Evaluating Racial Fairness in Credit Granting Models

• Vinicius Martins Goulart UNIFESP
• Lilian Berton UNIFESP

DOI: https://doi.org/10.5753/eniac.2026.31049

Resumo

Automated credit models risk propagating socioeconomic disparities. This paper evaluates the trade-off between predictive accuracy and algorithmic fairness in Decision Trees, Random Forests, and XGBoost using four frameworks: (i) Baseline, (ii) Fairness through Blindness, (iii) Counterfactual Fairness, and (iv) In-processing Mitigation (Exponentiated Gradient). Results reveal that tree-based models easily exploit proxy variables under blind conditions, rendering simple race removal ineffective. Conversely, in-processing mitigation strictly enforces demographic parity, actively mapping the critical tradeoffs between financial risk discrimination and algorithmic equity.

Referências

Agarwal, A., Beygelzimer, A., Dudík, M., Langford, J., and Wallach, H. (2018). A reductions approach to fair classification. In International Conference on Machine Learning, pages 60–69. PMLR.

Barocas, S. and Selbst, A. D. (2016). Big data’s disparate impact. California Law Review, 104:671.

Breiman, L. (2001). Random forests. Machine Learning, 45:5–32.

Chen, T. and Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, pages 785–794.

Cornacchia, G., Anelli, V. W., Narducci, F., Ragone, A., and Di Sciascio, E. (2023). Counterfactual fair opportunity: Measuring decision model fairness with counterfactual reasoning. Technical Report arXiv:2302.08158, arXiv.

Das, S., Stanton, R., and Wallace, N. (2023). Algorithmic fairness. Annual Review of Financial Economics, 15:565–593.

dos Santos Porphirio, J., dos Santos, D. J., Azevedo, S., and Berton, L. (2025). Analyzing data bias through data augmentation: A case study in financial data. In Encontro Nacional de Inteligência Artificial e Computacional (ENIAC), pages 950–961. SBC.

Hlongwane, R., Ramaboa, K. K. K. M., and Mongwe, W. (2024). Enhancing credit scoring accuracy with a comprehensive evaluation of alternative data. PLOS ONE, 19(5):e0303566.

Ivanov, A. (2026). Fairness calibration in credit scoring via counterfactual perturbation and group-wise regularization. Technical report, Preprints.org.

Kozodoi, N., Jacob, J., and Lessmann, S. (2022). Fairness in credit scoring: An assessment of optimization algorithms and resource allocation. Expert Systems with Applications, 207:117951.

Kusner, M. J., Loftus, J., Russell, C., and Silva, R. (2017). Counterfactual fairness. In Advances in Neural Information Processing Systems (Rules of NeurIPS), volume 30.

Mehrabi, N., Morstatter, F., Saxena, N., Lerman, K., and Galstyan, A. (2021). A survey on bias and fairness in machine learning. ACM Computing Surveys (CSUR), 54(6):1–35.

Pessach, D. and Shmueli, E. (2023). Algorithmic fairness. In Machine Learning for Data Science Handbook, pages 867–886. Springer International Publishing.

Rabonato, R. T. and Berton, L. (2025). A systematic review of fairness in machine learning. AI and Ethics, 5(3):1943–1954.

Santos, K. L. d. (2026). Parcelado: dinâmicas de consumo na periferia. Editora Fósforo, São Paulo.

SD, N. (2026). Structural gender bias in credit scoring: Proxy leakage. Technical Report arXiv:2601.18342, arXiv.

Wecchi, M. and Berton, L. (2026). Algorithmic fairness and bias mitigation in financial artificial intelligence: scoping review. AI and Ethics, 6(4):409.