How will the geography of unemployment affect the 2012 elections?
The jobless rates in five battleground states of 2008 are, on average, 1.46 percentage points higher than the average across the other 45 states. Does this mean that President Barack Obama’s re-election chances are worse than the weak state of the national economy would indicate?
Statistical scholars of elections, such as the economist Ray Fair of Yale University, have long noted the connection between national economic conditions and the re-election of incumbents. State-level conditions matter, too, and areas facing particularly poor economic conditions are particularly prone to punish an incumbent, even when he or she had little to do with the downturn.
Because we elect our president state by state, an incumbent’s chances depend on whether swing states feel particularly prosperous, and whether hardship is concentrated in up-for-grabs constituencies. The current weak economy continues to have profoundly different effects throughout the country. About half the states, 24, are in one of two extreme groups: recession-racked regions that still have double-digit unemployment rates, or areas posting solid comebacks, with unemployment below 7 percent.
As the Bureau of Labor Statistics’ helpful map illustrates (Figure 1, attached), the states showing a recovery from the recession are in three clusters: the resource-rich American heartland, nimble New England and the states around Washington.
Unfortunately for Obama, with the exception of Virginia, the regional recoveries won’t deliver many electoral votes for him, because those interior states are solidly Republican and he is already counting on most of New England anyway.
Unemployment is still highest in three regional clusters near Florida, Michigan and California. The Golden State has lost much of its luster, with an 11.9 percent unemployment rate that is exceeded only by the 13.4 percent rate in neighboring Nevada. That is more than double the level in states such as Nebraska and New Hampshire.
Each state’s economy depends on the complexities of local industries and entrepreneurs, but one simple variable does a reasonably good job of explaining current unemployment rates across states. The next figure (Figure 2, attached) shows the relationship between unemployment rates in September 2011 and the number of adults with high school degrees in 2009. As the share of adults with a high school degree increases by 5 percentage points, the unemployment rate drops on average by about 1.7 percentage points.
One explanation for this is that in September 2011, the unemployment rate for high school dropouts was 14 percent and the unemployment rate for college graduates was 4.2 percent. But even these dramatic gaps can’t explain the strong state-level correlation between dropout rates and unemployment. On a national scale, an increase of 5 percentage points in the share of the population with a high school degree means unemployment should drop by less than 0.5 of a percentage point, which is far smaller than the observed correlation.
Another explanation for the strong skills-unemployment connection is that if you live in an area with a higher proportion of less-skilled neighbors there will be fewer people to give you a job or pay for your services. A third explanation is that the correlation between unemployment and dropout rates may reflect the fact that less skilled workers came to places such as Nevada and California because of the false dreams that fueled the housing bubbles in those states.
According to this view, the presence of abundant numbers of high school dropouts isn’t the cause of unemployment, but an indicator of past overexuberance, which has left a legacy of jobless construction workers and vacant businesses.
State-level skills are also correlated with support for Obama in 2008. At the individual level, the president was actually most successful with high school dropouts, and his Republican opponent, Senator John McCain of Arizona, did best with people with some college education.
But as the next figure shows (Figure 3, attached), Obama’s vote share increased, on average, by about 12 percentage points, as the share of the population with a college degree increased by about 10 percentage points.
So skilled states, which have less unemployment, voted for Obama. Shouldn’t that mean that jobless rates are highest where voters most strongly opposed the president? Actually, no.
The next figure (Figure 4, attached) shows the far less perfect relationship between support for Obama in 2008 and unemployment today. When we statistically estimate the relationship, it seems to be a curve that first rises and then falls with support for the president.
There is a cluster of highly educated pro-Obama states that are doing reasonably well. The 10 states with the highest shares of adult college graduates all went for Obama and have an average unemployment rate of 7.4 percent. But the highest jobless rates are, on average, not in the rock-ribbed Republican states but rather in more closely contested areas.
Less-educated Republican states are experiencing unemployment rates below the levels predicted by their college graduation rates, while Democratic states with increased high school dropout rates, such as Michigan and California, are experiencing unusually high unemployment levels.
Several Republican states, such the Dakotas, have few high school dropouts, even though they also have fewer people with more advanced degrees. At the same time, several Democratic states, such as New York and California, have plenty of high school dropouts, as well as an abundance of the overeducated. Statistically, high school graduation rates matter most for unemployment.
Yet California and Texas have about the same share of adult high school dropouts -- but the unemployment rate in Texas is far below the national average, and California’s is far above it. The populations of Nevada and Louisiana have similar skill sets, but Obama-supporting Nevada is doing terribly, while McCain-supporting Louisiana is holding its own.
Perhaps the success of many less educated Republican states reflects that party’s appeal in natural-resource-rich states that are now doing well. Perhaps the Republican distaste for taxes and regulations has enabled these states to thrive.
It may just be coincidence, but the recession seems to be hitting the moderately educated battleground states that went Democratic in 2008 -- Florida, North Carolina and Ohio -- with far more severity than ultra-Republican areas.
There is a silver lining for the president. Even if those three swing states vote Republican in 2012, it won’t be enough for that party to win the White House. Once you look beyond those states, the local economics in slightly more pro-Obama states, such as Virginia, Iowa and New Hampshire, are far more robust.
These subnational economic currents seem likely to create a close election, but the strength in those slightly more Democratic states may provide the president with a sort of breakwater, a set of voters who supported him in 2008 and are now experiencing relatively good economic conditions.
(Edward Glaeser, an economics professor at Harvard University, is a Bloomberg View columnist. He is the author of “Triumph of the City.” The opinions expressed are his own.)
To contact the writer of this article: Edward Glaeser at firstname.lastname@example.org.
To contact the editor responsible for this article: Max Berley at email@example.com.