Since the 2022 general election, every Brazilian polling section is scheduled to open at 08:00 and close at 17:00 Brasília time. In local time this moved the nine-hour voting window one hour earlier (07:00-16:00) in the UTC-4 states and two hours earlier (06:00-15:00) in Acre and eleven western Amazonas municipalities, leaving its length unchanged. Using municipal turnout from the Superior Electoral Court (TSE), general elections 2010 to 2022 (5,566 municipalities), a pre-registered two-way fixed-effects difference-in-differences compares round-1 presidential turnout in 2022 with its pre-reform mean (2010, 2014 and 2018). The registered primary estimate of the difference in this change per hour of earlier local window is +0.16 percentage points (pp; wild cluster bootstrap 95% interval [-1.79, +1.72] pp, p = 0.81, 27 state clusters). Under the pre-registered reading rule this is inconclusive rather than a precise null: effects of up to about 1.8 pp per hour in either direction cannot be excluded. The comparison of 2022 with 2018 alone (the 2022 event-study coefficient) is +0.69 pp per hour (p = 0.46). Separate one-hour and two-hour coefficients, the 2026 general election, the 2024 municipal election, round 2 and several specifications give intervals that include zero, and a within-Amazonas comparison of two-hour and one-hour municipalities gives +0.92 pp (permutation p = 0.38). Pre-reform trends differ between the groups (the two-hour group rose by 4.2 pp from 2010 to 2018 while the one-hour group fell by 3.1 pp), so the estimates cannot be read as the effect of the reform. In a pre-registered descriptive outcome, the share of Amazonas polling sections closing at least 30 minutes after the scheduled local close rose from 8% (2018) to 60% (2022) in the two-hour municipalities and from 18% to 38% in the one-hour municipalities; in Acre it fell from 16% to 8%.
No takes yet. Share an insight, caveat, or question.
Keisuke Watanabe (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: