The Democrats' White Flight

New data show two broad trends: the exodus of whites from the Democratic Party and America's divergent views along racial lines

By any standard, white voters' rejection of Democrats in November's elections was daunting and even historic.

Fully 60 percent of whites nationwide backed Republican candidates for the House of Representatives; only 37 percent supported Democrats, according to the National Election Poll exit poll conducted by Edison Research. Not even in Republicans' 1994 congressional landslide did they win that high a percentage of the white vote.

Moreover, those results may understate the extent of the white flight from the Democratic Party, according to a National Journal analysis of previously unpublished exit-poll data provided by Edison Research.

The new data show that white voters not only strongly preferred Republican House and Senate candidates but also registered deep disappointment with President Obama's performance, hostility toward the cornerstones of the current Democratic agenda, and widespread skepticism about the expansive role for Washington embedded in the party's priorities. On each of those questions, minority voters expressed almost exactly the opposite view from whites.

Much can change in two years--as Obama's own post-2008 odyssey demonstrates. These results, however, could carry profound implications for 2012. They suggest that economic recovery alone may not solve the president's problems with many of the white voters who stampeded toward the Republican Party last year. "It comes down to that those voters are very skeptical of the expansion of government," says Colorado Republican Party Chairman Dick Wadhams, a veteran strategist. "The voters who went with Obama in 2008 did not know what they were going to get with that vote. Now that they've seen the health care bill, the stimulus bill, the bailout, the cap-and-trade proposal--issue after issue, they don't like what they see."

That resistance could, in turn, increase the pressure on Obama to accelerate the generation-long transformation of the Democratic electoral coalition that he pushed forward in 2008. With so much of the white electorate, especially working-class whites, dubious about the president's direction, to win a second term he will likely need to increase turnout and improve his showing among the groups that keyed his 2008 victory--minorities, young people, and white-collar white voters, especially women. In 2012, Obama may be forced to build his Electoral College map more around swing states where those voters are plentiful (such as Colorado, North Carolina, and even Arizona) and less on predominantly blue-collar and white states such as Ohio and Indiana that he captured in 2008.

David Axelrod, Obama's chief political strategist, said in an interview that "it would be a mistake to take exit polls from a midterm election and extrapolate too far" toward 2012. Conditions--and the composition of the electorate--will change a great deal by then, he said. But he acknowledged that Obama must "reset" the public perception about his view of government's role. Axelrod, who plans to return to Chicago next month to help direct the president's reelection campaign, also made it clear that he sees as a "particularly instructive" model for 2012 the case of Democratic Sen. Michael Bennet in Colorado, who won his contest last fall by mobilizing enough minorities, young people, and socially liberal, well-educated white women to overcome a sharp turn toward the GOP among most of the other white voters in his state.

Given the trends among the white electorate evident in these exit-poll findings, that formula might represent Obama's most promising path to a second term. Because the 2010 elections dealt such a heavy blow to the Democrats' old models of electoral success, the imperative of electoral transformation is looming ever larger for the president. "He has to make an effort to reclaim some of the lost [white] vote," says Simon Rosenberg, president of NDN, a Democratic analysis and advocacy group. "But he's got to push the new electorate harder."


After Election Day, several media outlets released exit-poll data breaking down the contrasting level of support among white and minority voters for Republican and Democratic congressional candidates. But they did not publish results that separated by race the responses to questions that measured attitudes about Obama's performance, the state of the economy, the national agenda, and the way voters described their own ideology. It was those additional race-specific results that National Journal recently purchased from Edison Research, the organization that conducts the exit surveys. These polls provide an unusually valuable lens through which to assess such attitudes, because surveyors interview many more respondents (17,504 in the national survey this year) than in a typical poll.

From every angle, the exit-poll results reveal a new color line: a consistent chasm between the attitudes of whites and minorities. The gap begins with preferences in the election.

