A General Synthetic Control Framework of Estimation and Inference

Project Outcomes Statement

The primary goal of this project was to develop synthetic control methods that enable applied researchers to evaluate the impact of policy interventions and other treatments in situations where traditional experimental approaches are not feasible. These methods are widely used in fields such as economics, public policy, and public health to assess interventions like new laws, public health initiatives, or external shocks such as natural disasters.

Synthetic control methods are particularly useful for estimating causal effects in cases where conducting a randomized controlled trial (RCT) is impractical, such as when interventions are implemented at a large scale (e.g., state-level regulations). A synthetic control represents an estimate of what the outcomes for treated units would have been in the absence of the intervention. This control is constructed as a weighted average of untreated units, chosen to closely match the characteristics of the treated units. By comparing actual outcomes to the synthetic control, researchers can more accurately assess the true impact of policies, events, or shocks.

The outcomes of this project include: (a) published, forthcoming, and submitted papers on synthetic control estimation and related methods, with two publications in the Journal of the American Statistical Association; (b) graduate student training through research assistantships, collaborations, and coursework, as well as seminars aimed at disseminating the research findings and providing training. For instance, the PI delivered tutorials on synthetic control estimation, now available online via the Chamberlain Online Econometrics Seminar and the NBER Methods Lectures.