Partial Identification of the Average Treatment Effect with Stochastic Counterfactuals and Discordant Twins

TL;DR

This paper introduces a new method for partially identifying the average treatment effect using stochastic counterfactuals and discordant twin outcomes, providing bounds through constrained optimization in observational data.

Contribution

It develops a novel approach combining stochastic counterfactuals and discordant twins to improve partial identification of causal effects in observational studies.

Findings

Provides bounds for the average treatment effect

Uses discordant twin outcomes as evidence for randomness

Demonstrates applicability with three examples

Abstract

We develop a novel approach to partially identify causal estimands, such as the average treatment effect (ATE), from observational data. To better satisfy the stable unit treatment value assumption (SUTVA) we utilize stochastic counterfactuals within a propensity-prognosis model of the data generating process. For more precise identification we utilize knowledge of discordant twin outcomes as evidence for randomness in the data generating process. Our approach culminates with a constrained optimization problem; the solution gives upper and lower bounds for the ATE. We demonstrate the applicability of our introduced methodology with three example applications.