After two years of a punishing recession, minority support for House Democrats sagged in this election to the lowest level recorded by exit polls in the past two decades, according to calculations that Alan Abramowitz, a political scientist at Emory University, provided to National Journal. The Hispanic vote for Democrats in House races slipped to 60 percent, compared with about two-thirds for Obama in 2008 (although some Hispanic analysts say that other data indicate a better showing for Democrats last year). But even so, a solid 73 percent of all nonwhite voters--African-Americans, Hispanics, Asians, and others--backed Democratic House candidates in the midterm election, according to the new analysis.

Meanwhile, Republicans, with their 60 percent showing, notched the party's best congressional result among white voters in the history of modern polling. Media exit polls conducted by Edison Research and its predecessors have been tracking congressional elections for about three decades. In no previous exit poll had Republicans reached 60 percent of the white vote in House races. The University of Michigan's National Election Studies, a biennial pre- and postelection poll, is another source of data on voting behavior dating to 1948. Republicans had never reached 60 percent of the congressional vote among whites in any NES survey. Only in the NES surveys had Democrats reached that 60 percent congressional support level among white voters: in their 1974 post-Watergate landslide and in Lyndon Johnson's 1964 rout of Barry Goldwater.

November's gap between the voting preferences of whites and minorities was at the wider end of the range over the past two decades but it wasn't the absolute widest. More striking was the disparity between the two groups' views on other questions with implications for the 2012 election.

First among those was Obama's performance. Exactly 75 percent of minority voters said they approved; only 22 percent said they disapproved. Among white voters, just 35 percent approved of the president's performance, while 65 percent disapproved; a head-turning 49 percent of whites said they strongly disapproved. (Those whites voted Republican last fall by a ratio of 18-to-1.)

The racial gulf was similar when voters were asked whether they believed that Obama's policies would help the nation in the long run. By 70 percent to 22 percent, minorities said yes; by 61 percent to 34 percent, whites said no. On election night, much attention focused on the exit-poll result that showed voters divided almost exactly in half on whether Congress should repeal the comprehensive health care reform legislation that Obama signed last year or should preserve or even expand it. But that convergence obscured a profound racial contrast. The vast majority of minority voters said they wanted lawmakers to expand the health care law (54 percent) or maintain it in its current form (16 percent), while only 24 percent said they wanted Congress to repeal it. Among white voters, the sentiments were almost inverted: 56 percent said that lawmakers should repeal the law, while much smaller groups wanted them to expand it (23 percent) or leave it alone (just 16 percent).

"The issues we'll burnish are ones that will resonate better with some of these [disaffected  white] voters." --White House political strategist David Axelrod

The gap was also wide in attitudes about two fundamental tenets. Minorities were almost exactly twice as likely as whites to say that life would be better for the next generation than for their own; whites were considerably more likely to say that it would be more difficult. And on a question measuring bedrock beliefs about the role of government, the two racial groups again registered almost mirror-image preferences. Sixty percent of minorities said that government should be doing more to solve problems; 63 percent of whites said that government is doing too many things that would be better left to businesses and individuals.

The irony in these results is that minorities expressed more faith in both the future and the government than whites did, even though the recession has hit minority communities harder. Rodolfo de la Garza, a political scientist at Columbia University who studies Hispanics' attitudes, says that part of the explanation is that whites found the downturn more psychologically wrenching because more of them (especially white-collar whites) had expected to make a steady ascent up the economic ladder. More minority workers hold marginal positions in the private economy, he says, so they were less likely to be shocked by the severity of the downturn--and more likely to turn to government, rather than the private sector, to help survive it. "They didn't lose money on Wall Street; they had shitty jobs, if they had jobs, so where would they look to if not the [government]?" de la Garza asked.

Polls have consistently shown that whites, by contrast, have aimed more of their economic frustration at government than at corporations. That reversed a warming toward government activism during President Bush's second term that helped drive the Democratic breakthroughs in the 2006 and 2008 elections. Obama and the Democratic Congress expanded government's role across a wide range of issues precisely at the moment when white voters' confidence in Washington hit rock bottom. That collision partly explains the force of the backlash in November. "This is a fundamental transformation [of attitudes] going back to where it was before 2006 and 2008," Democratic pollster Mark Mellman says. "Part of it was occasioned by the economy [since 2008]; part of it was occasioned by the response to the economy. People felt government did a lot of big things that were inappropriate. They felt government took care of the big guys--and not me."


