Anderson Economic Group Work

Who Wins if Voter’s “Pocketbooks” Prevail in Presidential Election?

East Lansing, Michigan.  October 22, 2012 –Economists Patrick Anderson and Ilhan Geckil recently released their pre-election “pocketbook” prediction of the 2012 presidential elections, using their award-winning, election-predicting pocketbook model. The model, which  explains about 3/4 of the variation in the popular vote in American presidential elections over most of the past century, predicts “pocketbook” voters will favor the challenger over the incumbent this year.

The model considers five categories of variables representing almost all the economic information a voter will have prior to the November election. Both state and national models indicate a likely margin of 3 points, with a standard error of +/- 2.5 points. This is based on data for the following:
1.     Income—change in income is a pure “pocketbook” indicator; more income typically means more votes. This is a negative for 2012, and annual real growth has only been around 2% in recent years.
2.    Employment—growing unemployment unnerves voters and typically means fewer votes. This is a strong negative in the upcoming election despite some recent improvements as unemployment has been high throughout the incumbent’s term.
3.    Inflation—higher inflation (or deflation) affects voters negatively. This is a neutral factor for the 2012 election, with current annual inflation running at a modest 2.8%.
4.    War—voters rally around the commander in-chief in a full-scale war; they have a different view of limited wars. This is also a neutral factor for 2012 as the long-running war in Afghanistan has spanned multiple presidential administrations.
5.    Third-party—these candidates often affect the election but rarely win major states. This is also a 2012 neutral as no third-party is expected to obtain a large share of votes.
The “pocketbook” model historically explains 21 of the last 24 popular vote outcomes in U.S. presidential elections and correctly predicted the popular vote winner in the 2008, 2004, and 2000 elections. The elections where the model did not predict the popular vote winner were 1952 (Eisenhower), 1960 (Kennedy), and 1976 (Carter).
This model has been presented at the Hauenstein Center for Presidential Studies, first in 2004 and again on February 7, 2012. They also describe the model in the article Pocketbook Predictions of Presidential Elections (2004), which was published in the journal of Business Economics. The article earned the Edward A. Mennis award for the authors.
Details on the 2012 version of the model are available here.
About Anderson Economic Group  – Anderson Economic Group, LLC specializes in business valuation, public policy, and industry analysis and has offices in Chicago, Illinois and East Lansing, Michigan. Patrick Anderson, the firm’s CEO, is one of the nation’s foremost experts on how private businesses and the economy affect elections. He has written over 100 published works including “Pocketbook Issues and the Presidency” (2004) and “The Value of Private Businesses in the United States” (2009).
CONTACT:   Patrick Anderson, Anderson Economic Group, LLC.   517.333.6984