A new study published on arXiv analyzes over 1.7 million speeches from the U.S. Congress between 1973 and 2024 to understand the relationship between emotional expression and legislative effectiveness. Using a transformer-based classifier, researchers identified eight discrete emotions and found that emotional expression in speeches has increased over time. The study indicates that certain emotions, like enthusiasm and pride, are positively correlated with legislative effectiveness, while anger shows a negative association. Additionally, emotional valence and diversity were linked to greater effectiveness, whereas emotional intensity was linked to less effectiveness. AI
IMPACT This research demonstrates how AI can be used to analyze nuanced aspects of political communication and their impact on real-world outcomes.
RANK_REASON Research paper published on arXiv detailing a computational analysis of legislative speeches. [lever_c_demoted from research: ic=1 ai=0.7]
- Anger
- arXiv
- disgust
- enthusiasm
- fear
- grief
- Hope
- Joy
- Pride
- transformer-based emotion classifier
- United States Congress
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →