The post-convention national polls mostly show John McCain with a small lead over Barack Obama. But what's been happening in the states? I've been looking at the post-convention state polls at realclearpolitics.com, pollster.com, and fivethirtyeight.com and find some significant differences from pre-convention polls. They tend to suggest that the battlefield is shifting, with more states within McCain's reach and fewer within Obama's.
Some caution is in order: We're talking about only one or two polls in some states but as many as eight in ultracritical Ohio. I haven't included the Zogby Internet polls in my analysis. I've rounded off the averages in each state to full percentages (and rounded 0.5s downwards for both candidates), and I'm reporting the difference between the McCain percentage and Obama percentage. Here's my analysis:
The big industrial states. Michigan and Pennsylvania are Obama +2, Ohio is McCain +3. In each case, McCain is 1 point better than Bush's final percentage against Kerry in each state. An old rule of American politics is that economic distress moves voters toward Democrats. Michigan, Ohio, and western Pennsylvania are in economic distress. But they haven't moved toward the Democratic nominee, as compared with 2004. The old rule isn't operating. Here's another possible rule. When voters see that tax increases aren't producing a better economy, they don't move toward a Democratic nominee who is proposing higher taxes, even though he says they'll hit only the rich. In Michigan, the Democrats (with a few turncoat Republicans) raised taxes in 2007; in Ohio, the Republicans (with some Democratic support) raised taxes before 2006. Those tax increases haven't helped those states' economies, not so as you'd notice, though they've helped members of public employees unions. McCain was running much worse than this in pre-convention polls in Pennsylvania and somewhat worse in Michigan. His convention bounce gives him a good chance to win the electoral votes of Pennsylvania (21) and Michigan (17), while leaving him in pretty good shape in Ohio (20).
The new marginals. Obama has been running consistently better than John Kerry or Al Gore in Colorado and Virginia, states that have had comparatively vibrant economies and have also seen influxes of young voters, who tend to be heavily pro-Obama. Just look at all those singles rental apartments and loft-like condos in Arlington and Alexandria and LoDo in Denver. Colorado comes out of the conventions as +1 Obama, Virginia as +1 McCain. In both cases, the average is depressed by one poll that shows the state going the other way. Colorado (9 electoral votes) and Virginia (13) are still very hotly contested ground.
The northern tier. The Obama campaign had hoped to be competitive in some northern tier states: the Dakotas, Montana, and Alaska. Pre-convention polls provided some reasonable basis for this hope. Post-convention polls don't. Alaska, unsurprisingly, is McCain-Palin +27. Montana is McCain +11, North Dakota McCain +14, South Dakota +17. More importantly, Minnesota is just Obama +1, Wisconsin Obama +3, Washington Obama +4, Oregon Obama +7. So scratch 12 electoral votes as plausible Obama targets and add 38 electoral votes as plausible McCain targets (or, excluding Oregon, 31 electoral votes). This is a big change, and it remains to be seen if later polls will show these states to be as close as the relatively few polls we've seen so far do.
The western odd ducks. Nevada is McCain +2. New Mexico, in a shift from pre-convention polls, is McCain +2 (but that's only one poll). These states were seriously contested in 2000 and 2004 and look to be again in 2008.
The South. Florida is McCain +5; it was Bush +5 in 2004. North Carolina is McCain +11; it was Bush +12 in 2004 (despite the presence on the Democratic ticket of the now happily forgotten John Edwards). But two North Carolina polls show McCain way ahead (+17 and +20); two others show him, as did most pre-convention polls, narrowly ahead (+3 and +4). I have more respect for the polling firms showing the big McCain margins, but this state still bears watching. Georgia, where Obama has sent scads of organizers, is McCain +16.
The Northeast. One poll shows New Hampshire Obama +6 (Zogby Interactive has McCain ahead there): inconclusive. Three polls show New Jersey as Obama +6; it was Kerry +7 in 2004. Astonishingly, one poll shows New York as Obama +5, but this is Siena, which seems to have a lot more undecideds than other New York polls, which have shown Obama well above 50 percent. The New Jersey and New York numbers may tempt the McCain campaign to start advertising on New York City media. I suspect this is a temptation that will and should be resisted, for the time being.
There are a lot of states with no post-convention polls, including interesting ones like Indiana and (if only because of its 55 electoral votes) California. My overall conclusion is that the playing field has shifted in favor of McCain. He seems competitive now, where he arguably wasn't before the conventions, in Pennsylvania (21 electoral votes), Wisconsin (10), Minnesota (10), Washington (11), and maybe Oregon (7): a total of 59 electoral votes, all carried by John Kerry and Al Gore. Obama no longer seems competitive in North Dakota (3), Montana (3), and Alaska (3): a total of 9 electoral votes.
Or to look at it another way, from Bush's 2004 electoral vote total of 286, you now have to subtract Iowa (7), which is Obama +12 in the latest Des Moines Register poll, and maybe Colorado (9), Virginia (13), and New Mexico (5), which gets the Republican total down to 252. Or to 247, if you include Nevada (5). But in the northern tier there are 63 more electoral votes within reasonable reach of McCain in the northern tier and New Hampshire. And maybe he wants to start looking at New Jersey (15). I see Obama as competitive or leading in states with 338 electoral votes (granting him the 27 in Florida, which looks to me increasingly unlikely). I see McCain as competitive or leading in states with 342 electoral votes. Advantage shifting toward McCain.