Causal Inference: The Mixtape (mixtape.scunning.com)

Scott Cunningham’s textbook "Causal Inference: The Mixtape" is now available online, offering a practitioner-focused introduction to the toolkit social scientists use to answer “what causes what.” The site combines conceptual explanations with concrete modeling techniques and runnable code in both R and Stata, using real-world policy examples (e.g., minimum wage effects, early childhood education, distribution of malaria nets) to show how to move from messy observational data to credible causal conclusions. The resource also links to "Mixtape Sessions" for those who want to learn directly from the author. For the AI/ML community, this is a compact, applied primer on methods that bolster model interpretability, counterfactual reasoning, and robust policy evaluation—areas where purely predictive approaches fall short. The mix of conceptual framing, applied econometric techniques, and reproducible code makes it practical for ML practitioners interested in causal estimation, domain adaptation, or algorithmic fairness to incorporate principled causal thinking into experiments and deployment. Whether you’re a researcher wanting defensible effect estimates or an engineer building systems that must reason about interventions, the Mixtape provides an accessible bridge from theory to code.