The Impact of the #MeToo Movement on Language at Court -- A text-based causal inference approach

TL;DR

This paper investigates how the #MeToo movement influenced court language regarding sexual violence, using causal inference and text analysis to detect subtle shifts and delays in language change across U.S. courts.

Contribution

It introduces novel indicators and combines causal inference with advanced text quantification methods to analyze the movement's impact on judicial language.

Findings

No clear immediate change detected in court language due to #MeToo

The movement may have accelerated language evolution with a significant delay

Potential heterogeneity in effects based on judge's gender and political views

Abstract

This study assesses the effect of the #MeToo movement on the language used in judicial opinions on sexual violence related cases from 51 U.S. state and federal appellate courts. The study introduces various indicators to quantify the extent to which actors in courtrooms employ language that implicitly shifts responsibility away from the perpetrator and onto the victim. One indicator measures how frequently the victim is mentioned as the grammatical subject, as research in the field of psychology suggests that victims are assigned more blame the more often they are referred to as the grammatical subject. The other two indices designed to gauge the level of victim-blaming capture the sentiment of and the context in sentences referencing the victim and/or perpetrator. Additionally, judicial opinions are transformed into bag-of-words and tf-idf vectors to facilitate the examination of the evolution of language over time. The causal effect of the #MeToo movement is estimated by means of a Difference-in-Differences approach comparing the development of the language in opinions on sexual offenses and other crimes against persons as well as a Panel Event Study approach. The results do not clearly identify a #MeToo-movement-induced change in the language in court but suggest that the movement may have accelerated the evolution of court language slightly, causing the effect to materialize with a significant time lag. Additionally, the study considers potential effect heterogeneity with respect to the judge's gender and political affiliation. The study combines causal inference with text quantification methods that are commonly used for classification as well as with indicators that rely on sentiment analysis, word embedding models and grammatical tagging.