A general approach for predicting the behavior of the Supreme Court of the United States

• Published in: PloS one Vol. 12; no. 4; p. e0174698
• Main Authors: Katz, Daniel Martin, Bommarito, 2nd, Michael J, Blackman, Josh
• Format: Journal Article
• Language: English
• Published: United States Public Library of Science 12.04.2017; Public Library of Science (PLoS)
• Subjects: Algorithms; Appeals; Artificial intelligence; Bayesian analysis; Behavior; Biology and Life Sciences; Casts; Classification; Colleges & universities; Complexity; Computer and Information Sciences; Computer simulation; Construction; Court decisions; Decision trees; Dental materials; Ecology and Environmental Sciences; Engineering; Engineering and Technology; Environmental law; Federal court decisions; Forecasting; Forest management; Fraud; Historical account; Human rights; Humans; Informatics; Information processing; Information retrieval; Intellectual property; Learning algorithms; Leaves; Lenses; Machine Learning; Markov chains; Mathematical models; Natural language (computers); Natural language processing; Physical Sciences; Physics; Political science; Politics; Powers and duties; Prediction markets; Predictions; Protocol (computers); R&D; Research & development; Research and Analysis Methods; Science; Social Justice - trends; Social Sciences; Speech; Statistical mechanics; Statistics; Supreme Court Decisions; Supreme court justices; Taxation; Technology; Television; United States; United States; United States > US; Chicago Illinois
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0174698

Building on developments in machine learning and prior work in the science of judicial prediction, we construct a model designed to predict the behavior of the Supreme Court of the United States in a generalized, out-of-sample context. To do so, we develop a time-evolving random forest classifier that leverages unique feature engineering to predict more than 240,000 justice votes and 28,000 cases outcomes over nearly two centuries (1816-2015). Using only data available prior to decision, our model outperforms null (baseline) models at both the justice and case level under both parametric and non-parametric tests. Over nearly two centuries, we achieve 70.2% accuracy at the case outcome level and 71.9% at the justice vote level. More recently, over the past century, we outperform an in-sample optimized null model by nearly 5%. Our performance is consistent with, and improves on the general level of prediction demonstrated by prior work; however, our model is distinctive because it can be applied out-of-sample to the entire past and future of the Court, not a single term. Our results represent an important advance for the science of quantitative legal prediction and portend a range of other potential applications.