Updated at 5:01 p.m. on Jan. 21.
America's increasing racial diversity made it possible for Barack Obama to comfortably win the presidency last year while capturing only 43 percent of the white vote, according to exit polls. And if the minority share of the vote continues to grow roughly as fast through 2012 as it has since 1992, Obama could win re-election with even less support from whites.
But even so, the depth -- and breadth -- of the decline in his support among whites ought to be giving Democrats heartburn. That's especially so because a falloff in minority participation during the midterm election makes it likely whites will likely cast a slightly larger share of the vote in November than the 74 percent they did in 2008. In each of the three major elections since 2008 -- the GOP's wins in the 2009 New Jersey and Virginia gubernatorial races and Republican Scott Brown's stunning upset victory Tuesday in the Massachusetts Senate contest -- the GOP candidate ran notably better among whites in their state than John McCain did against Obama in 2008.
In the latest Allstate/National Journal Heartland Monitor poll, released last week, Obama's approval rating among whites stood at just 38 percent, down from 53 percent in the first survey last April. In the new poll, the share of whites who strongly disapprove of his performance (39 percent) is double the portion of whites who strongly approve (19). Last April, whites who strongly approved of his performance outnumbered whites who strongly disapproved by 33 percent to 25 percent.
In one sense, the trend among whites isn't surprising. While Obama's approval among African-Americans is virtually unchanged in Heartland Monitor polls since last April, it has declined among Hispanics. Tough times hurt a president everywhere: As John F. Kennedy might have phrased it, a receding tide lowers all boats.
Obama's difficulties with whites are broad-based. Since he emerged as a national candidate, non-college white men have been the toughest group for him, and they still are. His approval rating among them stands at just 34 percent in the latest survey, down even from the meager 39 percent of their votes he won in 2008. More blue-collar white men strongly disapprove of Obama (39 percent) than approve of him at all.
Though the Obama campaign had high hopes for non-college white women, the so-called waitress moms, they didn't give Obama many of their votes in 2008 (he carried just 41 percent of them). But in last April's Heartland Monitor survey, they appeared ready to give him a second look, with 53 percent approving; he now draws just 37 percent among these women.
Obama's numbers are also bleak among college-educated white men. These men, admittedly, usually prefer the GOP: Obama won just 42 percent of them in 2008, which was actually more than Bill Clinton, Al Gore or John Kerry managed. In the new survey, Obama's approval rating among this group stands at just 38 percent, and the share of them who strongly disapprove of his performance has reached a head-turning 49 percent.
But for Obama the most troubling trend is his trajectory among college-educated white women. This is the most Democratic-leaning group in the white electorate: Obama won 52 percent of their votes last November, and the Democratic nominee carried them in three of the four presidential elections before that. But Obama's approval rating with these well-educated, generally socially liberal women has slumped from 64 percent in last April's Heartland Monitor to 46 percent in the latest survey.
Perhaps even more important are the skeptical reviews from all segments of the white electorate, including college-educated white women, about the longer-term impact of Obama's policies. For instance, when asked how the president's agenda would affect people like them, 32 percent of college-educated white women said it would increase their opportunities to get ahead, far less than the 45 percent who said his agenda would diminish their opportunities. That's a sharp decline just from last September, and actually worse than Obama's showing on that question among non-college whites. (College white men are the most pessimistic group, with 26 percent expecting their opportunities to increase and 43 percent expecting them to decrease.)
Likewise, asked about the impact of Obama's economic policies over the past year, 53 percent of college-educated white women agreed they had "run up a record federal deficit while failing to end the recession or slow the record pace of job losses," compared to 36 percent who thought the president's policies had helped "avoid an even worse economic crisis and are laying the foundation for our eventual economic recovery." That's also a substantial decline just from last September, and again not much different than the results among the other key segments of the white electorate. (College-educated white men are especially dubious on this question, too, picking the negative choice by a nearly 2-1 margin.)
