Mismeasure of economics

-November 23, 2015

The resounding failure of economic models to provide accurate guidance in determining when, where, how, and why an economy will flounder has affected us all. Economists have tried myriad models from game theory to the bandit model but none provide the predictive power of other complex models like computational fluid dynamics or climate models.

Source: Pieria

Since the 19th century, mathematical approaches have been used to model economies. Mathematics encourages prediction, of course, but doesn’t assure accuracy. People seem to enjoy discussing market “forces,” but they’re not really forces. Physical forces don’t evolve with culture; economic tendencies do.

Economic models make a host of unjustified assumptions:
  • people think rationally (I know, try to say it with a straight face)
  • people act in their own best interest (somehow overlooked the desires for revenge and altruism)
  • everyone has access to all accurate information (ever try to get in on an IPO?)
  • investors act independently of each other (please)

Since around 2010, “nudge” economics has experienced isolated success in nudging economic behavior at small local scales. For example, printing tax bills on the red paper that scary collection agencies use improves response by several percent. Great successes grow from small ones; nudge economics is good stuff, but it won’t predict the behavior of global markets any more than I can predict the weather in Japan with my weather station in Petaluma.

The overriding hope among economists is that models of complex systems—think nonlinear dynamics, chaos, swarming, emergence—can combine collective behavior, evolutionary biology, neuroscience, anthropology, and sociology into an effective simulation. The simulation, one hopes, could then divine the behavior of global markets in time to avert tragedy much the way that weather simulations can warn pilots and sailors of impending doom.

There are at least two overriding differences between predicting the behavior of a complex physical system like a planet’s climate and a complex cultural system like an economy: first, the underlying forces that drive climates don’t change with the winds of desire and, second, the very act of predicting the climate doesn’t alter the climate itself.

Let me suggest that predicting economies may be fundamentally impossible. Not impossible in the sense that it’s too hard, but impossible in principle; impossible the way it’s impossible to measure both momentum and position with arbitrary accuracy. I suspect that there is an unavoidable Heisenberg-like uncertainty principle between individual people and their culture.

In physics, Heisenberg’s Uncertainty Principle embodies the fact that the experimenters cannot separate themselves from the experiment. Any measurement we make changes the universe. The HUP puts a concrete limit on our objectivity; we are always part of the system that we observe and every observation changes that system.

Now, back to economics. Say you come up with a model of the economy that can accurately predict economic performance. What would you do with that ability? You might be tempted to make a few dollars, right? I mean if you can predict economic performance, then you’ve won the game. The house has nothing on you. You’re not counting cards, you have x-ray vision and know every players’ hand.

But the instant you use that knowledge to peer into your economic crystal ball and make an investment, your model has to adapt to that change in the system. Once you start using your model, eventually, your model must either account for your use of it, or fail. My money’s on the latter because positive feedback loops are nonlinear beasts that have a propensity for going chaotic.

The exciting part, is that we’re likely to find out how close we can get to modeling a big economy without that destructive feedback.

Also see:

Loading comments...

Write a Comment

To comment please Log In