Oct 08 2012

Poll Turnout Models Are Confused

Published by at 9:38 am under All General Discussions

Update: Ed Morrissey has some similar observations over at Hot Air. Ed catches something important that I missed:

That’s even more true when one takes into consideration the sample breakdown in this poll.  It has a D/R/I of 38/30/32, for a D+8.  That’s more Democratic than 2008?s D+7, which took place in a cycle with much more Democratic enthusiasm than this poll demonstrates.  If Obama trails in a D+8 poll by 6 among the extremely likely voters with four weeks to go, he’s in deep trouble — and his debate performance certainly won’t boost him.

That is really a stunner, but goes back to what I have been saying. The old turnout model with a ridiculous +D sample is the only way you get a close race. If you adjust for the intensity in the sample itself, you get a Romney blow out. So it would seem the polls are skewed, in that they refuse to accept the change being seen in the sample, and continue to pretend it is 2008. – end update

Early in an election year pollsters have little choice but to seed their polling analyses with turnout models that reflect historic percentages by groups. This application of historic turnout is used because in normal elections periods the electorate is vary stable in its voting patterns. But in times of upheaval and concern, the voting patterns can shift. As we have seen since President Obama has been elected and produced a string of failures and questionable results (i.e., Solyndra), the electorate has been increasingly rejecting the path the Liberal Democrats set this country on 4 years ago. So in my mind historic turnout models are pretty useless.

That is evident in the surprise Governor elections of 2009, when the GOP swept Christie and McDonnell into office in NJ and VA (respectively) with historic win margins. It was true in early 2010 when Scott Brown won a surprise special election in MA to fill Ted Kennedy’s Senate Seat (special elections rarely send political messages, but this one did). It was demonstrated in spades during the 2010 midterm elections when the Democrats were wiped out in a historic drubbing. And it was true in 2011 and 2012 when the left failed to upend Wisconsin Senate and Gubernatorial elections when fighting Scott Walker.

For those who can connect dots, the pattern is clear.

The pattern is also much clearer today with the release of a brutal Battleground Poll. There are two turnout models being provided in this poll (something I wish every pollster would do to show the probable ranges of turnout dynamics, not just the historic model result). First, the historic model result:

A new POLITICO/George Washington University Battleground Tracking Pollof likely voters shows Obama ahead of Mitt Romney 49 percent to 48 percent nationally, a statistical tie and a percentage point closer than a week ago.

So it is a tie? Not very likely. At least these folks went into ‘the tabs’ and began to assess some confidence in the typical turnout model answer. And what they discovered is a blow to Team Obama:

Among those extremely likely to vote, Romney actually leads Obama 52 percent to 46 percent. That’s up from a 2-point lead last week. Obama led 50 percent to 47 percent among this group three weeks ago.

Emphasis mine. In three weeks we are supposed to believe the political dynamic shifted from +3 Obama to +6 Romney – a shift of 9% towards the GOP? Honestly, it could be true given Benghazi and Obama’s debate performance. I think other forces are in play here, especially a quiet but determined Tea Parry voting bloc who has been keeping politics out of their lives until the voting begins. And it has begun.

Irregardless, what this second result clearly shows is if you re-jigger the turnout model to represent group intensity (what the turnout model would change to from historic trends), then the result is dramatically different. This is a clear example of historic trends vs current pulse of the electorate.

And this makes a lot of sense when you think about it. The trend since 2008 has been historically unique turnouts, resulting in historic wins for the GOP. So if you blindly assume the Dem turnout success will be as big or bigger (to wipe out Obama’s erosion with independents) than 2008, then you can eek out a close race.

But if you look at how things have shifted after 4 years of brutal economic recession and failed government trickle down economics, then you get a Romney blow out.

There are clear details inside this poll which provide further confirmation of a historic blow out. First, partisan intensity:

Only 73 percent who support Obama say they are “extremely likely” to vote, compared to 86 percent who back Romney. Likewise, 84 percent of Republicans say they are extremely likely to vote, compared to 76 percent of Democrats.

