Was Bolsonaro’s 2018 electoral victory an institutional accident?

Abstract

Presidential Elections are critical moments for polyarchical systems, particularly in contexts of high social tension. In this regard, the 2018 presidential election in Brazil, which used a two-round system, is a significant case study. Intriguingly, the most divisive candidates went to the second round. Was this an institutional accident? Pairwise and positional voting procedures embody different generalizations of a majoritarian credo that underpins such elections. The paper mobilizes both perspectives and, using representative survey data, reconstructs the top four preferences of the Brazilian electorate a week before the election. The analysis reveals that the electoral winner, Jair Messias Bolsonaro - despite being a divisive candidate - was a Condorcet Winner and a Borda Winner. Conversely, the second-round loser, Fernando Haddad, was a Condorcet Loser. Thus, Bolsonaro’s victory in 2018 was not an institutional accident. The paper also explores possible alternative scenarios under different feasible sets of candidates through simulations, contributing to a deeper understanding of the role of decision procedures in critical junctures.