A Technocratic Zeitgeist: How Do Elected Representatives Use Technocratic Legitimation in Parliamentary Speeches?

Technocracy can be just as threatening to representative democracy as populism, some have argued. Whether that is technocracy in the shape of experts making value-based decisions or national politicians conceding authority to international bureaucratic unelected institutions, it undercuts basic democratic norms such as popular representation and accountability of decision-makers. In party government as we know it, democratic norms such as representation, accountability, and legitimacy have been entrusted to elected parliamentary assemblies. Parliaments are expected to contain the plurality of perspectives and opinions present in societies. And political parties are the actors who then perform the linking function between the people and those who govern. But what if elected members of parliament no longer perform that function? Just as some MPs take a populist stance, it is not inconceivable that others rely on technocratic legitimation for their preferred policies. Both of these rhetorical tactics undercut the role MPs and political parties are assumed to perform in a traditional form of party government. Peter Mair called attention to this possibility by distinguishing between parties focussing on one of two tasks, responsiveness or (governing) responsibility. To move from the normative to the empirical, this paper asks the question “To what extent do elected representatives use technocratic legitimation in their speeches in parliament?”. We hypothesize that technocratic legitimation is more easily available to MPs from certain parties (governing, long-established, mainstream centre) than from others (opposition, challenger or new, fringe). Another set of hypotheses investigate the link between party ideology and policy area. The codebook is based on established theoretical work on technocracy in order to make not just empirical inferences but also to highlight the normative implications for the quality of representative democracy. We use a computational text-based approach to take advantage of the data availability. A first test used prompt-based zero-shot data annotation (i.e. no fine-tuning and no instruction) on Dutch parliamentary data (2009-2019). This showed that known technocratic dimensions such as elitism, a scientific approach, and antipluralism were present. the Netherlands is a country where ‘good governance’ considerations have generally trumped representative concerns, making it a good test case.