About That PPP Poll In Ohio

Posted by in Blog on October 15, 2012 8:41 am / 1 comment

Democratic hearts are all aflutter today with the release of a new poll from the Public Policy Institute showing Barack Obama with a sizable lead over Mitt Romney of 51-46 percent in Ohio.

Such results do not look good for Romney, but when you dig a little deeper into the numbers, they just do not make any sense.

First, let’s go back to the exit polling data from 2008. Barack Obama won Ohio by 5 points over John McCain. In doing so, Obama:

1. Won 89% of the Democratic vote.

2. Won only 8% of the Republican vote.

3. Won amongst Independent voters, 52 to 44 percent and that is big.

In the current PPP poll, Obama:

1. Has support of 90% of the Democratic vote – This is virtually unchanged from 2008.

2. Is getting 11% of the Republican vote – This is a mind boggling number. Somehow or another, PPP was able to find more Republican support for Obama in 2012 than he received in 2008.

3. Obama is down amongst Independents 50-45 percent – This represents a huge 13 point shift from 2008.

Even if the 11% of GOP support is to be believed (and it is nearly impossible to believe), it’s hard to see how a 13 point shift amongst Independents results in Obama winning Ohio with the same spread as in 2008 (Obama won Ohio 52-47).

According to PPP, the anomaly is explained because of early voting results:

The key finding on this poll may be how the early voters are breaking out. 19% of people say they’ve already cast their ballots and they report having voted for Obama by a 76-24 margin. Romney has a 51-45 advantage with those who haven’t voted yet, but the numbers make it clear that he already has a lot of ground to make up in the final three weeks before the election.

Emphasis mine. These “early voter” results remind me a lot of the exit polls in 2004 that were leaked showing John Kerry winning the election easily.

Even Nate Silver says something is off:

We will see.

 
 

1 Comments

  1. Tom Dougherty says:
    October 15, 2012 at 4:50 pm

    Jay, regardless of their *early voter* explanation the poll was way short on sampling of Indies and their own crosstabs, as you cite, have Romney +5 there. Nate Silver is also right when he looks at early voters vs. likely voters and observes that such a disparity in the numbers is accurate. The numbers, per my analysis, is Obama +1.

Post a Comment