Monday 16 July 2012

Peter Bowbrick vs. Amartya Sen

Back in 1986,  Peter Bowbrick showed that Amartya Sen's work on Famines was mischievous nonsense. Famines occur because there isn't enough food. The Bengal famine wasn't caused by better paid workers in the Cities eating a lot more rice thus driving up its price and causing poor people in the villages to starve. When people get more money, they buy less rice and more fish and vegetables and butter and so on. Nobody can eat six times as much rice as they did before getting a pay rise.

The Bengal famine, like most famines, was caused by a fall in the supply of food. The Muslim League Govt. in Bengal believed, as Sen still does, that famines aren't caused by a shortfall of food. That's why they didn't introduce rationing but simply carried on lining their own pockets. When Wavell became Viceroy, he had to twist the arm of the Govt. of Bengal to stop obstructing famine relief for their own petty political reasons.

Sen issued a sort of obfuscating rejoinder to Bowbrick- at that time without a PhD- but failed to address the issues he had raised. Instead, he mocked Bowbrick for claiming that official figures were subject to a 3000% error! However, what Bowbrick had said was that, in that context, the difference between two unreliable figures gave an error of up to 30 times the quantum stated. That is perfectly reasonable. If I have a 10% error in both Income and Expenditure, my Surplus may be 10,000,000% off the mark. I think I'll have a 1 Paisa surplus but it turns out I'm 10,000 Rupees in the hole.
This is Sen at his sneering, sneaking, best-

Sen has just brought up the 3000% error in a disingenuous way which suggests that Bowbrick is so stupid that he thinks the officials thought supply or demand might be 30 times less or more than the outcome. He then says 'There is no doubt that all such figures are subject to possible errors'- what a humble little humbug it is! but primly adds, like a genteel maiden Aunt whose nieces are arguing about what size of dildo they should order online for her birthday, 'I shall not comment on the possibility of a 3000% error!'
However, the margin of error in the change in the shortfall- which is what affects the change in the price for a good in inelastic demand, like rice in Bengal- could very well be of the order of 30 or a 1000 of what he had calculated. If Sen were really an Economist it would be his duty to comment. But,  he isn't an Economist. Just a pi-jaw merchant, a surfer of availability cascades, a sneering, sneaking, careerist.
 Academia and the deeply corrupt Anti-Povery racket welcomed Sen's work and ignored Bowbrick. After all, if famines are about there not being enough food, then the solution is not far to seek. Get in the scientists and the technocrats to find ways to grow more food and get it distributed properly. Boring stuff fit only for Agronomists and Engineers and the odd bureaucrat who doesn't know his place. Academic careers- which consist of recycling your old dissertation again and again till finally even Death backs off from you under the impression you aint yet brain-dead- can't be made of such stuff . Some Borlaug or Swaminathan might just fix the problem once and for all.

Why is Sen considered a great Economist? Or, to put it another way, wouldn't it be great if Economists were more like Bowbrick?

1 comment:

  1. To clarify: a good crop forecast might lie within the limits of the true figure plus or minus 30%. The DIFFERENCE between two such forecasts (and Sen's argument relies entirely on differences) has a much wider error and can easily be thousands of times the true difference. So Sen's claim that the difference of 17% was completely RELIABLE which is what I was attacking is about as wrong as you get.

    The statisticians at the time were agreed that the forecasts Sen quotes were not good but were 'totally meaningless'. Worse, there were strong biases, so they did not observe crop failures. So, after the famine, the great Indian statisticians developed the systems we use now. They did it to stop hunger, not as academics concerned only to improve their citation indices.
    Peter@Bowbrick.Eu

    ReplyDelete