Fiche du document
16 février 2022
- 2202.08426
- Econometrica 91 (2023) 465-491
- doi: 10.3982/ECTA20720
Sujets proches
Clinical reports Cases, clinical reports, statistics Statistical data Cases, clinical reports, statistics Informatics Units Measurement, Units of Statistical data Statistical analysis Statistical methods Statistical science Linear regression Regression modeling Analysis, Regression Predictions Papers Philosophy--Methodology Learning, Machine Learning process Economic theory Political economy Scholarship Learning Use of computing informatisationCiter ce document
Jiafeng Chen, « Synthetic Control As Online Linear Regression », arXiv - économie, ID : 10.3982/ECTA20720
Métriques
Partage / Export
Résumé
This paper notes a simple connection between synthetic control and online learning. Specifically, we recognize synthetic control as an instance of Follow-The-Leader (FTL). Standard results in online convex optimization then imply that, even when outcomes are chosen by an adversary, synthetic control predictions of counterfactual outcomes for the treated unit perform almost as well as an oracle weighted average of control units' outcomes. Synthetic control on differenced data performs almost as well as oracle weighted difference-in-differences, potentially making it an attractive choice in practice. We argue that this observation further supports the use of synthetic control estimators in comparative case studies.