Do Voters Punish Political Instability? Attribution, Electoral Subversion, and Material Disruption in Peru

Download

Abstract

Do voters punish political instability at the ballot box? We assemble a district-level electoral panel for Peru (2006–2026) spanning two episodes at attribution’s poles: Fuerza Popular’s 2021 offensive of over a thousand petitions to annul tallies after Keiko Fujimori narrowly lost the runoff to Pedro Castillo and ran again in 2026, and the paralysis of Lava Jato megaprojects, a shock with no local culprit. Comparing targeted with profile-matched control districts, clustered by province, we find no differential runoff penalty (seven hundredths of a point, robust across thirty-six specifications). We rule out a penalty above five points under conservative trends but not smaller ones. A first-round withdrawal is inseparable from the left’s collapse, and the material shock moves nothing. The runoff coordinates voters against the left, so the offensive costs its author little, which helps explain why subversion can pay. Deterrence then falls to courts more than the electorate.

Keywords

Electoral accountability, Political instability, Electoral subversion, Democratic backsliding, Two-round elections, Difference-in-differences, Peru, Lava Jato

Data

The district-level replication data are available on the Data page. All tables are geographic aggregates built from public ONPE, JNE/INFOgob, and INEI sources, with no personal data.

Citation

Chávez Padilla, Carlos César. 2026. “Do Voters Punish Political Instability? Attribution, Electoral Subversion, and Material Disruption in Peru.” Working paper.