mlsanityData to 7 Oct 2026

Why Fair Labels Can Yield Unfair Predictions: Graphical Conditions for Introduced Unfairness

arXiv 2202.10816 ↗First posted 22 Feb 2022cs.LGcited by 10

Not on the map. Why?

Abstract (arXiv)

In addition to reproducing discriminatory relationships in the training data, machine learning systems can also introduce or amplify discriminatory effects. We refer to this as introduced unfairness, and investigate the conditions under which it may arise. To this end, we propose introduced total variation as a measure of introduced unfairness, and establish graphical conditions under which it may be incentivised to occur. These criteria imply that adding the sensitive attribute as a feature removes the incentive for introduced variation under well-behaved loss functions. Additionally, taking a causal perspective, introduced path-specific effects shed light on the issue of when specific paths should be considered fair.

Key facts. Cited by 10 papers in the map (data through 7 Oct 2026). Builds on 19 of its 44 references that are in the map. First posted on arXiv 22 Feb 2022.

Builds on

19 of 44 references.

References · 19

  1. 2017Counterfactual Fairnesscited by 802
  2. 2016Inherent Trade-Offs in the Fair Determination of Risk Scorescited by 582
  3. 2017Men Also Like Shopping: Reducing Gender Bias Amplification using Corpus-level Constraintscited by 420
  4. 2013Direct and Indirect Effectscited by 276
  5. 2020Fairness in Machine Learningcited by 194
  6. 2017Avoiding Discrimination through Causal Reasoningcited by 191
  7. 2018Path-Specific Counterfactual Fairnesscited by 179
  8. 2017Fair Inference On Outcomescited by 143
  9. 2019Towards a Critical Race Methodology in Algorithmic Fairnesscited by 124
  10. 2016A causal framework for discovering and removing direct and indirect discriminationcited by 71
  11. 2013Causal Networks: Semantics and Expressivenesscited by 67
  12. 2019PC-Fairness: A Unified Framework for Measuring Causality-based Fairnesscited by 67
7 more
  1. 2018Causal Reasoning for Algorithmic Fairnesscited by 51
  2. 2021Agent Incentives: A Causal Perspectivecited by 51
  3. 2018Learning Optimal Fair Policiescited by 49
  4. 2019Causal Modeling for Fairness in Dynamical Systemscited by 46
  5. 2019Fair Data Adaptation with Quantile Preservationcited by 29
  6. 2019A Causal Bayesian Networks Viewpoint on Fairnesscited by 19
  7. 2021Gradual (In)Compatibility of Fairness Criteriacited by 8

Built on by

Cited by year

Citing papers per year: 2022: 5; 2023: 4; 2024: 1.

10 papers.

  1. 2022Trustworthy Graph Neural Networks: Aspects, Methods and Trendscited by 88
  2. 2022Discovering Agentscited by 28
  3. 2023Reasoning about Causality in Gamescited by 20
  4. 2022What-is and How-to for Fairness in Machine Learning: A Survey, Reflection, and Perspectivecited by 18
  5. 2023Causal Fairness under Unobserved Confounding: A Neural Sensitivity Frameworkcited by 10
  6. 2022Consistent Range Approximation for Fair Predictive Modelingcited by 7
  7. 2022A Complete Criterion for Value of Information in Soluble Influence Diagramscited by 5
  8. 2023On Imperfect Recall in Multi-Agent Influence Diagramscited by 4
  9. 2023Unfair Utilities and First Steps Towards Improving Themcited by 0
  10. 2024Characterising Interventions in Causal Gamescited by 0