Inference for an Algorithmic Fairness-Accuracy Frontier – Francesca Molinari скачать в хорошем качестве

Inference for an Algorithmic Fairness-Accuracy Frontier – Francesca Molinari

Computer Science Seminar Series November 11, 2025 “Inference for an Algorithmic Fairness-Accuracy Frontier” Francesca Molinari, Cornell University Algorithms are increasingly used to aid with high-stakes decision making. Yet, their predictive ability frequently exhibits systematic variation across population subgroups. To assess the trade-off between fairness and accuracy using finite data, we propose a debiased machine learning estimator for the fairness-accuracy frontier introduced by Liang, Lu, Mu, and Okumura (2024). We derive its asymptotic distribution and propose inference methods to test key hypotheses in the fairness literature, such as (i) whether excluding group identity from use in training the algorithm is optimal and (ii) whether there are less discriminatory alternatives to a given algorithm. In addition, we construct an estimator for the distance between a given algorithm and the fairest point on the frontier, and characterize its asymptotic distribution. Using Monte Carlo simulations, we evaluate the finite-sample performance of our inference methods. We apply our framework to reevaluate algorithms used in hospital care management and show that our approach yields alternative algorithms that lie on the fairness-accuracy frontier, offering improvements along both dimensions. Francesca Molinari is the H. T. Warshow and Robert Irving Warshow Professor of Economics and a professor of statistics and data science at Cornell University. She received her PhD from the Department of Economics at Northwestern University after obtaining a BA and master's in economics at the Università degli Studi di Torino, Italy. Her research interests are in econometrics, both theoretical and applied. Her theoretical work is concerned with the study of identification problems and with proposing new methods for statistical inference in partially identified models. In her applied work, she has focused primarily on the analysis of decision making under risk and uncertainty. She has worked on estimation of risk preferences using market-level data and on the analysis of individuals' probabilistic expectations using survey data. Molinari is a Fellow of both the Econometric Society and the International Association for Applied Econometrics. She is currently serving as joint managing editor at the Review of Economic Studies and as a member of the board of editors at the Journal of Economic Literature.