Wednesday, October 17, 2007

Mechanism Design

There is a fascinating article over at Reason Online about the theory of "mechanism design" and how it won Leonid Hurwicz, Eric Maskin, and Roger Myerson the Nobel Prize in Economics.

Mechanism design is the study of how to create institutions that produce desirable outcomes whilst respecting privacy of the individuals the institution interacts with, and respecting the fact that individuals are self interested.

The "institution" in this context is taken as a simple mechanism that takes input from "agents" or individuals and produces some output.

This is very relevant to the UK today. Our current government has adopted a fairly authoritarian attitude towards the public. ID cards, DNA databases, and increased surveillance (2) are all aspects of this tendency towards advocating control and authority.

I don't have particularly strong feelings on any of these issues. I doubt very much whether any ID card biometric databases would be competently administered.

However I think it is worth studying how institutions (like the state) behave.

According to the article on Reason Online, it is fairly difficult to "design" such mechanisms from the ground up. The free market functions, but is prone to failures and interference. States function, but are also prone to failures and also tend to extend the influence of the state.

I am interested to see if any of these economic theories are ever applied in real life. Presumably one day we will stumble across an optimum method for dealing with scarcity (but, I suspect, not before scarcity itself has been reduced somewhat by technological factors), a sort of Economics 2.0.

No comments: