This was a problem first recorded by Aristotle: A category becoming a member of itself. In this case the category is a model of the economy or some part of the economy.
To what extent do economists climb inside their own models and contaminate their results? A social science equivalent of the uncertainty principle, examining the model changes it. But more prosaic since the disturbance may consist of some human being trying to use the model to game the system. The thought can't be original with me and my formal training consists of the three typical undergrad courses--macro, micro, and stats--so I'm definitely in way over my head but I do find this fascinating.
As soon as a functioning economic model escapes the laboratory it affects behavior in the real world and once a model is within itself it needs modification beyond the capabilities of its maker.
Sunday, March 01, 2009
Subscribe to:
Post Comments (Atom)
3 comments:
the phrase "climb inside their own models" is not precisely understood by me ...
i can see how people's egoic identity ossifies in the shape of an earlier discovery, and nothing new happens for the remainder of their lives .. is that similar?
discovered you on @pkedrosky's blog, interesting take, and style, but just outside my grasp ..
The answer was articulated better by a commenter on Paul's blog named Larsy:
Maybe the problem is that people take these models and construct positive feedback mechanisms for them. And positive feedback makes the system unstable and prone to oscillation, potentially causing the system to blow up.
In control systems, if the system starts behaving outside of the "normal" operating range of the model, the model changes (piecewise linear) or the system gets shut down.
I believe that popular models do indeed influence behavior, which provides an ironic sort of "validation." Take the Fibonacci retracement series used in technical analysis. Every book will say that it is not well understood why the series is somewhat successful in predicting price reversals. The books ignore the obvious answer: It works because people think it works. They buy and sell at the prices the model predicts, which is essentially the tail wagging the dog. Personally, I don't trust any model that isn't in the Victoria's Secret catalog.
Post a Comment