mlsanityData to 7 Oct 2026

The Fragility of Fairness: Causal Sensitivity Analysis for Fair Machine Learning

arXiv 2410.09600 ↗First posted 12 Oct 2024cs.LG cs.CYcited by 8

Not on the map. Why?

Abstract (arXiv)

Fairness metrics are a core tool in the fair machine learning literature (FairML), used to determine that ML models are, in some sense, "fair". Real-world data, however, are typically plagued by various measurement biases and other violated assumptions, which can render fairness assessments meaningless. We adapt tools from causal sensitivity analysis to the FairML context, providing a general framework which (1) accommodates effectively any combination of fairness metric and bias that can be posed in the "oblivious setting"; (2) allows researchers to investigate combinations of biases, resulting in non-linear sensitivity; and (3) enables flexible encoding of domain-specific constraints and assumptions. Employing this framework, we analyze the sensitivity of the most common parity metrics under 3 varieties of classifier across 14 canonical fairness datasets. Our analysis reveals the striking fragility of fairness assessments to even minor dataset biases. We show that causal sensitivity analysis provides a powerful and necessary toolkit for gauging the informativeness of parity metric evaluations. Our repository is available here: https://github.com/Jakefawkes/fragile_fair.

Key facts. Cited by 8 papers in the map (data through 7 Oct 2026). Builds on 39 of its 70 references that are in the map. First posted on arXiv 12 Oct 2024.

Builds on

39 of 70 references.

References · 39

  1. 2016Equality of Opportunity in Supervised Learningcited by 2,000
  2. 2018Model Cards for Model Reportingcited by 1,068
  3. 2017Counterfactual Fairnesscited by 802
  4. 2018A Reductions Approach to Fair Classificationcited by 585
  5. 2016Inherent Trade-Offs in the Fair Determination of Risk Scorescited by 582
  6. 2021Retiring Adult: New Datasets for Fair Machine Learningcited by 443
  7. 2019Measurement and Fairnesscited by 299
  8. 2021A survey on datasets for fairness-aware machine learningcited by 157
  9. 2018Fairness Under Unawareness: Assessing Disparity When Protected Class Is Unobservedcited by 94
  10. 2019Assessing Algorithmic Fairness with Unobserved Protected Class Using Data Combinationcited by 79
  11. 2020Fair Classification with Group-Dependent Label Noisecited by 67
  12. 2021Characterizing Fairness Over the Set of Good Models Under Selective Labelscited by 65
27 more
  1. 2021It's COMPASlicated: The Messy Relationship between RAI Datasets and Algorithmic Fairness Benchmarkscited by 61
  2. 2019Counterfactual Risk Assessments, Evaluation, and Fairnesscited by 59
  3. 2022Neural Causal Models for Counterfactual Identification and Estimationcited by 54
  4. 2021The Use and Misuse of Counterfactuals in Ethical Machine Learningcited by 52
  5. 2018Residual Unfairness in Fair Machine Learning from Prejudiced Datacited by 49
  6. 2020A Distributionally Robust Approach to Fair Classificationcited by 43
  7. 2022Causal Fairness Analysiscited by 42
  8. 2020Fairness in Risk Assessment Instruments: Post-Processing to Achieve Counterfactual Equalized Oddscited by 40
  9. 2021An Automated Approach to Causal Inference in Discrete Settingscited by 39
  10. 2020Fairness Evaluation in Presence of Biased Noisy Labelscited by 38
  11. 2020On Adversarial Bias and the Robustness of Fair Machine Learningcited by 36
  12. 2019The Sensitivity of Counterfactual Fairness to Unmeasured Confoundingcited by 35
  13. 2015Margins of discrete Bayesian networkscited by 26
  14. 2023Ground(less) Truth: A Causal Framework for Proxy Labels in Human-Algorithm Decision-Makingcited by 25
  15. 2021Assessing Fairness in the Presence of Missing Datacited by 24
  16. 2021Long Story Short: Omitted Variable Bias in Causal Machine Learningcited by 23
  17. 2021Fairness-Aware PAC Learning from Corrupted Datacited by 18
  18. 2020The Importance of Modeling Data Missingness in Algorithmic Fairness: A Causal Perspectivecited by 17
  19. 2022Selection, Ignorability and Challenges With Causal Fairnesscited by 16
  20. 2023Causal Fairness under Unobserved Confounding: A Neural Sensitivity Frameworkcited by 10
  21. 2013Instrumentality Tests Revisitedcited by 8
  22. 2013Quantifier Elimination for Statistical Problemscited by 8
  23. 2023Quantifying and mitigating the impact of label errors on model disparity metricscited by 7
  24. 2024Auditing Fairness under Unobserved Confoundingcited by 6
  25. 2020Fair Machine Learning Under Partial Compliancecited by 5
  26. 2023Results on Counterfactual Invariancecited by 4
  27. 2023Estimating the Likelihood of Arrest from Police Records in Presence of Unreported Crimescited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 4.

8 papers.

  1. 2025Bias Begins with Data: The FairGround Corpus for Robust and Reproducible Research on Algorithmic Fairnesscited by 2
  2. 2025Argumentative Debates for Transparent Bias Detection [Technical Report]cited by 1
  3. 2026Optimal Fair Aggregation of Crowdsourced Noisy Labels using Demographic Parity Constraintscited by 1
  4. 2025The influence of missing data mechanisms and simple missing data handling techniques on fairnesscited by 0
  5. 2025DiCE-Extended: A Robust Approach to Counterfactual Explanations in Machine Learningcited by 0
  6. 2026Falsifying Discriminant Validity of Predictive Algorithmscited by 0
  7. 2026Bayesian Sensitivity of Causal Inference Estimators under Evidence-Based Priorscited by 0
  8. 2026Long-term Fairness with Selective Labelscited by 0