Data-Driven Covariate Selection for Nonparametric and Cycle-Agnostic Causal Effect Estimation

TL;DR

This paper demonstrates that a local, data-driven covariate selection method based on conditional independence is valid for both cyclic and acyclic causal models, enhancing causal effect estimation in complex real-world data.

Contribution

It extends the theoretical guarantees of a covariate selection method to cyclic causal models, unifying cycle-agnostic causal effect estimation.

Findings

Method is sound and complete in cyclic models, not just acyclic.

Empirical validation shows reliable performance on synthetic cyclic data.

Invariance under $\sigma$-acyclification underpins the theoretical extension.

Abstract

Estimating causal effects from observational data requires identifying valid adjustment sets. This task is especially challenging in realistic settings where latent confounding and feedback loops are present. Existing approaches typically assume acyclicity or rely on global causal structure learning, limiting applicability and computational efficiency. In this work, we study a local, data-driven method for covariate selection based on conditional independence information. While this method is known to be sound and complete in acyclic causal models, its validity in the presence of cycles has remained unclear. Our main contribution is to show that these guarantees extend to cyclic causal models. In particular, our result relies on the invariance of conditional independence assertions under $\sigma$-acyclification. These findings establish a unified, cycle-agnostic perspective on covariate selection and causal effect estimation, showing that the method applies across cyclic and acyclic settings without modification. Empirically, we validate this on extensive synthetic data, showing reliable performance in cyclic causal models.