Since April 12, 2026, the Democracy Action Lab (DAL) has been conducting independent research into the administrative failures that affected the first round of Peru's presidential election, during which a significant number of polling stations in Lima opened hours after the legal start time. Drawing on more than 92,600 tally sheets reconstructed from the National Office of Electoral Processes (ONPE), scanned acts processed with multimodal AI models, and the official JNE list of late-opening stations, the team produced the most complete public database available on this election.

Our findings show that the late opening had a real and measurable effect on voter turnout in Lima — a reduction of between 2.5 and 5 percentage points — but that no empirically plausible scenario alters the order of the candidates advancing to the runoff. The administrative failure was serious in its own right: thousands of citizens were prevented from exercising their right to vote, and trust in Peruvian electoral institutions was further eroded. Ensuring this does not recur in the runoff and providing clear accountability for what happened on April 12, 2026, are obligations that stand independently of the findings on the final result.

The work is presented across three companion documents, all available above:

1. Working Paper (English) — Full academic version with methodology, data construction, identification strategies, robustness checks, and complete results.
2. Policy Brief (English) — Condensed summary of the findings for policymakers, electoral authorities, journalists, and the general public.
3. Amicus Curiae (Spanish) — Brief submitted to Peru's National Elections Jury (JNE), presenting the evidence directly to the body responsible for adjudicating the election.

This work was carried out with full academic independence and without funding from electoral campaigns or political parties.