How did Obama fare among white working-class voters? The answer depends on how one defines the term—a problem that has been a matter of contention in recent years.
The basic divide among scholars and analysts is how to define the working class—by income, by education level, or by some other measure. CNN’s exit polls, which were conducted in every state, give results in both the income and education categories, while taking account of the race of respondents.9 So, for example, among whites making less than $50,000 (25 percent of the electorate), Obama ran respectably, losing to McCain by just 51 to 47 percent nationally. However, among “whites—no college” (39 percent of the electorate), McCain won by 58 to 40 percent.
The results varied state by state, reflecting regional attitudes and the amount of resources Obama put into a given state. In Ohio, Obama won whites earning less than $50,000 (51 to 47 percent) and lost whites without college educations by ten points (54 to 44 percent). In Wisconsin, whose urban and rural districts were more pro-Obama than any of the states outside New England, Obama won non-college whites by 52 to 47 percent and whites earning under $50,000 by 60 to 39 percent. In New York, a blue state where Obama did not actively campaign, he did better than he did in Ohio but worse than in Wisconsin among those groups.
Predictably, the results in the South were starkly different. In Alabama, where Obama spent virtually no time or resources, he lost whites making under $50,000 by 87 to 11 percent and non-college whites by 90 to 9 percent. Even in North Carolina, a Southern state he visited frequently and in which he invested enormous resources, he lost non-college whites by 67 to 33 percent and whites under $50,000 by 57 to 42 percent. And in Virginia, where Obama both began and ended his election campaign, he lost non-college whites by 66 to 32 percent and whites earning less than $50,000 by the same 57 to 42 percent as he did in North Carolina.
The question raised by the recent debates over white working-class voters may thus be put as follows: Are all these voters, save Wisconsin’s, victims of “false consciousness,” or of “voting against their own self-interest”? If you are even casually familiar with this debate, these phrases will call to mind the author Thomas Frank, who argued exactly this in his 2004 best-seller What’s the Matter with Kansas?10 Frank evoked what he called “the Great Backlash” of working-class people against liberal elites. He described the takeover of the Chamber-of-Commerce, gin-and-tonic Republican Party in his home state of Kansas by religious working-class whites concerned passionately and almost exclusively with the repeal of Roe v. Wade. He asserted that Republican politicians and conservative proselytizers had, by insistent emphasis on alleged liberal turpitude and what he memorably called “the systematic erasure of the economic,” duped these voters into neglecting their own pocketbook concerns. A more concerted and populist effort on the Democrats’ part to appeal to these voters, Frank argued, would yield results and change our politics.
Even as the book was getting attention from many thousands of rank-and-file progressives, detractors both conservative and liberal attacked it. The general critique, which I partly shared, was that these folks may indeed have understood that they were voting against their own economic interests; they simply cared more deeply about cultural matters than economic ones. Maybe a country in which abortion was legal and homosexuality accepted was more distasteful to them than a country in which millionaires received most of the tax breaks. Rather than voting against their own economic self-interest, they were voting for their cultural self-interest, well aware that doing so would cost them economically.
No one challenged Frank quite as directly as Bartels. Sifting through the highly detailed National Election Studies (NES) data since the 1950s, he tested Frank’s arguments and prepared a paper for the meeting of the American Political Science Association in September 2005 called “What’s the Matter with What’s the Matter with Kansas ?” Naturally, the title, showing as it did a flash of wit to which we are unaccustomed in our social scientists, attracted a lot of attention. Bartels opened his paper by asking four questions:
• Has the white working class abandoned the Democratic Party?
• Has the white working class become more conservative?
• Do working class “moral values” trump economics?
• Are religious voters distracted from economic issues?
To each question, he answered: No.
Bartels has pressed these arguments since the publication of his paper. Frank has pressed back, with a stinging riposte he wrote and put on his Web site in late 2005, which read in part:
To begin with, consider the barren landscape of American politics as Bartels describes it—a featureless tundra swept of history, ideology, and any hint of the raw emotional resonance that everyone knows politics possesses. His NES America is not a place that I recognize. It might as well be the moon….11
“Everyone knows….” Social scientists don’t like that sort of talk. But Frank was describing a passion that he witnessed firsthand in a state he knew well. Passion is not always reducible to data points, and some of the phenomena Frank emphasized—the entry into working-class politics of legions of Christian conservatives, for one—are not reflected or precisely caught in voter surveys.
But Bartels’s conclusions have attracted wide attention and drawn a number of adherents. Here’s how he states them in Chapter 3 of Unequal Democracy, the chapter devoted to the white working-class question:
My analysis suggests that economic issues remain centrally important in contemporary American electoral politics, especially among “the people on the losing end” of the free-market system. Moreover, the political views of those people have changed rather little over the past few decades, while their support for Democratic presidential candidates has actually increased. If Republican electoral success is indeed a puzzle, the solution seems to have little to do with the cultural conservatism of the white working class.
He shows, for example, that since 1948, support for Democratic presidential candidates has decreased among upper- income whites but has actually increased among those in the lower brackets. Additionally, he shows that while Democratic party identification among lower-income whites has decreased since 1952, it’s decreased more sharply among whites with higher incomes.
But this is where Bartels’s definition of “working class” is open to question. It is based strictly on income and includes only those families at the bottom third of the income scale—those with incomes below $35,000. But the average family income in the United States in 2007, according to the Census Bureau, was $62,359.12 Surely many Americans of average income or slightly lower are working-class people. Ruy Teixeira and Alan Abramowitz, writing in Red, Blue, and Purple America, edited by Teixeira, offer a convincing critique. First of all, they assert, many low-income whites are not workers at all, but include those who are “retired, disabled, homemakers, or students.” Furthermore, they note, Bartels’s definition excludes “the very kind of workers who traditionally are most associated with that group,” i.e., are identified as workers. The average unionized blue-collar worker, they write, made $22.74 an hour in 2003, or about $45,000. Many augment that considerably with overtime. It’s pretty difficult to talk about a “working class” that doesn’t include the average auto worker, coal miner, or Teamster.
