This isn't an election, it's a political census. The Democratic primary process of 2008 is proceeding like a reading of The Almanac of American Politics. Here' s Ohio -- white working-class, no new jobs, young people fleeing to Chicago, California, or points east; advantage Clinton. Here's Maryland, upscale professionals and African Americans; advantage Obama. And here's Texas, huge Latino vote eclipses black vote while whites largely split; advantage Clinton.
This is the show-me-the-demographics-and-I'll-tell-you-the-winner Democratic presidential contest. Whites in all white states go Obama; whites in black-white states go Clinton. Older voters, Clinton; younger voters, Obama. Older white voters for whom nearly every change in the last 30 years has been for the worse -- that is, the Midwestern working class -- don't seem drawn to the candidate of change. It's all there in the statistics. Why bother to hold elections?
In fact, all but about 10 states have voted now, and only Wisconsin stands out for violating this rule. In Wisconsin, Barack Obama won the very same white working-class voters who elsewhere voted for Hillary Clinton. Electoral scholars will one day ponder what prompted this Wisconsin exceptionalism, but for our purposes, as we look ahead to those final 10 states, it's hard to imagine where Obama can duplicate that singular triumph, or where Clinton can win over the kind of voters who have been flocking to him. Can she snatch away Oregon? Improbable. Can he win Pennsylvania, where the mine-and-mill ghost towns -- inhabited, but antique -- between Philadelphia and Pittsburgh make Ohio's economy look like a veritable Silicon Valley by comparison? Not likely.
It would be nice if one of them did break through to the other side, did start winning voters out of the other candidate's base. That would give superdelegates some tangible achievement on which they could base their vote. Because if Florida (retirees) and Michigan (white working class) have primaries rescheduled for June, and Clinton wins them both by dint of demographics, then it's possible the delegate and popular vote counts may be nearly even at the close of the primary season. Which would put the superdelegates in a justifiable dither: If the primary contest is done and it comes out even, and if the dividing lines in the party aren't those of policy but those of identity -- what, dear God, is a superdelegate to do then? And how should the supes calculate the candidates' respective strengths against John McCain?
By two standards, Obama certainly seems the stronger. He certainly has greater appeal than Hillary to upscale independents, and he looks to be stronger in the previously Republican, now Democratic-trending states of the mountain West. By two other standards, Hillary looks more formidable. She clearly does better among those onetime (or heirs to) Reagan Democrats who've been coming back to the Democrats on economic issues. She also looks stronger in key Rust Belt states -- Ohio, Pennsylvania, Michigan.
Which brings us to one of the central facts of American politics over the past few decades: the deunionization of the industrial Midwest. The political value of unions to the Democrats in this region has been their ability to win voters over on the basis of economic issues whom the Democrats would otherwise have lost on foreign policy and cultural issues. On cultural and foreign-policy issues, the Midwestern white working class -- descendents of migrants from the South, from Appalachia, and from the Austro-Hungarian Empire -- is very conservative. Absent unions, they vote like their counterparts in the communities their ancestors left behind. Well, okay, not like Hungarians, necessarily. But very much like Southerners -- Southern whites, that is.
Where unions can reach these voters, even to this day, they make a huge difference -- more so in times of economic distress, such as these. Unions, of course, are nowhere the size in these states that they were half-a-century ago, but their political programs are a lot smarter, which can make up some of the difference. They certainly helped the Democrats win major victories in Ohio in 2006. They'll help the Democrats hugely in 2008.
But I have two fears. The first is that their numbers have shrunk so that they'll reach just a fraction of the voters they reached in decades past. That, however, is precisely why the AFL-CIO has put in place in these states its Working America program, which recruits union members not at the workplace but in a door-to-door canvass, and which clearly won tens of thousands of votes for the Democrats in 2004 and 2006 that they wouldn't have won otherwise.
My other fear can't be so readily allayed. It's that there's one issue on which the unions have never been that effective with the Midwestern working class: race.
In the late 1980s, Democratic pollster Stan Greenberg did a now-famous study of Macomb County, white working-class suburbs immediately north of Detroit that in 1960 had given John F. Kennedy his highest vote in any suburban county. Greenberg was investigating why Macomb had given Ronald Reagan such huge margins 20 years later, and what he discovered was that it all came down to race, to heavily black and increasingly disorderly Detroit, from which Macombers had fled. Macomb residents blamed the Democrats for the rising power of African Americans and for Detroit's decline. Accordingly, national Democrats could no longer win Macomb.
In the heyday of the United Auto Workers in the 1930s, 1940s, and 1950s in Detroit proper (from which whites had yet to flee) the UAW could produce an 80 percent vote for national Democrats on issues of the economy. But at the local level, in city elections, which were about policing and housing, the union backed progressives, and those candidates repeatedly lost. The solidarity of union members that existed on economic issues didn't extend to policing, where white members believed the role of the cops was to keep blacks in their place, and housing, where white members believed the role of the city was to keep blacks in their place.
Greenberg's study informed the 1992 Clinton campaign, including Clinton's condemnation of Sister Souljah. And while time has passed and younger white voters certainly have more liberal views on race than their elders, what you find in Ohio and Pennsylvania and Michigan are the elders, whom Hillary and Obama may well reach on economics, but for whom Obama's blackness may prove an insurmountable barrier.
So what's Obama to do? He can't very well downplay the Pennsylvania primary in any case, though demographically it's clearly Hillary's to lose. (The state's median age is the second highest in the land, after only Florida's.) He should play to his strength in Philadelphia and its suburbs, and in the new university-health care economy of Pittsburgh. But he should also make the hard slog through the old towns of old people who don't often see people who look like Obama and who trust them even less. He should talk to them about his work for displaced steel workers in Chicago, about the way in which Wall Street and both political parties abandoned people like them, about concrete projects to rebuild the state's infrastructure, about trade. (He should shoot Austen Goolsbee). Nobody expects him to win Pennsylvania anyway. And if he can have a respectable showing, or better, it's his best chance to win one on Clinton's terrain. Which is probably the one sure way to impress superdelegates desperate for benchmarks to which they can cling.
