Meaning
An econometric method constructs a counterfactual baseline by weighting unexposed units to approximate an exposed entity before an intervention. Synthetic control operates on a cardinal scale to measure treatment effects by comparing post-intervention trajectories against a weighted combination of control groups. This quantitative design minimizes interpolation bias through constrained optimization while avoiding extrapolation beyond observed domains.
Estimation Weights
Donor pool members receive fractional coefficients summing to unity under non-negativity constraints. Optimization algorithms minimize pre-intervention prediction error between the treated unit and the weighted aggregate across selected predictor variables. Restricted optimization vectors prevent extreme extrapolation during counterfactual construction.
Distance metrics evaluate pre-intervention fit quality across multiple temporal periods to ensure baseline stability.
Treatment Gap
Divergence calculations subtract the synthetic counterfactual value from the observed outcome of the treated unit after policy implementation. Temporal discrepancy vectors quantify treatment effects without relying on parallel trend assumptions required by traditional difference-in-differences models. Cumulative divergence over post-intervention periods reveals dynamic intervention responses over extended time horizons.
Donor Selection
Candidate control units undergo strict pre-screening to eliminate units affected by anticipation effects or unobserved shocks. Statistical restrictions exclude control entities showing extreme divergence from the pre-intervention trajectory of the treated unit. Variable weighting procedures assign zero coefficients to irrelevant donor characteristics during optimization iterations.
Data quality checks verify structural stability within the selected donor pool before counterfactual generation proceeds.
