Sequential deviation minimization for improving the estimation in synthetic control methods

The Synthetic Control Method (SCM) is widely used to evaluate causal effects of policy interventions in comparative case studies. However, conventional SCM often suffers from unstable weight estimation and lacks systematic inference procedures, which can undermine empirical reliability. This study proposes a new framework based on a Sequential Deviation Minimization (SDM) strategy to enhance both estimation and diagnostic assessment while improving practical feasibility. The SDM approach introduces a stable and interpretable weight estimation procedure, provides a diagnostic device for assessing covariate contributions through scale adjustment, and develops a cross-validation-based framework for evaluating prediction errors and intervention effects. The proposed method is illustrated through an empirical analysis of the impact of the 2011 Great East Japan Earthquake on registered dog numbers in a severely affected region, serving as an indirect indicator of social disruption. Simulation experiments further demonstrate the robustness and accuracy of the SDM framework. Overall, SDM offers a stable, interpretable, and extensible enhancement to SCM for applied causal analysis.