Speaking of Experts: Attitudinal Evidence on the Politics of Expertise from a Cross-National Deliberative Exercise

The role of experts in the governance of liberal democracies has become increasingly prominent in public debate over the past few decades. Once a rather obscure subject, mostly removed from political contestation, the question of what place and function non-elected experts should have in public-policy making is increasingly politicised, as a result of, among other things, the increasing relevance of anti-expert populist parties, as well as the string of crises that have hit Europe over the recent past, all of which have increased the public exposure of experts, as well as criticism of their potentially technocratic power. Accompanying the above transformations, a rapidly growing survey-based scholarship has investigated public attitudes on the expertise in public policy, gathering precious cross-national data on citizens’ stance on non-elected experts, as well as a number of correlates, such as socioeconomic factors, education, political trust, and national political circumstances, to mention a few. At the same time, survey research remains, by its nature, limited in the amount of information it can handle, as it conveys static pictures that necessarily flatten attitudes on simple variables and scales. As a result, much of the complexity of the politics of expertise is lost. As a way to address this gap, the paper presents an alternative and complementary approach to study attitudes on non-elected experts, which leverages textual data gathered in five mini-publics run between May and September 2023 in, respectively, in France, Germany, Italy, Poland, and the Netherlands. Taking national responses to Covid-19 as their starting point, each of the five mini-publics discussed and deliberated on the public and political role of experts for two days, focusing particularly on four interconnected aspects of the politics of expertise: a) the place of experts in the policy-making process; b) trust in expert-informed institutions; c) information and disinformation; d) and scientific communication. Taken together, they generated a corpus of approximately 18,000 spoken contributions from all participants. Using word embeddings and clustering methods, the paper describes ‘discourses on expertise’ prevailing in the five countries, in a way that adds depth and nuance to survey-based research. Our approach allows us to unpack the way in which non-elected experts and related matters are discussed and articulated in the five national settings, and determine, among other things, to what extent the four topics are structured along transnational frames as opposed to clustering in country-specific patterns.