Election 2012: What the MetroMonitor Says

The G.O.P. primary struggle is finally beginning to wind down (though nobody seems to have told the candidates!) and predictably the commentariat is moving on toward big-picture macro predictions about the national election.

On the agenda now are big, broad, sagacious opinions about how various issues like gas prices and the constitutionality of healthcare reform will affect the national vote.

The problem is, we don’t really have national elections in America. We have state ones, as Ezra Klein recently noted, and it is in a sub-set of those—the “swing states” that can tip either way—that elections are really won or lost. For that reason, it’s been interesting to read a flurry of posts and quotes lately by Klein, Ryan Avent, and my Brookings colleague David Damore (a contributor to a new Brookings book on political demography) properly focusing on the implications of swing state economic data and polling for fall vote outcomes. 

These guys don’t agree on what the data prognosticates but they’re at least looking in the right direction: toward ground-level dynamics in the states, with the notion that positive data may favor the reelection of President Obama.

Yet here’s a thought: States don’t really exist either, if you think about it. To understand swing states, that is, you need to look further down at what’s happening in their metropolitan areas, which concentrate massive pluralities of almost every state’s people and business activity (even “rural” Iowa’s). Specifically, you need to look at metropolitan economic trends—and not fixed data-points but trend data on job growth, the unemployment rate, and gross economic output. It is on changes on those indicators in the nine to 12 months prior to an election that political scientists say residents vote, with a positive direction favoring the incumbent.

So what do jobs, unemployment, and output trends look like in the major metros of the dozen swing states Gallup deemed most likely to decide the election: Colorado, Florida, Iowa, Michigan, New Hampshire, New Mexico, North Carolina, Ohio, Pennsylvania, Virginia, and Wisconsin?

As it happens, the release this week of the latest edition of Brookings’ quarterly Metro Monitor index of recovery trends provides a highly relevant dashboard for tracking the economic direction in the swing states’ major metros during the critical second half of 2011.

The dashboard is suggestive. Looking in aggregate first at the 34 large metropolitan areas that dominate the dozen swing states, the economic data appears generally—but not entirely—favorable for President Obama’s reelection campaign. 

In those 34 metros taken together, the unemployment rate fell faster during the year ended in December 2011 than it did nationally (by 1.4 percentage points in the metros compared to a 0.9 percent decline in the country as a whole). Likewise, the 34 swing-state metros started out with slower output growth than the nation for the first nine months of 2011 but caught up by the fourth quarter, when quarterly output growth from both geographies reached 0.7 percent—a high for the year. The rate of job growth in swing metros fell over the year to below the similarly-sagging national growth rate, but stayed positive. The initial takeaway, then, is that President Obama will be running in a collection of swing states where, on balance, the dominant metropolitan economy was getting better a little bit faster than it was elsewhere.

But as we say: All politics is local, or metropolitan, so to get more of a feel for the situation on the ground it’s necessary to look closer into the nation’s regions.

Begin with the Intermountain West. There, the three swing states of Colorado, Nevada, and New Mexico contain four metros: Colorado Springs and Denver; Las Vegas; and Albuquerque. Las Vegas is trending positive; Colorado Springs and Denver anemically positive; and Albuquerque ambiguously. Output growth accelerated across the region over the year, providing a positive directional. Unemployment fell over the year in all metros—strongly in Las Vegas and Albuquerque, modestly in Colorado metros. And job gains were positive and accelerating from the third quarter to the fourth in Denver and Las Vegas, while job losses accelerated in Albuquerque and hit Colorado Springs. Output growth accelerated in all four metros and met or exceeded the national growth rate in all except Albuquerque. We’d say this bodes well for the Obama reelect in Nevada, fairly well in Colorado, and leaves New Mexico a toss-up.

Turning to the Great Lakes things look more mixed. Metro trends in Iowa are strongly positive, in Michigan weakly positive, in Ohio and Pennsylvania ambiguous, and in Wisconsin weakly negative. The unemployment rate fell in every swing metro in the region over the year to December 2011—fastest in Ohio and Michigan. Output increased in all 17 Great Lakes swing state metros over both the third and fourth quarters of the year—accelerating in as many as saw slowings. Pittsburgh shined on jobs metrics as Pennsylvania’s other metros struggled across the board. However, 11 Great Lakes swing-state metros lost jobs in the fourth quarter, up from five in the third. Two quarters of job losses in Wisconsin at the end of the year darkened the mood there. Likewise, a negative turn in job growth in Michigan and four of Ohio’s seven major metros in the fourth quarter point to tough sledding for the incumbent.

As to Florida, by contrast, the trends are strongly positive. Unemployment rates fell by around 2.0 percentage points or more in seven of the swing state’s eight major metros. Job growth across Florida’s metros in 2011 was mostly consistent and above-average—six of the eight ended the year with faster job growth than the country as a whole—and all Florida metros were adding jobs come the fourth quarter except Lakeland. Output growth accelerated throughout the year in all metros except one, and the rate of output growth across Florida’s metros kept pace for the most part with the nation each quarter. Score Florida conditions as actually somewhat favorable for President Obama. 

Ditto New Hampshire and Virginia: However, by contrast, the situation appears ambiguous in North Carolina.

In sum, then, if the economic direction a year to nine months prior to the election matters, Brookings’ Metro Monitor suggests that the mood of metropolitan voters—a majority bloc in virtually all states—was on balance favorable at a crucial moment to the incumbent (all other things being equal) in about half the swing states; ambiguous in several more; and fairly negative in about a third of the states. In short, the electoral map looks ambiguous all around even at the local level. While the direction of these 34 diverse metropolitan economies may well be a key to understanding electoral dynamics in 12 states, the uncertain nature and pace of recovery in many places leaves the outcome of the election uncertain, for now, too.