Guest Post ... Ron Brownstein

College Women and Blue Collar Men:
Holding the Democratic Primary Nomination in Their Hands

Why is the Iowa Democratic presidential race so close? One key reason is that former Sen. John Edwards is executing a demographic squeeze on both Sen. Hillary Rodham Clinton and Sen. Barack Obama, the two candidates well ahead of him in the national polls.

From the outset of her campaign, Clinton in almost all surveys has run better among women than men and better among voters without a college education than those with college degrees or more. In most polls, in most places, Obama’s support has followed the reverse pattern-more male and particularly more college- educated. Edwards’ coalition hasn’t been as sharply defined: in almost all surveys, he has drawn about evenly from both men and women, and both better and less-educated voters.

The new Los Angeles Times/Bloomberg News poll in Iowa released Thursday shows this pattern largely persisting. But detailed data from the survey provided by Times Polling Director Susan Pinkus shows the dynamic also shifting in ways that carry important implications not only for Iowa but potentially other states to follow.

(Atlantic Media Political Director RON BROWNSTEIN)

These trends have produced starkly contrasting patterns of support for the two top contenders, Clinton and Obama. Clinton has been strongest in the place where her two advantages intersect: among women with less than a college education. Obama, inversely, has run best among college men-the point where his two advantages converge. The swing groups in the Democratic primary have been those conflicted between class and gender: non-college men, and college-educated women.

Among the non-college women-a group often described as waitress moms-Clinton still leads comfortably, if not quite as massively as she has in some earlier polls: the Times/Bloomberg survey found these voters dividing 41% for Clinton, 24% for Obama, and just 20% for Edwards. Among the college men, Obama is still strong, and Clinton still weak, but Edwards has emerged as a formidable competitor and is pressuring Obama in his stronghold: college men, in the survey, broke 31% for Obama, 29% for Edwards, and 15% for Clinton.

Looking just at Clinton and Obama, then, the New York Senator leads her Illinois rival by 17 percentage points with her strongest group (the non college women), while he leads her by 16 points with his strongest group (the college men). The Clinton camp can easily accept that trade because the non-college women outnumber the college men in the survey’s likely electorate by almost three-to-one.

But Clinton is struggling with the two key swing groups. At her highest points this fall, much of her gain in national surveys came from swelling support among college-educated women-many of whom were initially cool to her. In Iowa, those doubts have resurfaced: now she only attracts 34% of college-educated women, compared to 35% for Obama and 13% for Edwards.

In most places, in most surveys, Clinton also ran well through the fall among blue-collar men. But in the Times/Bloomberg Iowa survey, she attracted only 22% of them. That was still more than Obama (just 20%). But Edwards, with his biting populist message, now draws 35% of blue-collar Iowa men.

In effect, then, Edwards is siphoning away votes both Obama and Clinton probably hoped to attract here. Because Edwards is now stronger among men than women, he is hurting Obama (especially among college men). But because Edwards is stronger downscale than upscale, he is hurting Clinton (especially among the non-college men).
All of these cross pressures have produced a race far too close to call.

Another Times/Bloomberg survey in New Hampshire found similar patterns emerging among Democrats there too. Clinton led Obama by 17 percentage points (42% to 25%) among the non-college women, almost exactly the same advantage (40% to 22%) Obama enjoys over her among college men. (Her group again outnumbers his, though not as widely as in Iowa.) And, as in Iowa, the college women are divided almost evenly between them, with 30% preferring Obama, and 28% Clinton.

But in New Hampshire, Obama has opened a significant advantage among blue-collar men, leading Clinton 36% to 19%. If Obama could hold that advantage in New Hampshire, and extend it to other states, it would be critical because he has generally struggled with that group this year (and obviously still is struggling with them in Iowa).

In contrast to Iowa, Edwards runs third in New Hampshire with all four groups and doesn’t show much variation by gender or education; his support in the Granite State has reverted to the prairie-flat pattern evident in most national polls.

The momentum that the winners in Iowa and New Hampshire will acquire-and the deflation suffered by the losers-could scramble all of these patterns at least somewhat. But Clinton’s hold on non-college women, and Obama’s appeal to college-educated men, is now so consistent in polls that it doesn’t appear likely that anything short of a complete campaign collapse by either candidate will fundamentally dislodge it. That may leave the nomination in the hands of the college women and blue collar men that are displaying the greatest ambivalence about-and least durable attachments to-all of the leading contenders.

Ronald Brownstein.