Independent agencies, credible policies? Evidence from a large survey experiment study on the authorization of COVID-19 vaccine in four EU countries

It is a foundational assumption of regulation literature that endowing regulatory agencies with greater independence increases the credibility of their policies. This assumption has been central to the worldwide popularity of the agency model as the dominant model for organizing regulation, for both economic and social regulation. However, extant research provides us with little evidence for this key theoretical assumption. We put the independence–credibility assumption to rigorous empirical testing, focusing on the case of the approval of the Pfizer-BioNTech COVID-19 vaccine in the EU, following the European Medicines Agency’s (EMA) risk assessment. We pre-registered and conducted a large survey experimental study among approximately 5,000 citizens in four different EU countries (Netherlands, Sweden, Ireland and France). We conducted the study in the early stages of the vaccination process in the EU, only a few weeks after the approval of the vaccine. We tested, both observationally and experimentally, whether the degree to which EU citizens perceive EMA as independent from political stakeholders affects the credibility of its outputs, specifically, the credibility of its risk assessment of the vaccine and subsequently, citizens’ willingness to get vaccinated. We manipulated the perceived independence of the EMA by providing participants with information on EMA’s institutional independence from the EU political institutions and member states, against authentic, relatively neutral information on the agency. We confirmed the effectiveness of this manipulation by comparing the perceived independence of the agency across the different conditions. Our study’s fieldwork is in progress – most of the data has been collected prior to the submission of this abstract. Pre-registration link: https://osf.io/tp648/