# Enforcing Delayed-Impact Fairness Guarantees

arXiv: https://arxiv.org/abs/2208.11744 · HTML: https://mlsanity.com/paper/2208.11744/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 39 references that are in the map.
- First posted on arXiv 24 Aug 2022.

## Builds on (most cited first, 18 of 18)

- [Equality of Opportunity in Supervised Learning](https://mlsanity.com/paper/1610.02413/) (2016)
- [Fair prediction with disparate impact: A study of bias in recidivism prediction instruments](https://mlsanity.com/paper/1610.07524/) (2016)
- [A Reductions Approach to Fair Classification](https://mlsanity.com/paper/1803.02453/) (2018)
- [Inherent Trade-Offs in the Fair Determination of Risk Scores](https://mlsanity.com/paper/1609.05807/) (2016)
- [Algorithmic decision making and the cost of fairness](https://mlsanity.com/paper/1701.08230/) (2017)
- [Delayed Impact of Fair Machine Learning](https://mlsanity.com/paper/1803.04383/) (2018)
- [Towards Long-term Fairness in Recommendation](https://mlsanity.com/paper/2101.03584/) (2021)
- [Being Optimistic to Be Conservative: Quickly Learning a CVaR Policy](https://mlsanity.com/paper/1911.01546/) (2019)
- [Algorithms for Fairness in Sequential Decision Making](https://mlsanity.com/paper/1901.08568/) (2019)
- [On the Long-term Impact of Algorithmic Decision Policies: Effort Unfairness and Feature Segregation through Social Learning](https://mlsanity.com/paper/1903.01209/) (2019)
- [A Short-term Intervention for Long-term Fairness in the Labor Market](https://mlsanity.com/paper/1712.00064/) (2017)
- [Universal Off-Policy Evaluation](https://mlsanity.com/paper/2104.12820/) (2021)
- [From Fair Decision Making to Social Equality](https://mlsanity.com/paper/1812.02952/) (2018)
- [Best Arm Identification for Contaminated Bandits](https://mlsanity.com/paper/1802.09514/) (2018)
- [Achieving Long-Term Fairness in Sequential Decision Making](https://mlsanity.com/paper/2204.01819/) (2022)
- [Removing biased data to improve fairness and accuracy](https://mlsanity.com/paper/2102.03054/) (2021)
- [Bandit Learning with Delayed Impact of Actions](https://mlsanity.com/paper/2002.10316/) (2020)
- [Welfare and Distributional Impacts of Fair Classification](https://mlsanity.com/paper/1807.01134/) (2018)

## Built on by (most cited first, 8 of 8)

- [Baldur: Whole-Proof Generation and Repair with Large Language Models](https://mlsanity.com/paper/2303.04910/) (2023)
- [FaiREE: Fair Classification with Finite-Sample and Distribution-Free Guarantee](https://mlsanity.com/paper/2211.15072/) (2022)
- [Long-Term Fairness Inquiries and Pursuits in Machine Learning: A Survey of Notions, Methods, and Challenges](https://mlsanity.com/paper/2406.06736/) (2024)
- [Reinforcement Learning from Human Feedback with High-Confidence Safety Constraints](https://mlsanity.com/paper/2506.08266/) (2025)
- [FairSense: Long-Term Fairness Analysis of ML-Enabled Systems](https://mlsanity.com/paper/2501.01665/) (2025)
- [Long-term Fairness For Real-time Decision Making: A Constrained Online Optimization Approach](https://mlsanity.com/paper/2401.02552/) (2024)
- [Happiness as a Measure of Fairness](https://mlsanity.com/paper/2511.01069/) (2025)
- [Algorithmic Fairness Over Time: Advances & Prospects](https://mlsanity.com/paper/2610.15015/) (2026)

## Abstract (arXiv)

Recent research has shown that seemingly fair machine learning models, when used to inform decisions that have an impact on peoples' lives or well-being (e.g., applications involving education, employment, and lending), can inadvertently increase social inequality in the long term. This is because prior fairness-aware algorithms only consider static fairness constraints, such as equal opportunity or demographic parity. However, enforcing constraints of this type may result in models that have negative long-term impact on disadvantaged individuals and communities. We introduce ELF (Enforcing Long-term Fairness), the first classification algorithm that provides high-confidence fairness guarantees in terms of long-term, or delayed, impact. We prove that the probability that ELF returns an unfair solution is less than a user-specified tolerance and that (under mild assumptions), given sufficient training data, ELF is able to find and return a fair solution if one exists. We show experimentally that our algorithm can successfully mitigate long-term unfairness.