The well-educated women join the skeptics on a third key question: the impact of Obama's agenda on the country over the past year. Just 10 percent of these well-educated women say that the country is "significantly better off because of the policies Obama has pursued"; 41 percent say the country "is significantly worse off" because of those policies. The largest group (45 percent) say Obama's policies haven't produced significant results yet but are beginning to move the country in the right direction.
For Obama, there remains a potential majority between the 10 percent of these women who see progress already and the 45 percent who believe it is still possible. But the skepticism radiating from them is nonetheless striking: Obama's showing among the college-plus white women on this pivotal question is no better than his performance among the college-plus white men -- traditionally one of the toughest groups in the electorate for Democrats. The college-educated women actually render a more negative verdict than either non-college white men or women -- groups that provided Obama far less support against McCain.
The findings on these underlying questions are what separate the trend among whites and Hispanics. Obama's approval among Hispanics has also tumbled from 81 percent last April to 60 percent in the new poll. But Hispanics retain much more optimism than any category of whites about the potential long-term impact of his agenda. Hispanics, by a commanding 60 percent to 22 percent margin, still believe Obama's agenda will increase opportunities for people like them. And while they offer a more equivocal verdict on the impact of his economic policies over the past year (only a narrow plurality picks the positive choice), they are much sunnier about the long-term implications of his approach: 71 percent of them say that while his agenda hasn't produced big benefits yet, it is moving the country in the right direction.
There's also a clear racial divide on a question that measures bedrock political philosophy. Asked what role government should play in the economy, just 17 percent of Hispanics (and an even more meager 13 percent of African-Americans) agree with the Reagan-like sentiment that "in the current economic environment, government is not the solution to our economic problems; government is the problem." About four in 10 of both Hispanics and African-Americans take the opposite view that government must play an active role "in regulating the marketplace and ensuring that the economy benefits people like me." Roughly the same portion of both groups say that they would like government to play an active role but are not sure they can "trust government to do this effectively."
The breakdown among whites is very different. Just one-fourth of whites from all four groups -- college and non-college men and women -- align with the activist position. Far more back the Reaganite view that government is the problem: 48 percent of college-educated men, 44 percent of non-college men and a surprisingly high 44 percent of college-educated women. Only the waitress moms largely resist that perspective, with just 31 percent concurring. The remainder in each group are open to an active government but aren't sure it can be effective.
Whites are declining as a share of the electorate, and over the long term, the electorate's increasing diversity remains a great asset for Democrats at every level. The 2008 presidential election was the first in which minorities cast more than one-quarter of the vote (26 percent, to be precise), and Obama won a cumulative four-fifths of their ballots. If he roughly maintains that support and the minority share of the vote increases at about its rate over the past two decades (say, to 28 percent in 2012), the math says Obama could win re-election with less than 39 percent of the vote among whites.
It's not an experiment his political team is likely to be eager to attempt, though. And congressional Democrats have even more reason for concern about Obama's eroding standing among whites. Growing diversity has been a huge boon to them, too: The number of House districts where minorities constitute at least 30 percent of the population has roughly doubled, from one-fourth in the 1990s to one-half now. Those highly-diverse districts are the foundation of the Democrats' House majority: The party holds more than 70 percent of them, while Republicans maintain a narrow edge over Democrats in the less-diverse (or, put another way, more heavily white) districts.
Yet some of the Democrats' biggest gains since 2004 have come in predominantly white places; those Democratic gains have divided about evenly between districts in which the white population is mostly white-collar and well-educated (like the suburban Philadelphia seat that Democratic Rep. Joe Sestak is vacating to run for Senate), and districts crowded with the blue-collar whites most alienated from Obama (like the seats held by Reps. Walt Minnick in Idaho and Zack Space in Ohio).
If the economy doesn't improve, Democrats of all sorts will face difficult elections in November. But the threat is probably greatest for those representing predominantly white, largely downscale constituencies. The depth, persistence and intensity of the blue-collar resistance to Obama may virtually guarantee a long night for all but the most entrenched House Democrats whose districts require them to win support from a significant number of non-college whites. (The announced retirements of Democrats like Tennessee's John Tanner and Arkansas' Vic Snyder reflect a recognition of that likelihood.) The same will likely be true in Senate elections; just ask Arkansas Democrat Blanche Lincoln.
