This project examines whether the substantive content of legislation shapes contestation in the Council of the European Union. Although the Council is often described as a consensus-oriented institution, member states regularly contest legislative decisions through votes against, abstentions, or critical statements. Existing research has mainly explained this behaviour through member-state preferences, domestic politics, public opinion, and institutional factors. This project shifts attention to the content of the legislation itself and asks whether proposals involving the distribution of resources are more likely to generate contestation than more technical or administrative proposals.

The central argument is that politicisation does not affect all areas of EU decision-making in the same way. Legislative proposals that allocate financial resources, create uneven costs and benefits, or affect access to rights, markets, or social protections may generate clearer political stakes for member states. These proposals are more likely to create winners and losers and may therefore give governments stronger incentives to publicly signal disagreement. By contrast, proposals that are mainly technical, procedural, or administrative may be less likely to attract visible contestation.

To examine this argument, the project develops an LLM-assisted measure of legislative content. Instead of relying only on broad policy-area classifications or topic models, it uses large language models to code legislative proposals according to the extent to which they involve distributive consequences, technical regulation, implementation burdens, and uneven implications across member states. These measures will be linked to member-state voting behaviour in the Council, focusing on legislative decisions adopted under qualified majority voting.

The project contributes to the study of EU decision-making in two ways. Substantively, it develops a more direct account of how policy content shapes contestation by distinguishing between distributive and technical legislation. Methodologically, it explores how LLMs can be used to extract theoretically meaningful features from legislative texts in a transparent and validated way. The project is ongoing, and the conference paper will present the coding strategy, validation procedure, and preliminary evidence on the relationship between distributive legislation and contestation in the Council.