How are committees assigned in legislatures? This question has important implications for legislative organization and representation. Existing theories suggest that either the preferences or the expertise of legislators are the main factors that determine committee allocation. We test these theories using new data on committee allocation in the European Parliament (EP), a complex multiparty and multilevel system, for five legislative terms from 1999 to 2023. We employ gradient boosting, a machine learning technique, to explore the complex and nonlinear relationships between various predictors and committee allocation. We find that expertise, measured by previous voting membership, is the most important predictor of committee allocation in the EP. Other factors, such as preferences, preference alignment with the party group or national party delegation, and demographics, have only marginal effects. Our findings support the Informational Theory of committee allocation and challenge the Distributional and Partisan Theories.
This paper is part of the StREP project funded by the Norwegian Research Council grant 301021.