Midterm Elections: Richard Florida on Which Factors Drove Voting?

The conventional wisdom among pundits, pollsters and political analysts about the Republican victory in the midterms is that it represents a referendum on—and a stunning repudiation of—the Obama administration’s stewardship of the economy. “U.S. registered voters choose economic conditions by nearly a 2-to-1 margin over any of four other key election issues as the most important to their vote for Congress,” according to a Gallup analysis, a result that held “across all partisan groups.”

But the geographic patterns of Tuesday’s historic election results reveal a curious paradox. While the economy was clearly the voters’ No. 1 concern, economic conditions alone cannot explain why they cast their ballots as they did. A Wall Street Journal analysis of House races found that Democrats held onto their seats in congressional districts that were feeling the recession the worst. “Of the 25 congressional districts hit hardest by the recession—measured by joblessness, poverty rates and housing prices—16 are currently represented by Democrats. Fourteen of them won re-election despite the Republican tide.”

Economic factors did not drive state-wide races for Senate or governor, either. Democrats, for instance, held onto governorships in the blue states of New York, Massachusetts, and Maryland, and they won a victory in California even though it has taken a tremendous economic hit. Despite the massive Republican pickup in the House and smaller gains in the Senate and governors races, the American electoral map continues to reflect its long-held red vs. blue shading.

Columbia University's Andrew Gelman’s influential book Red State, Blue State, Rich State, Poor State sheds light on this conundrum. Rich voters trend Republican, Gelman and his colleagues found, while rich states trend Democratic. My own earlier analysis of polling data suggested that short-term economic factors like the unemployment rate or changes in housing values provided little explanation of state favorites for Senate or governor, while more deep-seated structural factors like income, social class, attitudes toward religion, and openness toward immigrants, as well as gays and lesbians were more likely to hold sway.

We witnessed no massive realigning of the electoral map, instead, America remains divided along the same political, cultural, and economic axes.

With the help of my colleague Charlotta Mellander, I took a close look at factors associated with the recession’s impact—like the change in unemployment and in housing prices since the onset of the crisis—that might have influenced voters. We also looked at income and several other structural variables. In The Emerging Democratic Majority, John Judis and Ruy Teixeira argued that Democrats have gained an advantage by adding the wealthier knowledge workers who cluster in urban centers to their historic base among poorer populations and minority groups. On the Red side of the divide, blue-collar working class voters have been shifting into the Republican column. Taking this into account, we examined the relations of work and class and partisan choice. Following Ronald Inglehart’s lead, we also looked at the relations between religious values, tolerance, and political preferences. Confining our analysis to state-wide Senate and gubernatorial races, we conducted a basic correlation analysis and compared the results for the current midterm. We also compared the mid-term pattern to the state-by-state vote for Obama and McCain in the 2008 presidential race. As always, we caution readers not to make too much of these findings. The size of the sample is small, and our analysis can only identify relationships among variables, and in no way implies causation. Still, a number of very interesting patterns emerge.

Here is what we found:

Despite all the hubbub about the economy, we found no evidence at all that short-term economic factors—unemployment and housing prices—significantly shaped state-wide voting patterns for either party. This is not to say that short-term economic factors did not matter at the margin: Clearly, election returns and exit polls clearly showed that many individuals shifted their 2008 Democratic vote to a Republican one in the midterms. But at the state level, deeper factors remained by far the dominant factor.

Income and class remain important, but less so than in the ’08 presidential race. Higher income states went for Obama, while lower income states went for McCain.

This trend continued to hold for Senate races, with higher income states voting Democrat and lower incomes states trending Republican, but not for gubernatorial races.

Class also continues to play a role, though its relation was less evident than it was in ’08. Obama took states where knowledge workers and the creative class—which makes up roughly a third of the workforce and includes workers in science and technology; business and management; law; arts, culture, media and entertainment; healthcare and education—comprise a larger share of the workforce, while McCain took blue-collar working class states. The creative class was more split in the midterms, and significant segments of it appear to have shifted from Obama and the Democrats to the Republicans. While a considerable change from ‘08, it is not that surprising, as many creative class voters tend to be independent and more candidate-centered. While the correlation between creative class states and Democrats remained positive, it was not statistically significant. Blue-collar working class states were positively associated with Republican Senate votes, and negatively associated with votes for Senate Democrats.

