and thus had to be misogynistic. It is because the media drives the overall narrative of India as an evil culture that an academic in Germany can come to such extremely biased and absurd conclusions. Even as a biased and negative narrative on India, may cause monetary and job losses in terms of reduced tourists, an even greater harm is done by pushing the message that Indian culture is inherently bad and the only way to improve is by Indians mimicking western societies.
The demonization of Indian society also has effects like diverting scarce government resources into propagating messages like “beti bachao, beti padhao”. The campaign implies that Indian parents are so bad that they have to be goaded to “bachao” their own child! This is an example of how a nation as large as India, with so many academics, and supposedly vibrant media has failed to do proper research and has naively accepted narratives that demonize Indian parents. This is a good example of how India still must gain its intellectual freedom and confidence.
The allegations that Indians in general are indulging in massive abortions by utilizing pre-natal sex-selection techniques appears to be flawed, and dependent on very limited and flawed statistical evaluation. There are multiple natural reasons – such as race, nutrition, age-of-parents at time of conception, hormonal issues, birth order etc. that may affect sex-ratio. Most of these potential influences on sex-ratio are clearly out of the reach of government policy. At the very least, the Indian government must not succumb to claims based on bad science but, must put in efforts to identify if there is indeed a problem. In addition, at the ground level, the only thing that the Indian government can immediately and effectively do, is to ensure that all children, especially girls, get access to good nutrition and education and stay in school longer.
- Anderson, Siwan and Ray, Debraj (2010): “Missing Women: Age and Disease”, Review of Economic Studies ;77, 1262–1300 http://www.econ.nyu.edu/user/debraj/Papers/AndersonRay.pdf
- Sen AK. (1992) “Missing women”, BMJ; Vol 304:587-58.
- Sen AK (2003) Missing women revisited, BMJ.; 327(7427): 1297–1298. http://www.ncbi.nlm.nih.gov/pmc/articles/PMC286281
- Jha, Prabhat et al, (2006) Low male-to-female sex ratio of children born in India: national survey of 1·1 million households, www.thelancet.com Published online January 9, 2006 http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(06)67930-0/abstract
- Jha, Prabat et al, (2011) Trends in selective abortions of girls in India: analysis of nationally representative birth histories from 1990 to 2005 and census data from 1991 to 2011, Lancet; 377: 1921–28 http://www.thelancet.com/journals/lancet/article/PIIS0140-6736(11)60649-1/abstract
- Richard Horton (2015), Offline: What is medicine’s 5 sigma?, Lancet; vol 385 http://www.thelancet.com/pdfs/journals/lancet/PIIS0140-6736(15)60696-1.pdf
- Gallup Poll (2011), Americans Prefer Boys to Girls, Just as They Did in 1941 http://www.gallup.com/poll/148187/americans-prefer-boys-girls-1941.aspx
- Coale AJ (1991) Excess female mortality and the balance of the sexes in the population: an estimate of the number of “missing females.” Population and Development Review v;17:517-23
- Mary E. John (2014) “Sex Ratios and Gender Biased Sex Selection: History, Debates & Future Directions”, Published by UN Women http://asiapacific.unfpa.org/sites/asiapacific/files/pub-pdf/Sex-Ratios-and-Gender-Biased-Sex-Selection.pdf
 Quote from Anderson Ray: “There is evidence that male offspring are more susceptible to death in utero than the female, so maternal malnutrition could be linked to a lower sex ratio at birth (Andersson and Bergstrom, 1998). With better health care, it is to be expected that the sex ratio at birth will rise as more male fetuses survive. Klasen and Wink (2003) attempt to correct for this; some reflection immediately shows that such a correction must increase the number of missing women at birth.
 Richard Horton, Offl ine: What is medicine’s 5 sigma?, vol 385, April 11, 2015
Featured Image: Wion News