Cain moves to 3rd place, Perry falls in GOP2012

Herman Cain has moved into third place as Rick Perry continues to fall in the race for the Republican Presidential nomination.

In the aftermath of Cain’s victory in the Florida straw poll, and Perry’s problems in debates, the GOP field is showing some sharp movement, up for Cain, down for Perry and improvement for now-front runner Mitt Romney, whose trend estimate is currently as high as it has ever been since January 2010. Romney had dropped off significantly in August and early September but has rebounded as Perry has fallen.

This is the second surge Cain has enjoyed since April. He had an early rise in support in late-May and early June, reaching 17 points in one PPP poll June 9-12. His subsequent drop has reversed since September 1, with a rapid rise in the last two weeks.

Cain’s strong performance in a Fox News poll 9/25-27 and a new ABC/WP poll done 9/29-10/2 reflects both his strong performance and warm reception in Florida and the difficulty GOP voters are having finding an alternative to Mitt Romney. Romney’s support has fluctuated but remained between 15 and 25 percent based on my trend estimates. He’s had difficulty breaking above 25, leaving many Republican voters trying one alternative and then another.

Meanwhile, Rick Perry’s decline continues with Fox and ABC/WP both finding his support now down to 17%, with a trend estimate at 18%. Still in second place, barely, but with no recent good news to arrest his fall.

And we still await at this hour the decision of Chris Christie on whether to enter the race. The ABC/WP poll put his support at 10% when included with all other current and potential candidates, a fair showing but certainly not an overwhelming mandate to enter.

Methodological Note: The ABC/WP poll presents several results for candidate support, for both all GOP and GOP Leaning respondents and for just registered voters among GOP and GOP leaning respondents. They also initially include Palin and Christie, and the allocate their supporters among current candidates. I use the registered voter sample with only announced candidates. This is also the practice at Real Clear Politics. In general, I try to use the most likely voter results published by the pollster, but as in this example there are often several different results from a single poll among which to choose. The Post story today, for example, uses the GOP+Leaners but not filtered for registered voters. Legitimate differences in judgement may lead to small differences in which results to present.