Then, there is the independent vote – the home of the Tea Party Libertarian movement:

Romney now leads among independents by 16 points, 51 percent to 35 percent. This is up from 4 points last week.

+12 in on week? Again, I suspect something more than a debate is at work here. It is possible the debate gave the insurgent 2010 voter reason to reengage and begin answering pollsters again. Or maybe the facade dropped and more people are answering honestly. Who knows, but a +16 with independents does not create a tie at the top.

The pollsters think this independent edge is offset by monolithic support from minority voters, yet they also say this:

While 82 percent of whites (who break for Romney by a 15-point margin) say they’re “extremely likely” to vote, only 71 percent of African-Americans and 70 percent of Latinos do. And just 68 percent of 18-to-29-year-olds, another key Obama constituency, put themselves in the “extremely likely” to vote category.

My gut feeling tells me the massively unemployed 18-29 year old group either sits this one out or votes against Obama – being the hardest hit economically by Obama-Reid-Pelosi mistakes. So take this argument with a grain of salt.

Gallup is also seeing a ‘surprise’ jump since the debate debacle. Expect more erosion to come. Obama will be walloped in the foreign policy side over Benghazi. That completely avoidable disaster begs the question – who in their right mind removes security forces while an Embassy is being targeted, hit and threatened? There is no defense of this, and Obama will once again be back on his heels fumbling.

6 responses so far

6 Responses to “Poll Turnout Models Are Confused”

  1. lurker9876 says:

    Politican Yard signs?

    I’m seeing Romney signs in my neighborhood. How many Obama signs in my neighborhood? ZERO!

    How about you?

  2. lurker9876 says:

    So the explanation of pulling out three security teams out of Bengazhi was that their mission was complete?


    A really great cover up or excuse!

  3. lurker9876 says:

    The mainstream media promised us that Obama will do better in the next two presidential debates AND win them.

    Nope. Obama ain’t gonna do any better and will lose miserably.

    It’s about time that we finally have someone that is really good and intelligent to stand up to Obama. Romney is tha’ man!

  4. archtop says:

    We’ve discovered a secret tape with Obama’s preparation for the next debate on it! Subject: foreign policy…

    Moderator: Mr. Obama, what is your opinion on Syria?

    Obama: I killed Bin Laden!

    Moderator: Uh, about the tensions between Iran and Israel, do you…

    Obama: I killed Bin Laden!!

    Moderator: OK. Let’s switch gears, what makes your approach to foreign policy different from Mr. Romney’s?

    Obama: Uh..Err…I don’t think before I shoot…I mean, shoot before I think…and I killed Bin Laden!

    Moderator: Any rebuttal Mr. Romney?

    Romney: Nope. Please…let the President speak some more…he can even have some of my time!

    Obama: Where’s the ^&*$% teleprompter??!


  5. Redteam says:

    I have little doubt that more voters will vote for Romny than Empty Chair, I’m just worried about how many votes are tallied for each. Kinda like in Venezuela where apparently the challenger got more votes than chavez but somehow the votes got lost and when they were found they actually favored Chavez. Isn’t that strange?
    If we can just keep all the dead people away from the polls, Romney will win. Empty Chair will get close to 100% of the dead voters.

  6. I strongly suspect we are looking at the execution of a deliberate campaign strategy by the Romney campaign.

    Pew shifted from reporting “all voters” to “likely voters” at the beginning of October like it does each election year.

    The other polls are following suit and it all happened at the same time as bump from Romney’s very good debate performance.

    A lot of Republicans and Very Right Wing Conservatives were hitting Romney for giving in on the media types in the debates.

    It looks to me, that for the Romney Campaign in the debate negotiations, the _timing_ of the 1st Debate to be with those usual election poll change overs was the goal, not the getting Republican leaning media talking heads.

    The Romney Campaign bet it all on a Romney Debate performance destroying Obama’s image…and it seems to have worked.

    And it worked because Romney spent more time on debate prep than any other Presidential candidate in history going in to face a President he knew to be thin skinned and coddled by the media.

    The books on this Presidential campaign by and for the political pros will be darned interesting.