Part IV · Quasi-Experiments for Public Policy

Module 12. Synthetic Control Methods

Build a weighted, data-driven counterfactual for a single treated region and test it with placebos.

Environmental policy: provincial cap-and-trade; Texas prison expansion about 6 hours3 lessons

Module progress0/3 lessons · 0%

Lessons

Synthetic Control by Convex Optimisation

· Explain why a single treated unit needs a weighted, data-driven control group
· Solve for synthetic control weights with scipy (non-negative, summing to one)
· Evaluate pre-treatment fit and read the post-treatment gap as the effect

120 min 5 Colab cellsVideo coming soon

Inference for Synthetic Controls: Placebos in Space and Time

· Run in-space placebo tests by re-assigning treatment to every donor
· Compute post/pre RMSPE ratios and a permutation p-value
· Run in-time placebos with a fake treatment date

120 min 6 Colab cellsVideo coming soon

Synthetic Control on Real Data: Texas Prison Expansion

· Prepare a state-year panel for synthetic control
· Match on pre-treatment outcomes and predictors, with predictor weights
· Estimate the effect of Texas's 1993 prison-capacity expansion on Black male incarceration

120 min 6 Colab cellsVideo coming soon