Friday, 18 November 2011

The foundations of economic models

How important are foundations to what economists do? Take, for example, the way economists think about competition. Milton Friedman has written,
Of course, competition is an ideal type, like a Euclidean line or point. No one has ever seen a Euclidean line – which has zero width and depth – yet we all find it useful to regard many a Euclidean volume – such as a surveyor’s string – as a Euclidean line. Similarly, there is no such thing as “pure” competition. Every producer has some effect, however tiny, on the price of the product he produces. The important issue for understanding and for policy is whether this effect is significant or can properly be neglected, as a surveyor can neglect the thickness of what he calls a “line”. The answer must, of course, depend on the problem.
In other words there is a trade-off between the usefulness of our models and the foundations of those models. Jean Tirole has made this point with regard to incomplete contracts models.
A methodological divide may have developed in our profession in recent years between those who advocate pragmatism and build simple models to capture aspects of rea;ity, and others who wonder about the foundations and robustness of these models [...]
Personally I’m more in the build simple models that help us understand the issues we are dealing with camp rather than worry about getting all the foundation exactly right before using models. To take Tirole’s example of incomplete contracts models. We know that there are issues with the standard, property rights, story told about why contracts are incomplete. The reliance of the property rights approach to the firm on a theory of incomplete contracts that assumes unforeseen or indescribable states of the world as a way of generating incompleteness has lead to a number of criticisms, the most significant of which is the Maskin and Tirole (1999) critique. Maskin and Tirole argue that information which is observable to the contracting parties can be made verifiable (to a third party) by the use of ingenious revelation mechanisms. The contracting parties write into their contract a game which when played gives the appropriate incentives for them to truthfully reveal their private information in equilibrium. This undermines the non-verifiability approach to incomplete contracts.

But such issues have not prevented the property rights approach from becoming a workhorse in the theory of the firm literature or from being applied in many other areas such as the analysis of firms’ internal organisation; firms’ financial decisions; the costs and benefits of privatisation; and the organization of international trade between inter- and intra-firm trade.

So the question remains, Does this lack of proper foundations for some economic modela really matter? Or is getting answers to important questions, regardless of foundations of the model, enough?


Anonymous said...

Wouldn't a model that actually helps you understand some economic point be one that is capable of half-way decent predictions? Seems to me that a model incapable of making any workable predictions is not a model that actually leads to understanding.

in this sense, all models need to be tested againstyt reality, and if they fail (and many economic models fail miserably at making half-way decent preedictions) then perhaps the problem is untenable assumptions at the base.

Unlearningecon said...

'Assume a spherical cow'

Engineers and scientists know the difference between a useful assumption and a non-useful assumption. The majority of scientific assumptions simply eliminate a known variable - the same cannot be said for economics.

Assuming perfect competition and using it as a starting point is a bad idea in an economy dominated by oligopoly.