Presidential Polls: Obama, Romney Will Win, Lose

by Bob Grundfest

The polling in the presidential race over the past week has led to some interesting analysis, and depending upon your view of the numbers both candidates can claim to be winning, which also means that both are losing.

The latest RealClearPolitics composite shows the president with a small lead, but his aggregate score of 46.4 is rather low for an incumbent and gives Mitt Romney many opportunities to catch and surpass him. A case in point is an NBC News/Wall Street Journal poll out Tuesday showing Mr. Obama with a 49-43% lead. The problem is that the poll oversamples Democrats by 11 points, 46-35%. My sense is that Democrats will not turn out in November in those numbers, so the poll is not as representative as it first appears. The other issue is that the poll uses registered voters (RV) as opposed to likely voters (LV), which at this point in the campaign might be worthless.

Another poll released on Tuesday by PPP/Daily Kos/SEIU shows the race to be tied at 46%, which would be great news for Romney if not for some troubling poll internals. First, it assumes that Romney will attract 17% of the African-American vote and 39% of the Latino vote. By November, that might be possible, but other polls show President Obama garnering support from those groups that approaches 80-90%, so this poll might be a bit optimistic. Further, the poll oversamples Democrats by 7%, which, again, is possible, but it also shows that 32% of respondents identify themselves as members of the Tea Party. That’s high. Finally, this is an RV poll using automated technology and did not call cell phones. All of this adds up to a poll that has a lot of holes in it for both sides to peek through and spin as they please.

The national tracking polls give us little to go on, with Rasmussen giving Romney a 1 point lead and Gallup giving that same lead to Obama.

The real action is with state polls, and for the most part these have been good news for the president. A Rasmussen poll of Michigan gives Obama a 48-42% lead using likely voters, and a Survey USA/KSTP poll  of likely voters in Minnesota has Obama up 46-40%. Good news for Obama, but it shows a decline in his support from previous Minnesota polls which could be a sign of slippage or statistical noise. Your choice, it seems, depending upon which way you lean. A SurveyUSA poll of Florida using likely voters has the president with a 48-43% lead, but the same release showed Republican Senate candidate Connie Mack leading Democrat Bill Nelson by 48-42%. That kind of ticket splitting doesn’t make sense, so we’ll need another Florida poll to sort that out.

There has also been some considerable punditry about where this race is heading despite the polling. Two schools of thought seem to be controlling the debate. One says that the negative ads that Obama is running in swing states, combined with an improving economy in those states, will lead to his reelection. The other thought is that Obama has spent almost $100 million dollars on negative ads and only has a small lead on Romney and no poll has him at 50% or above on the national level (some state polls do have both candidates with sizable leads depending on the state). Both camps are looking for any significant movement in the polls, but won’t likely see much until mid-August when Romney unveils his pick for Vice President and the conventions get under way in earnest.

Of course, it’s the states that matter most in our electoral system and Obama is going to want to lock down as many of his 2008 wins as possible, and the above discussion suggests that he’s moving in that direction. Still, polls are close in Ohio, Florida, North Carolina (where Romney leads) and Pennsylvania. This is where the election will be won or lost and Romney has yet to advertise seriously in any of those states (he can’t until after the GOP convention).

Stay tuned for more as the polls are released.

Post a Comment

Your email address will not be published.