Perhaps the most disturbing socio-economic imbalance in today?s India is her skewed sex ratio. For children below six years of age, the 2001 census reports a national sex ratio of 927 girls per 1,000 boys, one of the worst gender survival differences in the world. The problem is most acute in states with a higher share in agriculture and, perhaps surprisingly, among middle class and upper caste households. The subject has expectedly received considerable attention from academic researchers for several years, but many important questions remain. A recent research paper by Tarun Jain, a doctoral student at the University of Virginia, therefore throws some much-needed light on the matter.
The issue appears deceptively simple at first glance. Everyone knows that families covet sons for religious and cultural reasons and social ills like dowry make the preference even starker. But how do they make it happen? Media and literature are replete with sensational stories of female feticide and infanticide. These effects, while sadly true, do not explain most of the sex bias. Researchers estimate that the likely extent of sex-based abortion runs in hundreds of thousands ? shockingly large but still about a 100th of the tens of millions necessary to make a material difference in the national sex ratio. Similarly, female infanticide is more complex. A significantly higher proportion of new-born girls survive the first month of their lives than boys. Girls really seem to lose out to boys in the race for survival largely during the age group of 7-36 months. So the prime cause of the distorted sex ratio is neither feticide nor infanticide, but simply neglect of the girl child. There is evidence that the average Indian girl child gets less nutrition, immunisations, and medical attention than the average boy. Conclusion: parents discriminate, right? Hardly news; but there may be another answer that has more to do with how averages work.
The average girlchild in India simply has more siblings than the average boy. The reason for this is the well-known family planning phenomenon of using ?stopping rules?. In other words, couples are often likely to continue to have children, subject to affordability considerations, till they have a desired number of sons. So the last child is more likely to be a boy than a girl. Notice that this in itself does not skew the sex ratio, since every new pregnancy is about as likely to produce a girl as a boy. However, it does mean that the ?average? girl has more siblings than the ?average? boy. Therefore, even with perfectly non-discriminating parents, the average girl will have a smaller share of family resources, including childcare, than the average boy, something that may explain the evident neglect.
So the next question is why do people have these ?stopping rules? and son-preference? This may appear like a dumb question given our socio-cultural norms, but economists tend to believe that these norms are more the intermediate effects than the causes of economic phenomena. More importantly, why are agricultural and middle class and upper caste households particularly prone to this behaviour?
Here Tarun suggests a rather ingenious answer. Farmers like to keep their land within their family lineage. Therefore, in dividing land among his sons, a patriarch is likely to give a larger share of his land to the son who has produced more grandsons for him, after taking into account the ?elder son? effect. This would maximise the chances of keeping the land in the family line. The sons, in turn, know this and therefore compete with one another in producing boys ? and keeping them alive ? and hence adopt their ?stopping rules?.
Sounds like a typical economist?s pipedream? Well, consider the evidence. In a sample of agricultural households, each boy produced by a particular son appears to increase the son?s land inheritance by a small but statistically significant amount. But do the sons take this into account when having children? The probability of having another child declines by 34% for every boy, a son has but only by 13% for every girl he begets. However, the probability goes up by 4% for every boy one of the son?s competing brothers begets. Births of nieces produce no such effects. Clearly the sons are keeping scores and trying to play catch up in the game of having boys. As further evidence, this crosseffect of arrival of nephews enhancing chances of a new pregnancy appears only if the patriarch owns agricultural land and only while he is alive.
So, clearly the connection works. Does it explain the entire sex bias? Probably not. But it shows we need to focus on inheritance and property laws and fixing the customs there as part of our strategy to correct it.
Rajesh Chakrabarti teaches Finance at the Indian School of Business, Hyderabad