Digitalised descriptions of criminal behaviours contained in court verdicts and their use in sentencing research: Vast opportunities – what dangers?

Quantitative sentencing research in Europe is constrained by administrative datasets containing very limited information about offence seriousness. This article explores a largely untapped but rich data source describing behaviours: digitalised descriptions of behaviours contained in court verdicts. Drawing on comparative analysis across European countries, the study demonstrates the availability, detail, and digital accessibility of these descriptions. It outlines two complementary approaches for extracting meaningful data using recent advances in natural language processing – variable extraction and story-based categorisation. The article proposes a new behaviour-based classification system for sentencing research and evaluates the benefits, limitations, and data protection challenges of using court verdicts as open data. The findings underscore the transformative potential of these descriptions for sentencing research, comparative criminology, and policy development in Europe.