A Latent Factor Panel Approach to Spatiotemporal Causal Inference

2025-09-13

AI Summary

To address severe bias in causal effect estimation arising from unmeasured spatiotemporal confounders—particularly their nonstationarity—this paper proposes a causal inference method based on a latent factor panel model. The method posits a factor confounding assumption, wherein shared latent factors jointly capture the influence of unmeasured confounders on both exposure and outcome; causal effects are point-identified by imposing structural constraints on spatiotemporal interference. Innovatively integrating panel modeling, spatiotemporal effect decomposition, and matrix dimensionality reduction, it relaxes restrictive assumptions of conventional approaches—namely, confounder stationarity or spatial smoothness. Simulation studies demonstrate that our method substantially reduces omitted-variable bias compared to spatial smoothing models and standard panel benchmarks. Empirically applied to estimate the impact of prenatal PM₂.₅ exposure on birth weight in California, it validates both efficacy and practical utility.

Technology Category

• Reasoning under Uncertainty: Causality
• Machine Learning: Causal Learning
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Algorithms and analysis for incomplete, noisy, or partially observed Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Unmeasured confounding can severely bias causal effect estimates from spatiotemporal observational data, especially when the confounders do not vary smoothly in time and space. In this work, we develop a method for addressing unmeasured confounding in spatiotemporal contexts by building on models from the panel data literature and methods in multivariate causal inference. Our method is based on a factor confounding assumption, which posits that effects of unmeasured confounders on exposures and outcomes can be captured by a shared latent factor model. Factor confounding is sufficient to partially identify causal effects, even when there is interference between units. Additional assumptions that limit the degree of spatiotemporal interference, reasonable in most applications, are sufficient to point identify the effects. Simulation studies demonstrate that the proposed approach can substantially reduce omitted variable bias relative to other spatial smoothing and panel data baselines. We illustrate our method in a case study of the effect of prenatal PM2.5 exposure on birth weight in California.

Problem

• Addressing unmeasured confounding in spatiotemporal observational data
• Developing latent factor model for causal inference with interference
• Reducing bias in estimating PM2.5 exposure effects on birth weight

Innovation

• Latent factor panel approach
• Spatiotemporal causal inference
• Shared latent factor model