Monday, June 08, 2009

Cheung's Theorem


This is from Professor Steven N S Cheung's latest column, more here.

Several implications follow:

1. In situation where the problem of information asymmetry is serious, and where the government intends to intervene , we should expect more officials to act in ways that contravene public interests, like soliciting bribe, defining rights in such a way which would benefit their supporters...etc

2. It cautions against those who automatically ask for the visible hand to step in when information asymmetry prevents the market from discharging its ususal function.

The same asymmetry information which negatively affects the operation of the market also encourages or provides a better cover for officials to benefits themselves instead of working for the public interests. So the results generated by government intervention might be worse than that delivered by the market.

Furthmore, even if the incentive problem articulated by Professor Steven Cheung is resovled, say all officials are Angels, you would still encounter the Hayekian type knowledge problem. For officials would need to know exactly when to intervene and how to intervene. As the Professor mentioned, "但他们要懂得分辨哪些工作他们要做,哪些应由市场处理". We differ on this point because the Professor thinks the officials would do a pretty good job in this area if the incentive problem articulated by him could be ignored. I am far less sanguine about the officials' ability to resolve the knowledge problem.



Even officials can resolve the knowledge problem, either by themselves or some measures (e.g. consumer council's 格價特工), I think the cost-effectiveness is in doubt.

Gary M C Shiu said...

I disagree. The knowledge problem, the source of which is dispersed information among each and every individual, is not something that could be resolved WITHOUT the market.

On the supply of information side, you need incentives like profit for the information holder to reveal what he or she knows.

On the demand side, users of information can rely on the price signals which embody the knowledge revealed in the actions of participants whom the users do not and will know. How can they be so sure. Well, if the suppliers of info are motivated by profit/self-interests, they would have the icentives to reveal what they really know.

Yes, of course you would come back with problems like information asymmetry, deliberation distortions of knowledge one knows in order to gain an upper hand in transactions. True, those things happen.

But then we also know there are other mechanisms, created and maintained by competitive forces in the market to resolve those problems which arise in certain markets with perculiar characteristics. Reputation, signalling and other measures are means which can resolve those problems.

I think some Mechanism Design theorists have written stuff on how one can design scheme to solicit truthful info from individuals, kind of a substitute for the market. But I am not too familiar with that literature. I doubt that would work.