Estimating Causal Effects Under Interference with a Probabilistic Exposure Model

The Fall 2022 SDS Seminar Series continues on Friday, September 23rd from 2:00 p.m. to 3:00 p.m. with Dr. Nathan Wikle  (Postdoctoral Scholar in the Department of Statistics and Data Science at the University of Texas at Austin). This event is in-person, but a virtual option will be available as well.

Title: Estimating Causal Effects Under Interference with a Probabilistic Exposure Model

Abstract: Causal inference for environmental data is often challenging due to the presence of interference: outcomes for observational units depend on some combination of local and nonlocal treatment. One common solution includes the specification of an exposure model, in which treatment assignments are mapped to an exposure value; causal estimands of the local and spillover effects of treatment are defined through contrasts of the local treatment assignment and the exposure value. Notably, the exposure model is often defined via a network structure, which is assumed to be fixed and known a priori.  However, in environmental settings, treatment interference is often dictated by complex, mechanistic processes that are both stochastic and poorly represented by a network.  In this work, we develop methods for causal inference with interference when deterministic exposure models cannot be assumed or are unknown. We offer a Bayesian model for the interference mapping which we combine with a flexible non-parametric outcome model to marginalize estimates of causal effects over uncertainty in the structure of interference. To illustrate our methodology, we analyze the effectiveness of air quality interventions at coal-fired power plants in reducing two adverse health outcomes in Texas — pediatric asthma emergency department visits and Medicare all-cause mortality.