Algorithmic Fairness and Structural Injustice: Insights from Feminist Political Philosophy

Current fairness metrics miss structural injustices: The paper argues that algorithmic fairness research relies primarily on mathematical metrics rooted in distributive justice (who gets what), but this approach fails to address structural injustices—systemic patterns of disadvantage that harm vulnerable groups beyond just unequal distribution of resources.

Feminist philosophy offers overlooked solutions: Feminist political philosophers, particularly Iris Marion Young, have developed critiques of distributive justice that could strengthen algorithmic fairness work. The paper demonstrates that these perspectives have been largely neglected in the algorithmic fairness field despite their relevance to real-world harms.

High-stakes decisions require broader accountability: Algorithms used in consequential domains like bail recommendations, medical resource allocation, and mortgage lending can perpetuate structural inequalities that mathematical fairness metrics alone cannot detect or prevent, suggesting current approaches are insufficient for protecting vulnerable populations.

"Responsible algorithmic fairness" is needed: The paper proposes moving beyond traditional fairness metrics toward a "responsible algorithmic fairness" paradigm that accounts for structural injustices, implying that meaningful progress requires rethinking how we define and measure fairness in algorithmic systems.