Fast Threshold Tests for Detecting Discrimination

Emma Pierson, Sam Corbett-Davies, Sharad Goel
Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, PMLR 84:96-105, 2018.

Abstract

Threshold tests have recently been proposed as a useful method for detecting bias in lending, hiring, and policing decisions. For example, in the case of credit extensions, these tests aim to estimate the bar for granting loans to white and minority applicants, with a higher inferred threshold for minorities indicative of discrimination. This technique, however, requires fitting a complex Bayesian latent variable model for which inference is often computationally challenging. Here we develop a method for fitting threshold tests that is two orders of magnitude faster than the existing approach, reducing computation from hours to minutes. To achieve these performance gains, we introduce and analyze a flexible family of probability distributions on the interval [0, 1] – which we call discriminant distributions – that is computationally efficient to work with. We demonstrate our technique by analyzing 2.7 million police stops of pedestrians in New York City.

Cite this Paper

BibTeX
@InProceedings{pmlr-v84-pierson18a, title = {Fast Threshold Tests for Detecting Discrimination}, author = {Pierson, Emma and Corbett-Davies, Sam and Goel, Sharad}, booktitle = {Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics}, pages = {96--105}, year = {2018}, editor = {Storkey, Amos and Perez-Cruz, Fernando}, volume = {84}, series = {Proceedings of Machine Learning Research}, month = {09--11 Apr}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v84/pierson18a/pierson18a.pdf}, url = {https://proceedings.mlr.press/v84/pierson18a.html}, abstract = {Threshold tests have recently been proposed as a useful method for detecting bias in lending, hiring, and policing decisions. For example, in the case of credit extensions, these tests aim to estimate the bar for granting loans to white and minority applicants, with a higher inferred threshold for minorities indicative of discrimination. This technique, however, requires fitting a complex Bayesian latent variable model for which inference is often computationally challenging. Here we develop a method for fitting threshold tests that is two orders of magnitude faster than the existing approach, reducing computation from hours to minutes. To achieve these performance gains, we introduce and analyze a flexible family of probability distributions on the interval [0, 1] – which we call discriminant distributions – that is computationally efficient to work with. We demonstrate our technique by analyzing 2.7 million police stops of pedestrians in New York City.} }
Endnote
%0 Conference Paper %T Fast Threshold Tests for Detecting Discrimination %A Emma Pierson %A Sam Corbett-Davies %A Sharad Goel %B Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2018 %E Amos Storkey %E Fernando Perez-Cruz %F pmlr-v84-pierson18a %I PMLR %P 96--105 %U https://proceedings.mlr.press/v84/pierson18a.html %V 84 %X Threshold tests have recently been proposed as a useful method for detecting bias in lending, hiring, and policing decisions. For example, in the case of credit extensions, these tests aim to estimate the bar for granting loans to white and minority applicants, with a higher inferred threshold for minorities indicative of discrimination. This technique, however, requires fitting a complex Bayesian latent variable model for which inference is often computationally challenging. Here we develop a method for fitting threshold tests that is two orders of magnitude faster than the existing approach, reducing computation from hours to minutes. To achieve these performance gains, we introduce and analyze a flexible family of probability distributions on the interval [0, 1] – which we call discriminant distributions – that is computationally efficient to work with. We demonstrate our technique by analyzing 2.7 million police stops of pedestrians in New York City.
APA
Pierson, E., Corbett-Davies, S. & Goel, S.. (2018). Fast Threshold Tests for Detecting Discrimination. Proceedings of the Twenty-First International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 84:96-105 Available from https://proceedings.mlr.press/v84/pierson18a.html.

Related Material