Motivated Reasoning in the Field: Polarization of Prose, Precedent, and Policy in U.S. Circuit Courts, 1891-2013

Abstract

This study explores politically motivated reasoning among U.S. Circuit Court judges over the past 120 years, examining their writing style and use of previous case citations in judicial opinions. Employing natural language processing and supervised machine learning, we scrutinize how judges' language choices and legal citations reflect partisan slant. Our findings reveal a consistent, albeit modest, polarization in citation practices. More notably, there is a significant increase in polarization within the textual content of opinions, indicating a stronger presence of motivated reasoning in their prose. We also examine the impact of heightened scrutiny on judicial reasoning. On divided panels and as midterm elections draw near, judges show an increase in dissent votes while decreasing in polarization in both writing and citation practices. Furthermore, our study explores polarization dynamics among judges who are potential candidates for Supreme Court promotion. We observe that judges on the shortlist for Supreme Court vacancies demonstrate greater polarization in their selection of precedents.

Keywords: motivated reasoning, judicial behavior, machine learning