Measured both geographically and demographically, these new exit-poll results show that Democrats maintained openings in only slivers of the white electorate. In House elections, the bottom fell out for Democrats in both the South (where they won just 24 percent of whites) and the Midwest (37 percent). The party remained relatively more competitive along the coasts, capturing 46 percent of white voters in the East and 43 percent in the West.

A separate National Journal analysis of the results from exit polls in Senate elections found similar trends. Edison Research conducted exit polls last year in 26 Senate races; in 19 of them, the Democratic Senate candidate won a smaller share of the white vote than President Obama captured in the state two years earlier. Democratic Senate candidates carried a majority of white voters in just seven races and reached 45 percent of the vote in only two more. Except for West Virginia, those states were all near an ocean (or, in Hawaii's case, in one).

Democrats have been losing support among blue-collar white voters since the 1960s, but in this election, they hit one of their lowest points ever. In House campaigns, the exit poll found, noncollege whites preferred Republicans by nearly 2-to-1 with virtually no gender gap: White working-class women--the so-called waitress moms--gave Republicans almost exactly as many of their votes as blue-collar men did.

These blue-collar whites expressed profound resistance to Obama and his agenda. Just 30 percent of them said they approved of the president's job performance (compared with 69 percent who disapproved). Two-thirds of them said that government is doing too many things. An approximately equal number said that Obama's agenda will hurt the country over the long term. Only about one-fifth of these voters said that the stimulus had helped the economy, and 57 percent wanted to repeal the health care law--even though they are uninsured at much higher rates than whites with more advanced education.

"The significance of the tea party is that it is not a situational vote." --Jeff Bell, American Principles Project

In Senate races, the story was no better for Democrats: They won majorities of white voters who don't have a college education in just three states and garnered at least 45 percent in only two more. Even Democratic Sens. Barbara Boxer of California and Michael Bennet of Colorado, each of whom ran well among upscale whites, won only about one-third of working-class white voters. In Wisconsin, those blue-collar whites doomed Democratic Sen. Russell Feingold: He carried most minority voters and a thin 51 percent of college-educated whites, but he was crushed among working-class whites, who gave him only 40 percent of their votes.

Merle Black, a political scientist at Emory University, says that blue-collar disaffection from Democratic candidates reflects not only immediate economic distress but also a longer-term process of alienation from the party. "The noncollege whites ... see themselves as a declining minority within the national Democratic Party, where they have very little control or influence on the policies," he says. "The party is controlled by the coastal elites and nonwhites, and that is a very different kind of Democratic Party" than a generation ago.

Compared with 2008, Democrats lost ground among college-educated whites as well, but they maintained more support in this group than among blue-collar whites. Democratic Senate candidates won at least half of the votes of college-educated whites in 10 races and at least 45 percent in two others. Almost all of those states are along the East or West coasts or in the Upper Midwest, the regions that have been the foundation of the Democrats' Electoral College map since Bill Clinton's time. In heartland states such as Arkansas, Missouri, Ohio, and even Illinois, Democratic support cratered among college-educated whites.

White-collar men and women also parted ways much more significantly than their blue-collar counterparts did. College-educated white men backed Republican House candidates and registered negative views of Obama's job performance as overwhelmingly as blue-collar whites did. College-educated white women, though not immune to these trends, displayed more resistance. Although traditionally the most liberal portion of the white electorate, even these women cooled toward Democrats last year. In contrast to the majority support they provided Obama in 2008, they voted 55 percent to 43 percent for Republicans in 2010 House races. In the exit poll, most of them agreed that government was trying to do too much, and a slim majority of them said they wanted Congress to repeal the health care law.

In key Senate races, however, especially in culturally more liberal states, these women backed Democrats in substantial numbers. Both Bennet and Boxer, for instance, carried about three-fifths of this bloc, which proved essential to their victories. Obama's popularity among these college-educated women deteriorated, but in the exit polling, 45 percent of them still said they approved of his performance, far higher than the rate among most other whites.