Religious orientation remains a key pivot point in America’s cultural and political divide. In 2008, more-religious states went for McCain and less-religious states went for Obama. This pattern continues to hold for the Senate, though not for governor’s races. Religion is positively associated with both Republican votes for Senate and negatively associated with Democrat votes for Senate. (Our religion variable is from Gallup polls that ask individuals if religion is an important part of their everyday life.)

From Tom Tancredo in Colorado to Carl Paladino in New York, we're constantly reminded that immigration and gay rights remain significant wedge issues in American politics. (We used the percentage of immigrants and gays and lesbians in states as proxy measures for openness.) Salient in 2008, openness toward gays and lesbians and toward immigrants, were again among the most important factors in state partisan patterns. States with higher percentages of gays and lesbians and higher percentages of immigrants went for Obama in 2008 while those with lower percentages went for McCain, and these trends also continue to hold. Immigrants appear to have a more substantial relation with votes in Democratic states, while gays and lesbians have a more noticeable relation in Republican states. States with larger percentages of immigrants were more likely to vote Democratic in both Senate and governor races. The percentage of immigrants was negatively associated with Republican votes for Senate but not significantly associated with Republican votes for governor. The percentage of gay and lesbian residents in a state was negatively associated with Republican votes for both Senate and governor, and it was positively associated with the Democratic votes for Senate, but not significantly associated with Democratic votes for governor.

But the strongest factor of all in our analysis was the red-blue pattern itself. States that voted for Obama in ’08 tended to elect Democrats to Senate and governor, while those that went for McCain again went for Republicans. The correlations were significant for both Senate and governor races across both parties, though they were about twice as powerful for Senate races.

The upshot of this could not be more clear. We witnessed no massive realigning of the electoral map, instead, America remains divided along the same political, cultural and economic axes. Richer states are still more likely to be Democratic and poorer ones Republican. But it's about more than just money. The creative class might have split its vote to some degree, but working class states continue to trend red, while states with higher percentages of immigrants and especially gays and lesbians continue to tack Democratic.

Of course, economic conditions do play a role in elections and this one is no exception. Obama benefited strongly from the support of the creative class in 2008; economic conditions have considerably tempered their enthusiasm this time around. And "throw the bums out" inevitably takes a greater toll on the party in power.

But a discernible political pattern remains, one that is etched in a deep class divide that is rooted in the very structure of our economy. It’s not just that more educated, higher income, knowledge workers prefer to live in denser cities and metros areas on the coasts. The logic of an idea-driven economy generates innovation and productivity by concentrating them there. But just as the economy benefits from the concentration of human capital and the creative class in tighter, spikier geographic areas locations, our political system penalizes it, as evidenced by the big swaths of red in the less densely populated states and the smaller specks of blue on America’s new electoral map.

And this political reality handicaps the nation’s ability to address the very serious economic problems it faces. In previous periods of economic crisis and transformation, like the Long Depression of the late 19th century and the Great Depression of the 1930s, America benefited from the " critical realignments" that were long ago identified by Walter Dean Burnham, which recast the electorate to create more stable governing and policy coalitions. These political realignments shift the power balance between the parties and, in doing so, provide the political underpinnings for the major changes in public policy that are needed to help the nation adjust to structural economic change.

Though our economy is currently in the midst of a similar great reset, our politics today reflect what Burnham called an “unstable equilibrium.” In fact, the overlay of class and geographic divides, combined with Washington’s inability to get much done, creates an especially volatile partisan cycle.

Democratic anger at Bush motivated massive voter enthusiasm in the ‘06 and ‘08 cycles among Democrat-leaning groups. The same kind of mobilization was apparent not just in the Tea Party but in Republican-leaning groups this cycle. According to a November 2 Gallup poll, 63 percent of Republicans surveyed reported that they were more enthusiastic than usual about voting—an all-time high. Americans are most enthusiastic about voting when they feel the least empowered—it is hardly an inspiring picture. And the cycle is chronically unstable. No sooner is a new administration or a new congressional majority in place than anger begins to mount on the other side and the cycle begins again.

The consequences of this gridlock-backlash-gridlock cycle extend far beyond politics, paralyzing America’s ability to deal with the deep and fundamental economic issues it faces. Just when the United States needs bold, forward-looking leadership which can develop broad efforts to renew the economy, upgrade jobs, spur innovations, and address mounting inequality, it is stymied by a volatile political system propelled by anger and backlash, leaving it with gridlock and inertia.

Richard Florida is Director of the University of Toronto’s Martin Prosperity Institute and author of The Great Reset, published this month by Harper Collins.