Causal Effect

Also known as · Causal Inference

The change in yy caused by a one-unit change in xx with everything else held fixed. It is a counterfactual quantity, so it is never directly observed in the data.

When to use

Lecture 1 sets up the goal of econometrics as estimating this effect; use it to explain why a raw correlation is not enough and what a randomised experiment or regression tries to recover.

See Lec 1.

Appears in

No references yet.