It’s more than a little frustrating trying to follow Democrats’ analysis of social classes in this country. Most of the time now, there are only two classes – the rich (very precisely defined as those with at least $250,000 in annual family income) and the middle class, which includes everybody else. But in the analysis of elections a “working class” shows up, one which is invariably “white” and, it seems, predominantly male.
Most Democrats, and especially the more progressive ones, know that moving the white working class away from its decades-long lopsided loyalty to the Republican Party is crucial to achieving a long-term governing majority. But instead of appealing to this demographic electoral block directly, it seeks to lump them in with what Dems think is a universally beloved “middle class.” This is a tactical mistake, as in many working-class precincts calling somebody “middle class” is meant as a put down and an insult – somebody who doesn’t live “real life,” lacks common sense, and yet thinks they’re “all better.” Believe me, I’ve been on the front end of this insult, sometimes deservedly so.
Of all the ways of defining class in America the one that gets the least attention is how people self-identify – that is, what class people see themselves as being in. In exit polls, for example, you get a choice of “White, Black, Hispanic/Latino, Asian, or Other” in defining your race. There is no such question for class. Rather, pollsters ask questions about education and income, and then analysts assign people to various classes based on the analysts’ own definitions. As is often pointed out on this site, the one national survey that consistently asks people to identify themselves by class has for decades found about 46% self-identify as “working class” and another 46% as “middle class.” Nobody has any idea how voters who see themselves as working class have actually voted — ever.
Over the last decade, through what has often been a rich debate among political scientists, journalists, political operatives, and statisticians, the presence or absence of a bachelor’s degree has come to be used as a marker identifying voters as either “working class” or “middle class.” Because having a bachelor’s degree correlates pretty strongly with having a professional or managerial job and because these jobs correlate with higher incomes, this is a serviceable marker for “middle class.” Likewise, because the two-thirds of jobs that are not professional or managerial usually do not require bachelor’s degrees and have lower average incomes, the absence of a bachelor’s degree is a good-enough way of locating the “working class” among voters. Until exit-pollsters provide voters with a range of choices on class, as they do now for race, this education marker is the best we can do in measuring how social class affects voting.
Problem is that in the last two elections, these two broad classes voted almost exactly the same way. In 2008 both “college graduates” and “no college degree” voters voted for Barack Obama by a margin of about 53% to 46%, whereas both groups in 2010 voted 52% to 46% for Congressional Republicans. So, there was a big swing in the last two years, but both the working class and the middle class swung exactly the same way and to the same degree. Thus, class by itself seems not to affect how people vote.
If, however, you measure class along with race, then class matters a bit more. Neither class of whites gave Obama a majority in 2008, but middle-class whites gave him 47% of their vote, while working-class whites gave him only 40% of theirs. Meanwhile, among non-white voters (lumping together all “Black, Hispanic/Latino, Asian and Other” voters), there was a similar degree of difference by class but in the opposite direction – working-class non-whites gave Obama a larger majority (83%) than middle-class non-whites (75%). A similar race-class pattern occurred in the 2010 Congressional elections, with working-class whites giving Republicans 62% while middle-class whites gave them 57%, whereas working-class non-whites were more decisively Dem at 77% than middle-class non-whites at 71%.
Two conclusions emerge from this breakdown:
One is that race matters way more than class. In fact, very few large groups of whites have voted majority Democratic at the national level for decades. Using only the exit polls, which do not cover all possible groupings, the only whites who gave Obama a national majority in 2008 were Jews (83%), whites with “no religion” (71%) or “other religion” (67%), and 18-to-29-year-olds (54%) – though it is important to add that Obama won white majorities in 19 states and in the Northeast as a whole.
The other conclusion is that the single largest race-class grouping, the base of the base of the Republican Party in America, is working-class whites. Even though declining as a proportion of the electorate (as non-whites increase faster in the population and as more whites get bachelor’s degrees and are, therefore, no longer considered “working class”), working-class whites are still almost two of every five voters, and until 2010 they had been voting in the neighborhood of 60/40 for the GOP in national elections.
In parts of the country outside the South, however, the white working-class, like whites in general, has been drifting toward the Democrats over the past few decades, culminating in the 2008 election when, for example, Obama won majorities of white workers in 14 states and got into the high 40s in four others. That drift was reversed big time in the 2010 Congressionals. According to the guru on these matters, Ruy Teixeira: “The most significant shift against the Democrats [in 2010] occurred among the white working class. Congressional Democrats lost this group by 10 points in both 2006 and 2008. Yet that deficit ballooned to 29 points in 2010.”
That’s a huge move toward Republicans who were against saving the American auto industry and who voted against infrastructure investments and jobs, (very) partial bailouts of state governments, extensions of unemployment insurance, and health care reform and tax policies that benefit working-class whites more than any other race-class grouping (in absolute numbers though not proportionately). And this massive swing occurred nowhere more strongly than in the Great Lakes states, including strong union states Pennsylvania, Ohio, Michigan, Illinois, and Wisconsin.
What accounts for this swing of previously Democratic white working-class voters in 2010 will be the subject of my next blog. Until then, I can do no better than recommend that all Democrats look at a conservative Republican’s class analysis of “Midwest at Dusk.”
Jack Metzgar, Chicago Center for Working-Class Studies