Is the incumbent curse the incumbent's fault? Strategic behavior and negative incumbency effects in young democracies

- Studies based on regression discontinuity designs detect an “incumbent curse” in young democracies.
- We note that RD studies depend on the analysis of close elections, which induce players to strategically in ways that undermine potential incumbency advantages.
- Data from Mexico shows an incumbency curse in municipal elections but not in elections for national legislators.
- Municipal data show that the vote share of second-loser parties drops drastically after a close election and that the magnitude of the curse varies in states governed by copartisan governors.
- Failure to account for these mechanisms risks exaggerating the size of the incumbent curse.