Two radically different story lines are emerging in the presidential race, depending on what kind of poll you look at.
If you look at the national-level data, Barack Obama seems to be underachieving. In the latest Gallup daily tracking poll, the presumptive Democratic nominee holds a scant two-point edge over John McCain. The margin is also two points in Rasmussen’s daily poll—which also shows a dead-even race when "leaners" are factored in. Some other recent polls have been a little more favorable to Obama, but the combined weight of the available national data strongly suggests that Obama, despite his personal popularity and the enormous built-in advantages his party enjoys this year, is locked in a much closer race than he should be.
But if you ignore the national numbers and instead consider individual state polls, a realigning landslide suddenly seems to be within Obama’s reach. In state after state, he’s performing far better than John Kerry did in 2004, and numerous Republican bastions are seemingly in play. Consider Indiana, which George W. Bush won by 21 points in 2004 and which lasted voted for a Democrat 44 years ago—and which Obama leads by one point in the most recent survey. Or North Carolina, which Bush carried by 12 points in ’04 but where the latest poll has Obama within three. And so on. In North Dakota, the race is tied. In South Dakota, Obama trails by just four. Ditto for Alaska, perhaps the most Republican state in the union. He also leads in Montana and Colorado and in all but one recent survey in Virginia.
And the trend isn’t just evident in red states. In states where Kerry eked out victories last time around, polls now give Obama sizable leads. Kerry nearly fumbled away Minnesota (a three-point nail-biter), but Obama has a 17-point advantage in the most recent poll. Wisconsin and New Hampshire were photo-finishes in ’04, but Obama has opened a double-digit lead there. Plus, Obama is running ahead in states that Kerry barely lost, like Iowa (by an average of seven points), New Mexico and Nevada.
On top of all this, Obama is performing as well as any Democratic nominee is supposed to in the biggest blue states—California, New York, Illinois, New Jersey and Massachusetts—and leads (in some cases substantially) in every recent swing state except Florida, where the average of recent polls gives McCain a three-point edge.
There are some traditionally Republican states where Obama is performing at a more typical (for a Democrat) level, like Utah, Alabama and Tennessee, but overall at the state-by-state level he seems positioned to win the November election going away. So how is it—with Obama so close to McCain in so many red states and so far ahead in all of the big blue states—that the national polls show such a close race? With all of the dramatic strides Obama is making in individual states, shouldn’t his national margin be much wider?
One tempting thought is that the national polls might seem so different because many of the red states where Obama is overachieving are so small. So while there might be palpable movement in his direction in, for instance, North Dakota (which accounts for 0.2 percent of the U.S. population), it’s possible that in a national survey of 500 voters, only one North Dakotan—or maybe even none—is actually interviewed.
But if we take the average result from recent polls in each state and weight each state according to its share of the national population, we get an overall national result that’s entirely consistent with current national polling: Obama 46.2 percent, McCain 42.7—a 3½-point race. So there really is no inconsistency between the close national horse race and Obama’s clearly superior position in individual state polls.
The most obvious explanation for this is the large number of undecided voters included in most polls, which makes it tough for either candidate to break 50 percent in most states right now. In South Carolina, for example, Obama is clearly running better than Kerry did (or Al Gore, for that matter) and trails McCain by just six points in an average of that state’s most recent polling. But as surprisingly close as the race is, Obama’s raw number—39 percent support, on average—is nothing new for a Democrat in the state (Kerry finished with 41 percent in ’04).
The same is true in many other states, red and blue. Obama leads by an average of 17 points in dark-blue New York, but he’s only averaging 53 percent of the vote there (while Kerry took 58 in ’04). He’s slaughtering McCain in California, but only averaging 53 percent support there. And he’s opened leads in Ohio, Michigan and Pennsylvania, but also isn’t securing the level of support that Kerry had in those states on Election Day four years ago.
This doesn’t mean that Obama is in trouble in any of these blue states or that the number of undecided voters is unusually high. It’s simply a partial explanation for how seemingly solid polling data in individual states can translate into lukewarm national numbers.
But things get more revealing if we take the numbers a step further and try to adjust the current state averages to account for the voters who are now undecided or threatening to vote for third-party candidates. For the sake of this exercise, let’s award the undecided/third party vote in each state proportionally, based on the current average levels of support for both candidates. For instance, the Massachusetts average now has Obama leading, 52 to 36.2 percent. If we adjust that proportionally, Obama ends up with 58.96 percent to McCain’s 41.04 percent. Do this for all 50 states, and Obama ends up with 51.98 percent of the national popular vote, with McCain at 48.02 percent.
This doesn’t exactly look like a landslide, and yet in all but four states, Obama’s final number would be an improvement—substantial in many cases—over Kerry’s ’04 performance. In some cases, this means trimming 30-point Kerry deficits in dark red states to 20 or 15 points, a nice accomplishment that won’t change the bottom line in those states. But in other cases, it means cutting 15-point Kerry losses in half (or more) and moving within theoretical striking distance in a state. What’s striking about this data is that just about all of the improvement in individual states from ’04 is on the Democratic side. McCain may end up holding on to the traditional red states that now seem in doubt, but he’s not threatening in any of the traditional blue states.
This all shows us two things. For one, even if Obama’s surprising standing in red states endures through November, it won’t mean he’s a shoe-in on Election Day. If he comes close without flipping any of them over, McCain would still have a chance in the Electoral College and Obama would not automatically score a runaway victory in the national popular vote. In that sense, the current national polls that show a tight race are spot on.
But the individual state polls that seem so rosy for Obama aren’t misleading, either, in the sense that they reflect the potential for an Obama landslide. Obama is only flirting with the possibility now—his prospects may fade by Election Day—but he has the potential to win over a handful of states the other party has long counted on winning. McCain doesn’t.
In other words, the race really is close, and McCain has the potential to win. But only Obama has the potential to win big.
Follow Steve Kornacki via RSS.