George Fisher was partner in a firm of auctioneers. He was well enough off to send his son, Ronald, to Harrow in 1904. But he lost his wealth soon after. Ronald worked hard on his studies, and got a scholarship in mathematics to Gonville and Caius College, Cambridge, in 1909. After graduation, he did not get a university job; he earned a living teaching in undistinguished schools, doing casual farming, and writing articles.
He had an early interest in sex selection; in 1915, he wrote in Eugenics Review on the evolution of sexual preference – that is, on parents’ preference for male or female children. In 1919, he got a job in Rothamstead, an agricultural research station. He was a friend of Leonard, the last son of Charles Darwin, and dedicated to him a book he wrote in 1930, The Genetical Theory of Natural Selection. In 1943, he was appointed professor of genetics in Cambridge.
Amongst other things, he formulated Fisher’s principle. In simple language, it means: Suppose that a certain population gives birth to fewer males than females because it has a high proportion of families that have more female than male children. Then an average male would have more females available to mate with, and would have more children. So couples that are genetically disposed to have more boys would produce more grandchildren. That would increase the proportion of males; over a few generations, the shortage of males would be rectified, and the sex ratio would come back to 1:1.
The data do not support this principle. There are 1.07 percent more males than females in the world. Sex ratios can be extreme in small regions with high migration of one sex: for instance, the migration of males for work has given the United Arab Emirates an excess of males and Kerala an excess of females. But if we exclude small countries, Russia and five of its east European neighbours have sex ratios under 0.85 – that is, they have 17 per cent more females than males. At the other end are six countries in the Arabian peninsula. They are rich because of their oil, and have imported a lot of men to work while the locals have a good time. And close to them are India and China, with about 7 per cent more males than females.
More males leave India to study or work abroad than females; but India is too populous for this to affect its sex ratio much. It is common knowledge that this ratio is low because of female foeticide. The subject is generally not discussed in public; we find it embarrassing. Ravinder Kaur is the only scholar to have studied it seriously. But last year, Sylvia Barcellos, Leandro Carvalho and Adriana Lleras-Muney asked in an NBER paper, whether Indian parents treated boys and girls differently. They showed that families with one child spent an hour – 30 per cent – more time on it if it was a boy; larger families spent half an hour more on a boy. A boy was also likely to be breastfed longer, vaccinated or given vitamin supplements. In sum, Indian parents mollycoddled boys.
Now, The Economic Journal has published an interesting study by Analia Schlosser and Luogia Hu. They pooled data from three rounds of the National Family Health Survey, covering more than a million and a half people, and ran regressions on them. Their calculations give a more nuanced picture.
For instance, one question that families were asked was: how many sons and daughters would you like to have? From which Schlosser and Hu calculated a preferred male-to-female ratio. One would have expected it to be highly correlated with the actual sex ratios amongst children. It was in Punjab; but it was not in Madhya Pradesh, where parents equally prefer boys.
Ultrasound technology makes it easy to determine the gender of an unborn child. It was introduced in 1990. The proportion of boys increased more rapidly after 1990, even more so in northern, son-preferring people and amongst richer people who could afford ultrasound. But Schlosser and Hu ran separate regressions on figures before and after 1990, and derived the conclusion that couples have a strong preference for at least one son, but not an aversion to having daughters.
Sex selection is not the only consequence of sex preference; even after birth, girls are discriminated against: fed less, given less medical attention, not sent to school and withdrawn earlier from school. Schlosser and Hu ran regressions on the weight of children, and inferred that the underfeeding of girls went down after ultrasound became available. They eliminated the possibility that this happened only in better-off families. They also found that breastfeeding of girls lasted longer after 1990. In other words, sex selection reduced the subjective surplus of girls in parents’ minds, and reduced neglect of girls that were born. There was, however, no improvement in girls’ mortality rate. That was intriguing. Schlosser and Hu made a number of guesses, none of which is convincing.
Thus, families’ preference for boys did not weaken after the invention of ultrasound technology. Families with a stronger boy preference continued to have more boys; the technology only made it easier for them to exercise the preference. But families with girls became smaller. And the number of older siblings of girls dropped. In other words, the families which did not practice abortion and continued to have children until they got a boy could now stop after having fewer children; the size of families with girls went down, and girls presumably had less competition from siblings and got better care.
Ultrasound sex selection has other consequences that are not covered by the data of Schlosser and Hu: amongst them, a lot of men cannot find wives, wives are imported from strange, faraway regions, and informal polygamy has returned. It may have other consequences that have not been investigated or are not being talked about: for instance, increase in prostitution, greater instability in marriage, or greater frequency of rape. There is also the progressive decline in fertility and in family size and reduction in population growth, which may or may not have something to do with sex selection. It has been a major technological innovation, and is having a complex impact on our society: it is a huge experiment in a national laboratory. Demographers are writing scholarly articles on it; but we need to ask broader questions. We have had a census registrar who has produced voluminous statistics for almost a century and a half; he still occupies valuable space in Lutyens’ Delhi. But he finances or supports hardly any research.
How much we could get from figures has been illustrated by Schlosser and Hu. I wish Indian social scientists would follow their lead and do illuminative research. The academic world has been invaded by Hindutwit amateurs in an aggressive, intolerant frame of mind. A few public intellectuals are fighting them in newspapers and on television channels; but that is not the battleground where the war will be won or lost. One way of surviving the onslaught is to do low-profile, uncontroversial research. Demographic sociology is one field that offers good scope for such work.
Image credits: blogs.unicef.org.uk