Fairness in Algorithmic Decision Making: An Excursion Through the Lens of Causality

Abstract

As virtually all aspects of our lives are increasingly impacted by algorithmic decision making systems, it is incumbent upon us as a society to ensure such systems do not become instruments of unfair discrimination on the basis of gender, race, ethnicity, religion, etc. We consider the problem of determining whether the decisions made by such systems are discriminatory, through the lens of causal models. We introduce two definitions of group fairness grounded in causality: fair on average causal effect (FACE), and fair on average causal effect on the treated (FACT). We use the Rubin-Neyman potential outcomes framework for the analysis of cause-effect relationships to robustly estimate FACE and FACT. We demonstrate the effectiveness of our proposed approach on synthetic data. Our analyses of two real-world data sets, the Adult income data set from the UCI repository (with gender as the protected attribute), and the NYC Stop and Frisk data set (with race as the protected attribute), show that the evidence of discrimination obtained by FACE and FACT, or lack thereof, is often in agreement with the findings from other studies. We further show that FACT, being somewhat more nuanced compared to FACE, can yield findings of discrimination that differ from those obtained using FACE.

Secure and Robust Machine Learning for Healthcare: A Survey

“…In predictive healthcare, the absence of causal relation can raise questions about the conclusions that can be drawn from outcomes of DL models. Furthermore, fairness in decision making can better be enforced through the lens of causal reasoning [117], [118]. The estimation of the causal effect of some variable(s) on a target output (e.g., target class in multi-class classification problem) is important to ensure fair predictions.…”

On Consequentialism and Fairness

“…Beyond the criteria mentioned in section 5, numerous other fairness metrics have been proposed, such as procedural fairness (Grgić-Hlača et al, 2016 ) and causal effects (Madras et al, 2018 ; Khademi et al, 2019 ). Meanwhile, other papers have emphasized that simply satisfying a particular definition of fairness is no guarantee of the broader outcomes people care about, such as justice (Hu and Chen, 2018b ).…”

Design Principles for Signal Detection in Modern Job Application Systems: Identifying Fabricated Qualifications

“…Further, the system evaluation design mitigated machine learning bias by using observed behaviors (the observed act of deception and observed verbal/nonverbal behaviors during responses) to train classifiers. As the system expands to predict constructs beyond deception, such as interview performance or cultural fit, future research must incorporate methods to alleviate bias (e.g., gender or race discrimination) caused by training data [41]. This bias is the result of "garbage-in, garbage-out" effects; using training data derived from subjective human interpretation naturally includes the inherent biases of humans.…”