As we move forward into 2010, we’re going to be fed one helluva lot of polls.
Rasmussen. Gallup. PPP. And on and on and on.
Each pollster has their own methodology. Their own formula to determine who is polled and why.
One of the largest variables in determining a poll’s sample is to set a standard of partisanship. If you oversample Democrats, Republicans or Independents, the accuracy of your poll becomes less valid. So, how exactly does the country identify itself?
Rasmussen had some surprising numbers that it released on Sunday:
Currently, 35.5% of American adults view themselves as Democrats. That’s down from 36.0 a month ago and from 37.8% in October. Prior to December, the lowest total ever recorded for Democrats was 35.9%, a figure that was reached twice in 2005.
The number of Republicans inched up by a point in December to 34.0%. That’s the highest total for Republicans since December 2007, just before the 2008 presidential campaign season began.
That means that among likely voters, the partisan gap is a measly 1.5%. For reference, previously Quinnipiac has used a partisan gap of 9 points to poll the Ohio gubernatorial race. Do I expect Quinnipiac to all of a sudden change their partisan breakdown down to 1.5%? Of course not. They have their own way of determining Party ID of the electorate. But with consistent polling from Rasmussen that shows a shrinking partisan gap, a gap in other polls above 6 or so should be met with a sense of skepticism.
Another point of interest as we move forward is when respected pollsters like Quinnipiac, Gallup and PPP begin to focus more on likely voters rather than all registered voters. That will provide a more clear picture of where we stand as we head into November.
Keep an eye out.