Electoral cycles in government employment: Evidence from US gubernatorial elections

Abstract

Incumbents may opportunistically design policies increasing employment before elections or postpone cuts until afterwards. I investigate electoral cycles in public sector employment around US gubernatorial elections. Exploiting staggered gubernatorial election cycles across states, I use both county fixed effects models and a geographic discontinuity design that compares neighboring counties at state borders with a difference in gubernatorial election cycles. Consistent with manipulation, state and local government employment per capita are higher leading up to elections; afterwards, employment abruptly returns to normal. Political and spatial heterogeneities are investigated, including by election competitiveness, term limits, incumbent party affiliation, and ideological alignment between the incumbent and the state legislature or local citizens. Differences across types of government employment, and private sector employment, are also explored.