Presidential Polls 2012: Latest CBS NYT Quinnipiac Poll Shamelessly Skews for Obama


With less than a week to go until the election, the time has come for most major polling firms and organizations to release their final polling results and predictions. A great deal of the polling samples have become more reflective of the electorate as it exists today, significantly tightening the race. However, it is somewhat refreshing to see that the latest CBS/NYT/Quinnipiac Poll is doubling down and providing us with a poll that has as much basis in reality as unicorns, flying pigs, and a Michael Bay movie where women are dressed normally.

The latest CBS/NYT/Quinnipiac Poll is a last hurrah for those that either still believe that Democratic turnout in 2012 will resemble turnout in 2008, or are desparately trying to convince others that it will happen. Rather than accurately reflecting reality, it is much more likely that these polls a symbolic attempt to by these organizations to collectively thumb their nose at a country that seems to be increasingly less enamored with President Obama every day. 

Most pollsters make sure that the last polls they put out before an election are fine-tuned, so that they can try and reflect the final results as accurately as possible. Well, thank god that CBS, the New York Times and Quinnipiac didn't buckle under the pressure!

This poll shows Obama leading in the key swing states of Ohio, Florida and Virginia. In Ohio, Obama leads Romney, 50% to 45%. In Florida, he is ahead 48% to 47%, and in Virginia, he leads Romney 49% to 47%. How did this poll manage to pull this off? Easy. The samples are so skewed that the pollsters behind it should be ashamed of themselves.

Before liberals start celebrating, a closer look at the samples and projected turnout is necessary. The samples used here do not just count on the Democrats' turnout being as high as it was in 2008; they also make the assumption that the Republican turnout will be even lower than it was four years ago! That is simply not going to happen. Not only has Mitt Romney actually led in early voting, but Gallup has also projected that this year we will see the highest GOP turnout for any election, ever.

This poll completely ignores the exit polls from 2010, instead choosing to focus on those from 2008, when Democrats had a good year. Here is a breakdown of the D/R/I exit polling for 2008 and 2010 for these three states, as well as the sample that was used for the 2012 predictions. (Virginia's middle numbers were taken from 2009.) 

FL: 2008 - 37/34/29, 2010 - 36/36/29, 2012 - 37/30/29

VA: 2008 - 39/33/27, 2009 - 33/37/30, 2012 - 35/27/30

OH: 2008 - 39/31/30, 2010 - 36/37/28, 2012 - 37/29/30

So, with this data, we can assume that the D+7 and two D+8 samples are what they believe the turnout will be. Now, here is a chart that shows the voter enthusiasm for these same three states:

Are we supposed to expect that all of these new GOP voters, who are more enthusiastic and have appeared in greater numbers more recently, will just sit at home? Did they all suddenly die off last week? There is no bloody way that Democrats will be out voting Republicans by 7 or 8 percentage points on Election Day. Even if one were to compromise and assume it will be closer to even, even according to this poll Romney still enjoys a massive lead over Obama with independent voters.

In 2008, President Obama led with independents in all three states. Four years ago, he was ahead +7 in FL, +8 in OH, and +1 in VA. Now, Romney is the one that enjoys the clear advantage, leading with independents +5 in FL, +6 in OH and +21 in VA. 

After learning that Romney has led Obama in early voting by 7 points, the only hope Democrats have is to assume that they will vote in larger numbers on Election Day. While the infamous 538 blog still predicts that Romney will be crushed, and the Real Clear Politics average shows a statistical dead heat, neither of these two account for the oversampling of one group or the undersampling of another. Averaging a bunch of polls that predict more Democrats than Republicans turning out will of course indicate an Obama victory. 

Some people will trumpet the fact that Obama and Romney are tied in the last Gallup Poll at 48% each. However, that is poll of registered voters. When it comes to likely voters, Romney leads Obama 51% to 46%. In the latest Rasmussen poll, Romney leads Obama, 49% to 47%. 

As for some of those swing states listed above? Some polling firms have packed up and left, stating that there is no point in staying because Romney is going to clean house. Independent polls in Florida and Virginia also show Romney in the lead, while others in Ohio show a razor-thin race. 

But wait! There's more!

Hot on the wires now is a new PPP Poll that shows Obama leading in Ohio, 50% to 45%. This kind of statistic can be taken to the bank, provided that one believes that 45% of the voters in Ohio are Democrats. The D/R/I on this one? 45/36/19. 

Yeah, that makes sense. The Democrats will add 6 points to their 2008 turnout, while most of the independent voters will phone it in. Okay. Sure. Anything to try and tilt the RCP average and give Chris Matthews a thrill up his leg. I didn't think it could get more ridiculous than the above mentioned CBS/NYT/Quinnipiac poll, but thankfully, Public Policy Polling did not disappoint!