A mental model is a simplified representation of how something works. Economists rely on them constantly — not because they’re accurate, but because they’re useful.

“All models are wrong, but some are useful.” — George Box

The Cobb-Douglas function is a good example: it doesn’t perfectly describe any real economy, but it generates correct predictions about factor shares and growth patterns often enough to justify its use.


What makes a good model

A useful mental model in economics tends to be:

  1. Tractable — you can derive implications from it without heroic assumptions
  2. Falsifiable — it makes predictions that could, in principle, be wrong
  3. Portable — it applies across contexts beyond the one it was built for

The adaptive systems perspective adds a useful layer: good models should also describe how the system responds to shocks, not just its equilibrium state.


The danger of misapplied models

The same properties that make a model useful — simplicity, tractability — also make it dangerous when applied beyond its domain. Treating a partial-equilibrium model as a general one, or assuming constant returns to scale in a context with strong complementarities, produces systematically wrong conclusions.

Knowing a model’s boundaries is as important as knowing the model itself.


Notes

This note is a stub. Expand with specific examples from growth theory, behavioral economics, and institutional analysis.