Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
119 tokens/sec
GPT-4o
56 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
6 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Performativity and Prospective Fairness (2310.08349v2)

Published 12 Oct 2023 in cs.CY

Abstract: Deploying an algorithmically informed policy is a significant intervention in the structure of society. As is increasingly acknowledged, predictive algorithms have performative effects: using them can shift the distribution of social outcomes away from the one on which the algorithms were trained. Algorithmic fairness research is usually motivated by the worry that these performative effects will exacerbate the structural inequalities that gave rise to the training data. However, standard retrospective fairness methodologies are ill-suited to predict these effects. They impose static fairness constraints that hold after the predictive algorithm is trained, but before it is deployed and, therefore, before performative effects have had a chance to kick in. However, satisfying static fairness criteria after training is not sufficient to avoid exacerbating inequality after deployment. Addressing the fundamental worry that motivates algorithmic fairness requires explicitly comparing the change in relevant structural inequalities before and after deployment. We propose a prospective methodology for estimating this post-deployment change from pre-deployment data and knowledge about the algorithmic policy. That requires a strategy for distinguishing between, and accounting for, different kinds of performative effects. In this paper, we focus on the algorithmic effect on the causally downstream outcome variable. Throughout, we are guided by an application from public administration: the use of algorithms to (1) predict who among the recently unemployed will stay unemployed for the long term and (2) targeting them with labor market programs. We illustrate our proposal by showing how to predict whether such policies will exacerbate gender inequalities in the labor market.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (59)
  1. The gender unemployment gap. Review of Economic Dynamics, 30:47–67, 2018. doi: 10.1016/j.red.2017.12.005.
  2. Algorithmic profiling of Job Seekers in Austria: How Austerity Politics Are Made Effective. Frontiers in Big Data, 3, 2020. doi: 10.3389/fdata.2020.00005.
  3. Machine bias: There’s software used across the country to predict future criminals. and it’s biased against blacks, 2016. URL https://www.propublica.org/article/machine-bias-risk-assessments-in-criminal-sentencing.
  4. Interventions over Predictions: Reframing the Ethical Debate for Actuarial Risk Assessment. In Sorelle A. Friedler and Christo Wilson, editors, Proceedings of the 1st Conference on Fairness, Accountability and Transparency, volume 81 of Proceedings of Machine Learning Research, pages 62–76. PMLR, 2018.
  5. Fairness and Machine Learning: Limitations and Opportunities. MIT Press, 2023.
  6. Fabian Beigang. On the Advantages of Distinguishing Between Predictive and Allocative Fairness in Algorithmic Decision-Making. Minds and Machines, 32(4):655–682, 2022.
  7. Improving fairness in criminal justice algorithmic risk assessments using optimal transport and conformal prediction sets. Sociological Methods & Research, 2021.
  8. Fair Risk Algorithms. Annual Review of Statistics and Its Application, 10(1):165–187, 2023. doi: 10.1146/annurev-statistics-033021-120649.
  9. A Systematic Review of the Gender Pay Gap and Factors That Predict It. Administration & Society, 49(1):65–104, 2016. doi: 10.1177/0095399716636928.
  10. Giuliano Bonoli. The Political Economy of Active Labor-Market Policy. Politics & Society, 38(4):435–457, 2010. doi: 10.1177/0032329210381235.
  11. Measurement invariance versus selection invariance: Is fair selection possible? Psychological Methods, 13(2):75–98, 2008. doi: 10.1037/1082-989x.13.2.75.
  12. Bundesagentur für Arbeit. Statistik der Bundesagentur für Arbeit Berichte: Blickpunkt Arbeitsmarkt –Die Arbeitsmarktsituation von Frauen und Männern. Nürnberg, May, 2023.
  13. Counterfactuals for the Future, 2022.
  14. What Works? A Meta Analysis of Recent Active Labor Market Program Evaluations. Journal of the European Economic Association, 16(3):894–931, 2018. doi: 10.1093/jeea/jvx028.
  15. Alexandra Chouldechova. Fair Prediction with Disparate Impact: A Study of Bias in Recidivism Prediction Instruments. Big Data, 5(2):153–163, 2017. doi: 10.1089/big.2016.0047.
  16. Priority to unemployed immigrants? A causal machine learning evaluation of training in Belgium. Labour Economics, 80:102306, 2023. doi: 10.1016/j.labeco.2022.102306.
  17. New public management and the rule of economic incentives: Australian welfare-to-work from job market signalling perspective. Public Management Review, 20(8):1186–1204, 2017. doi: 10.1080/14719037.2017.1346140.
  18. Counterfactual risk assessments, evaluation, and fairness. In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. ACM, 2020. doi: 10.1145/3351095.3372851.
  19. Causal modeling for fairness in dynamical systems. In Hal Daumé, III and Aarti Singh, editors, International Conference on Machine Learning, volume 119 of Proceedings of Machine Learning Research, pages 2185–2195. PMLR, 2020.
  20. Fairness Is Not Static: Deeper Understanding of Long Term Fairness via Simulation Studies. In Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency, FAT* ’20, page 525–534. ACM, 2020.
  21. Statistical profiling in public employment services. OECD Social, Employment and Migration Working Papers, (224), 2019. doi: 10.1787/b5e5f16e-en.
  22. Using Artificial Intelligence to classify Jobseekers: The Accuracy-Equity Trade-off. Journal of Social Policy, 50(2):367–385, 2020. doi: 10.1017/s0047279420000203.
  23. Fairness through awareness. In Proceedings of the 3rd Innovations in Theoretical Computer Science Conference. ACM, 2012. doi: 10.1145/2090236.2090255.
