I read with some interest (and a few chuckles) the front page diaryby Jed Lewison, which describes the growing notion among conservatives that "the polls" are a product of a vast, left-wing, socialist conspiracy designed to kill the will of the Republican faithful.
That desparate Republicans can so easily toss aside science & methodology while breaking out the tinfoil hats is not surprising. Pathetic and sad, but not surprising. As Bill Clinton might say, I don't know whether to laugh or cry.
A look at Florida, however, throws cold water on this theory. If anything, it kinda points in the other direction. Since July 1, 2012, there have been 26 head-to-head polls conducted of the state by a variety of firms. Of those 26 polls, President Obama has led in all by 9 of them.
And of those 9 polls in which Obama has trailed, he has trailed by an average of just 2 points. Moreover, these 9 polls were conducted by just four pollsters: Rasmussen, Gravis Marketing, Purple Strategies, and AIF/McLaughlin. Now, prior to 2012 I had never heard of the latter three outfits. Rasmussen, of course, has been around and mocked for years now.
First, here's a list of all polls conducted in the state of Florida since July 1, 2012 according to the folks at RealClearPolitics:
Washington Post 9/19 - 9/23 769 LV 4.5 51 47 |
Obama +4 |
PPP (D) 9/20 - 9/23 861 LV 3.3 50 46 |
Obama +4 |
ARG 9/20 - 9/22 600 LV 4.0 50 45 |
Obama +5 |
Miami Herald/Mason-Dixon* 9/17 - 9/19 800 LV 3.5 48 47 |
Obama +1 |
WeAskAmerica* 9/18 - 9/18 1230 LV 2.8 49 46 |
Obama +4 |
Purple Strategies 9/15 - 9/19 600 LV 4.0 47 48 |
Romney +1 |
FOX News 9/16 - 9/18 829 LV 3.0 49 44 |
Obama +5 |
Gravis Marketing 9/15 - 9/16 1728 LV 2.5 47 48 |
Romney +1 |
Rasmussen Reports 9/12 - 9/12 500 LV 4.5 48 46 |
Obama +2 |
NBC/WSJ/Marist 9/9 - 9/11 980 LV 3.1 49 44 |
Obama +5 |
AIF/McLaughlin (R) 9/9 - 9/10 600 LV 4.0 47 50 |
Romney +3 |
SurveyUSA 9/7 - 9/9 596 LV 4.1 48 44 |
Obama +4 |
Gravis Marketing 9/2 - 9/2 1288 LV 2.7 47 48 |
Romney +1 |
PPP (D) 8/31 - 9/2 1548 LV 2.5 48 47 |
Obama +1 |
CNN/Time 8/22 - 8/26 776 LV 3.0 50 46 Obama +4 |
Obama +5 |
Gravis Marketing 8/20 - 8/20 728 LV 3.8 46 50 |
Romney +4 |
CBS/NYT/Quinnipiac 8/15 - 8/21 1241 LV 3.0 49 46 |
Obama +3 |
Rasmussen Reports 8/15 - 8/15 500 LV 4.5 43 45 |
Romney +2 |
Purple Strategies 8/13 - 8/14 600 LV 4.0 47 48 |
Romney +1 |
CBS/NYT/Quinnipiac 7/24 - 7/30 1177 LV 3.0 51 45 |
Obama +6 |
PPP (D) 7/26 - 7/29 871 LV 3.3 48 47 |
Obama +1 |
SurveyUSA 7/17 - 7/19 647 LV 3.9 48 43 |
Obama +5 |
Purple Strategies 7/9 - 7/13 600 LV 4.0 45 48 |
Romney +3 |
Mason-Dixon* 7/9 - 7/11 800 LV 3.5 46 45 |
Obama +1 |
Rasmussen Reports 7/9 - 7/9 500 LV 4.5 45 46 |
Romney +1 |
WeAskAmerica 7/1 - 7/2 1127 LV 2.9 46 45 Obama +1 |
Obama +1 |
|
So there you have it: President Obama has run ahead of Romney consistently over the last 3 months, with only four polling organizations -- most of them unproven -- keeping the aggregates relatively close.
Rasmussen Reports is of course well known around these parts. He's most famous for missing badly in the 2010 elections, as Nate Silver of fivethirtyeight.com noted:
The 105 polls released in Senate and gubernatorial races by Rasmussen Reports and its subsidiary, Pulse Opinion Research, missed the final margin between the candidates by 5.8 points, a considerably higher figure than that achieved by most other pollsters. Some 13 of its polls missed by 10 or more points, including one in the Hawaii Senate race that missed the final margin between the candidates by 40 points, the largest error ever recorded in a general election in FiveThirtyEight's database, which includes all polls conducted since 1998.
Moreover, Rasmussen's polls were quite biased, overestimating the standing of the Republican candidate by almost 4 points on average. In just 12 cases, Rasmussen's polls overestimated the margin for the Democrat by 3 or more points. But it did so for the Republican candidate in 55 cases -- that is, in more than half of the polls that it issued.
Rasmussen's 2008 results were more accurate, although as has been discussed here and elsewhere there appears to be a tendency to "hone in" toward the concensus as election day grows closer. Also, the Rasmussen tracking poll is at odds with most national polling showing a much tighter race today. Either Rasmussen is right and everyone else is wrong, or his methedology is biased.
In August, while diving into a Gravis poll of Ohio that placed Obama up by a slim 1-pt margin, Nate Silver of fivethirtyeight noted the firm appeared to have a house effect of 4-5 points in favor of Republicans:
A one-point lead isn’t much, and Mr. Obama has gotten some better numbers than that in Ohio. So why does this qualify as good news for him? Because this firm has had Republican-leaning results in the other states that it has polled, putting Mr. Romney up by 2 points in Florida, 1 point in Colorado and 17 points in Missouri, making it several points more Republican-leaning than the consensus of surveys in those states. Once the model adjusts for the firm’s “house effect,” it treats Mr. Obama’s nominal 1-point lead as being the equivalent of a 4- or 5-point lead instead. Thus, Mr. Obama’s chances of winning Ohio rose somewhat based on the survey.
Indeed, as recently as
this week Gravis has a poll putting Obama with a tiny 1-point lead in Ohio, which looks curious when sandwiched between a Washington Post poll and Fox News poll putting the president up 8 and 7 points, respectively. Gravis Marketin's Florida polling is equally out of whack.
I don't know much about Purple Strategies, other than it's run by Republican and Democratic consultants, Alex Castellanos and Steve McMahon. It doesn't have much of a track record yet as it started up in 2010. The other outfit, AIF/McLaughlin, is some kind of Republican business group out of Florida. They've done 2 polls of the state, both showing Mittens up 1 point.
So there you have it: In Florida, at least, the only polls that show Romney with any kind of advantage (albeit a small one) are Republican-skewed polls or polls with little or no track records.
Florida is going blue, again. And Romney is toast.
Don't let the science-hating Replublicans tell you otherwise.