Mastering 'Metrics: The Path from Cause to Effect

The best introduction to causal inference I’ve encountered. Angrist and Pischke manage to make econometric methods both intuitive and rigorous, using real-world examples that stick with you.

The “Furious Five” - randomized trials, regression, instrumental variables, regression discontinuity, and differences-in-differences - are presented not as abstract techniques but as practical tools for answering causal questions. Each method is motivated by a compelling example before diving into the mechanics.

What makes this book exceptional is its focus on the intuition behind each method and when to use them. The authors are honest about limitations and assumptions, helping readers develop good judgment about when these tools are appropriate.

The writing is engaging and sometimes humorous, making what could be dry material surprisingly enjoyable. The connection to their more advanced text “Mostly Harmless Econometrics” provides a clear path for deeper study.

Essential reading for anyone doing data science who wants to move beyond correlation to actually understanding cause and effect.