Rajeev Srinivasan’s Innumeracy
Published by Ravikiran Rao January 23rd, 2004 in Economics(I am taking this back. See here)
This is Bash-a-Mallu-whose-views-you-otherwise-agree-with Week on Anarcaplib. Today’s victim is Rajeev Srinivasan (via JK)
Now, Rajeev graduated from IIT Madras, where presumably he picked up powers of analysis, and with a Marketing MBA from Stanford, where I suppose he learnt spin to argue persuasively. Unfortunately these tendencies seem to have come into conflict in this article.
Apparently Goldman Sachs have come up with a model that tries to predict the growth of countries, and the model predicts spectacular growth for India. Now to test the model (in Rajeev’s own words) they applied the model to the world of 1960 to see how well the model would have predicted actual growth since 1960.
This is fairly routine. Say you come up with a model that says: If a country’s government spending is x% of the GDP, it’s corruption index is y and the prevalance of college education among people is z%, then the country’s GDP growth is some function of the three factors f(x, y, z) (and that translates to say 8% if you plug in the actual numbers). How do you test if your model is correct and complete? Apply it to the past and see if your model correctly predicts the growth. To the extent that your model predicts correctly what happened in the past, you have confidence that it will behave in the future.
Clear?
It turns out that the model predicts 7.5% growth for India whereas our actual growth for the period 1960-2000 was 4.5%.
So what does Rajeev do? He treats the 3% difference not as an error in the model’s prediction, but as penalty that India had to pay because of Nehru’s policies. He also coins a grandoise term the “Nehru Penalty” which, he hopes will replace the misleading term “Hindu Rate of Growth”
Now, I agree that Nehru’s policies were terrible and we would have grown much, much faster if we had followed capitalist policies. That much, much higher figure might well be 7.5% or it may be even more. But just because I agree with someone’s conclusions doesn’t mean that I can overlook faulty logic.
I haven’t read the report and I am relying entirely on Rajeev’s article. So it is possible that the report is indeed doing what Rajeev claims it is doing, i.e. projecting what would have been the GDP growth rate if Nehru had followed the policies it advocates. But that is unlikely. Comparing one hypothetical number to another hypothetical number is a rather poor way to test a model. Even if this is the case, Rajeev is guilty of poor wording.
I don’t think the author is dumb. It is just that like everyone he is wired to be combative, not logical. Me, I am an exception. I wouldn’t have made a good lawyer.
6 Responses to “Rajeev Srinivasan’s Innumeracy”
- 1 Trackback on Apr 27th, 2004 at 4:52 pm
- 2 Trackback on Oct 13th, 2004 at 6:50 pm
In India Unbound, Gurcharan Das writes that Nehru’s ideas were a product of his times. The real fault lay with Indira Gandhi (and imho, the Janata opposition) which refused to jettison an ideology which was long past its due date.
I believe sticking on to the socialist dogma when even Communist China was “u-turned” by Deng Xiaoping and the Asian Tigers were well on their way was more the cause of the depressed growth rate (and institutionalised defeatism)
Abt theories, models and conclusions, the article looks like Rajeev started with the conclusion and then took the model as proof. But, the idea why the model’s accuracy is not being questioned is because for all the other countries in the study, the model is accurate to about 0.1 or 0.2 of a percentage, except India. While this result should be instigating more investigation on the factors responsible, just Nehruvian policy or Indra Gandhi’s policy cannot take the blame. They might be one of the many factors, and unless we know what the other factors are or truely prove that its only politics that is the main factor for this discrepency, the study does not help India much.
I think this Mallu bashing is an effect of the I-am-going-to-lose-Blog-Mela syndrome.
JK,
President, Online Poll Riggers Inc.
Three conclusions that I can draw from ur response to Rajeev’s article:
1. You are a non-IITian(so am I..). Thats why you tend to pick on what his background is rather than the article first.
2. I think he has a well-explained logic than what you do, in 3 lines. His logic seems to be reasoned out better and is more acceptable.
3. “Nehru Penalty” was an apt term and that was the basis of his article, and he explained it quite clearly.