Poll Denialism has subsided a bit since Romney began making gains last last week, but here it is again from John Nolte at Breitbart.com, in a form whose perversity really needs to be understood:
In 2008, President Barack Obama won the independent vote over John McCain by a margin of eight points, 52-44. This morning, a new Battleground Poll has Mitt Romney massacring Obama among indies by a whopping 16 points, 51-35.
That’s a 24-point swing among independents since 2008, a group that makes up anywhere from a quarter to a third of voters, and yet Battleground still has Obama in the lead 49-48…?
But if I’m skeptical of those bottom-line numbers, our journalist overlords who have chosen to palace guard instead of question will declare me a “truther.”
Put aside for a moment that Nolte is conflating 2008 exit poll numbers for self-identification with those of a current GWU-Battleground poll, with no examination of how the question is asked and how hard respondents are pushed to choose a label. And forget if you can that exit polls almost invariably show lower percentages of indies because nearly all respondents have just voted for a partisan candidate. There’s a more fundamental misunderstanding here.
At the height of the Poll Denialism uprising, there were two arguments you heard to support the idea the polls were cooked. The one that got most attention was that pollsters were allowing too many people to identify themselves as Democrats, and thus should “weight” the numbers to reflect the “correct” proportions based on past elections. But the second was that if Mitt was winning among independents, he couldn’t be trailing overall! The second is Nolte’s argument.
The two issues are, of course, inter-related: what happened after 2008 was that Republican self-identification took a significant dive, in part because the Tea Party Movement persuaded many of the most conservative voters in the electorate–certainly those least inclined to vote for Barack Obama in 2012–to self-identify as independents, so as to help persuade the Republican Establishment not to take them for granted. Presto change-o, this shift in self-identification simultaneously produced a larger Democratic plurality over Republicans while tiling independents significantly in a Republican direction. The net effect on likely candidate preferences was nil. Yet now both of these phenomena are being used to suggest there’s a secret Romney majority in the electorate that pollsters don’t want to admit.
It’s not real complicated. If part of your party’s actual base begins denying its partisanship, it will have a perverse effect on the spectrum of self-identification. If you have issues with that, Mr. Nolte, blame your Tea Party buds.
