News your connection to The Boston Globe

Who are you calling working class?

The scholarly food fight over how to define the lunch pail class and what Democrats can do to win back their affections - if, that is, they ever lost them

(Getty / Stone)

IF THERE'S ONE thing Republican and Democratic strategists agree on, it's that white, working-class Americans have been fleeing the Democratic Party in droves. The statement seems beyond dispute-just part of the political landscape.

One of the most trenchant elaborations of this view came in the 2004 bestseller ''What's the Matter with Kansas? How Conservatives Won the Heart of America," by the essayist and historian Thomas Frank. He argued that ''the poor," ''the weak," and ''the victimized" were increasingly throwing their lot in with Republicans, swayed by social issues into voting against their economic self-interest. How else to explain why McPherson County, Nebraska, among the poorest in the United States, gave 80 percent of its vote to President Bush in 2000?

Through reportage, historical anecdote, and much scabrous wit, Frank argued that big-business Republicans were gulling these voters, dangling issues like gay marriage and abortion before them during campaigns but doing little to address these issues once elected. What they did implement were economic policies helping the rich. ''If you make over $300,000," Frank wrote, ''raise a glass sometime to those indigent High Plains Republicans as you contemplate your good fortune."

In two recent papers, however, the Princeton political scientist Larry Bartels-who has a barbed pen himself-has attempted to refute some of the assumptions behind Frank's thesis. Last September, Bartels presented a paper at the American Political Science Association meeting in Washington that asked these questions: Are working-class whites abandoning the Democratic Party? Has the working class become more conservative? Within the working class, do ''moral values" trump economics? Wielding data from the National Election Study survey on presidential races from 1952 to 2004, Bartels answered: No, no, and no-and set off a debate that has been echoing in political circles ever since.

In a response published in December on his website, Frank accused Bartels of ''blowing off" the phenomenon of conservative populism, a move he called ''folly on a magnitude that not even a political scientist can measure." His chief complaint, among many, was that Bartels's definition of working class was bizarrely flawed. This month, Bartels presented an essentially rewritten version of his paper at Harvard, responding to Frank's complaints. Though the tone of these salvos might suggest it, the continuing dispute is not a left-right one. Bartels describes himself as a longtime nonpartisan nonvoter who now leans toward the Democrats, because of growing concerns about economic inequality. The high-octane rhetoric instead seems to derive from the importance of what's at stake: an understanding of what constitutes the working class, how its views have changed, and whether the Democratic Party needs to reinvent itself in order to appeal to those voters.

. . .

The dispute has overtones of a food fight between the humanities and the social sciences. Frank has written that Bartels's numbers-and-surveys approach saps all the culture, history, and ideology-not to mention all the vim and life-from politics, while Bartels has dismissed Frank's arguments as the product of a shoot-from-the hip pundit culture, even though Frank holds a doctorate in history from the University of Chicago.

Yet not all empirically minded academics agree with Bartels. Adam Berinsky, an associate professor of political science at MIT, says he finds Bartels's papers ''very compelling," but Alan Abramowitz, a political science professor at Emory, says, ''If [Bartels's] point is that nothing has changed, or that Democrats are doing better than ever among the white working class-I doubt even Howard Dean would agree with that."

That's not quite what Bartels is saying, but it's not all that far off: In his first paper, Bartels found that if you define white working-class voters as those whose income is in the lowest third nationally (families earning $35,000 or less), they have only grown more loyally Democratic in presidential elections since 1976. From 1952 to 1972, considering only the two-party vote, 46 percent of white voters in the bottom third of income voted for Democratic presidential candidates, as did 47 percent in the middle third, and 42 percent in the top third. From 1976 through 2004, however, 51 percent of whites in the bottom third went for the Democrat, as did 44 percent of those in the middle, and only 37 percent at the top.

Far from turning against Democrats, inflamed by the ''hallucinatory appeal" of cultural issues, as Frank put it, the working class has embraced the Democrats. The better off are the ones who are shifting away.

Bartels's paper had other data purporting to show that economic issues still determined the votes of low-income Americans, but in his response, Frank zeroed in on ''a mistake so basic it effectively negates [Bartels's] entire effort": defining the working class as those in the bottom third of family income. ''He's talking about the poor in the guise of talking about the working class," says Ruy Teixeira, a pollster in Washington, DC, who agrees with Frank on this point. Rick Perlstein, author of ''Before the Storm: Barry Goldwater and the Unmaking of the American Consensus" (2001) is harsher: ''He tortured the data badly."

Frank didn't precisely define his working class in his book, but he now suggests that a better definition might be Americans lacking a college degree. He notes that a family consisting of an autoworker, a secretary, and their four kids-a family that seems working-class in common-sense terms-might earn $70,000 a year, which would inch them into the top third of family income. And among Americans without a college degree, the challenge the Democrats face does seem immense. As Teixeira points out, Bush beat Kerry 61 percent to 38 percent among those voters, according to one exit poll.

There are problems with this definition, too, of course: A significant majority of Americans lack a college degree, for example, and plenty of them (Bill Gates springs to mind) are quite well off. Bartels still thinks that income is a better measure of class, but in his second paper, he decided to take the fight to Frank's own turf, defining the white working class as white voters lacking college degrees. While Kerry indeed did poorly within this group, Bartels says his performance wasn't too far off the historical average for Democrats. Graphing the two-party results in Presidential elections from 1952 to 2004, Bartels found that the Democrats lost only 6 percentage points to the Republicans among whites lacking college degrees. And one region in particular is driving that shift: the South.

Within the South, the shift from Democrat to Republican for this group has amounted to 20 percentage points. Outside the South: 1 percentage point.

Considering margins of error, he concludes that the decline of white support for Democrats is ''entirely attributable to the demise of the Solid South as a bastion of Democratic allegiance." The realignment of the South after the Democratic party embraced civil rights is an important story, but it's not the same thing as hardhats bailing on the Democrats nationwide.

And what of the increasing importance of social issues to working-class voters? Using voters' self-descriptions from the National Election Study survey of 2004, Bartels notes that when white voters without college degrees ranked social and economic issues in order of importance to them, gun control, at fifth out of 15, was the highest rated social issue. Abortion languished at 13th. Overall, economic issues carried 50 percent more weight than social issues.

The conclusion Bartels draws is that all the talk of the need to reinvent the Democratic Party is a ''ludicrous overreaction." ''The basic pattern of electoral support is pretty even, and both parties ought to think of themselves as fighting hard at the margins," he says.

Frank, meanwhile, is annoyed that the Bartels paper, which he has already eviscerated once, has become a moving target as a result of Bartels's latest revision. He writes in an e-mail that he won't read the new paper or comment on it until it is published. (It's been accepted by the Quarterly Journal of Political Science.) ''When he has decided what he really, truly wants to stand behind, then I will consider responding," Frank says. When and if Frank and other political analysts do respond, Democratic activists tired of their losing ways will surely be reading closely.

Christopher Shea's column appears in Ideas biweekly. E-mail

Today (free)
Yesterday (free)
Past 30 days
Last 12 months
 Advanced search / Historic Archives