Skip to content

Micro Versus Macro

Why do macro models look so different from micro models?

A caricature:

  • A macro model:
    • Has 7 parameters which are calibrated to match 7 data moments.
    • Has decision rules that only depend on a few state variables.
  • A micro model:
    • Has 700 parameters which are estimated so that the model replicates the complete life histories of everyone in a dataset.
    • All decisions a person ever makes in life are modeled.
    • There are preference shocks everywhere.
    • Decision rules depend on everything the researcher can observe.

Why do researchers in the same profession, often studying the same question, write down models that look so different?

I think the answer is that applied micro economists view models very differently from macro economists.

In macro, a model is a proof of concept. It is a reasonable model with reasonable parameters that demonstrates that a particular mechanism can be big.

In micro, a model is a close approximation of the true data generating process (that's why there is so much emphasis on hypothesis testing). The quantitative answers are taken seriously.

Last updated: 2016-Dec