About Those "Rigged" Polls

My column from this morning partially debunked this GOP nonsense about the allegedly intentional skew in polls to favor Obama. Here's a little more research on the subject.

I decided to go back and look at polls in previous election cycles. I started looking at some 2010 polls, where I happened across a Times poll from September that listed many of its results going back to 1992, as the Times periodically does. This was fascinating.

Go look at page 28 of this. As you will see, the Times asked poll respondents roughly every couple of months from 1992 until the September 2010 date of this poll whether they considered themselves to be a Democrat, a Republican, or an Independent.

Results:

In 151 Times polls over the 18-year period, Democrat beat Republican 148 times. The average Democratic margin in those 151 polls is +7 (6.853, if you wanna get technical). So over nearly two decades of Times polls, the Democratic advantage is seven points.

That's one heck of a long conspiracy! Actually, it reflects how people feel about the two parties at a given moment, which is what the sampling numbers are telling us now. You can look at the history. Republican ID was nearest equal footing during Dubya's good days, notably at the beginning of the Iraq war. The first GOP advantage since they started asking this question appeared, bang, mid-April 2003, right around when the war was "won." Even then it was just by a point, 31-30. The three Republican advantages were one point, one point, and two points.

The biggest pro-Democratic bulge came in early 2009, when Obama was new and riding high. The spread hit 18 points twice then and was routinely double digits throughout 2009.

And so we circle back to today. Just as throughout those 18 years, a certain percentage of people tell pollsters they "consider themselves" to be A, B, or C depending in part upon the prevailing winds, so it is now. Fewer people are telling pollsters they consider themselves Republican now than even a month ago, having watched both conventions and that video tape. So far from being a conspiracy, these numbers just reflect degrees of warmth toward each party at any given moment.

I doubt that we'll equal 2008's Dem + 7 advantage, and I certainly don't think Obama is going to win Ohio by 10 points. But there will be some kind of Democratic advantage, in all likelihood. As these numbers show, there almost always has been.