Path to Victory

So to my disappointment but no particular surprise, McCain is enjoying his own convention bounce and it appears that this race is reverting to its pre-convention status, which is to say a virtual tie. Now a lot can happen in the next two months, and a swing either way will render any analysis obsolete. But if there are no major changes, how are things looking likely to shake out?

My take: Obama has the inside track to victory, though not the way you might expect. Here's how.

First of all, Florida and Ohio are, as always, "tipping point states" for Obama. If he wins those, he probably does not need much more help beyond the core democratic states. But they are (at best) toss-ups right now. Does that mean the outcome is obliged to boil down to those two states? Well, the Obama team sees a bigger map, and so do I. How can Obama realistically win without OH and FL?

I have the 2004 map at the top for reference, and we'll use that as a baseline. The first requirement for Obama is to hold all the Kerry states. Is he doing this? So far, I would argue yes. Michigan is the big prize that McCain will make a play for, but Obama has held a consistent lead here and I don't see it flipping, especially in this economy. PA also is an enticing target for McCain which is likely to be a huge waste of his limited resources. New Hampshire voters are notoriously unpredictable and have liked McCain in the past; this state could prove an Achilles' heel for Obama in a close race.

The next goal for Obama is to hold the states he's looking likely to flip based on the current status. These are Iowa and New Mexico. 538.com currently gives Obama a 90% likelihood of taking IA, and 87% in NM, so these are not "safe" per se, but they look very good for our man. This takes Obama up to 264 EVs

So the final task for Obama if to find his winning margin. The most likely place for that, interestingly, seems to be either Colorado or Virginia. Both have been historically red states which are trending blue. Both have recently revitalized state democratic parties. Colorado has a democratic governor, is highly likely to have two democratic senators next year, and has democratic majorities in the US house delegation and in both state houses. Even absent a bounce from the DNC having been held in Denver, this is looking like highly fertile ground for Obama. Virginia might be a tougher nut, but is certainly also within reach. It has a democratic governor and will also have two democratic senators in 2009, so democrats can clearly win statewide races here. But the wins so far have been very narrow, and as part of the old Confederacy, it's not clear how much Obama's race will impact votes: there is a large African-American vote, but Obama has been very weak in the rural, white-dominated areas. It's highly winnable, but I think that if Obama takes VA, then he's probably already won OH and the point is moot, whereas I can easily see Obama winning CO and losing OH.

There are certainly other opportunities for Obama, some of them quite surprising -- MT, ND, NC, IN, and NV. Again, like VA, I view these as states which Obama will probably only win if the election is swinging widely for Obama, and as such these probably are not going to be tipping point states. However, expect Obama to spread the field and campaign heavily in these states, forcing McCain to play defense. If the race is close, it is possible that one of these states could provide Obama with an unexpected path to 270, though that would be truly bizarre.

It's just that kind of year.

Oh, and if you're not already reading it, you really should add 538.com to your list of daily reads. The most sophisticated and useful statistical analysis of polling data I have come across anywhere.