Polls: Romney Down One, or Up FIve?

Posted: Oct 29, 2012 11:27 AM

As the 2012 election clock winds down toward zeroes across, poll-watching becomes an excruciating (and sometimes contradictory) daily (if not hourly) spectator sport.  Three national polls out today show a close race, though the tea leaves may hold some positive signs for Mitt Romney.  Let's run through the results: Rasmussen's daily tracker has tightened to 49/47 for Romney.  Obama has been bouncing between 46 and 47 for days on end, while his challenger has crept just above 50 percent on several occasions.  Today Romney dips into the high 40s, while Obama remains static.  As for the organization's "swing state" tracker, Romney's lead is holding steady:

In the 11 swing states, Mitt Romney earns 50% of the vote to Obama’s 46%. Two percent (2%) like another candidate in the race, and another two percent (2%) are undecided. Romney has now led for 12 straight days with margins of four to six points most of that time.  

WaPo/ABC News' latest survey shows Romney leading the president by the slimmest of margins, 49/48.  The better news for Romney is that the poll's internals suggest a more robust lead than a mere point.  For starters, the sample is an indefensible D+7 -- with Republicans turning out at a significantly lower rate than 2008.  Ed Morrissey runs the math:

If Romney wins independents by 15, and the gender gap by 2, how can he only be leading by one overall? Simple — the Post/ABC sample has a declining number of Republicans in its sample. Today’s rolling three-day average has a D/R/I of 35/28/34, which would put Republicans at seven points below their 2010 turnout and five points below their 2008 turnout. Two days ago, the sample was D+4 at 34/30, but now it’s D+7. The toplines are simply not credible in relation to the internals. If Romney is winning independents outside the margin of error and has neutralized the gender gap, Obama can’t possibly be within one point overall.  

The Politico/GWU Battleground poll has Obama inching back ahead by one point (49/48) -- but again, the internals tell a slightly different story.  Romney leads by double-digits among independents (50/40), which improves upon Obama's eight-point win with indies in 2008, en route to a seven-point national victory.  There's also the enthusiasm gap, which favors the GOP ticket by a considerable margin:

...the GOP nominee maintains a potentially pivotal advantage in intensity among his supporters. Sixty percent of those who support Obama say they are “extremely likely” to vote, compared to 73 percent who back Romney. Among this group, Romney leads Obama by 9 points, 53 to 44 percent. By any measure, the race is neck-and-neck: 43 percent say they will “definitely” vote Romney, compared to 42 percent who say the same of the president.  

Romney has a 13-point intensity advantage, yet both he and Obama have strikingly low totals among "definite" supporters.  Forty-two percent is a weak level of hard support for an incumbent president, but Romney doesn't do much better.  Obama benefits from a D+4 that under-samples Republicans (though not as badly as WaPo/ABC News' partisan split) But here's where things get really interesting.  The bipartisan GWU/Battleground poll has also released a "vote election model," which takes demographics, voter intensity and other data into account.  When they run these numbers, Romney emerges with a five-point national lead:

The bipartisan Battleground Poll, in its “vote election model,” is projecting that Mitt Romney will defeat President Obama 52 percent to 47 percent. The poll also found that Romney has an even greater advantage among middle class voters, 52 percent to 45 percent. The poll’s election model takes into account variables including voter intensity, age, and education, and voters who are certain in their vote. The race “remains very close in the surface,” Goeas said, “but the political environment and the composition of the likely electorate favor Governor Romney.”  

I would love to get a peek under the hood of the Battleground poll's methodology.  They show a one-point lead for Obama "in the surface," but project a relatively comfortable Romney win.  Hence my "excruciating and contradictory" description above.  If you're trying to make heads or tails of the turnout models, I'd strongly encourage you to read this from NRO's "numbers cruncher" Josh Jordan.  In short, both Gallup and Rasmussen have shown Republicans pulling even with Democrats in terms of partisan identification among the 2012 electorate at large.  This is a major shift from 2008, when Democrats held a substantial lead in the party ID sweepstakes:  

In the current tracking poll, Gallup finds that the 10-point advantage for Democrats has now turned into a 1-point Republican advantage. The current party breakdown is now 35 percent Democrats, 36 percent Republicans, and 29 percent independents. And just in like 2008, that 1-point advantage increases when independents are asked which party they typically lean to, with 49 percent identifying as Republicans and 46 percent as Democrats. That number backs up the trends in other polling showing Romney leading among independents by large margins ... While Team Obama loudly declares that Gallup has to be an outlier, there is one other poll that has been tracking party affiliation every day as well: Rasmussen. Just like Gallup, Rasmussen runs a daily tracking poll with about 1,500 respondents included in the partisan-affiliation breakdown. In 2008, Rasmussen found Democrats with a 7.1 percentage point advantage in turnout, which was a perfect prediction of the Democratic-turnout margin on Election Day. In September of 2012, Rasmussen has Republicans now edging Democrats by 2.6 points, with a split of 34 percent Democrats, 37 percent Republicans, and 29 percent independents.

Perhaps this phenomenon, coupled with their own data on independents, has prompted the Battleground pollsters to project a Romney win one week from tomorrow.

UPDATE - Gallup's daily tracker shows a five-point Romney lead, 51-46.