28.6 Other Advances

L. Sun, Ben-Michael, and Feller (2023) Using Multiple Outcomes to Improve SCM

• Common Weights Across Outcomes: This paper proposes using a single set of synthetic control weights across multiple outcomes, rather than estimating separate weights for each outcome.

• Reduced Bias with Low-Rank Factor Model: By balancing a vector or an index of outcomes, this approach yields lower bias bounds under a low-rank factor model, with further improvements as the number of outcomes increases.

• Evidence: re-analysis of the Flint water crisis’s impact on educational outcome.

References

Sun, Liyang, Eli Ben-Michael, and Avi Feller. 2023. “Using Multiple Outcomes to Improve the Synthetic Control Method.” arXiv Preprint arXiv:2311.16260.