A Systems Thinking Approach to Algorithmic Fairness

TL;DR

This paper introduces a systems thinking framework for modeling algorithmic fairness, integrating causal graphs, machine learning, and policy considerations to address bias and fairness trade-offs.

Contribution

It presents a novel approach that combines systems thinking with causal inference and machine learning to better understand and address algorithmic fairness issues.

Findings

Models bias using causal graphs within a systems framework

Links AI fairness to political and legal contexts

Supports policymakers in understanding fairness trade-offs

Abstract

Systems thinking provides us with a way to model the algorithmic fairness problem by allowing us to encode prior knowledge and assumptions about where we believe bias might exist in the data generating process. We can then encode these beliefs as a series of causal graphs, enabling us to link AI/ML systems to politics and the law. This allows us to combine techniques from machine learning, causal inference, and system dynamics in order to capture different emergent aspects of the fairness problem. We can use systems thinking to help policymakers on both sides of the political aisle to understand the complex trade-offs that exist from different types of fairness policies, providing a sociotechnical foundation for designing AI policy that is aligned to their political agendas and with society's shared democratic values.