Presidential voting and the local variability of economic hardship

Wendy K.Tam Cho, James G. Gimpel

Research output: Contribution to journalArticlepeer-review


We examine variations in the impact of several components of economic hardship on the 2008 presidential vote by county. High gas prices, mounting foreclosures, and rising unemployment all enhance the Democratic vote share in areas critical to winning an Electoral College majority. Using Geographically Weighted Regression (GWR), however, we show the varying impact of these forces, controlling for previous Democratic voting, race, age, and income. Economic problems do not produce anything like a uniform response, and not merely because they are geographically uneven in their intensity. Some populations hit by economic downturn would not have voted for the incumbent's party under any circumstances, while others supported the in-party in spite of hard times. Even so, the combined weight of rising jobless claims and escalating foreclosures was sufficiently unsettling in key states to make for an early call on Election Night.

Original languageEnglish (US)
Article number1
Issue number1
StatePublished - Apr 9 2009


  • Economic voting
  • Geographically weighted regression
  • Political geography
  • Presidential elections
  • The 2008 presidential election

ASJC Scopus subject areas

  • Sociology and Political Science
  • Social Sciences(all)


Dive into the research topics of 'Presidential voting and the local variability of economic hardship'. Together they form a unique fingerprint.

Cite this