T C A Srinivasa-Raghavan: The Gandhi family superstition

OKONOMOS

"When can a superstition persist in the face of rational learning?"

While trawling the Net for interesting research, for some reason I was thinking about Congressmen and their blind faith in the Gandhi family which has persisted despite the fact that, since 1977, the family has lost the party its traditional electoral base, made common cause with political opponents, subscribed to an economic theology that defies common sense, caused the party to lose two general and countless state level elections, and worst of all, treated party members the way officers of the East India Company used to treat their Indian servants.

Quite serendipitously, I came upon the following sentence. "By a superstition we mean a belief which is objectively false. But when can a superstition persist in the face of rational learning?" I read on, and found a paper* by two economists called Drew Fudenberg and David K Levine, both of Harvard.

Having asked why people suspend belief, they give us a possible explanation "" by reverting to one of my favourites, the Code of Hammurabi whose essence our own Dr Manu heavily borrowed for his Smriti.

Hammurabi had an interesting way of establishing guilt: just throw the accused in the river. If he (or she) survives, innocence would have been proved. The superstition underlying this was that "the guilty are more likely to drown than the innocent."

But, ask the authors, if innocence can be proved by surviving in the river, why won't people believe that the guilty will be struck dead by lightning?

Well, sorry about this, but apparently it all has something to do with the "partial characterisation of the outcomes that arise as the limit of steady states with rational learning as players become more patient" and "... a mechanism that uses superstitions two or more steps off the equilibrium path, such as 'appeal by surviving in the river,' is more likely to persist than a superstition where the false beliefs are only one step off of the equilibrium path."

In short, more knowledge diminishes superstition "" or ought to, anyway.

So then the issue really becomes how many experiments a rational learner must perform in order to start believing that his old beliefs were wrong. Or, what will it take for Congressmen to dump the family?

The following observation is absolutely fascinating. "Rational but very impatient learners will only play 'greedy' strategies that maximise current payoff so that steady states with impatient rational learners can exhibit a wide range of false beliefs."

So even rational people, who want quick gains will play the game in such a way that it persuades others who are less rational to subscribe to superstitions. The last Congress prime minister and, indeed, the current one, both fall in this category.

The secret of their success lies in the prior beliefs of the younger generation. If, for instance, the younger generation has been taught that if it commits a crime, lightning will strike it, it will first try to see if this true. When after a few crimes, nothing happens, or when it discovers that lightning strikes are independent of whether it commits crimes, it will stop believing. Or, what is the same thing, why Congressmen will begin to think that winning or losing an election might be occurrences independent of their faith in the Gandhi family.

But to arrive at this conclusion, you require repeated iteration. This is not always possible.

In the Hammurabi case, you need not just criminals, but also accusers. "...the young commit crimes and are then accused and then punished so they learn that crime does not pay, and as they grow older stop committing crimes."

But "" and here lies the rub "" accusers enter the game only after a crime has been committed. Since very few crimes are committed, they don't get to play very often, and "infrequent play reduces the value of experimentation, because there will likely be a long delay before the knowledge gained can be put to use."

The only way out for young accusers is to make false accusations, which they will not normally make. So the result is that "they will never learn that the river is as likely to punish the innocent as the guilty."

When you think about it, the Congress party is exactly like this "" a lot of old buffers persuading the young to hang on to the Gandhi family myth. The young don't get opportunities to see that they are being taken for a ride and are too scared to call the bluff, and the myth persists.

*Superstition and Rational Learning, Harvard Institute of Economic Research, Discussion Paper No. 2114
http://post.economics.harvard.edu/hier/2006papers/2006list.html