Through which channels can elites influence public support for EU crisis policies? We identify two channels: cueing ("words", i.e. how politicians talk about the EU influences citizens’ policy attitudes) and policy legacies (memories of "deeds", i.e. policy interventions of the EU influencing such policy attitudes). In a pre-registered experiment in 8 EU countries, our paper asks whether cueing and policy legacies may be alternative mechanisms or, on the contrary, must complement each other in the EU to modify citizens’ positions on policy. We look at a crucial policy which may have a lasting legacy: the adoption of the pathbreaking NGEU package during COVID. We operationalize policy legacy with past policy satisfaction and by presenting individuals with a memory prime about the positive and negative effects of the policy. We operationalize cues by randomly assigning individuals to conditions where they are exposed to messages from their national government in support or against the policy. We then contrast and compare the effects of the satisfaction prime and the government cues on support for a new policy of fiscal solidarity in the EU, which forms our dependent variable. Finally, we consider heterogeneity in our treatment effects: specifically, we look at ideology, trust in the European Union and views on European integration as well as identity, and finally territorial differences.
Our design allows us to estimate the impact of policy legacy effects based on individual, micro-level, attitudes towards previous policy, as well as differentiate between self-reinforcing and self-underming legacy mechanisms that result in wanting to maintain the status quo or enhance the policy. We contribute to the policy legacy literature by showing how attitudes towards past policies interact with elite cues in a multilevel polity.