My Court, My Emotions: Predicting the Outcomes of Emotional Arousals in the ECHR

• In Event: New Approaches to Measuring Emotions

Abstract

It has recently been hypothesized in the political science literature that judges convey their preferences during public hearings (Dietrich et al., 2019). Using the European Court of Human Rights (ECHR) as our case study, we show that the leanings of judges can be revealed from oral arguments by extracting visual and audio cues. This paper analyzes 152 webcasts of hearings corresponding to more than 200 hours of video recordings from 2007 to 2022 that show judges in conversation with different plaintiffs, such as government officials and individuals. We then use emotional expressions and the level of emotional arousals to predict judicial decisions. Our video data is further enriched with a unique set of manually collected features used as controls which provide added robustness to the analysis. The research empirically contributes to the field by being the first study to methodologically study emotional arousals of judges from ECHR (i), and by combining image and audio data to predict leanings of judges in high courts (ii). The results show that visual and audio cues carry information that text-based judgments do not possess.

Authors

• Cantay Caliskan, University of Rochester

• Ezgi Siir Kibris