Swing State Math: Obama Has a Tough Task Ahead of Hm

Job creation is the focus of this election. Like a bad hangover, persistently high unemployment has plagued the economy even though official accounts declared the recession over in June 2009. When the October jobs report suggested the unemployment rate had finally fallen below 8 percent, President Obama touted the success of his administration's jobs' programs, prompting Governor Romney to joke at the Al Smith dinner that the President had reminded people, "You're better off now than you were four weeks ago". Obama will be hoping that the November report - the final estimate before Election Day - will add to the narrative he's selling that the labor market is improving thanks to his policies. But how much do bad labor conditions really affect elections?

After months of campaigning, there are really only nine states in play-Colorado, Florida, Iowa, Nevada, New Hampshire, North Carolina, Ohio, Virginia and Wisconsin. With the October drop in the national unemployment rate, seven of these so-called swing states also experienced declines in their unemployment rate, and two remained steady. Is this good news for President Obama?

The Bureau of Labor Statistics maintains data on local unemployment rates, so to analyze the impact of local unemployment rates on elections, we collected data for battleground states since 1976. Any state won by either party by less than a 5 percent margin was considered a swing state to provide a good picture of the real battlegrounds in the nine elections considered.

With few exceptions, in nearly all elections the average unemployment rate in the battleground states won by the incumbent party has tended to be lower than the average unemployment rate in the states won by the challenger, and the unemployment rates in incumbent-won states tend to be drifting downward. Basic statistical analysis suggests that these differences are economically significant. In other words, high unemployment is bad for incumbents.

In 1980, Ronald Reagan beat incumbent President Jimmy Carter in 12 of the 16 battleground states, 9 of which had an unemployment rate above the national average. The average unemployment rate in the states won by Jimmy Carter was lower than the national average, except in West Virginia. Every swing state in 1980 experienced an increase in unemployment prior to the election in 1980, with an average increase of 1.5 percent in these states.

In 1984, Reagan beat Walter Mondale in a landslide, leaving only three battleground states. Massachusetts and Rhode Island, both won by Reagan, had unemployment rates that had fallen by more than 1.5 percent in the election year, while Minnesota, the only state won by Mondale, experienced a rise in its unemployment rate by 0.2 percent during Reagan's first term. Overall, all three battleground states were below the national average for unemployment, but the rise in Minnesota's unemployment rate might have made Minnesotans reluctant to say Reagan had made them better off than they were four years earlier.

In 1988, George H.W. Bush won in states with an average rate of 5.3, while Dukakis won in states with an average rate of 6.0. Given the nationwide average of 5.4, the Democratic nominee stole states that had unemployment rates above the national average, while the defending Republicans may have benefited from the perception in the seven swing states they won that they had turned the economy around. Interestingly, unemployment rates in swing states on Election Day 1988 were 2.5 percent lower on average than they were in 1984, but job growth in the five states won by Dukakis had slowed considerably from January to the election in 1988.

When Clinton defeated Bush in 1992, George W. Bush defeated Kerry in 2004, and Obama defeated McCain in 2008, the incumbent consistently carried a majority of swing states with unemployment rates below the national average, and the challenger consistently won swing states with rates above the national average. The only two exceptions to this trend were Clinton's re-election, where battleground state unemployment was marginally higher than the national average, and the 2000 election, where swing states with unemployment rates marginally different than the national average split between Bush and Gore.

In this election, four swing states with 59 electoral votes - Colorado, Florida, Nevada, and North Carolina - have unemployment rates above the national average, while five swing states with 51 electoral votes - Iowa, New Hampshire, Ohio, Virginia, and Wisconsin - have rates below the national average. Five of these nine states - Florida, Iowa, Nevada, North Carolina, and Ohio - have seen drops in their unemployment rates this year, but only one - Ohio - has seen its unemployment rate fall since Obama's election. On average, the unemployment rate has risen by 1.6 percent in this year's swing states since Election Day 2008.

While voters make decisions based on a variety of issues, it is possible that the margin of victory in battleground states might be determined by specific labor market conditions within those states. In reality, with 12.1 million people looking for work, the highest national unemployment rate in the fall of an election year since World War II, six of nine swing states with unemployment rates above 7 percent, and all swing states with unemployment rates above 5 percent, President Obama has a tough task ahead of him