Fair Without Leveling Down: A New Intersectional Fairness Definition
This addresses fairness for intersecting sensitive groups in ML, but is incremental as it refines definitions without introducing new methods.
The paper tackles intersectional group fairness in classification by proposing a new definition, α-Intersectional Fairness, to address shortcomings in existing measures, and shows that popular fair ML approaches fail to improve over a baseline, revealing a 'leveling down' effect where fairness gains come from degrading best group performance rather than improving worst group performance.
In this work, we consider the problem of intersectional group fairness in the classification setting, where the objective is to learn discrimination-free models in the presence of several intersecting sensitive groups. First, we illustrate various shortcomings of existing fairness measures commonly used to capture intersectional fairness. Then, we propose a new definition called the $α$-Intersectional Fairness, which combines the absolute and the relative performance across sensitive groups and can be seen as a generalization of the notion of differential fairness. We highlight several desirable properties of the proposed definition and analyze its relation to other fairness measures. Finally, we benchmark multiple popular in-processing fair machine learning approaches using our new fairness definition and show that they do not achieve any improvement over a simple baseline. Our results reveal that the increase in fairness measured by previous definitions hides a "leveling down" effect, i.e., degrading the best performance over groups rather than improving the worst one.