Causal inference in economics and marketing

• Hal Varian
• https://www.pnas.org/content/113/27/7310.short

Causal inference, Causal inference with observational data, Machine learning

Basic identify of Causal inference:

Outcome for treated - Outcome for untreated = $$Outcome for treated - Outcome for treated if not treated$$ + $$Outcome for treated if not treated - Outcome for untreated$$ = Impact of treatment on treated + selection bias

This tells us that the crucial quantity to estimate is between the actual outcome and the counterfactual for the treated group.