Even in the tide of discontent that propelled almost all voters toward Republican candidates, relatively more of well-educated white women remained loyal to Democrats. The same was true among all young white voters. Fewer of them backed Democratic congressional candidates than voted for Obama in 2008, but whites under 30 gave Democrats a much higher share of their vote than did older whites. Those two groups--young people and college-educated women--are the splintering foundations on which Obama will likely have to build any hope of a recovery in the white electorate for 2012.


These emphatic 2010 results represented another shovel of earth on the grave of the New Deal electoral coalition, centered on working-class whites, that long anchored Democratic politics. But the decline of that coalition began long before Obama or House Speaker Nancy Pelosi. No Democratic presidential candidate since Jimmy Carter in 1976 has captured as much as 45 percent of white voters, according to exit polls. And not since 1992 have whites given half or more of their votes to Democratic congressional candidates. The erosion has been especially pronounced among the white working class: No Democratic presidential nominee since 2000 has won more than 40 percent of its votes.

Despite that decline, Democrats have survived, and at times thrived, by building a new coalition. They have won the overall popular vote in four of the past five presidential elections, and they recaptured Congress in 2006 with a coalition that now revolves primarily around young people, minorities, and college-educated whites, especially women. That so-called coalition of the ascendant offers Democrats long-term advantages because all of those groups are growing as a share of the population.

Minorities, most important, more than doubled their share of the vote from 12 percent in 1992 to 26 percent in 2008. In his victory that year, Obama won only 43 percent of the white vote (and merely 40 percent among noncollege whites). Yet he captured a larger share of the overall popular vote than any Democratic nominee since Lyndon Johnson in 1964 by winning 80 percent of that growing pool of nonwhite voters, along with majorities among whites under 30 and college-educated white women.

But if 2008 demonstrated the possibilities of that new alignment, the 2010 election demonstrated its limits. It has proven to be a boom-and-bust coalition because turnout in midterm elections usually declines modestly among minorities and sharply among young people; both groups fell off even more than usual in 2010, producing an older and whiter electorate that compounded the GOP's advantage. "We have gotten to the point where we have two different electorates: presidential and nonpresidential," says veteran Democratic consultant Bill Carrick of California.

Equally significant, although racial diversity is spreading and education levels are rising, these trends are not evenly distributed across the country. As a result, the Democrats' coalition of the ascendant is much more potent in coastal states than in most interior states still dominated by white voters, many of them older and working-class. In 137 House districts, at least 80 percent of the population is white; after November, Republicans control a crushing three-fourths of those seats. And, as Feingold discovered, there are not enough minority and well-educated white voters to win Senate races in many interior states if Democrats cannot remain competitive among blue-collar whites.

Finally, Democrats also discovered last year that they cannot rely on cultural affinity alone to hold most well-educated whites who become dissatisfied with the party's economic performance. Some of the most ominous midterm results for Obama came in Pennsylvania and Illinois, where white-collar white voters who had been crucial to his victories two years earlier flocked to the GOP's Senate nominees.

Partly because the minority share of the vote will almost certainly rise again in 2012, Obama probably won't need to match his 2008 percentage of the white vote to win a second term. But all of these considerations suggest that he and the party's congressional candidates must nonetheless improve on their historically low 2010 showing to avoid further losses in 2012. "At the levels of [white discontent] you are talking about, no amount of surge voting [from minorities and young people] is going to overcome that," says Mike Podhorzer, deputy political director of the AFL-CIO.

So one critical question is how much of the white disaffection from Democrats evident in 2010 is rooted in irrevocable ideological alienation and how much will dissolve if the economy improves. According to veteran conservative strategist Jeff Bell, all signs suggest that Obama has permanently antagonized much of the white electorate (nearly half of which this year identified itself as conservative in the exit poll). "The significance of the tea party is that it is not a situational vote," says Bell, the policy director at the American Principles Project, a right-leaning advocacy group. "They are going to be militant even if, or when, the economy improves.... It's significant if you have more voters who are willing to vote with the conservative coalition regardless of what's going on with the economy."

