Using Supervised Learning Methods to Measure Women's Rights" An Analysis of the Language of Reproductive Rights Briefs

Abstract

The battle over reproductive rights has moved to the forefront of politics, and although the Supreme Court plays an important role, the debate there is understudied. This project looks at the language surrounding reproductive rights in amicus curiae briefs.Amicus curiae briefs are important sources of legal arguments for Justices, exposing them to a specific vocabulary. Brief writers develop a strategy of how to frame reproductive rights in order to achieve a goal: their preferred disposition. When doing this, under what conditions do writers consider women’s rights persuasive? In order to answer this question, I used a naive Bayes classifier to classify all 615 amicus curiae briefs submitted for Supreme Court cases on reproductive rights based on their legal arguments. Results show that brief authors prefer arguments based on others’ rights instead of women’s rights. The subject of a case affects this distribution, and when the decision of a case was pro-reproductive rights, authors had higher rates of using women’s rights arguments. Pro-reproductive rights authors are more likely to use arguments based on women’s rights. Brief authors steadily used more women’s rights arguments as time went on.