GOP Surge in Party ID?

Pollster Scott Rasmussen finds what looks like a significant shift in party identification, with Republicans rising 2 points in December to 34, as against 36 for Democrats. This looks a lot more like the 37-37 split recorded in the 2004 exit poll than the 38-31 Democratic advantage in December 2006.

Here are Rasmussen's month-to-month numbers:

Or consider these December numbers:

Another way to look at it: Republicans were running just under parity from January 2004 to January 2006. Democrats jumped to a 4-point lead in July 2006 and maintained it or ramped it up as high as 7 points from July 2006 to November 2007. Then we're back to near-parity in December 2007.

Yes, it's just one month. It may not be sustained. Indeed, the dynamics of the presidential campaign could change these numbers a lot. But party identification ordinarily doesn't change very much, so small changes are different. Rasmussen's numbers in retrospect forecast the Republican defeat in November 2006.

He observes, "It is startling to note that the Democrats have lost two thirds of the partisan advantage since taking control of Congress." Could this be some kind of turnaround?