Escaping the Impossibility of Fairness: From Formal to Substantive Algorithmic Fairness

Abstract Efforts to promote equitable public policy with algorithms appear be fundamentally constrained by the “impossibility of fairness” (an incompatibility between mathematical definitions fairness). This technical limitation raises a central question about algorithmic fairness: How can computer scientists and policymakers support reforms algorithms? In this article, I argue that promoting justice requires reforming methodology fairness. First, diagnose problems current for fairness, which call “formal fairness.” Because formal fairness restricts analysis isolated decision-making procedures, it leads impossibility models exacerbate oppression despite appearing “fair.” Second, draw on theories substantive equality from law philosophy propose an alternative methodology, “substantive takes more expansive scope analysis, enables escape provides rigorous guide alleviating injustice algorithms. sum, presents new direction away “fair” toward evaluations whether how in practice.