Working Paper: NBER ID: w28885
Authors: Eli Benmichael; Avi Feller; Jesse Rothstein
Abstract: The synthetic control method (SCM) is a popular approach for estimating the impact of a treatment on a single unit in panel data settings. The "synthetic control" is a weighted average of control units that balances the treated unit's pre-treatment outcomes and other covariates as closely as possible. A critical feature of the original proposal is to use SCM only when the fit on pre-treatment outcomes is excellent. We propose Augmented SCM as an extension of SCM to settings where such pre-treatment fit is infeasible. Analogous to bias correction for inexact matching, Augmented SCM uses an outcome model to estimate the bias due to imperfect pre-treatment fit and then de-biases the original SCM estimate. Our main proposal, which uses ridge regression as the outcome model, directly controls pre-treatment fit while minimizing extrapolation from the convex hull. This estimator can also be expressed as a solution to a modified synthetic controls problem that allows negative weights on some donor units. We bound the estimation error of this approach under different data generating processes, including a linear factor model, and show how regularization helps to avoid over-fitting to noise. We demonstrate gains from Augmented SCM with extensive simulation studies and apply this framework to estimate the impact of the 2012 Kansas tax cuts on economic growth. We implement the proposed method in the new augsynth R package.
Keywords: Synthetic Control Method; Causal Inference; Ridge Regression
JEL Codes: C21; C23; E62; H71
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| ASCM (F53) | viable alternative when traditional SCM fails to achieve a good pretreatment fit (L99) |
| ASCM (F53) | correct for bias due to imperfect pretreatment fit (C22) |
| good pretreatment fit (C52) | small bias estimated by ASCM (C20) |
| poor pretreatment fit (C52) | significant divergence of ASCM from SCM estimates (C59) |
| ridge ASCM estimator (C51) | weighted average of control unit outcomes (C29) |
| ASCM (F53) | outperform alternative estimators in terms of bias and root mean squared error (C51) |
| ASCM (F53) | estimate the impact of the 2012 Kansas tax cuts on economic growth (H29) |
| choice of hyperparameter in ridge ASCM (C52) | balance bias and variance (C46) |