Synthetic Difference in Differences

@article{Arkhangelsky2018SyntheticDI,
  title={Synthetic Difference in Differences},
  author={D. Arkhangelsky and Susan Athey and David A. Hirshberg and Guido Imbens and Stefan Wager},
  journal={NBER Working Paper Series},
  year={2018}
}
We present a new perspective on the Synthetic Control (SC) method as a weighted least squares regression estimator with time fixed effects and unit weights. This perspective suggests a generalization with two way (both unit and time) fixed effects, and both unit and time weights, which can be interpreted as a unit and time weighted version of the standard Difference In Differences (DID) estimator. We find that this new Synthetic Difference In Differences (SDID) estimator has attractive… 
A Design-Based Perspective on Synthetic Control Methods
Since their introduction in Abadie and Gardeazabal (2003), Synthetic Control (SC) methods have quickly become one of the leading methods for estimating causal effects in observational studies in
The Augmented Synthetic Control Method
TLDR
The main proposal, which uses ridge regression as the outcome model, directly controls pre-treatment fit while minimizing extrapolation from the convex hull and bound the bias of this approach under a linear factor model and shows how regularization helps to avoid over-fitting to noise.
Synthetic Control with Time Varying Coefficients: A State Space Approach with Bayesian Shrinkage
Synthetic control methods are a popular tool for measuring the effects of policy interventions on a single treated unit. In practice, researchers create a counterfactual using a linear combination of
On the Use of Two-Way Fixed Effects Regression Models for Causal Inference with Panel Data
TLDR
It is proved that the multi-period difference-in-differences estimator is equivalent to the weighted 2FE estimator with some observations having negative weights, implying that in contrast to the popular belief, the 2FE estimation does not represent a design-based, nonparametric estimation strategy for causal inference.
Why Synthetic Control estimators are biased and what to do about it: Introducing Relaxed and Penalized Synthetic Controls
This paper extends the literature on the theoretical properties of synthetic controls to the case of non-linear generative models, showing that the synthetic control estimator is generally biased in
A Bayesian Alternative to Synthetic Control for Comparative Case Studies
TLDR
A Bayesian posterior predictive approach to Rubin's causal model, which allows researchers to make inferences about both individual and average treatment effects on treated observations based on the empirical posterior distributions of their counterfactuals, and a Bayesian shrinkage method for model searching and factor selection.
Combining Matching and Synthetic Control to Trade Off Biases from Extrapolation and Interpolation
The synthetic control method is widely used in comparative case studies to adjust for differences in pre-treatment characteristics. A major attraction of the method is that it limits extrapolation
Impact of Regression to the Mean on the Synthetic Control Method: Bias and Sensitivity Analysis.
TLDR
This work develops a novel correction for RTM bias that allows for valid inference and shows how this correction can be used in a sensitivity analysis to reanalyze data concerning the effects of California's Proposition 99, a large-scale tobacco control program, on statewide smoking rates.
A Penalized Synthetic Control Estimator for Disaggregated Data
TLDR
This work proposes a synthetic control estimator that penalizes the pairwise discrepancies between the characteristics of the treated units and the Characteristics of the units that contribute to their synthetic controls.
On the Assumptions of Synthetic Control Methods
Synthetic control (SC) methods have been widely applied to estimate the causal effect of large-scale interventions, e.g. the state-wide effect of a change in policy. The idea of synthetic controls is
...
...

References

SHOWING 1-10 OF 91 REFERENCES
Balancing, Regression, Difference-in-Differences and Synthetic Control Methods: A Synthesis
In a seminal paper Abadie et al (2010) develop the synthetic control procedure for estimating the effect of a treatment, in the presence of a single treated unit and a number of control units, with
Generalized Synthetic Control Method: Causal Inference with Interactive Fixed Effects Models
Difference-in-differences (DID) is commonly used for causal inference in time-series cross-sectional data. It requires the assumption that the average outcomes of treated and control units would have
The Augmented Synthetic Control Method
TLDR
The main proposal, which uses ridge regression as the outcome model, directly controls pre-treatment fit while minimizing extrapolation from the convex hull and bound the bias of this approach under a linear factor model and shows how regularization helps to avoid over-fitting to noise.
Synthetic Controls and Weighted Event Studies with Staggered Adoption
TLDR
This paper proposes "partially pooled" SCM weights that instead minimize both average and state-specific imbalance, and combines traditional fixed effects methods to obtain an augmented estimator that improves over both SCM weighting and fixed effects estimation alone.
Inference on average treatment effects in aggregate panel data settings
TLDR
A K-fold cross-fitting procedure to remove the bias induced by regularization is proposed and demonstrates an excellent performance in simulation experiments, and is taken to a data application, where it is re-evaluate the economic consequences of terrorism.
Robust Synthetic Control
TLDR
The condition under which the fundamental assumption in synthetic control-like approaches holds holds is established, i.e. when the linear relationship between the treatment unit and the donor pool prevails in both the pre- and post-intervention periods.
Revisiting Event Study Designs
A broad empirical literature uses "event study" research designs for treatment effect estimation, a setting in which all units in the panel receive treatment but at random times. We make four novel
Combining Matching and Synthetic Control to Trade Off Biases from Extrapolation and Interpolation
The synthetic control method is widely used in comparative case studies to adjust for differences in pre-treatment characteristics. A major attraction of the method is that it limits extrapolation
Difference-in-Differences with Multiple Time Periods
TLDR
This article considers identification and estimation of treatment effect parameters using DID with (i) multiple time periods, (ii) variation in treatment timing, and (iii) when the "parallel trends assumption" holds potentially only after conditioning on observed covariates.
Synthetic Control Method: Inference, Sensitivity Analysis and Confidence Sets
Abstract We extend the inference procedure for the synthetic control method in two ways. First, we propose parametric weights for the p-value that includes the equal weights benchmark of Abadie et
...
...