Scott Brown's victory in Tuesday's Massachusetts Senate race captures that risk. Conclusive answers on the demographic pattern in the voting are not available because exit polls were not conducted for the race. One election-day survey conducted for the Democratic group Women's Voices, Women Vote showed little difference between the preferences of whites with and without a college education.
Other data, though, suggest Coakley's difficulties were most pronounced among blue-collar whites. A separate election-day survey by the Republican polling firm Fabrizio, McLaughin & Associates did not report results by education, but found that Brown ran even with Coakley in union households. The final private tracking poll for Brown reached essentially the same result on union households, and also found him with a much larger lead among all voters without a four-year college degree than those with one. Most emphatically, an election night survey of Massachusetts voters conducted for the AFL-CIO by the Democratic firm Hart Research Associates found a huge education gap among whites. In that survey, Coakley actually led Brown 51 percent to 49 percent among whites with a college education. But among non-college whites, Brown crushed her 63 percent to 37 percent, according to Guy Molyneux, who conducted the poll.
Analysis of town-level election results reinforces those findings. Brown did best in smaller towns in northeastern, central and southern Massachusetts beyond Boston and its affluent suburbs. In many of the state's well-educated, white-collar communities -- such as Brookline, Lexington, Weston, Newton, and Cambridge -- Coakley matched or exceeded the Democratic share of the vote in the 2006 and 2002 gubernatorial races. (In Brookline, for instance, she equaled the 75 percent Gov. Deval Patrick captured in his comfortable 2006 victory; in Lexington she matched his 65 percent; and in Cambridge she bested his 82 percent by three percentage points.) But she declined, sometimes substantially, from the recent Democratic gubernatorial vote in many prominent blue-collar places. In Fall River, she finished 12 percentage points behind Patrick's total; in Lowell 7 points; in New Bedford 11 points; in Taunton 14 points; in Fitchburg 16 points. These findings fit with a New York Times computer analysis which concluded that the towns in which Coakley's vote declined the least from Obama's 2008 performance tended to have many more college graduates than the state overall.
The precedent for this pattern of blue-collar distress is the GOP landslide in 1994, which swept Republicans to control of both chambers in Congress. The explosion in the Republican vote from 1992, when Democrats won 258 House seats, to 1994, when Republicans won 230, came almost entirely among blue-collar whites. Exit polls showed that the Republican share of the total House vote actually declined slightly from 1992 to 1994 among minorities and remained almost unchanged among college whites (rising only from 54 percent to 56 percent, a statistically insignificant difference). But Republicans surged from 47 percent of the vote among non-college whites in 1992 to 61 percent in 1994, powering their historic breakthrough.
A similar GOP blue-collar surge in 2010 could take a sharp bite out of the Democratic majority. But it isn't as threatening to Democrats now as in 1994 because those blue-collar whites are a smaller share of the electorate. In 1994 they accounted for almost exactly half of all votes. In 2008, they cast about four in 10 ballots; though the figure could rise slightly in 2010, it almost certainly won't approach the 1994 level. Conversely, even if minorities decline from their record share in 2008, they will cast a much larger share of the vote in 2010 than 1994 (when they were just 13 percent of the total). Republicans might make some gains with minorities compared to their dismal 2008 showing, but it's virtually certain Democrats will still carry a comfortable majority of their votes.
These calculations suggest that to take back the House this year, Republicans will almost certainly need to post significant gains not only among blue-collar whites (as in 1994) but also among better-educated whites. Put another way, to prevent a difficult election from becoming a disastrous one, Democrats will need to limit their losses primarily to the downscale places and establish a fallback line of defense in more upscale districts. In that effort, Democrats probably can't count on much help from college-educated men, who are palpably moving toward the GOP. Which means that to avoid the worst in November, Democrats must reassure the college-educated women now expressing so much skepticism about their performance -- or at least convince those women that they would like what Republicans are offering even less.
National Journal reporter Jim Barnes and researcher Cameron Joseph contributed to this report.