Parole candidates with similar characteristics can receive different outcomes. Prior research has shown a high degree of arbitrariness in California parole-release decisions. Sociologist AJ Alvero, legal scholar and economist Ryan Sakoda, legal scholar Kristen Bell, and data scientist Jake Searcy will use machine learning tools to better understand the decision-making process during parole proceedings in California. They will analyze transcripts from California parole-release hearings from 2007 through the present for their study.