CBS Poll For VA Makes No Sense

I am not one of those “Ignore the polls!” type of people. The people who shout that were probably the same ones who were saying it a week before election day in 2008 when it was obvious Obama was going to win and win easily.

That being said, there is nothing wrong with being skeptical about polls and the numbers should be dug into at every level to see if the numbers reflect the current political environment and how they line up with the 2008 results.

Therefore, the state poll released today by Quinnipiac, CBS and the NY Times for VA (CO and WI were also included) makes little sense.

Barack Obama won in 2008 under the best of circumstances, politically. Bad economy? Check. Unpopular wars? Check. Weak Republican candidate? Check. A Democratic candidate promising big changes? Check. Large money advantage? Check.

In Virginia, Obama became the first Democrat since Lyndon Johnson in 1964 to win that state. He won 53% of the vote to McCain’s 47%. It would take the absolute best of circumstances four years later for him to duplicate those numbers.

Except it’s not even close. The economy is still bad, but it’s his economy. Yes, Bin Laden was killed on his watch, but he now has a Middle East disaster on his hands. Romney is not the best Republican candidate, but unlike McCain, he’s willing to throw a punch. Hope and change has become, “It could be worse!” and Romney has the advantage this time with the money.

Let’s look first at the exit polls from 2008 for Virginia. This is from CNN, who always have great exit polling data to look at.

Obama beat McCain amongst Democrats 92% to 8%. No surprise there. McCain also won 92% of the Republican vote. Amongst independents, Obama narrowly won 49% to 48% over McCain.

The party breakdown – Democrats, 39% – Republicans, 33% – Independents 27%. In summary, Obama enjoyed a +6 advantage amongst voters that identified themselves as Democrats over Republicans.

The latest CBS/NY times poll has Obama at 50% and Mitt Romney at 46%. Usually, when a candidate breaks that 50% barrier, the challenge is in trouble.

But when you dig into these numbers, it makes less and less sense. Let’s break it down first by party affiliation in terms of support. Obama is up 95% to 4% amongst Democrats. Romney leads Obama amongst Republicans 95% to 3%. Romney leads Obama amongst Independents 53% to 42%, an 11 point advantage.

So that’s where the head scratching starts. You think, “How can Obama be down by 11 points to Romney amongst Independents, but still be ahead by 4?”

It’s in the party identification. In 2008, in VA it was +6 for Democrats which was in line nationally. Obama won by over 7% and Gallup reported that in 2008, the party identification for Democrats was 36% and for Republicans it was 28% or +8. If we fast forward to 2012, that big advantage is gone. In looking at what Gallup has now (and ignore the latest number that shows Democrats +8 again as it reflects numbers during the the Democratic National Convention. That was reflected in the Gallup daily tracking poll for a day or two when Obama led Romney 50% to 43% but the race is essentially tied again so that advantage is gone. I remind you to look at the trend, not one particular result) the advantage averages around +3 for Democrats.

With that in mind, we turn to the party ID breakdown in this latest poll:

Democrats, 35% – Republicans 24% – Independents 36% – A +11 advantage for Obama

How did they arrive at these numbers? Again, Obama won in 2008 under the best of circumstances for him. Virginia has open primaries so there are no registration numbers to track, but it just seems odd that in opposition to all other indicators out there, Obama has increased his advantage in Virginia. In addition, nobody can say with a straight face that Romney will win Independents by 11 points and still lose overall by 4 points.

This is not the first poll in Virginia that uses what seems to be outlandish party identification numbers. If anybody has any insight as to how they are arriving at these numbers, please let me know.