Short answer: almost certainly not.
It is one of the oldest maxims in following the polls, and it applies for polls you like every bit as much as poll you don't like: when a poll is released that is markedly different than everyone else, it is pretty unlikely that everyone else is wrong.
If I have one beef with the universal declaration of today's Bloomberg Poll as an outlier, it is that I rarely hear the same conclusion drawn about a certain polling outfit's daily presidential tracking poll, though it, too, has been close to double digits away from the critical polling mass from time to time.
On to the numbers:
PRESIDENTIAL GENERAL ELECTION TRIAL HEATS:
NATIONAL (Gallup Tracking): Romney d. Obama (47-45)
NATIONAL (Rasmussen Tracking): Romney d. Obama (47-45)
NATIONAL (Selzer for Bloomberg Media): Obama d. Romney (53-40)
NATIONAL (YouGov): Obama tied with Romney (44-44)
IOWA (We Ask America—R): Obama d. Romney (45-44)
MICHIGAN (We Ask America—R): Romney d. Obama (45-43)
MONTANA (Rasmussen): Romney d. Obama (51-42)
NEBRASKA (Garin-Hart-Yang for Project New America): Romney d. Obama (52-40)
WASHINGTON (PPP): Obama d. Romney (54-41)
WISCONSIN (Marquette Law School): Obama d. Romney (49-43)
DOWNBALLOT POLLING:
(2014) CO-SEN (PPP): Sen. Mark Udall (D) 47, Bill Owens (R) 43; Udall 48, Mike Coffman (R) 39; Udall 48, Jane Norton (R) 38; Udall 48, John Suthers (R) 38; Udall 49, Tom Tancredo 39; Udall 49, Doug Lamborn (R) 36, Udall 50, Ken Buck (R) 35
NE-SEN (Garin-Hart-Yang for Project New America): Deb Fischer (R) 52, Bob Kerrey (D) 38
PA-12 (Global Strategy Group for Critz): Rep. Mark Critz (D) 46, Keith Rothfus (R) 36
WA-SEN (PPP): Sen. Maria Cantwell (D) 51, Michael Baumgartner (R) 36
WI-SEN (Marquette Law School): Tommy Thompson (R) 49, Tammy Baldwin (D) 41; Baldwin 44, Mark Neumann (R) 44; Baldwin 45, Jeff Fitzgerald (R) 39; Baldwin 45, Eric Hovde (R) 36
WI-SEN—R (Marquette Law School): Tommy Thompson 34, Mark Neumann 16, Eric Hovde 14, Jeff Fitzgerald 10
A few thoughts, as always, await you just past the jump...
There are good reasons to discount today's Bloomberg Poll, but the pollster they employ is not one of them. Selzer has one of the best track records in the game, and there has been more than one occasion where Selzer was the only pollster to actually hit the fairway (most famously, the 2008 Democratic caucuses in Iowa).
That said, there is ample reason to believe that this one is not an occasion where their polling is ahead of the curve, so to speak. Digging into the sample composition, there are some issues that seem problematic.
Interestingly, the partisan composition here is not out of line. It is 38 percent Democratic, 33 percent Republican, 29 percent Independent or refused to state. That is in-between the partisan compositions of the 2008 and 2010 exit poll data, which puts it in the realm of "reasonableness".
What seems markedly less reasonable are the racial and age breakdowns in the poll.
This poll sample was 67 percent white. The 2008 electorate, according to the exit polls, was 74 percent white. In 2010, it was 77 percent. I would expect that number to be less than 74 percent this year, as the nonwhite population grows and with an enlarged presidential electorate, but all the way down to 67 percent? I am very skeptical.
Furthermore, there were considerably more voters under the age of 30 (22 percent) than there were aged 65 or older (16 percent). In 2008, a historically strong year for the youth vote, that split was 18/16. In 2010, a fairly crappy year for the youth vote, it was 12/21.
Given how strongly Republicans perform with old, white folks (McCain took that cohort 58-40 in 2008), it doesn't take a Romney staffer to note how a younger, more diverse sample might put a thumb (or a whole fist) on the scale. Much like the Gallup poll's sample may well be tweaking numbers to the detriment of the president (as the always excellent Mark Blumenthal noted this weekend), this poll may be a tweak in the opposite direction.
In other polling news:
- Post-recall Wisconsin, as it happens, doesn't look dramatically different than pre-recall Wisconsin. Such is the verdict of the latest Marquette Law School poll, which shows that the presidential race falls between MLS' previous two polls there (Obama leads in this one by six), and that Tammy Baldwin is competitive with everyone on the Republican side save for Tommy Thompson, who enjoys an eight-point edge. Indeed, it was here that the MLS poll actually buttresses a point from the House of Ras (though they totally contradicted Ras' poll on the presidential race in the Badger State). One of Rasmussen's findings was that Thompson was a considerably stronger general election candidate than anyone else the GOP had to offer. The Marquette poll, without a doubt, confirms that finding. However, another thing Marquette confirms is that Thompson is nowhere near a lock to win the nomination. For a former governor, and the biggest name in his party for two decades in the state, to be polling at 34 percent in a Republican primary is nothing short of pathetic. However, the thing that might ultimately save Thompson's bacon is the fact that there are three Republicans vying for the nomination. If you combine the vote of the three GOPers, the "anti-Tommy" vote would beat Thompson 40-34. But, with all three in the mix, someone needs to collapse, and someone needs to surge, in order to really make Thompson sweat out the primary season.
- Anyone looking for a sign of how high the odds are for a GOP takeover of Ben Nelson's Nebraska Senate seat need know only this fact: Garin-Hart-Yang is a Democratic pollster. That makes that fourteen point deficit look even worse, when placed in context.
- Speaking of Democratic pollsters, we get another Democratic poll that provides the DCCC with good news. The GOP state lege in Pennsylvania did everything they could to screw the Democrats in redistricting, and one of their crowning achievements was shoehorning Democratic Reps. Jason Altmire and Mark Critz into one district. What's more: it was a 54-45 McCain district in 2008. That said, Critz dropped an internal poll today showing him up 10 on Republican Keith Rothfus. That's despite the point that, according to the Critz campaign, the president trails Mitt Romney in the district by double digits. Given that Obama did not do particularly well here in 2008, this is entirely plausible. It doesn't make it any less obnoxious that Critz made his big "I'm too independent to go the Democratic Convention" spiel this week, but it does provide some context for that decision.