Atanu Dey On India's Development

Sex selection in a Second-Best World

Niket in a comment raised the issue of the skewed sex ratio in the context of population control. To my mind, the differential preference for boys over girls is a consequence of overpopulation as well. If the population problem were to be addressed, the skewed sex ratio problem will also be addressed. For my views on the causes and consequences of the skewed sex ratio, check two earlier blog entries. The first, The Skewed Sex Ratio where I wrote:

A little reflection on the facts leads one to conclude that the skewed sex ratio is a consequence of other underlying facts such as resource constraints, exhorbitant cost of dowry for getting daughters married, female illiteracy, and so on. Poor families have severe resource constraints, ranging from calories to clothing to education. If sons have a greater net present value (due to their future earning capacity), girls are disadvantaged in the share that they get of the limited resources.

It all boils down to the fact that this is a second-best world. There are multiple problems which conspire to create the skewed sex ratio. Merely addressing the effect leads to idiotic policy recommendations such as banning the determination of the sex of a foetus. One unforunate consequence of that ban could well be the increase in the number of new-born female infants killed, or worse still, chronic neglect of the unwanted girl child.

In a followup post, The Lop-sided Sex Ratio (revisited), I explained why I believed that selective abortion of female foetuses was better than the neglect of female children.

My position is that the fact is that some people value female children less than male children. This is a lamentable fact but a fact nonetheless. I did not dictate that people value girls less. I am taking that as given and (at least for the present) unalterable fact. Breast beating may feel good but will do little to alter that fact. Altering that fact would be an end that all right-thinking people devoutly wish for. It may take a few generations. Until then, what is the most humane way to deal with the problem. Do millions of unwanted girl children have to suffer inhuman neglect? Can society protect the rights of children with as much gusto as the protection of foetuses? Which is the lesser evil: the aborting of female foetuses or the terrible fate of an unwanted girl child?

I don’t support bans of abortions because in second-best world, it may be a welfare improving intervention. Of course, I should explain what a “second-best world” means before people jump down my throat for suggesting something so horrific as a second-best world.

I define a system to be “first-best” if the system is free from all defects or distortions, that is, an ideal system. In most real (as opposed to ideal) systems, there are multiple distortions and that makes them “second-best”. In the presense of multiple distortions, policy recommendations or interventions that are welfare improving in first-best situations, may not be welfare improving. In fact, applying first-best prescriptions to second-best systems could lead to significant worsening of the system. So what is one to do in a second-best world? If one could, the best thing to do is to remove all the distortions simultaneously and move to a first-best system. Otherwise, it may be necessary to actually introduce additional distortions to improve the system in some cases.

What I am saying above is an extension of what in the international trade literature is called the theory of the second best. Refer to Jagdish Bhagwati’s poineering work in that regard. I believe that the theory of the second best has a much wider applicability than just in international trade. Indeed, I will go so far as to say that the theory of the second best should be an essential tool in every thinking person’s mental toolkit.

A rough example would be the irradiation of people suffering from cancer. You would not recommend heavy doses of radiation for healthy people. Now if you take the stance that radiation is bad (a policy for first-best cases) and rigidly apply that in the case of a cancer patient (a second-best case), you could potentially hurt the person instead. What you did was introduce a distortion to counter some other distortion.

Another rough example: hand a pair of crutches to a person who is not lame, and you reduce his mobility. Remove the crutches from a person with a broken leg, and you reduce his mobility. Remove both distortions — the broken leg and the crutches — simultaneously and you have improved the system. Remove less than all of the distortions and you could make the system worse off.

Aborting healthy fetuses is a bad thing in a first-best world. A first-best world is one where children are not neglected, where resources are plentiful, where there is no overcrowding, where every pregnancy is wanted and there are no accidents, where women have full control of their reproductive choice, where women are not raped or coerced, … the list goes on. But it is not a first-best world. It is a second-best world and hard choices may have to be made.