FORECASTING THE 2012 ELECTION WITH STATE-LEVEL ECONOMIC INDICATORS: A POSTMORTEM

Abstract

Our State-Level Economic Model incorrectly forecast the outcome of the presidential election. The model's projections were based on the two-party vote in the prior election, the status of the incumbent party in the election/reelection cycle, national and state unemployment rates, and changes in real per capita personal income levels for state residents. Even before the Republican nomination process began, President Obama's success in the 2008 campaign coupled with the electoral advantage typically enjoyed by first-term incumbents bode well for his reelection chances. However, by the arrival of the fall campaign, unemployment remained high and income growth was stagnant, meaning that the president faced significant economic headwinds. Our estimates indicated that Governor Romney would carry most of the battleground states. We were wrong. President Obama carried all but one of them.