Campaign econometrics

White House/Pete Souza

President Barack Obama and former Massachusetts Gov. Mitt Romney talk in the Oval Office following their lunch, Nov. 29, 2012.

Article Highlights

  • Models that predict elections using only economic variables have an almost perfect track record.

    Tweet This

  • As weak as the recovery has been, the econometrics suggested that it was strong enough to reelect an incumbent.

    Tweet This

  • Fundamentals of the economy were a boost to the president. Maybe no strategy or candidate could have overcome them.

    Tweet This

This article appears in the December 17, 2012, issue of National Review.

Since the presidential election, Republicans have offered numerous explanations for their defeat. These range from President Obama’s superior get-out-the-vote organization to demographic changes, and almost everyone seems to agree that Republicans must make changes in their party if they are to win the presidency again.

But was the election really so surprising? Ever since the pioneering work of Ray Fair at Yale found a clear link between economic conditions and election outcomes, economists have known something: Models that predict elections using only economic variables have an almost perfect track record. When the economy is improving, incumbents tend to win. When it is worsening, they tend to lose.

The fact is, though the recovery from the recent recession has been long and tepid, conditions have slowly but steadily improved over the last year. This has been especially true in some swing states, such as Ohio, where the unemployment rate currently sits at 6.9 percent, down from 8.3 percent last October.

As weak as the recovery has been, the econometrics suggested that it was strong enough to reelect an incumbent. The most sophisticated extension of Fair’s model has been tracked for almost a decade by Moody’s Analytics. The Moody’s model uses state-level economic data to predict the outcome of the presidential election in each state, and then aggregates the information to make a prediction for the Electoral College. In February, Moody’s published the nearby electoral map. Comparing its analysis with the actual outcome, one sees that the model called every state correctly except Florida.


Poll numbers fluctuated throughout the campaign, but Moody’s projections after the February article showed a steady improvement for Obama as the economy inched forward. The only change in its prediction over this period was Florida’s flipping back and forth between the candidates. The analysis narrowly favored Obama in Ohio, something  Moody’s confirmed with a county-level model that predicted that the counties having the strongest economies would go almost uniformly for Pres ident Obama.

While it may be tempting to treat the results of this election as a referendum on the Republican party, this model suggests that the fundamentals of the economy were a decisive boost to the president. It may be that no strategy, and no candidate, could have overcome them.

Kevin Hassett is the director of economic policy studies at AEI

Also Visit
AEIdeas Blog The American Magazine
About the Author

 

Kevin A.
Hassett

What's new on AEI

Making Ryan's tax plan smarter
image The teacher evaluation confronts the future
image How to reform the US immigration system
image Inversion hysteria
AEI on Facebook
Events Calendar
  • 25
    MON
  • 26
    TUE
  • 27
    WED
  • 28
    THU
  • 29
    FRI
Wednesday, August 27, 2014 | 3:00 p.m. – 4:15 p.m.
Teacher quality 2.0: Toward a new era in education reform

Please join AEI for a conversation among several contributors to the new volume “Teacher Quality 2.0: Toward a New Era in Education Reform” (Harvard Education Press, 2014), edited by Frederick M. Hess and Michael Q. McShane. Panelists will discuss the intersection of teacher-quality policy and innovation, exploring roadblocks and possibilities.

Event Registration is Closed
No events scheduled this day.
No events scheduled this day.
No events scheduled this day.
No events scheduled this day.
No events scheduled this day.
No events scheduled this day.
No events scheduled this day.