For want of a theory

Imagine that Manmohan Singh and Sonia Gandhi were students of game theory. Would that have enabled them to avoid the dreadful splattering of eggs on their faces? Perhaps not, because a knowledge of this arcane branch of economics is neither a necessary nor a sufficient condition for displaying courage in a game of chicken. But it might help them analyse better what went wrong in their face-off with Prakash Karat. It is not too late even now. They can ask someone to give them a short lecture on the work of the three winners of the Bank of Sweden prize for economics "" known somewhat misleadingly as a Nobel ""this year. The main effort came from 90-year-old Leonid Hurwicz. It was built upon later by Eric Maskin and Roger Myerson.

How would Mr Hurwicz's work have helped the Prime Minister and his party boss? Quite simply by alerting them to the dangers of believing those who don't tell the truth and of disbelieving those do. In their case, it was the latter that laid them low: they did not believe that anyone could be as ideologically obdurate as Prakash Karat was, to the extent of bringing down a government which, for all practical purposes, he controls. Mr Hurwicz's singular insight, which led to what is boringly called mechanism design theory, was something any railwayman or air traffic controller can tell you, namely, that in a system which depends on communications for optimal outcomes, if even one person economises on the truth, the outcome will not be optimal. This becomes particularly true in a system that relies primarily on private information. Although Mr Hurwicz did not have them in mind, the theory applies just as well to spy rings, mafias, triads, etc.

A political coalition like the UPA or the NDA, it can be argued, is just another form of such secretive systems. Thus, the UPA mechanism would have been "incentive-compatible if it is a dominant strategy for each participant to report his private information truthfully". Sharad Pawar, Lalu Prasad and M Karunanidhi were, in all probability, telling the truth that they would not ally with the Congress if it went for early elections. Sonia Gandhi believed them but, as we saw, not Karat. Result: she did not get what Mr Hurwicz called an incentive compatible mechanism. This suggests the problem that Mr Hurwicz identified "" possibly in the context of the private communications between the US and the USSR when thermo-nuclear war was a clear possibility if one side lied. This was that no incentive scheme, regardless of how smart it is, can ensure that people always tell the truth. If you know that there is a non-negative probability that someone may be lying, even if only just a bit, you will make mistakes with your own strategy. This is what happened to Sonia Gandhi and Manmohan Singh. They disbelieved Prakash Karat and believed their allies, none of whom, in this game of private communications, may have been telling the whole truth. The original mistake they made was in not setting up the game in such a way that everyone is forced to tell the full truth and only the truth. Therefore, bluffs were to determine the outcome: remember the Prime Minister's dare, "Let them do their worst!". First Mr Hurwicz and then Messrs Maskin and Myerson showed that such mechanism designs are no good. What a pity that the Prime Minister was a trade economist and not a game theorist.