Why the new CBS/NYT/Quinnipiac polls are meaningless
I am not a poll truther, indeed when Romney was trailing in the polls and the trendline for him was bad in September, I warned conservatives to take those polls seriously and stop trying to claim that every poll was part of some secret plot to undermine Romney. That having been said, someone needs to explain to me how — according to the new NYT/CBS/Quinnipiac poll — Romney leads among independents in Virginia by 21 points but is somehow losing the state to Obama by 1 point. Is there a single sober person who has a turnout model for Virginia that would allow Obama to overcome a 21-point deficit among independents? I don’t think so. If Romney wins indies by 21 points on election night, he carries VA by 5 points.
In 2008, independents made up 27% of the Virginia electorate and Obama won them by 1 point en route to a 6 point win. NYT/CBS/Quinnipiac says Romney leads by 21 points among independents in Virginia today. Yet, somehow their poll shows Obama actually ahead in Virginia by 1 point. For the sake of argument, lets just pretend 2008 turnout turnout model, a model most analysts believe overstates Obama’s numbers, is the turnout model for Virginia in 2012. Even by the 2008 turnout model Obama simply can not lose independents by 21 points and win the state.
Its not just Virginia, in Ohio the Quinnipiac poll shows Romney ahead among indies by 6 but losing the state by 5 points. Again, if you assume the 2008 turnout model - the most advantageous to Obama as humanly possible - this simply defies logic. Obama won independents in Ohio in 2008 by 8 points, independents made up 30% of the Ohio electorate in 2008. If Romney leads independents in Ohio by 6 points, and even assuming the 2008 Obama super turnout model, how in God’s name is Romney losing by 5 points? Simply put, it is not possible.
I am not a tin foil hat libertarian, I don’t believe that everything is part of some vast conspiracy, and that includes these polls. What I do believe, however, is that these polls are at odds with themselves and with any reasonable turnout model for this election. I question any media outlet that puts faith in a poll that includes such huge internal contradictions and that isn’t based on any plausible turnout model.
I don’t know whether Obama or Romney is ahead for sure in Ohio, though I suspect Obama has a slight lead. I don’t know whether Obama or Romney is ahead for sure in Virginia, though I suspect Romney has a slight lead. What I do know for sure and with 100% certainty is that the CBS/NYT/Quinnipiac polls don’t give us any idea of who is ahead in either state.