By applying computational linguistics tools to the analysis of US federal district courts’ decisions from 1932 to 2016, I quantify the rise of economic reasoning in court cases that range from securities regulation to antitrust law. I then relate judges’ level of economic reasoning to their training. I find that significant judge heterogeneity in economics sophistication can be explained by attendance at law schools that have a large presence of the law and economics faculty. Finally, for all regulatory cases from 1970 to 2016, I hand code whether the judge ruled in favor of the business or the government. I find that judge economics sophistication is positively correlated with a higher frequency of pro-business decisions even after controlling for political ideology and a rich set of other judge covariates.