The latest poll by NPR and its bipartisan polling team [pdf] shows President Obama with a 7-point lead among likely voters nationally and a nearly identical lead of 6 points in the dozen battleground states where both campaigns are spending most of their time and money.
But the poll also finds former Massachusetts Gov. Mitt Romney very much within striking distance of the incumbent as the two men begin a series of three debates Wednesday in Denver. More than 80 percent of respondents said they planned to watch the first televised clash Wednesday and one in four said the debate could influence their vote.
The poll found 51 percent of the likely voters planning on or leaning toward a vote for the president, with 44 percent voting for or leaning toward his challenger. In the battleground sub-sample, the numbers were 50 percent Obama and 44 percent Romney. Those numbers were slightly better for the president than his job approval rating in the poll. Nationally, the president was at 50 percent approval (46 percent disapproval), but in the battleground he was at 48 percent approval and 49 percent disapproval.
Battleground voters were also more downbeat about the direction of the country. Asked whether things were generally going in the right direction or "pretty seriously off on the wrong track," 59 percent in the battleground said wrong track and just 36 percent said right direction. That gap of 23 points was only 16 points on the same question in the national sample.
The poll of 800 likely voters was conducted over the final five days of September by Stan Greenberg of Democracy Corps and Whit Ayres of Resurgent Republic. About a third of those polled live in the 12 states considered in play for the Nov. 6 election: Colorado, Florida, Iowa, Michigan, Nevada, New Hampshire, New Mexico, North Carolina, Ohio, Pennsylvania, Virginia and Wisconsin.
Ayres, the Republican half of the team, noted that the actual electorate in November may not have as many Democrats as this NPR poll's likely voter sample, which he called "a best-case scenario" for the president's party.
"When you sample voters over time, you inevitably get varying proportions of Democrats and Republicans in the sample. It's nothing nefarious, just the vagaries of sampling," Ayres said. "This sample ended up with seven points more Democrats than Republicans. In 2008, there were seven points more Democrats than Republicans in the electorate, according to exit polls, But in 2004, there were equal numbers of Democrats and Republicans."
If this year's voters were to split evenly again between the two major parties, Romney would have an advantage. The NPR poll found him a 4-point favorite among independents.
Most observers expect this year's party ratio to be somewhere between the Democratic tilt of 2008 and the even split of 2004 (which recurred in the midterm elections of 2010). Stan Greenberg, the Democratic member of the polling team, said polling this year has generally found fewer people self-identifying with the GOP.
"They're moving into the independent category," Greenberg said, "where also if you look at the brand position of the Republican Party and Democratic Party, the Republican Party favorability has been dropping throughout this whole period."
But while the ranks of independents are growing, that does not imply a large number of undecided voters with five weeks left to Election Day. The pollsters found only 2 percent calling themselves undecided. Moreover, only 11 percent of the president's supporters and 15 percent of Romney's said they might still change their minds.
"We have a very polarized electorate where people go to their tribal corners and fight it out," Ayres said. "But in an election this close, even a point or two could make a difference."
The poll indicates that the Republican challenger has a tall order to fill Wednesday night and in the remaining weeks, as he has fallen behind on issues such as taxes, the economy and Medicare.
Ayres said that meant his party's nominee needed "to paint a compelling picture for a better economic future, and explain why his emphasis on small businesses and private sector solutions is more likely to succeed than Obama's emphasis on governmental and public sector solutions."
Ayres also noted that when his firm tested the statement "Obama's economic plan is working and we need to stay the course" versus the opposing option "Obama's economic plan is not working and we need to try something different," the latter choice was easily the more popular. But when the statements were altered to emphasize Romney's video quotation about "the 47 percent who don't pay taxes," the results were different.
"That reinforces the argument that Obama cannot win a referendum on his economic record," Ayres added. "The only way he can win is to so thoroughly trash Romney that he becomes an unacceptable alternative."
The NPR poll, like others in recent weeks, showed half the electorate giving Romney an unfavorable personal approval rating. Ayres said that was the other imperative for Romney in the debates: "He needs to come across as knowledgeable and compassionate about people who are hurting in this economy. ... If he does that, then he will help to close this gap."
Democratic pollster Greenberg maintained that efforts to make the election a referendum on the economy had been under way for months and had yet to take hold. Nonetheless, he said, the president cannot afford to sit on his current lead.
"He's got to decide on one thing that he wants to communicate here," Greenberg said. "My guess is he'll want to communicate a presidential but not arrogant empathetic style. He's got to focus in a way that seals the deal."