Local Causal Discovery for Structural Evidence of Direct Discrimination
This work addresses the problem of robust fairness analysis for policy and algorithmic design in complex domains, offering an incremental improvement over existing causal discovery methods.
The paper tackles the challenge of identifying causal pathways of unfairness in algorithmic decision-making by introducing LD3, a local causal discovery method that uncovers structural evidence of direct discrimination. It shows that LD3 is more time-efficient and returns more plausible results than baselines, with execution times reduced by factors of 46 to 5870 on real-world data.
Identifying the causal pathways of unfairness is a critical objective for improving policy design and algorithmic decision-making. Prior work in causal fairness analysis often requires knowledge of the causal graph, hindering practical applications in complex or low-knowledge domains. Moreover, global discovery methods that learn causal structure from data can display unstable performance on finite samples, preventing robust fairness conclusions. To mitigate these challenges, we introduce local discovery for direct discrimination (LD3): a method that uncovers structural evidence of direct unfairness by identifying the causal parents of an outcome variable. LD3 performs a linear number of conditional independence tests relative to variable set size, and allows for latent confounding under the sufficient condition that all parents of the outcome are observed. We show that LD3 returns a valid adjustment set (VAS) under a new graphical criterion for the weighted controlled direct effect, a qualitative indicator of direct discrimination. LD3 limits unnecessary adjustment, providing interpretable VAS for assessing unfairness. We use LD3 to analyze causal fairness in two complex decision systems: criminal recidivism prediction and liver transplant allocation. LD3 was more time-efficient and returned more plausible results on real-world data than baselines, which took 46$\times$ to 5870$\times$ longer to execute.