You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Don't Panic--reassurances From A Poll Neurotic

If you're the kind of person who obsesses over minor poll movements (and, if you're reading this, you almost certainly are), there's a good chance you noticed the slight tightening yesterday in Pennsylvania and Virginia, as well as the national tracking polls, which Nate covered in his daily write-up. Herewith, a couple reasons I'm not worried:

1.) The polls were always going to tighten a bit as McCain and the GOP made their final push. Among other things, the GOP was almost certainly hoarding cash for a final advertising blitz in key states. Watching TV yesterday in Washington (which is the same media market as Northern Virginia), I noticed a lot more McCain ads than I'd seen in a while. In Pennsylvania, we're even getting reacquainted with the good Reverend Wright.

Why am I not losing my s**t? McCain needed to steadily close the gap over the previous week to get close enough for this late movement to matter. That's why I was sweating the tightening I noticed early last week. But it didn't take. Though my national tracking poll average (minus Battleground and Investors' Business Daily, which lag too much for our purposes), dropped to 6.85 yesterday from 7.25 on Saturday and 7.33 on Friday, it's still up a half-point from a low of 6.28 on Wednesday. (Ignore the magnitude at this point, which depends on which average I use. It's the changes that matter.)

2.) As for Pennsylvania--which is basically McCain's only shot--yes, you'd rather see the poll numbers widening than narrowing. And it's admittedly less than ideal to see Wright starring in a new ad campaign. On the other hand, you've got to keep the following in mind: There are about 4.5 million registered Democrats in Pennsylvania, about 3.2 million Republicans, and just over one million voters who are either unaffiliated or belong to other parties. (Click here for complete registration statistics.) In order to carry the state, McCain either needs to run up a massive lead among the unaffiliateds and flip a chunk of Democrats, or he needs to flip a substantial number of Democrats. There just aren't enough Republicans out there to win any other way.

But here's the problem: A huge number of those Democrats and unaffiliateds either voted in April's primary or thought about doing so, meaning they're already familiar with Jeremiah Wright (who made more than a cameo in the run-up). Which is to say, Wright's probably baked in the cake for most of these people; the Democrats and Independents who tell pollsters they support Obama are supporting him despite Wright, not because they're unaware of him. So McCain's unlikely to win many of them via last-minute introductions.

Having said that, I think there will be more late defections, and that the final margin will be close in Pennsylvania (under five points). The absence of early voting there may be the biggest reason, since it denied Obama a chance to bank votes before McCain's last-minute push, and since he only has one day to get so many people to the polls in Philly and the 'burbs. The campaign and polling sites just might not manage to squeeze them all in. (Though I'm guessing they'll get pretty close.) For what it's worth, that logic also applies to Ohio.

3.) Speaking of early voting, that's the final reason I'm feeling relaxed today. A lot of commenters have made some version of the following argument: Yes, there's theoretically a way Obama could get to 270 without Pennsylvania. But if he loses Pennsylvania, things will have gone so catastrophically wrong it's hard to imagine him winning the other states he needs.

There's something to that, but I actually disagree. Because of the absence of early voting in Pennsylvania, I could see Obama winning Colorado, Nevada, New Mexico, Florida, and possibly North Carolina even if he loses the Keystone State. (The one place its hard to see him winning if he craters in Pennsylvania is Virginia, because there's isn't much early voting there, either.)

The early voting numbers in those five other states are simply gaudy--see this invaluable website--and there are a lot of ways to get to 270 with those states but without Pennsylvania and Virginia. For example, the Kerry states plus Iowa, Colorado, New Mexico, Nevada, and North Carolina get you to 272. Throw in Florida--which I think Obama will carry--and it becomes a route.

Granted, all that could change if there's some unexpected external event (you know what I'm talking about, to paraphrase the eternally-wise Michael Goldfarb), or if millions of voters are simply lying to pollsters about who they're supporting. Short of that, Obama is extremely likely to be our next president.

--Noam Scheiber