Axelrod, not surprisingly, disagrees. He said he did not consider it unexpected that working-class white voters, in particular, turned so harshly against Democrats this year, "because they have borne the absolute brunt" not only of this downturn but the longer-term stagnation in living standards. But with the economy at least stabilizing, Axelrod contends, Obama  will have an opportunity to define himself less in reaction to crisis and more through issues of his own choosing that could appeal more to whites (and other voters) who have cooled toward him since 2009.

One example is the president's recent declaration that the United States faces a new "Sputnik moment" that demands a more systematic strategy to compete with international economic powers such as China and India. Over the next two years, Axelrod added, Obama will return more consistently to other themes from his celebrated 2004 Democratic convention speech and his 2008 campaign, such as overcoming partisan divisions, reforming Washington, and molding government's "important but limited role" in American life. "We have to reclaim our fundamental message equities from 2008," Axelrod says. "The issues we'll burnish are ones that will resonate better with some of these [disaffected white] voters, because we'll have an opportunity to choose them."

To the extent the economy rebounds, that would also boost Obama with some of the white voters who embraced the GOP in 2010. But short of a roaring financial recovery, many analysts in both parties believe that Obama will find it difficult to fully reconnect with most of the white voters who have drifted away from him. "I think a large majority of those voters are gone for good; I don't know what he can do to change their impression of his view of government," Wadhams, the Colorado GOP chairman, says.

But Wadhams quickly adds that Obama might be able to persuade some of those voters to support him anyway in 2012 if Republicans select a nominee they find unacceptable, particularly on social issues. Wadhams has painful recent experience with that phenomenon: Despite widespread dissatisfaction with Washington, Bennet won reelection to the Senate last fall partly because so many white-collar Colorado suburbanites (especially women) found Ken Buck, his tea party-infused Republican opponent, too conservative on abortion and other issues. "If our presidential nominee in 2012 ... appears too extreme on abortion or gay marriage or some other social issue, there's a slice of the electorate that clearly could go back to Obama," Wadhams worries.

Axelrod is thinking in similar terms. In broad strokes, he argues, Obama will benefit in 2012 because the election will be framed less as a referendum on the nation's direction and more as a choice against a Republican alternative. "The hardest thing in politics is to be measured against yourself," he said. But in 2012, "these voters, and all voters, will be faced with a choice. And I view that as an opportunity."

More specifically--and perhaps more revealingly--Axelrod also has his eye on the Colorado example, where the exit poll found that Bennet lost blue-collar white women by double digits and blue-collar white men by more than 2-to-1. Yet he prevailed by amassing strong support from young people, Hispanics, and other minorities; holding his deficit among college-educated white men to single digits; and routing Buck among college-educated white women. A similar formula, Axelrod suggests, could be available to Obama in 2012, especially if the Republican presidential primary process, as he expects, tugs the eventual GOP nominee toward the right. "The Bennet thing was particularly instructive," Axelrod said. "They made a big effort there not only among Hispanics but women. The contrast he drew with Buck was very meaningful. That's why I say the gravitational pull of those Republican primaries is going to be very significant."

The importance of the Colorado model is that it suggests a potential path to a second term for Obama even if he regains only modest ground among white voters. In the interview, Axelrod rejected the idea that the Democrats' difficulties among blue-collar whites will force the reelection campaign to downplay metal-bending states such as Ohio, Indiana, and Wisconsin where those voters dominate. But without a major revival among working-class whites, winning such states will be difficult for Obama. That would increase the pressure on his campaign to prevail in swing states that fit the Colorado mold, with large numbers of minorities and well-educated whites.

This list would include a few states already in the Democratic presidential coalition (particularly Pennsylvania, which reverted toward the GOP this year) but would focus mostly on those at its periphery, including Virginia, North Carolina, Florida, Nevada, Arizona, and Colorado itself. If Democrats can't soothe much of the white discontent that cost them their House majority, the Ivy League-educated, mixed-race Obama will need to win more of the states defined by the same titanic social forces that he embodies: growing diversity and rising education levels. Even more than in 2008, Obama's 2012 map may revolve around states that see a face like his when they look toward their future.