Mostly harmless econometrics

Mostly Harmless Econometrics provides an empiricist's companion to key econometric tools, emphasizing linear regression for statistical control, instrumental variables for natural experiments, and differences-in-differences methods that exploit policy changes.

It addresses causal questions such as the effects of class size on learning or education on wages, while covering extensions like regression-discontinuity designs and quantile regression.

The authors argue that fancier techniques are often unnecessary and potentially harmful, using simple, practical methods illustrated with empirical examples relevant to social sciences.