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 which leverages some 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%.

When you think about it, not a lot has changed in the legal world from the days of To Kill A Mockingbird to the latest John Grisham thriller. Sure, literature snobs may insist that Atticus Finch's flawless moral heroism should never be compared to the conflicted protagonists of contemporary legal page-turners, but in terms of the substance of how lawyers do their lawyering, the fundamentals have barely changed in 80 years, from the career track of a young lawyer to the set-up of a law firm. The same cannot be said of virtually any other profession. Indeed, the legal industry seems more dusty than dynamic; the robes and wrinkles that mark those at the top of the field hardly scream modernity. But change is afoot, as a couple of powerful market forces are driving law firms to adopt modern corporate efficiency.