Causal Bias Quantification for Continuous Treatments

TL;DR

This paper introduces a new method to quantify and analyze causal bias in continuous treatments using structural causal models, enabling better causal inference and regularization of predictive models.

Contribution

It extends causal effect definitions to continuous treatments, characterizes causal bias, and provides an efficient estimation method under certain conditions.

Findings

Bias expression is zero iff causal effect is identifiable.

Method effectively quantifies causal bias in various settings.

Causal regularization improves predictive model performance.

Abstract

We extend the definition of the marginal causal effect to the continuous treatment setting and develop a novel characterization of causal bias in the framework of structural causal models. We prove that our derived bias expression is zero if, and only if, the causal effect is identifiable via covariate adjustment. We show that under some restrictions on the structural equations, the causal bias can be estimated efficiently and allows for causal regularization of predictive probabilistic models. We demonstrate the effectiveness of our method for causal bias quantification in various settings where (not) controlling for certain covariates would introduce causal bias.