36 Pages Posted: 28 Nov 2015 Last revised: 24 Oct 2016
Date Written: October 21, 2016
This paper presents a novel algorithmic approach to the problems of inconsistency and bias in legal decision making. First, we propose a new tool for reducing inconsistency: "Synthetic Crowdsourcing Models" (“SCMs”) built with machine learning methods. By providing judges with recommendations generated from statistical models of themselves, such models can help those judges make better and more consistent decisions. To illustrate these advantages, we build an SCM of release decisions for the California Board of Parole Hearings. Second, we describe a means to address systematic biases that are embedded in an algorithm (e.g., disparate racial treatment). We argue for making direct changes to algorithmic output based on explicit estimates of bias. Most commentators concerned with embedded biases have focused on constructing algorithms without the use of bias-inducing variables. Given the complex ways that variables may correlate and interact, that approach is both practically difficult and harmful to predictive power. In contrast, our two-step approach can address bias with minimal sacrifice of predictive performance.
Keywords: Judicial Decision Making, Machine Learning, Prediction, Forecasting, Judgmental Bootstrapping, Policy Capture, Inconsistency, Criminal Justice, Judgment and Decision Making, Parole
Suggested Citation: Suggested Citation
Laqueur, Hannah and Copus, Ryan, Synthetic Crowdsourcing: A Machine-Learning Approach to the Problems of Inconsistency and Bias in Adjudication (October 21, 2016). Available at SSRN: https://ssrn.com/abstract=2694326 or http://dx.doi.org/10.2139/ssrn.2694326