Between-round mobilization and the composition of Congress: a counterfactual with Brazilian electoral data for 2026

Abstract

The article estimates how the Brazilian Chamber of Deputies and Senate elected in 2026 would differ if what each presidential finalist's campaign did differently between rounds had happened in the first round. Using municipality-by-zone official returns, ecological inference and the official seat rule, a preregistered counterfactual separates turnout from persuasion and treats both campaigns alike. Combined, turnout moves seats within the noise of the electoral formula, and even full coattails among vote switchers keep the effect below 10 seats in the baseline. Alone, only the Lula arm reaches that threshold, but not when each pole is restricted to the finalist's own party or federation. A very different Chamber would require loyal voters to change their legislative vote, on which the runoff says almost nothing. The replicable method links between-round flows to seats.