In presidential polling, the whole must eventually equal the sum of its parts. If a candidate has a consistent lead on the state level, then it will eventually show up in national polls. The opposite is also true; if a candidate takes a sharp decline in national polls, then that will also be reflected on the state level. Last week, Nate Silver noted the extent to which that hasn't been true of this election. Nationally, the race is a near-tie between Barack Obama and Mitt Romney. But in state polls, Obama maintains a clear lead over the Republican nominee. By calculating a value called the "implied national vote"-the difference between the current polling average and the 2008 results-Silver finds that based on state polling, you should expect a 3- to 4-point lead for Obama in national surveys.
This isn't to say that the race is actually less close than it looks-the state polls might end up overstating Obama's strength-but to offer some context for the continued discrepancy between state and national numbers, as demonstrated by the latest joint poll from Quinnipiac University, CBS News and the New York Times. The new surveys show Obama leading in three states-Florida, Pennsylvania, and Ohio-by substantial margins. In Florida, the president takes 51 percent to Romney's 45 percent. In Ohio, it's 50-44. And in Pennsylvania, it's an 11- point difference, 53-42.
For the Obama team, this result is even better because it comes from a survey of likely voters. Traditionally, likely voter polls favor Republicans over polls that sample registered voters; the demographic characteristics that make you likely to vote-white, older, higher income, higher education-are also the ones that make you more likely to vote Republican. The icing on top is that, according to Silver, these polls imply a substantial lead for Obama, even as national polls continue to show a close race.
Because of the vagaries inherent to polling of every kind, we shouldn't take these as definitive signs of where the race is (the Quinnipiac polls, in particular, might oversample Democrats). But at the very least, they convey the degree to which Obama is probably doing a little better than national polls suggest. For Romney, whose rough July ended with an even rougher trip abroad, this highlights the extent to which August is crucial for his presidential bid. If he can't catch up to Obama now, odds are good that he'll stay behind him throughout the fall.
For more polling information, go to The Prospect's 2012 election map.
