Predicting human rights compensation amounts at european court
How Much is a Human Right Worth? ECtHR-NPD: A Benchmark for Predicting Non-Pecuniary Damage Awards
Computation and Language
Summary
Figuring out how much money a court should award for human rights violations is tricky because there is no fixed formula. The authors created a large dataset of past European Court of Human Rights cases including the amounts awarded when no rules set those amounts. They tested many computer models to see if they could predict these awards based on case information, finding even advanced language models struggled to get some details right. This shows the problem is difficult and current AI methods have room for improvement.
What this means in practice
- •For legal technology developers: Build tools that predict compensation amounts in human rights cases using historic court decisions and case details from the ECtHR-NPD dataset.
- •For insurance claim analysts: Estimate non-pecuniary damage awards to improve assessment accuracy for claims involving human rights violations.
Authors
Yanyi Pu, Damian A. Gonzalez-Salzberg, Zheng Yuan, Nikolaos Aletras
Abstract
Existing legal benchmarks cover diverse tasks, while continuous monetary remedies remain comparatively underexplored. We introduce ECtHR-NPD, to the best of our knowledge, the first benchmark for predicting non-pecuniary damage (NPD) awards at the European Court of Human Rights (ECtHR) from case information when no statutory formula or explicit calculation rule determines the amount. ECtHR-NPD contains 14,575 cases with case-level awards in nominal euros, chronological splits, and a protocol separating target construction from model input. We evaluate a battery of methods, including constant predictors, gradient-boosted trees, retrieval methods, fine-tuned encoder language models (LMs), prompted decoder LMs, and knowledge-augmented agents. Our results show that more sophisticated LM and agentic approaches do not consistently outperform the strongest feature-based baseline. All model families struggle to identify zero awards and to calibrate high-award predictions, with further degradation on the Challenging test view, making ECtHR-NPD a challenging testbed for current state-of-the-art open-weight and proprietary LMs.