  24. Runaway Feedback Loops in Predictive Policing. Proceedings of Machine Learning Research, 81:1–12, 2018.
  25. Active labour market policies for the long-term unemployed: New evidence from causal machine learning, 2021.
  26. Ben Green. Escaping the Impossibility of Fairness: From Formal to Substantive Algorithmic Fairness. Philosophy & Technology, 35(4), 2022. doi: 10.1007/s13347-022-00584-6.
  27. A Short-term Intervention for Long-term Fairness in the Labor Market. In Proceedings of the 2018 World Wide Web Conference on World Wide Web. ACM Press, 2018. doi: 10.1145/3178876.3186044.
  28. Yaowei Hu and Lu Zhang. Achieving Long-Term Fairness in Sequential Decision Making. Proceedings of the AAAI Conference on Artificial Intelligence, 36(9):9549–9557, 2022.
  29. 50 Years of Test (Un)fairness. In Proceedings of the Conference on Fairness, Accountability, and Transparency. ACM, 2019. doi: 10.1145/3287560.3287600.
  30. Downstream Effects of Affirmative Action. In Proceedings of the Conference on Fairness, Accountability, and Transparency. ACM, 2019. doi: 10.1145/3287560.3287578.
  31. Fairness in Algorithmic Profiling: A German Case Study, 2021.
  32. Avoiding discrimination through causal reasoning. In I. Guyon, U. Von Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, editors, Advances in Neural Information Processing Systems, volume 30. Curran Associates, Inc., 2017.
  33. Equality-minded treatment choice. Journal of Business & Economic Statistics, 39(2):561–574, 2019. doi: 10.1080/07350015.2019.1688664.
  34. Inherent Trade-Offs in the Fair Determination of Risk Scores, 2016.
  35. The Child Penalty Atlas. National Bureau of Economic Research, 2023. doi: 10.3386/w31649.
  36. Heterogeneous Employment Effects of Job Search Programs. Journal of Human Resources, 57(2):597–636, 2020. doi: 10.3368/jhr.57.2.0718-9615r1.
  37. Distributive Justice and Fairness Metrics in Automated Decision-making: How Much Overlap Is There?, 2021.
  38. Counterfactual fairness. Advances in neural information processing systems, 30, 2017.
  39. Inequality-Averse Outcome-Based Matching. 2023. doi: 10.31219/osf.io/yrn4d.
  40. Are active labor market policies (cost-)effective in the long run? Evidence from the Netherlands. Empirical Economics, 60(4):1719–1746, 2019. doi: 10.1007/s00181-019-01812-3.
  41. What is the value added by caseworkers? Labour economics, 14(2):135–151, 2007.
  42. Algorithms and Decision-Making in the Public Sector. Annual Review of Law and Social Science, 17(1):309–334, 2021. doi: 10.1146/annurev-lawsocsci-041221-023808.
  43. Delayed Impact of Fair Machine Learning. In Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence. International Joint Conferences on Artificial Intelligence Organization, 2019. doi: 10.24963/ijcai.2019/862.
  44. To predict and serve? Significance, 13(5):14–19, 2016.
  45. Daniel Malinsky. Intervening on structure. Synthese, 195(5):2295–2312, 2018.
  46. Fair When Trained, Unfair When Deployed: Observable Fairness Measures are Unstable in Performative Prediction Settings, 2022.
  47. Algorithmic Fairness: Choices, Assumptions, and Definitions. Annual Review of Statistics and Its Application, 8(1):141–163, 2021. doi: 10.1146/annurev-statistics-042720-125902.
  48. The Nature of Long-Term Unemployment: Predictability, Heterogeneity and Selection. National Bureau of Economic Research, 2023. doi: 10.3386/w30979.
  49. A Classification of Feedback Loops and Their Relation to Biases in Automated Decision-Making Systems. In Equity and Access in Algorithms, Mechanisms, and Optimization. ACM, 2023. doi: 10.1145/3617694.3623227.
  50. Performative prediction. In International Conference on Machine Learning, pages 7599–7609. PMLR, 2020.
  51. "We Would Never Write That Down": Classifications of Unemployed and Data Challenges for AI. Proceedings of the ACM on Human-Computer Interaction, 5(CSCW1):1–26, 2021. doi: 10.1145/3449176.
  52. Modelling the long-term fairness dynamics of data-driven targeted help on job seekers. Scientific Reports, 13(1), 2023.
  53. Fairness and Abstraction in Sociotechnical Systems. In Proceedings of the Conference on Fairness, Accountability, and Transparency. ACM, 2019. doi: 10.1145/3287560.3287598.
  54. Correcting Underrepresentation and Intersectional Bias for Fair Classification, 2023.
  55. Fair Policy Targeting. Journal of the American Statistical Association, pages 1–14, 2023. doi: 10.1080/01621459.2022.2142591.
  56. The Effectivnesness of Active Labor Market Policies: A Meta-Analysis. Journal of Economic Surveys, 33(1):125–149, 2018. doi: 10.1111/joes.12269.
  57. Algorithmic Unfairness through the Lens of EU Non-Discrimination Law. In 2023 ACM Conference on Fairness, Accountability, and Transparency. ACM, 2023. doi: 10.1145/3593013.3594044.
  58. Fairness in Learning-Based Sequential Decision Algorithms: A Survey. In Handbook of Reinforcement Learning and Control, pages 525–555. Springer International Publishing, 2021. doi: 10.1007/978-3-030-60990-0_18.
  59. How do fair decisions fare in long-term qualification? Advances in Neural Information Processing Systems, pages 1–13, 2020.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (2)
  1. Sebastian Zezulka (2 papers)
  2. Konstantin Genin (3 papers)
Citations (2)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now