During convention season, the polls temporarily provide a less accurate picture of the race as voters sway back and forth on either side of their eventual preference. But this week, the polls are becoming more and more predictive of the eventual outcome with every passing day. History suggests that we should expect the lingering effects of the DNC to steadily diminish to the point of elimination over these next few days: Put differently, by Friday: we’ll be able to start assessing whether Obama’s post-DNC boost was a temporary bounce or a resilient bump.
If Obama’s four point lead persists through the week, Obama should be considered a very strong favorite for reelection. While it might seem that the heart of the campaign is still to come, the candidate leading two weeks after the in-party convention has gone onto win the popular vote in every presidential election since Truman's come from behind victory in 1948. And the only other comebacks were also staged by candidates or incumbents who ascended to the presidency or the nomination following resignation, assassination, or a late decision not to seek reelection. Truman, Ford, and Humphrey gains are hard to analogize to other contests involving incumbents, given that the candidates were forging their own electoral coalitions for the first time out of the fragmented remnants of older ones. And although Romney is behind by just four points, the race has been unusually stable and Obama may be at or very near 49 percent of likely voters, which means Romney would need to win an extremely large share of an unusually small number of undecided voters to overcome a seemingly tiny deficit.
There are signs, however, that Obama’s bounce might be fading. As observed on Saturday, Obama’s advantage in the trackers has steadily shrunk and Friday's state polling wasn’t exactly consistent with a big Obama lead. Over the weekend, Obama fell to just a 3 point advantage in the Gallup tracker. Similarly, Obama trails by 1 in the Rasmussen poll, compared to a 3 or 4 point lead after the DNC.
But while there wasn't much polling over the weekend, the polls suggest that Obama’s bounce continues, even if it has fallen slightly off of its post-DNC peak. Although Obama’s 3 point lead in Gallup is far less than the 7 point lead he held last week, he’s still 2.5 points above Gallup’s longer term average and 2 points above his pre-convention showing. Additionally, Obama ticked up to 50 percent in Gallup’s approval tracker, suggesting that he’s maintained his elevated standing in their most recent samples. And while Obama appears to have receded to Rasmussen’s longer term average, Obama actually leads once leaners are included, 49-48—which would also be about 2.5 points above Rasmussen’s longer term average. In general, I’m agnostic about whether it’s better or worse to include leaners at this stage, but as a Rasmussen subscriber, my look at the internals tells me this is an instance where “leaners” is probably more representative, at least right now. So the tracking polls are both consistent with a shrinking but still existent bounce. We’ll see whether Obama stabilizes with a 2 or 3 point bounce (good for a 3 or 4 point lead among likely voters in an average of national polls) or if his numbers continue to decline.
Similarly, the state polls were generally consistent with a persistent Obama bounce In Virginia, Pennsylvania, New Jersey, and Kentucky, Obama was performing near or just slightly off 2008-levels. The results in Pennsylvania and New Jersey aren’t surprising, given other polls, but we haven’t seen much from Kentucky and a 14 point race there might be a little surprising. And while Kentucky isn’t a swing state, it’s actually a state where the polls are pretty useful. Parts of southeastern Ohio and southwestern Virginia tend to track along with Kentucky, and many have argued that Obama was likely to suffer additional losses in the heart of Appalachia, given the region’s demographics, the so-called War on Coal, and Obama’s performance in the primaries. So if SurveyUSA is correct and Obama hasn’t fallen off in the Blue Grass state, it could be a sign that Romney won’t capitalize on his opportunities in southeastern Ohio and southwestern Virginia. Perhaps with help from a bit of resilience in southwestern Virginia, PPP shows Obama up by 5 in the Commonwealth--a showing that's pretty consistent with PPP's other post-DNC polling.
Finally, Obama has a big lead in Germany. If it were a state it would hold 92 or 93 electoral votes under current law, and every current state would lose about 20 percent of their own electoral votes. However, that would probably prompt Congress to expand the size of the House of Representatives beyond 435 seats. If the size of the House was proportionately expanded, Germany would be worth 116 electoral votes. If that seems small, that’s because it assumes that Germany accedes to the Union as 1 state. If the 16 German states acceded separately, then North Rhine-Westphalia's 29 electoral votes would rival New York and Florida while Nate Silver would need to rename his blog something like SixEightyFour.