Class is a function of many considerations. Income is one, but so are educational attainment, type of occupation, and what we might call subjective self-definition (and the sets of cultural values, likes, and dislikes that follow from that). If we look back at the total numbers of the Obama vote I have mentioned—in Ohio and Wisconsin, for example—we find that in each case, Obama did better, usually far better, among whites with low income than among whites with low education. The reason for this would seem to be, as Teixeira and Abramowitz suggest, that the low-income category includes some groups that were pro-Obama, like students and entry-level workers in certain lower-wage sectors that tend to be liberal—schoolteachers and employees of nonprofit organizations, for example. By contrast, those with lower education levels tend to be more culturally conservative, more Republican—and, let’s face it, less open to the idea of a black president. This suggests, I would think, that education is a better guide to class than income.
Teixeira and Abramowitz completed their chapter earlier this year and made it public in the spring, right around the time that the question “Can Obama win working-class whites?” was all the rage. They used a combination of income and occupation data that seems a more reliable basis for a definition of working-class whites. Doing so, they noted that Bill Clinton won these voters by 1 percent in both of his races, Al Gore lost them by seventeen points, and John Kerry lost them by twenty-three points. They argued that to win the White House, the Democrats would have to get the Kerry deficit “down to around 10–12 points to achieve a solid popular vote victory.”
Obama did not do this. After the election, Teixeira e-mailed me the following information. By his definition, using the “whites—no college” category, not the income category, as his baseline, Obama lost the white working class nationally by eighteen points. Obama’s performances among whites with and without college educations were very similar to Michael Dukakis’s in 1988. The key difference, though, is that the size of the white working-class vote—as defined by “no college”—is down 15 percent from 1988, while the size of the white college-educated vote is up 4 percent. Teixeira and his coauthor’s prediction was wrong, he told me, because of their too-cautious assumptions about white college-graduate support for Obama and the size of the minority turnout.
This, finally, suggests the crucial lesson: for all the attention lavished on white working-class voters, we now have reached the point where they don’t matter as much anymore. The white population, around 68 percent today, is expected to be 61 percent by 2020 and 50 percent by 2050. In his chapter on race and immigration William Frey of the Brookings Institution shows significant differences in the “purple” states—those in which one or the other party won by less than 10 percent in 2004. The fast-growing purple states, such as Colorado and Nevada, have smaller white populations and considerably larger Latino populations than slow-growing purple states, such as Florida, which are older and whiter. Indeed, to read through the seven informative essays in Red, Blue, and Purple America—devoted to such subjects as changes in family structure, the growth of new types of suburbs, the “clustering” by which demographic groups tend more and more to live together, and the rise of the more liberal “Millennial” generation (twenty-five and younger)—is to see that almost every major demographic trend favors the Democratic Party over the long term.
The challenges to the Republican Party, in these trends and in this year’s results, are obvious. The party has to seek votes in the categories that are growing. To do that it has to moderate its views on immigration, race, and certain cultural issues. As David Frum and other conservatives have argued, it also needs to take economic equality at least a little more seriously. But institutionally, it is hard pressed to do any of these things. Its most loyal rank-and-file adherents are its shock troops of the religious right. At the same time economic royalists, to use Franklin Roosevelt’s term for the wealthy supporters of tax cuts and unregulated markets, run the party in Congress and Washington. If the Republicans can’t shake their vestigial Reaganism, then President Obama and the Democrats, as the political analyst Ed Kilgore put it, “may have a large window of opportunity to build a majority against an opposition party that’s drunk on the locusts and wild honey of the political wilderness they inhabit.”13
Obama, of course, has to deal with his own rambunctious supporters while facing a drastic economic situation. The Onion, the satirical newspaper, captured the situation well, as it often does, with its November 5 headline: “Black Man Given Nation’s Worst Job.” But if the campaign taught us anything, it should have taught us that he has the wisdom and the patience to exceed expectations. A realignment may not yet be upon us, but 365 electoral votes against a war hero have bled some of a liberal’s natural pessimism out of me.
—November 18, 2008
Go to www.cnn.com/ELECTION/2008/results/polls, and click around for full results and exit polls nationwide and by state. CNN did not break down exit-poll results in the income and education categories by race in 2004, making it difficult to compare Obama's performance among these groups against Kerry's.↩
Frank's response, which I read at the time, seems to have disappeared from his Web site, www.tcfrank.com. This quote is taken from a short excerpt that appears on the site of the History News Network at hnn.us/roundup/entries/19729.html.↩
See Ed Kilgore, "The Anatomy of Conservative Self-Deception," The Democratic Strategist, November 13, 2008.↩
Go to www.cnn.com/ELECTION/2008/results/polls, and click around for full results and exit polls nationwide and by state. CNN did not break down exit-poll results in the income and education categories by race in 2004, making it difficult to compare Obama’s performance among these groups against Kerry’s.↩
Frank’s response, which I read at the time, seems to have disappeared from his Web site, www.tcfrank.com. This quote is taken from a short excerpt that appears on the site of the History News Network at hnn.us/roundup/entries/19729.html.↩
See Ed Kilgore, “The Anatomy of Conservative Self-Deception,” The Democratic Strategist, November 13, 2008.↩