The big picture: Obama continues to hold a modest but clear lead nationally and in critical battleground states like Ohio, Virginia, and Florida.
This is alot, so let’s digest it in segments.
Good News For Obama
Nearly every poll was consistent with a modest Obama lead nationally of about 4 points and a meaningful edge in the electoral college. Obama hit fifty percent in three national polls and led by four or more in four of the five national surveys. Big twenty-point leads in New York and California, the two largest blue states, probably provide much of the margin, but a growing lead in the lean-blue states like Michigan also helped. EPIC-MRA found Obama up by 10 in Michigan, up from a 3 points prior to the convention.
But more importantly, Obama appears to be faring well in perhaps the three most critical battleground states: Ohio, Virginia, and Florida. Obama got decent results from two Rasmussen polls showing Obama with a narrow advantage in Ohio and Florida. A 1 or 2 point lead isn’t big, but it’s a little better for Obama given Rasmussen’s GOP-lean and a worse showing in August.
The bigger splash came from three surveys conducted by NBC/WSJ/Marist in Ohio, Virginia, and Florida, which showed Obama leading by substantial 7, 5, and 5 point margins among likely voters, respectively. Perhaps as importantly, Obama was at 49 or 50 percent in all three states.
The underlying numbers were just as troubling for the Romney campaign. In all three states, Romney’s comeback chances start with gaining the support of undecided voters, but just 25 percent of undecided voters in Florida, Ohio, and Virginia held a favorable impression of Romney, compared to 43 percent for Obama. Now, Obama’s 43 percent tally is hardly impressive, but it would appear more than high enough to block Romney from sweeping every undecided voter in these three states, which is essentially what it would take for him to fight to a dead-heat. Overall, a prohibitively low 40 percent of Ohio voters have a favorable impression of Romney.
Romney’s issue isn’t just that undecided voters don’t like him. They don’t seem to have much confidence in his ability to handle the office more generally. Just 9 percent of undecided voters thought that Romney would be "good" on foreign policy compared to 47 percent who thought Obama would do a good job. 9 percent strikes me as unusually low—I suspect 9 percent of undecided voters would volunteer that I would do a good job on foreign policy, simply by having name introduced.
At this point, Romney’s chances hinge on the NBC/Marist polls just simply being wrong. If Romney is actually down half a dozen points in these three states, it might be too late to make a comeback. Over the summer, a whole presidential campaign’s worth of advertisements was spent on advertising and voters probably have unusually well-defined views of the candidates. That’s especially true for the incumbent president, who holds 49 or 50 percent of the vote in each of these states. The poll strongly suggests that Romney can’t just count on a flood of undecided voters, which couldn't even put Romney over the top,
And if these polls are right, Romney doesn't have a route to the presidency without these three states. Obama starts out with 237 electoral votes and he’s clearly ahead in New Hampshire and Nevada, which would give him 247 electoral votes. From there, a victory in Florida on its own, or a win in Ohio or Virginia coupled with an additional state (Wisconsin? Colorado?) would get Obama over the top. In fact, Obama could plausibly lose all of these states and win the presidency, provided he captures Iowa (where polls have shown a tight race) in addition to Colorado and Wisconsin, where Obama is probably a modest favorite, although there isn’t yet any post-convention polling from the Badger State.
Now fortunately for Romney, these are the best state-level polls for Obama since the election, so they might not be perfectly representative of the race in these three states. Even if one includes these NBC/Marist polls, Obama leads by a far less daunting 3.6 points in Ohio and 2 points in Florida in an average of post-convention polls, although that Florida number might be a little low. Overcoming a 3.6 point deficit is much more achievable than seven, especially since it’s more consistent with Obama’s 4 point lead nationally. Given the other polls, my presumption is that NBC/Marist shows Obama doing a little better than he actually is. But Romney’s chances would be imperiled if these results were confirmed by other surveys and a 3.6 point deficit in Ohio is still an uphill climb.
Good News for Romney
--After showing Obama ahead following the DNC, Romney finally reclaimed a lead in Rasmussen’s tracking poll. That could be a sign that Obama’s post-convention bounce is fading, and Gallup’s approval tracker also showed Obama dipping back to 49 percent. That said, Rasmussen weights for party-ID and some national polls show that Obama's movement was due to gains among likely voters. By weighting the likely voter universe, it's possible that Rasmussen's method has obscured some of Obama's gains.
--An AIF/McLaughlin poll showed Romney up by 3 points in Florida. That's an unequivocally good result, but it looks like an outlier, at least for now. McLaughlin is a GOP polling firm, and while you'll note that partisan polling firms are routinely included and cited on this blog, my philosophy on private polling is as follows: if the partisan polling confirms the public polling, it gives me more confidence in the public polling; but if the partisan polling is out of line with the public polling, I tend to defer to a consensus of public polls (especially if the partisan pollster doesn't regularly release results). In this case, the AIF/McLaughlin result looks like the outlier.
The other post-convention Florida polls are from Rasmussen, SurveyUSA, and NBC/WSJ/Marist, which show Obama up 2, 4, and 5, respectively. So I'm initially a little skeptical of the McLaughlin result: here's a partisan Republican polling firm showing the Republican performing a net-5 points better than the next best poll for the Republican candidate, and this firm doesn't usually release polls the public. And the next best poll happens to be Rasmussen, which has a bit of a GOP House Effect of its own. For now, I'm not focusing on this result until a non-partisan, public pollster shows something similar.
If the McLaughlin poll is excluded, Obama leads Florida by 3.7 points in post-convention polls, with 48.3 percent of the vote.
Doesn’t Cut Either Way
While there was plenty of good news for the president and a bit for Romney, other polls didn’t necessarily offer much to either side:
--Two polls showed Obama with a slight edge in Colorado, but on balance the polls suggested that Obama was doing worse in Colorado than he was nationally, especially since the more substantial 5 point lead came from a Democratic pollster. Obama won Colorado by 9 points in 2008, so one would probably expect him to be faring better than suggested by these polls if he was up by 4 points nationally. Now, ARG isn’t exactly a sterling pollster, so this result doesn’t necessarily indicate that Obama’s chances in Colorado are in jeopardy, but it’s not great news for Chicago.
--Finally, a WMUR/UNH poll showed Obama up by 5 in New Hampshire, another state that has leaned Democratic in early polls but hasn’t shown much post-convention movement. The poll has a lot of undecided voters, so I’m not going to read too far into it. It's worth noting that Obama's job approval rating is 52 percent among likely New Hampshire voters, which probably bodes well for how Obama would fare if undecided voters were pushed.
Let me conclude by observing a pattern: these are three overwhelmingly white, moderate states with a lot of independent voters.