The point of this post is to expand on, rather than supplant, Matt Hurley's post below on Ohio, because I think Matt really hits the nail on the head with regard to the problems Obama's candidacy is likely to encounter, and why his problems with white working class voters are likely to be a real problem come November. I just think the post doesn't show the full depth of Obama's problems.
This map expands upon Matt Hurley's map. The reddest counties are the counties that were won by Bush twice, by Mike Dewine in his losing 2006 Senate race, and by Ken Blackwell in his losing Governor's race. The bluest counties are the counties that voted Democratic in all four elections. The pure purple counties are the true swing counties, the counties that if the Republicans win, they most likely win the election, but if the Democrats win, they most likely win the election.
As you can see, as Matt pointed out, as you move West, the map gets redder. The key counties for Republicans are the counties in the east-central part of the state. If the Republican wins there, he wins the state. If he loses, he loses the state. And if he starts to lose the west-central counties, he loses the state, big time.
Now, let's see how Obama did in his matchup with Hillary. In this map I have adjusted for the statewide results, such that the purple counties are those where Hillary won with 50-60% of the vote. The dark red counties mean that Hillary won with 65%+ of the vote; the dark blue mean he won with 55%+ of the vote.
As you can see, he had deep strength in the counties that Democrats only lose in a blowout, like Cuyahoga (Cleveland), Montgomery (Dayton) and Franklin (Columbus). He also held his own -- though he still lost in absolute terms -- in the reddest counties, and overperformed in Hamilton county, where Democrats also typically perform poorly.
But in the purple counties, he got blown out. To look at this another way, look at this map. It shows the "basic" red-blue map in it, with counties where he carried 40% of the vote or less marked with crosshatches.
So what does this mean? Well, first off it means that, within the Democratic party, Obama starts out the weakest in the counties that a Democrat needs to worry about the most in order to win. More worrisome for the Obama campaign is that, if he is not popular with Democratic partisans in those counties, how would he fare with independents or weak Republicans?
Similar problems manifest in Pennsylvania and Virginia -- the counties that Democrats win when they typically win, Obama performs well. But the counties that Democrats lose when they lose narrowly, Obama performs abysmally in.
Now you may argue that his strength in the Northwest could mean that a Democrat can afford to underperform in the counties in the Southeast. Unfortunately for that argument, the counties in the West are heavily, heavily Republican. He could double Democratic turnout, and it wouldn't make a deny in Bush's 2004 percentages.
So we're left with what he will probably try to do: Jack up his numbers among African Americans, college students, and wealthy suburbanites. In short, he is going to try to build the McGovern coalition, but on steroids. Indeed I think this is exactly why Democrats like Pelosi are fighting so hard for him: If he can win, then they can ditch those blue dog Democrats who stand in the way of true progressive reform. Of course, the problem is that every Democrat since McGovern save Bill Clinton and Jimmy Carter has tried to win with some version of that formula, and hasn't performed particularly well. And the other problem is that maps and political alignments tend not to change quickly; rather, realignments occur gradually over time. But that is a story for another post.
Cross-posted at Race42008.com