Brendan Apfeld and Alex Branham have published an article in the Journal of Elections, Public Opinion and Parties.
Title: Campaign shocks and party support: evidence from Brazil’s 2014 presidential election
Abstract: This article investigates the effect of exogenous shocks during an election cycle on electoral outcomes. Specifically, we examine the impact of the unexpected death of a prominent candidate, Eduardo Campos, in the 2014 Brazilian presidential election on support levels for the three main parties. Did the effects die out relatively quickly, providing only a temporary “bounce” or dip in support levels? Or did they alter the fundamentals of the campaign environment and produce a lasting change or “bump” in support levels that lasted until the election? Our results show that while the shock did have short-term effects on all parties’ support, it was only the party that lost its leading candidate where any longer lasting shift in support is detected; we estimate that the party received around 11 percentage points more support than they would otherwise have garnered, had Campos not died. While this was not enough to secure victory, it shows that individual candidates should be understood as a “fundamental” feature of the campaign environment, any change in which is likely to have a lasting effect on voter behavior.