Irreconcilable Differences: Either The State Polls Are Divorced From Reality, Or the National Polls Are

Posted at Hot Air, Sean Trende crunches some numbers in an attempt to determine if the state polls and the national polls could both be right.

His conclusion: They can't both be right. One series of data must be wrong. (Or, actually: At least one series of data must be wrong. They could both be wrong.)

He attempts to discover if it's possible -- as some liberals analysts suggest -- that Romney might actually be ahead in national polls owing to overperformance in non-swing-states, lightly polled states, both Blue and Red. This theory postulates that Romney could be closing the gap in Blue states where he'll lose anyway, and running up the score in Red states where he'll win away, but is still stubbornly behind in swing states. Thus, his national poll lead, while nice, is owed to good performance in states where it really doesn't matter. Where it does matter -- the highly polled swing states -- he's behind.

But he crunches numbers to see if that sort of scenario is likely. His finding is that it just isn't likely that Romney could plausibly have high enough margins in noncompetitive states to give him the national vote lead while simultaneously losing the competitive ones.

Trouble is, he's not able to say which data set is more reliable. Though he does knock Nate Silver (not by name, but it's implicit) for the unwarranted assumption that of course the state polls, mostly conducted by firms that haven't been around all that long so we have little idea of their accuracy, must be right.

So this is what it is, ultimately: Both sides have reason to think they're ahead. Both sides, it seems, really do think they're ahead. Neither side is actually blowing smoke (as a losing campaign like McCain's did in 2008).

One side is going to be crushingly disappointed on Election Night.

Something just terrible is about to happen to somebody. I sure hope it's not us.

Now, I would take solace for my preferred take on things from the fact that everyone is now going to Wisconsin.

And when I say "everyone," you know who I mean?

I mean EV-ER-Y-ONE.

Ryan Stumps in Wisconsin; Clinton, Biden, Romney and Obama Follow Suit

That's kind of a good sign, but then, if Obama thought he was ahead in Ohio, and Wisconsin became Romney's only possible hope, it would make perfect sense for him to go up there to make sure Romney didn't create a new avenue of victory.

So that doesn't really clarify things.

From the beginning of this campaign, I've been thinking it comes down to one thing: 5% of the voters defecting from Obama, to either flip to Romney or, at least, simply not vote (or vote for a third party candidate).

Obama got 53% of the vote. If 5% of the voters -- about 10% of his total voter pile -- flip to the challenger or stay home, Romney should win.

Since the polls are conflicting on this point, let's look at newspaper endorsements.

Today came another flip: The Nashua (NH) Telegraph flipped from its 2008 endorsement of Obama to endorsing his challenger Romney in 2012. From that endorsement:

Four years ago, with little hesitation, we endorsed then-Sen. Barack Obama to become the 44th president of the United States, saying it was a time for “new leadership, a new approach to governing, a new way of conducting the people’s business.”

So the basic question facing The Telegraph editorial board when it met last week came down to this: Did the former Illinois senator do enough to live up to those admittedly high expectations to warrant a second term?

After several hours of spirited debate, not unlike conversations taking place in kitchens and living rooms across America, we reached a consensus that he had not. Perhaps more importantly, when we identified the key challenges facing the nation – jobs, the economy and the national debt – we concluded he was not the best candidate to meet them.

That person is former Massachusetts Gov. Mitt Romney, and we hereby endorse him to become the 45th president of the United States.

I've written earlier that these flips in newspaper endorsements are somewhat helpful for Romney, as a direct matter, as they will tend to sway some undecided voters.

But they're also important indirectly as proxies for, as the Telegraph says, the "conversations taking place in kitchens and living rooms across America." Editorial staffs aren't kitchens and living rooms, of course-- but they're actually more liberal, and more likely to buy into Obama's brand of progressivism than most families talking about the question.

Now, looking at newspaper endorsements, has Obama lost that game-changing 10% of his 2008 support?

Actually, looking just at newspaper endorsements, it's closer to twenty percent. Almost one in five of the major metro dailies which endorsed Obama -- most with a long, long history of endorsing Democrats for President -- have now flipped to support Mitt Romney.

I don't know what that says, exactly. Obviously I'm casting about for confirmatory data.

Still. If you've lost 20% of the liberal-leaning newspapers which enthusiastically supported you in 2008, and which endorse Democrats almost as a reflex, then you just can't be doing all that well with the voters.

posted by Ace at 04:59 PM

| Access Comments