Do — and why do — people interact with politicians on social media? Evidences from Brazilian state level elections

This article presents the results of the analyzes on Facebook campaigning of 140 candidates who ran for governor of the 27 Brazilian federation units during the state level elections. Following Vaccari and Nielsen (2013) analytical framework, we collected the total amount of Facebook posts’ context and interactions of all candidates for governor, to propose an advanced approach of their methodology. We describe the frequency distribution of attention that each candidate generated and use statistical regression to analyze the decisive factors for that amount of attention. In general, the results showed a very active digital campaigning. In the regression model, a competitive candidate and the candidate’s number of press citations were the most predictive factors. The study confirmed the results the previous literature has achieved in other contexts as well.