Principal Fairness for Human and Algorithmic Decision-Making
This addresses fairness issues in human and algorithmic decision-making, offering a novel causal perspective, but it is incremental as it builds on existing causal concepts without broad empirical validation.
The paper tackles the problem of fairness in decision-making by introducing principal fairness, a new criterion based on causal inference that prohibits discrimination among individuals similarly affected by the decision, differing from existing statistical and causality-based fairness definitions.
Using the concept of principal stratification from the causal inference literature, we introduce a new notion of fairness, called principal fairness, for human and algorithmic decision-making. The key idea is that one should not discriminate among individuals who would be similarly affected by the decision. Unlike the existing statistical definitions of fairness, principal fairness explicitly accounts for the fact that individuals can be impacted by the decision. Furthermore, we explain how principal fairness differs from the existing causality-based fairness criteria. In contrast to the counterfactual fairness criteria, for example, principal fairness considers the effects of decision in question rather than those of protected attributes of interest. We briefly discuss how to approach empirical evaluation and policy learning problems under the proposed principal fairness criterion.