Courts, including the U.S. Supreme Court, publish a shock number of opinions without divulging the author. These unsigned per curiam opinions were traditionally considered a means of deciding uncontroversial cases in which justices were all in agreement. Today, though, these anonymous opinions often dispose of highly controversial issues, frequently over significant disagreement among justices. Since 2005, the Roberts Court has disposed of at least 65 cases through unsigned per curiam opinions, and many cases have also included unsigned concurring or dissenting opinions. Political commentators, practitioners, historians and scholars have all decried this trend, arguing that it undermines judicial accountability.
The role of unsigned judicial opinions received renewed attention when the Supreme Court announced its controversial 2012 decision in National Federation of Independent Business v. Sebelius concerning the constitutionality of Obamacare—the 2010 Patient Protection and Affordable Care Act. Many experts had predicted that the Court would find Obamacare unconstitutional, and that the pivotal vote would come from Justice Anthony Kennedy. After the decision, there was speculation that Chief Justice John Roberts had switched his vote and that the formerly unsigned dissent (later attributed to Justices Kennedy, Scalia, Thomas, and Alito) had originally been the majority opinion authored by the Chief Justice. But it also has been hypothesized that the joint dissent was actually written by Justices Kennedy and Scalia.
A study published in the Stanford Technology Law Review by a group of scholars at the Massachusetts Institute of Technology and the Berkman Center for Internet & Society at Harvard University provides new evidence that the likely authors of the Obamacare joint dissent are Justices Kennedy and Scalia. Combining data from 568 Supreme Court opinions between 2005 and 2011, these scholars— including Professor Andrew Lo, a principal investigator at CSAIL, and William Li, a graduate student at CSAIL —have utilized machine learning algorithms to predict authorship of opinions that are unsigned or whose attribution is disputed. Applying those algorithms to the joint dissent in Obamacare, the study offers clear answers: “Justice Kennedy is the predicted author . . . [and] Justice Scalia has the second-highest probability.”
Read the full paper here: http://stlr.stanford.edu/pdf/algorithmicattribution.pdf. For more information on Lo’s work, please visit: http://www.csail.mit.edu/user/2532.