Wednesday, October 29, 2014

Lars P. Syll — Macroeconomic aspirations


More on micro foundations and realism. Microfoundations are supposed to add realism. They don't, because the assumptions of conventional modeling are unrealistic. Representative agent models are hopelessly simplistic and agent based model quickly get intractable unless they are kept very simple, too. Simplicity may be a virtue, but over-simplying in modeling isn't. And complexity is complex. 

The challenge is accounting for the complex as simply ("economically") as possible, while also "saving the appearances," as Aristotle put it. The danger lies in generalizing a special highly stylized case based on assumptions that make the model irrelevant to reality.

The basic challenge of science lies in successfully connecting models with reality. Many models can be devised to account for the same data. The generally accepted criteria are correspondence, consistency, elegance (simplicity, economy) and practicality. Preferring consistency and simplicity to correspondence and practicality doesn't cut it.

Lars P. Syll’s Blog
Macroeconomic aspirations
Lars P. Syll | Professor, Malmo University

No comments: