Last week's Pennsylvania primary demonstrated that Barack Obama is not unbeatable. This might sound a strange way to put it. Hasn't it always been true that Obama is beatable?

Well, consider an alternate reality in which Obama had won Pennsylvania. His people certainly thought long and deeply about this alternate reality--why else spend a staggering $12 million on one state's primary? Hillary Clinton would have dropped out. Obama would have shown that he can win white working-class votes in a big, diverse, populous state. Way back after the Iowa caucuses, he playfully observed that everywhere he goes becomes Obama country. What if, amid a deluge of ads, after spending the better part of six weeks crisscrossing Pennsylvania's white ethnic inner suburbs and rural counties, he had managed to turn them into Obama country? There'd be no denying that he had the political Midas touch.

The Obama campaign, a far shrewder, more effective, more creative operation than any we've seen in Democratic politics in years, didn't spend that extraordinary sum for laughs. One has to assume Obama's rapid-fire responses to Clinton's attacks on guns and security were a dry run for the general election. Yet he didn't win in Pennsylvania, even against Hillary Clinton's near-penniless campaign, full of mutinous senior advisers eager to jump ship, even with a media cheering section to urge him on.

Not only did Obama not expand beyond his core constituencies--as always, he was crushed among Catholics, an atypically big slice of Pennsylvania's Democratic electorate, and white working-class voters--he lost ground with affluent professionals, the group that has powered his historic fundraising success, with weekly churchgoers, and with the moderates who have until recently seen him as one of their own. He lost Greater Pittsburgh and the Philadelphia suburbs by wide margins, and he also lost the northeastern part of the state by a whopping 66 to 34 percent. In a new Brookings study of Pennsylvania's political demographics, William Frey and Ruy Teixeira identify this region, centered on Allentown, as key to the state's political future. If Pennsylvania's Northeast keeps trending Democratic, the state will become solidly blue. But if a Republican candidate can hold the line or make some modest gains with the region's white working class voters, the picture looks very different. And as it turns out, the GOP may have a candidate who can do just that in John McCain. As Hillary Clinton's campaign slow-marches to its unhappy end, she is offering lessons not only for how McCain can defeat Obama--she is pointing towards a possible bright future for the Republican brand. She's probably not thrilled about that. But before we get ahead of ourselves, it's worth considering the scale of the obstacles Republicans face.

Right now, head-to-head match-ups between McCain and Obama look encouraging for the Republican. Pollsters find the two are running neck-and-neck. This picture couldn't be more misleading, as the McCain camp has been reminding anyone who'll listen. Once the Democratic nomination is settled, there is every reason to believe McCain will fall behind. The vast majority of Clinton Democrats will rally round the flag, Obama will be able to train his fire on McCain, and a massive independent expenditure campaign designed to tie McCain to the Bush White House will get under way. Given that President Bush now has the highest disapproval ratings in the history of presidential approval ratings, and that much of McCain's strength derives from independent voters who now believe he'll take the country in a different direction, those matchup numbers could get ugly fast.

Meanwhile, the public mood is near toxic for the incumbent party. According to a new Pew Research Center survey on the middle class, 31 percent of Americans believe they are worse off than they were five years ago. Compare this with 25 percent in 1979, shortly before Ronald Reagan defeated Jimmy Carter with a campaign promise of hope and change. Only 41 percent feel better off. Among self-identified middle class voters, 79 percent are convinced it is tougher now than five years ago for middle class families to maintain their standard of living.

It's true that these numbers were fairly grim, even before President Bush's reelection: For example, only about 20 percent in 2003 rated the economy as excellent. But the ideological landscape is far less favorable to Republicans today. When Pew conducted its most recent comprehensive report on trends in political values in 2007, it found that the gap between Democratic and Republican partisan identification was a mammoth 15 percentage points, a sharp shift from parity in 2002. And as the economy flirts with recession, the social issues that supposedly drive America's Bitter Bloc into Republican arms may be overshadowed. An October 2007 survey, for example, found that 65 percent consider "Energy" (gas prices) very important to their vote, while only 22 percent felt the same way about same-sex marriage. In October 2004, the numbers were 54 percent and 32 percent respectively. (And gas prices are appreciably higher today than they were last fall.)

Then there is the small matter that the Republican coalition is shrinking. In a survey released this month, Teixeira and Alan Abramowitz chronicle the long-term decline of the white working class, a decline driven in large part by education and income gains but also by sharp growth in the Latino population. When Republicans win elections, they tend to do so on the strength of super-majorities of white working class voters. Democratic success in 2006 derived not from winning these voters but merely from cutting the Republican advantage. Among non-college-educated whites in households making $30,000 to $50,000 a year, Bush won 62 percent to John Kerry's 38 percent. The same group went for Republicans in 2004 congressional voting by 22 points. Among non-college-educated whites in households making $50,000 to $75,000, Bush won by 70 percent to 29 percent, and congressional Republicans by 32 points. In the 2006 midterms, by contrast, the first group supported Republicans and Democrats equally, and the second backed Republicans by only 21 points.

In Building Red America, Thomas B. Edsall described the Republican party as a coalition of winners. Blue-collar Republicans tended to be non-college-educated women and men in intact families, who had successfully adapted to economic change and were inclined to believe that the market economy worked for them. It is easy to see how a tough economic climate will shrink the winners' circle. This is particularly true in the hard-hit Great Lakes region, where McCain has a chance to perform well against Obama. A reprise of 2006 in the form of Republican underperformance among white working-class voters would doom McCain. He needs to keep his edge with these voters well in double digits to blunt the growing strength of Republican-unfriendly constituencies like less affluent college-educated whites, unmarried women, and nonwhite voters. That is, McCain needs to win a bigger share of a shrinking slice of the electorate. That won't be easy, particularly since domestic issues aren't his strong suit, as his disappointing performance in the Michigan primary showed. If he pulls it off, however, McCain will have a shot at winning states like Pennsylvania and Michigan.

If the old coalition is shrinking, can McCain break out of the Republican demographic box by winning more nonwhite voters? He has a better chance than any other Republican. The party has performed decently with some Latino constituencies, some Asian ethnic groups, and with small but significant minorities of culturally conservative black voters, as in Ohio in 2004. But in light of Obama's popularity among black voters, it seems safe to say that McCain doesn't have much of a shot with a constituency that has long been a Democratic stronghold. As for Latinos and Asian Americans, the so-called new minorities, the challenge is vexing because these groups are so diverse, and both are subject to what demographer William Frey calls "translation gaps." States with large Latino and Asian populations don't always have large Latino and Asian electorates, thanks to disproportionately high numbers of children under 18 and foreign-born residents who haven't yet acquired citizenship.

For example, 60 percent of Asian Americans are foreign born. Because Asian Americans tend to live in married households and have relatively high incomes, this ought to be a strong Republican constituency. That was once the case, when memories of the Cold War motivated Korean American and Vietnamese American voters to back hawkish Republicans, and it could happen again as culturally conservative foreign-born Asian Americans acquire citizenship and younger churchgoing Asian Americans reach voting age. Asian-American voters have nonetheless been trending Democratic, in no small part because they are clustered in heavily Democratic metropolitan areas. And assimilation is fast turning Asian Americans into a less rather than a more distinctive constituency from whites with similar class and educational backgrounds. You might say Asian Americans are assimilating into the Democratic-leaning tendencies of their college-educated white neighbors.

This leaves Latino voters. Writer and hip-hop historian Jeff Chang has suggested that Latino voters in California backed Clinton because they place a great deal of weight on community leaders who were courted aggressively by the Clinton machine. John McCain can't match Clinton's success in this regard, as the Latino political establishment is, with the exception of Florida's Cuban Americans, very nearly monolithically Democratic.

A number of analysts have attributed strong Latino support for Hillary Clinton to a deep-seated antagonism towards black Americans, an antagonism some Obama partisans have gone so far as to suggest has been stoked by the Clintons. There is no doubt some truth to this notion. Latinos and native-born blacks have clashed in urban politics, particularly in California where Latino political power has arguably surpassed that of black voters. But what if Latino voters are simply mirroring the preferences of similarly situated Anglo voters? Given that Latino voters by definition represent the most assimilated slice of the Latino population, it makes sense that, say, non-college-educated Latinos would parallel non-college-educated whites in preferring Clinton to Obama. And if that's true, it suggests that the demographic decline of the white working class is an illusion--it will be remade as an Anglo-Latino white working class, just as conceptions of whiteness grew to include previous waves of immigrants. Latino distinctiveness will likely endure on certain issues, particularly on immigration. Yet that distinctiveness will fade.

In a sense, Hillary Clinton's coalition of white working class and Latino voters represents a better path for the Democratic party's future than Barack Obama's coalition of social liberals and black voters, which, as John Judis has noted, resembles nothing so much as George McGovern's losing coalition of 1972. Granted, there are far more college-educated liberals now than there were a generation ago. But here's the thing--the McGovern coalition included all minorities, as though minority status were defining and rigid. To the extent Latino voters can be pried loose from neo-McGovernism, the whole enterprise collapses.

But it is by no means obvious that McCain can pry Latino voters lose, particularly in light of the tarnished state of the Republican brand. McCain broke with conservative Republicans to embrace a comprehensive immigration reform, but he's been forced to soft-pedal the issue. Moreover, he will never win Latino voters in an immigration liberalization bidding war with Democrats. Instead of focusing on immigration as such, McCain needs to appeal to Latinos as members of a broad, pan-ethnic group of working class strivers. In doing so, he could remove places like Nevada, Florida, and New Mexico from the swing state column and improve his standing in increasingly blue-trending Colorado. McCain's balanced and measured plan to tackle the mortgage crisis, his call for doubling the tax exemption for dependents, and his proposal for delivering more affordable health care all represent a promising start. But he needs to weave together these and other proposals into a compelling narrative that goes beyond rewarding our guys and punishing yours.

It's a tall order. Still, if McCain manages to pull off a victory, Republicans will owe a debt to the path blazed by the Hillary Clinton campaign in Pennsylvania--not that they would ever thank such an unlikely benefactor, and not that she would ever want to be thanked.

Reihan Salam is an editor at the Atlantic and a fellow at the New America Foundation. With Ross Douthat, he is the author of Grand New Party: How Republicans Can Win the Working Class and Save the American Dream (Doubleday), due out in June.

Next Page