We propose a generalization of the synthetic control method to a multiple-outcome framework, which improves the reliability of treatment effect estimation. This is done by supplementing the conventional pre-treatment time dimension with the extra dimension of related outcomes in computing the...