Modeling and Economics

By Ian

Interesting post up at Simulation: The Weblog. David Upton notes an article from the Mises Institute that does a rather usual (for the MI) routine on the limitations of data and econometrics.

Upton notes:

The problem is that sometimes the assumptions do work. Mathematical models and simulations can be used with amazing effectiveness to predict reality, to help us modify or improve it, and so on. At the moment, this is mostly in fields like engineering where the 'physical laws' are well known and the effects can be readily quantified.

Of course, it's the uncertainty around just those "physical laws" that makes a good deal of assumptions about economic activity so...well, fuzzy. But if I understand Upton's point, I think I side with him. My version is much shorter: "The reliance on data and math is bad compared to what?" I'm not convinced that the vast majority of economics education is laboring under a false idea of the usefulness of mathematics. By contrast, it doesn't sound like the Austrian School is winning converts based on the obvious superiority of their approach either.

The rest of Upton's post is also interesting as it talks about the iterative process of refining modeling techniques. Though I will say it reads like a shorthand version of Kuhn. As the theory keeps requiring more and more special exceptions the model that used to fit reality pretty well loses applicability, making room for the next step forward.

-- February 24, 2006 10:13 AM |

Comments

Buzzcut wrote:

The vangard of engineering at this time is "Six Sigma". You do a bunch of experiments, do data tricks like multiple regressions and such, and come up with you relationships in the data.

This is a lot different than engineering in the past, when there was a lot more time spent coming up with the actual mathematical model in a theoretical sort of way.

I can see how economics is much the same. Math models based on simplified theory were the way it was done in the past (maybe still is in some departments).

But then you get the freakonomics guy who is just mining data (although he's gone way beyond simple linear regressions, multiple or not).

The difference between engineering and economics is the quality of the data. I'm doing experiments where I can control the various inputs. Economists are using data on the economy collected by various governmental departments that are of dubious quality.

-- February 24, 2006 2:22 PM ∞

Post a comment