From Control to Treatment: Causal Forecasting Beyond the Observed Panel
Abstract
Conventional panel data methods recover retrospective counterfactuals within the observed horizon. Many policy decisions, however, are prospective: how would an untreated unit evolve beyond the observed panel under an intervention it has not previously experienced? We develop a causal forecasting framework that combines the counterfactual logic of synthetic controls with multivariate time-series extrapolation. Under a latent factor model, we impose low-rank structure on the treated-state time factors. This yields the Two-Way Synthetic Forecasting estimator, which learns cross-unit weights from pre-treatment outcomes and temporal dynamics from treated donors' post-treatment trajectories. We establish pointwise identification, finite-sample error bounds, consistency, asymptotic normality, and inference for prespecified linear summaries over fixed forecast horizons. Simulations support the theory, and an application to the opening of NFL stadiums during the 2020 season illustrates how the method can inform a prospective policy change using only information available at the decision date.
Related
- Latent-driven Imputation and Forecasting for partially observed Time Series (LIFTS)
- Dynamic factor and double PCA models for partially observed survival curves: Forecasting demand in short-term rental markets
- Causal Fingerprints Reveal Hidden Mechanisms in Partially-Observed Dynamical Systems
- Compiling and forecasting implied volatility surfaces via survey sampling, tempered inverse-probability sequence learning, and causal activity clocks
- Partially Observed McKean--Vlasov Control with Vanishing Observation Intensities
BibTeX
@misc{shen2026from,
title = {{From Control to Treatment: Causal Forecasting Beyond the Observed Panel}},
author = {Dennis Shen},
howpublished = {arXiv},
year = {2026},
month = aug,
doi = {10.48550/arxiv.2606.18512},
eprint = {2606.18512},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2606.18512},
}
RIS
TY - UNPB
TI - From Control to Treatment: Causal Forecasting Beyond the Observed Panel
AU - Dennis Shen
PY - 2026
DA - 2026/08/31
PB - arXiv
DO - 10.48550/arxiv.2606.18512
UR - https://arxiv.org/abs/2606.18512
AB - Conventional panel data methods recover retrospective counterfactuals within the observed horizon. Many policy decisions, however, are prospective: how would an untreated unit evolve beyond the observed panel under an intervention it has not previously experienced? We develop a causal forecasting framework that combines the counterfactual logic of synthetic controls with multivariate time-series extrapolation. Under a latent factor model, we impose low-rank structure on the treated-state time factors. This yields the Two-Way Synthetic Forecasting estimator, which learns cross-unit weights from pre-treatment outcomes and temporal dynamics from treated donors' post-treatment trajectories. We establish pointwise identification, finite-sample error bounds, consistency, asymptotic normality, and inference for prespecified linear summaries over fixed forecast horizons. Simulations support the theory, and an application to the opening of NFL stadiums during the 2020 season illustrates how the method can inform a prospective policy change using only information available at the decision date.
ER -