Probably more so if your chosen line of work is to measure public opinion.
Probably even more so if your company, less than two weeks ago, gave a lead to the candidate that you are now declaring hopelessly doomed.
And yet, somehow, that is precisely what happened last night, just minutes after the Tuesday edition of the Wrap went to press. It is, without question, one of the most mind-boggling utterances to exit the mouth of a pollster.
More on that later. For now, on to the numbers:
NATIONAL (Fox News): Romney 46, Obama 45 (LV); Obama 46, Romney 44 (RV)DOWNBALLOT POLLING:
NATIONAL (IBD/TIPP Tracking): Romney 49, Obama 44
NATIONAL (Ipsos/Reuters Tracking): Romney 45, Obama 44 (LV); Obama 44, Romney 41 (RV)
NATIONAL (Rasmussen Tracking): Romney 48, Obama 47
NATIONAL (YouGov): Obama 49, Romney 46 (LV); Obama 50, Romney 44 (RV)
FLORIDA (Univ. of North Florida): Obama 49, Romney 45 (LV); Obama 51, Romney 42 (Adults)
MAINE (Pan Atlantic/SMS): Obama 51, Romney 37, Others 1
MONTANA (Montana State University-Billings): Romney 49, Obama 35, Others 3
MONTANA (PPP): Romney 52, Obama 41
NEVADA (PPP): Obama 51, Romney 47
NEVADA (SurveyUSA): Obama 47, Romney 46, Others 3
NEW HAMPSHIRE (Rasmussen): Obama 48, Romney 48, Others 3
NEW MEXICO (Rasmussen): Obama 54, Romney 43, Others 2
PENNSYLVANIA (Rasmussen): Obama 51, Romney 46, Others 1
RHODE ISLAND (Brown University): Obama 58, Romney 32
WISCONSIN (Rasmussen): Obama 51, Romney 49
AZ-SEN (Harstad Research for the DSCC): Richard Carmona (D) 47, Jeff Flake (R) 43A few thoughts, as always, await you just past the jump...
AZ-SEN (Tarrance Group for the Flake campaign): Jeff Flake (R) 49, Richard Carmona (D) 43
ME-SEN (Pan Atlantic/SMS): Angus King (I) 50, Charlie Summers (R) 24, Cynthia Dill (D) 12
MT-SEN (PPP): Sen. Jon Tester (D) 45, Denny Rehberg (R) 43, Dan Cox (L) 8
NV-SEN (PPP): Sen. Dean Heller (R) 47, Shelley Berkley (D) 44, David Venderbeek (IAP) 4
NV-SEN (Rasmussen): Sen. Dean Heller (R) 48, Shelley Berkley (D) 45
OH-SEN (SurveyUSA): Sen. Sherrod Brown (D) 42, Josh Mandel (R) 38, Scott Rupert (I) 4
PA-SEN (Susquehanna Research--R): Sen. Bob Casey (D) 46, Tom Smith (R) 44
RI-SEN (Brown University): Sen. Sheldon Whitehouse (D) 59, Barry Hinckley (R) 30
MT-GOV (Montana State University-Billings): Rick Hill (R) 40, Steve Bullock (D) 38, Ron Vandevender (L) 2
ME-01 (Pan Atlantic/SMS): Rep. Chellie Pingree (D) 57, Jon Courtney (R) 24
ME-02 (Pan Atlantic/SMS): Rep. Mike Michaud (D) 52, Kevin Raye (R) 32
MI-06 (Myers Research for the O'Brien campaign): Rep. Fred Upton (R) 47, Mike O'Brien (D) 42, Christie Gelineau (L) 7
MT-AL (Montana State University-Billings): Steve Daines (R) 36, Kim Gillan (D) 23, Dave Kaiser (L) 3
MT-AL (PPP): Steve Daines (R) 43, Kim Gillan (D) 34, Dave Kaiser (L) 10
NY-19 (OnMessage for the NRCC): Rep. Chris Gibson (R) 47, Julian Schreibman (D) 39
OH-06 (Public Opinion Strategies for the NRCC/Johnson campaign): Rep. Bill Johnson (R) 47, Charlie Wilson (D) 39
RI-01 (Brown University): Rep. David Cicilline (D) 46, Brendan Doherty (R) 40, David Vogel (I) 7
RI-02 (Brown University): Rep. Jim Langevin (D) 49, Michael Riley (R) 32, Abel Collins (I) 5
Before we get to the serious stuff, let's dispense with the most oft-cited poll story of the last 24 hours, which was the utterly stunning declaration by Suffolk University polling director David Paleologos that his unit will not be polling the battleground states of Florida, Virginia, and North Carolina.
The reason? Because, in Paleologos' view, Obama cannot win any of them:
"Before the debate, the Suffolk poll had Obama ahead 46 to 43 [in Florida] in the head-to-head number,” Paleologos responded. “A poor place to be for a couple of reasons. Number one, his ballot test, his head-to-head number was below 47 percent before the debate, and it’s very, very difficult when you have the known quantity, the incumbent, to claw your way up to 50. So that was a very, very poor place for him to be."So, for those of you scoring at home, Paleologos is saying that they won't poll three battleground states, two of which are states that Suffolk polling showed Obama in the lead, because Obama can't win.
