Expertise on the Floor: When and Why Parties Invoke Science in Parliamentary Debate

Scientific expertise has become an increasingly visible component of contemporary policymaking, yet we still know surprisingly little about how elected representatives themselves mobilize science within political debate. While existing research has documented the growing institutional role of experts and advisory bodies, the everyday rhetorical use of scientific authority by political actors in parliament remains underexplored. In this paper, we examine when, how, and by whom scientific knowledge is invoked as a political resource in parliamentary discourse, focusing on the Dutch Tweede Kamer between 1994 and 2025. We conceptualize the scientization of politics not as a linear process in which evidence simply informs decision-making, but as a strategic and contested practice embedded in political communication. Drawing on scholarship in public administration, political sociology, and science and technology studies, we argue that scientific authority can serve multiple political functions: legitimating controversial decisions, signaling competence, shifting responsibility, closing or postponing debate, and contesting opponents’ claims. Importantly, once science enters political argumentation, it becomes subject to selection, framing, and challenge, generating reciprocal dynamics of scientization and politicization. Empirically, the study analyzes more than three decades of plenary parliamentary speeches, covering several governing coalitions and major policy debates in the Netherlands. The Dutch case is particularly informative due to its strong tradition of expert-informed governance combined with a highly fragmented, ideologically diverse party system. These conditions create strong incentives for parties to publicly justify their positions and to strategically mobilize or contest scientific authority in parliamentary debate. Methodologically, we develop a hybrid computational text analysis approach to detect and classify references to science in parliamentary speech. First, we construct a domain-specific Dutch dictionary capturing scientific terminology, research-related language, advisory bodies, and institutional knowledge infrastructures. Second, we expand detection using multilingual sentence embeddings to identify indirect or paraphrased appeals to scientific authority that escape keyword-based methods. This approach balances interpretability and recall, allowing us to distinguish explicit from implicit references to science. In a subsequent step, science-related sentences are classified according to their functional use in argumentation, distinguishing instrumental, legitimating, contesting, temporal, and responsibility-shifting roles. We link these measures to party characteristics (government versus opposition status, ideology, populist orientation), issue domains, and crisis periods. This allows us to test expectations about government–opposition asymmetries, issue ownership, and the distinctive ways populist actors engage with scientific authority. We hypothesize, among other things, that government parties more often use science to legitimate and close debate, while opposition and populist parties are more likely to invoke science to contest conclusions, emphasize uncertainty, or challenge expert institutions. We aim to advance empirical research on the scientization of politics and contributes to broader debates about expertise, democratic legitimacy, and political communication.