Unveiling Controversies and Contestation in EU Legislative Decision-Making: A Text Analysis Approach

Project idea

In this project, we study legislative decision-making in the European Union (EU), focusing on controversies and disagreements in the Ordinary Legislative Procedure (OLP). Some topics are more likely to create debate. Nevertheless, it is not easy to systematically identify the topics of legislative proposals or to understand which ones face the most opposition. While EU policy classifications, such as officially defined policy areas, provide a useful framework, they are often too broad to fully reflect the nuanced differences between specific policy initiatives. To address this, we aim to introduce a broader categorization of legislative proposals topics and find out which topics lead to more controversy during the decision-making process.

This project classifies legislative proposals using text analysis. First, we use EuroVoc thesaurus keywords linked to each proposal as a basic classification. To improve this, we apply unsupervised machine learning methods, such as topic modelling, to the full text of the proposals. This helps us group the proposals based on the language they use and better understand what each one is about. Then we investigate the association between the content of a proposal and the level of contestation it faces within and between the Council and the European Parliament (EP). This allows us to identify which topics are more likely to generate disagreement and political debate.