Why do some issues receive more interest from the public while others do not? This paper develops a theoretical and empirical approach that explains the degree to which issues expand from the elite to the public. We examine how candidates in the 2014 European Parliament elections talked about EU issues, in comparison to other political issues. We rely on data collected from Twitter and use a combination of human coding and machine learning to analyse what facilitates interactions from the public. We find that most political actors did not try to engage with the public about EU issues, and lack of engagement results in less interactions from the general public. Our findings contribute to understanding why EU issues still play a secondary role in European politics, but at the same time highlight what low cost communicational tools might be useful to overcome this expansion deficit.