IAPA Christmas Reading 2020

Machine learning for public policy: Do we need to sacrifice accuracy to make models fair?

Growing applications of machine learning in policy settings have raised concern for fairness implications, especially for racial minorities, but little work has studied the practical trade-offs between fairness and accuracy in realworld settings. This empirical study fills this gap by investigating the accuracy cost of mitigating disparities across several policy settings, focusing on the common context of using machine learning to inform benefit allocation in resource-constrained programs across education, mental health, criminal justice, and housing safety. In each setting, explicitly focusing on achieving equity and using our proposed post-hoc disparity mitigation methods, fairness was substantially improved without sacrificing accuracy, challenging the commonly held assumption that reducing disparities either requires accepting an appreciable drop in accuracy or the development of novel, complex methods.

Read the paper from Machine Learning Department and Heinz College of Information Systems and Public Policy, Carnegie Mellon University


Ethics by Design: An organizational approach to responsible use of technology

Published by the World Economic Forum, December 2020

Ethics by Design: An organizational approach to the responsible use of technology integrates key insights from psychology and behavioural economics with findings from market research and interviews with leaders of international organizations. It aims to help shape organizational design decisions to prompt better and more routine ethical behaviours. The report promotes an approach that focuses on the environments that can lead ordinary people to engage in more ethical behaviours rather than relying solely on their personal character. The report outlines steps and recommendations for organizational design that have proven to be more effective than conventional approaches such as compliance training and financial compensation.

Republished in accordance with the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International Public License