NBER WORKING PAPER SERIES BALANCING , REGRESSION , DIFFERENCE-IN-DIFFERENCES AND SYNTHETIC CONTROL METHODS : A SYNTHESIS

By Nikolay Doudchenko, Guido W. Imbens. 2016

Subfields: Statistics and Probability, Causal Inference & Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Statistics and Probability

In Causal Inference & Discovery

Citations in the corpus

The State of Applied Econometrics - Causality and Policy Evaluation (2016, neutral): “In the simple case where there are no characteristics Zi and the cause Wi is a binary indicator, the assumption that the cause is completely randomly assigned leads to the conventional Eicker-Huber-White standard errors (Eicker [1967], Huber [1967], White [1980]).”
Synthetic Control Methods: Never Use All Pre-Intervention Outcomes Together With Covariates (2015, methodology): “Thus, when using regression-based predictor weights, the resulting synthetic control unit will again be determined by solely optimizing the fit with respect to the outcome’s pretreatment values, an approach which is called “constrained regression” by Doudchenko and Imbens (2016).”
Large Dimensional Latent Factor Modeling with Missing Observations and Applications to Causal Inference (2019, unknown): “…individual treatment effects for each i and t which is novel in the literature on causal inference for panel data.25 In many empirical applications, the object of interest is a unit’s average treatment effect over time (Abadie et al., 2010, 2015; Doudchenko and Imbens, 2016; Li and Bell, 2017).”
Econometric Methods for Program Evaluation (2018, background): “Hsiao et al. (2012) and Doudchenko and Imbens (2016) propose variants of the synthetic control framework that extrapolate outside the support of the columns of X0.”
Placebo Tests for Synthetic Controls (2017, methodology): “Doudchenko and Imbens (2016) consider the use of regularization methods such as best subset regression or LASSO to estimate the SC weights.”
Using Synthetic Controls: Feasibility, Data Requirements, and Methodological Aspects (2021, background): “Doudchenko and Imbens (2016) consider an estimator that fits all pre-treatment outcomes for the treated, with weights that may be negative and may not sum to one, and allow for a constant shift in the level of the synthetic control estimator.”
On the Properties of the Synthetic Control Estimator with Many Periods and Many Controls (2019, methodology): “1 remains valid if we consider a demeaned SC estimator, as proposed by Ferman and Pinto (2019), which is numerically the same as including a constant in the minimization problem (3), as proposed by Doudchenko and Imbens (2016).”
Revisiting the synthetic control estimator (2016, methodology): “This is equivalent to a generalization of the SC method suggested in Doudchenko and Imbens (2016) which includes an intercept parameter in the minimization problem to estimate the SC weights.”
The Impact of Machine Learning on Economics (2018, background): “Doudchenko and Imbens (2016) propose using regularized regression to determine the weights, with the penalty parameter selected via cross-validation.”
The Augmented Synthetic Control Method (2018, methodology): “Doudchenko and Imbens (2017) also suggest a version of residualization that is equivalent to excluding the lagged outcomes X from the outcome regression.”
What’s trending in difference-in-differences? A synthesis of the recent econometrics literature (2022, unknown): “Recent work has begun to try to combine insights from the two literatures (e.g., Arkhangelsky et al., 2021; Ben-Michael et al., 2021, 2022; Doudchenko and Imbens, 2016).”
Synthetic Difference In Differences (2018, unknown): “Many approaches to SC settings, including Abadie, Diamond, and Hainmueller [2010, 2015], Doudchenko and Imbens [2016], Xu [2017], Athey, Bayati, Doudchenko, Imbens, and Khosravi [2017], Carvalho, Masini, and Medeiros [2018], Li and Bell [2017], can be thought of as either focusing on constructing ba”