Sunday, October 21, 2012

A Romney landslide?

Brian Cates has a very interesting post explaining how Romney might be ahead of Obama by 10 points, based on an analysis of the polling internals.  You may or may not agree with him, but he does a fine job of showing his work.  This bit is the core of his argument:
That 7 point turnout advantage in 2008 for Democrats was historic.  You mean to tell me many of these pollsters have been claiming they see numbers in their polls that leads them to conclude Democrats are going to TOP that turnout in 2012? 

Why, yes.  I am telling you that.  Because that's what they did.  

In September,  6 major pollsters claimed to see a D+7 advantage for Democrats.  3 of them  - Reuters, ABC/WaPo and Dem. Corps - claimed to see a Democrat advantage of 10 points or more.
It's plausible that this election will see a D+0 advantage (i.e. none at all).  Not guaranteed, but plausible.  If so, how does that translate?  Well, consider that even in the D+10 polls Obama didn't break 50% - in fact he hasn't broken 50% in around a year or so.  Now take away that D+10, making it D+0 and what do you get?  Low 40s - 42%, maybe 43% to Mitt's 52%.

When was the last election that say that sort of split?  It was 1988, when George H.W. Bush captured 426 Electoral Votes.  And Bush only won the popular vote 49-42.  The race this year will hinge on turnout, but I'm thinking that my prediction of Romney at 370 or so is understating the outcome.

6 comments:

The Czar of Muscovy said...

My math has Romney *already* at 252 votes as of today, with a likely 300 at least by the time of the election.

Your 370 might not be at all far-fetched.

New Jovian Thunderbolt said...

Don't get cocky

Borepatch said...

T-Bolt, this isn't cockiness. I've said repeatedly that a Romney win is very likely bad. Just calling things how I see them.

That said, if the turnout is D+0, this will be very, very bad for Democrat House and Senate candidates, and we're likely to see 52-47-1 R-D-I in the Senate. So long, Obamacare (the Supreme Court ruled it a tax, and the Senate rules don't allow the filibuster for a tax statute).

wolfwalker said...

Meh. Only one poll there is that matters: Election Day.

The rest -- not a word of it do I believe. It's MUCH too easy to rig a poll. Sample size, sample slant, selection method, weighting, wording of the questions...

SiGraybeard said...

Case in point is a Pew poll my son sent me a few weeks ago (9/20). They gave Romney a +8 advantage and said he had the biggest lead of any candidate since Clinton in '92. All but gave him the election.

So I read the fine print and find they polled 21% more Democrats than Republicans. 21% more and they only had a +8 margin? I'd figure they'd have at least +15 or 18 with that many more D than R.

As we approach the election, though, look for the polling companies to stop trying to influence the election and win the tag of "most accurate in '12", which helps get business next time around. Polls should get more accurate for that reason alone.

When all is said and done, though, Wolfwalker tells the absolute truth. The only poll that matters is the official one on 11/6.

Sean D Sorrentino said...

Yep, a crushing victory for Romney.

Here's my problem. People piss and moan about Romney being just the other face of the Gozer. We need to stop pretending that there is some acceptable candidate who we can elect. There isn't. There are those who will destroy us as fast as possible and those we can beat the hell out of until they do as we say. I want the type I can beat on.

Politics is hard. It's work every day to influence your reps and Senators to do what you want instead of what's easy and profitable. We keep trying to find some philosopher king who will rule us well instead of recognizing that it's up to us to put in the hard work of forcing them to do as we say.

Milton Friedman said it best

http://youtu.be/ac9j15eig_w

But making it profitable for the wrong people to do the right things is a great deal of work, and we are fundamentally lazy.