Masked Unfairness: Hiding Causality within Zero ATE
This work is significant for policymakers and practitioners in AI ethics, revealing a critical flaw in ATE-based fairness regulation that can allow substantial discrimination to persist undetected.
This paper introduces the "causal masking problem," a linear program designed to optimize an objective (e.g., profit) while maintaining a zero average treatment effect (ATE) between a protected attribute and a decision. The authors demonstrate that such optimization can lead to significant unequal treatment despite a zero ATE, highlighting that confounding drives the divergence between true and causally masked fairness.
Recent work has proposed powerful frameworks, rooted in causal theory, to quantify fairness. Causal inference has primarily emphasized the detection of \emph{average} treatment effects (ATEs), and subsequent notions of fairness have inherited this focus. In this paper, we build on previous concerns about regulation based on averages. In particular, we formulate the "causal masking problem" as a linear program that optimizes an alternative objective, such as maximizing profit or minimizing crime, while retaining a zero ATE (i.e., the ATE between a protected attribute and a decision). By studying the capabilities and limitations of causal masking, we show that optimization under ATE-based regulation may induce significant unequal treatment. We demonstrate that the divergence between true and causally masked fairness is driven by confounding, underscoring the importance of full conditional-independence testing when assessing fairness. Finally, we discuss statistical and information-theoretic limitations that make causally masked solutions very difficult to detect, allowing them to persist for long periods. These results argue that we must regulate fairness at the model-level, rather than at the decision level.