"So we’re looking at this polling data not only in Florida but in Virginia and North Carolina and it’s overwhelming."
And all this, apparently, because Obama is not above 47 percent. Of course, other polling units that don't have a pissload of undecided voters (as Suffolk appears to have) have Obama leading in Florida and Virginia, and have him at 49 and 50 percent.
The rest of the polling community has varied reactions, though unified in their mix of confusion or condemnation of the Paleologos declaration of future Romney victory. The most thoughtful response may well have come from PPP's Tom Jensen:
"I think he’s totally wrong about Obama’s prospects in those states, particularly Florida and Virginia, and I just think it’s really strange you’d go on national TV and make those declarations without having fresh polling in hand showing Romney ahead in those states. But nuance doesn’t count for much on primetime cable."The best response, however, has to come from SurveyUSA's Jay Leve:
"This guy from Suffolk is obviously a jackass."Heh. I don't know if I would go that far, but it does seem that Paleologos is either (a) a Romney cheerleader, or (b) someone with a very 2004 view of polling as a predictor of results (as in, the long discredited "all undecided votes go to the challenger" school of poll analysis). Which means that their future polling, including new data tomorrow in Nevada, is now going to be accompanied by a pretty good sized grain of salt. For Suffolk, this is damaging to their reputation, and all because of a gigantic unforced error by someone who seemed desperate, desperate, for Bill O'Reilly to like him.
Now, onto the serious stuff...
As has been the case all week, the big electoral/polling question has been the question of whether the first presidential debate had created an ephemeral, convention-esque bounce for Mitt Romney, or had honestly re-calibrated the race.
With another day of data, we are only slightly closer to an answer there. And that is because there is a fair amount of evidence for both arguments. To close out this Wednesday's Wrap, let us embark on a quick summary.
Signs that the post-debate "bounce" was a recalibration of the race in favor of Romney
1. National polling continues to move, however incrementally, in Romney's direction. Yesterday's polls (with a likely voter screen) averaged a Romney lead of 0.6 percent. Today's polls (again, with a likely voter screen) edged out to a Romney lead of 0.8 percent. That is not totally troubling yet, since only one of the tracking polls (Ipsos-Reuters) showed Romney gains even while shedding his best polling days (Oct 4-5). However, you cannot ignore the Fox News poll. Despite the network's crusade against all things Obama, the polling they've sponsored has actually been more than fair to President Obama. And this shows a dip for the president, despite being conducted after the Thu-Fri "Romney surge" days. That would hint that Romney's gains were not entirely transient.
2. Even the Obama campaign appears to be conceding that the polls have moved in Romney's direction, even in this week's internal polling. Major Garrett of National Journal reported today that Obama's internal polling has Virginia, Florida, and (perhaps most critically) Ohio "within the margin of error". If we bought previous media reports that internal polling for both sides post-DNC had Obama up in high single digits in the Buckeye State, that is some pretty serious erosion.
3. The one bright spot in the data today, Gallup's tracking poll, might be more the result of a marked shift in their methodologies than a shift in voter behavior. Gallup has increased the ratio of cellphones in their nightly tracking sample, and that would probably lead to a sample universe that is both younger and more politically diverse. In HuffPo, noted political scientist Alan Abramowitz already notices the difference.
Signs that the post-debate "bounce" is already starting to recede
1. Taking a polling average shows, as we noted here in the Wrap the other night, that it seems as if October 4th and October 5th were the pivotal pro-Romney days. Some excellent legwork from Nate Cohn (who is definitely worth reading) proves the case. He found that in the post-debate polls conducted primarily before the weekend, the average Romney bounce was 4.9 percent. For polls conducted after the weekend, the Romney bounce was down to 1.9 percent.
2. Obama still holds leads in most of the states that matter, and margins seem to, in most cases, have eroded only slightly from pre-debate averages. Yesterday's CNN poll in Ohio (Obama +4) was matched today with a new poll in Florida, that also showed Obama +4. Now, it is a new poll by a new pollster (the University of North Florida), and any poll taken over a week with a relatively small total number of respondents invites some skepticism. But, a lead is a lead, and beats the holy heck out of the alternative. Republican-affiliated Rasmussen also continued to give Obama leads in New Mexico, Wisconsin and Pennsylvania, albeit with a sharply reduced lead in the latter. And, oddly, New Hampshire is better in the Rasmussen poll for Obama than their previous poll, where Mitt Romney actually held the lead.
3. Gallup might've goosed their methodology, but a lot of folks would posit that this new methodology actually puts them closer to accurately representing the electorate. What's more, these should've been days where Obama's standing in the Gallup poll should have eroded, since it included the three pre-debate days that Gallup had already affirmed were quite good days for the president (Obama +5). What that means, in simple tracking poll parlance, is that Obama replaced good days with what must have been, given his slightly expanding margins, more good days. One has to imagine that, by the weekend, the subtraction of Romney's excellent Thursday and Friday tracking days will pad Obama's advantage among RVs, and may yet still give him an edge with their new LV screen.