Wednesday, October 31, 2012

The Great Independent Debate

Quinnipiac/CBS/New York Times release new swing state polls of the big three battlegrounds -- Florida, Virginia, and Ohio.

The surveys are already causing a stir, thanks to seemingly incompatible results.

For example, in the Virginia survey, Barack Obama leads Mitt Romney, 49%-47%. There's nothing controversial about that topline number.

It's not significantly outside the RCP average of the state, and I think -- considering the demographics of the state -- you have to consider Obama the slight favorite in Virginia.

But it's just terribly difficult to somehow match that with the fact that Romney leads by 21% among independents. Obviously, that seems completely incompatible with Obama up by 2%, overall.

Of course, the answer can be found in the sampling, which suggests an 8% Democratic turnout advantage, which is 2% more than in 2008 when the entire planet was aligned for Obama.

Defenders of the predictive nature of results like this suggest that more and more Republican-minded voters are now calling themselves "independents"; thus being independent means you're a disgruntled conservative of some sort -- maybe a tea partier even.

But that doesn't line up with the Virginia results.

Why?

Because in the same Quinnipiac poll taken just two weeks ago, Obama led among independents by 2%.

If Virginia independents were really a  bunch of tea partiers or a disproportionate number of disgruntled conservatives, then Obama would have never been leading with the group two weeks ago.

So not only does the "independents are mainly Republicans by a different name" explanation fall short here, but it also runs short of understanding what the Virginia electorate is. Independents in the state are more, well, independent than in other states.

In states that have shown Romney up big with independents since the beginning, I might be able to buy that it's just a proxy for party ID, but in Virginia -- thanks to its recent polling history -- it's a bridge too far.

UPDATE: Kristen Soltis makes another good point on this, just generally (h/t Josh Kraushaar).

I have a hard time believing GOPers are even LESS likely today than 08 to call selves Reps. Ppl forget how bad 06 + 08 were.