« Does Haggard Hurt? | The RCP Blog Home Page | Adwatch '06: Heather Wilson »

New Hampshire - Part III

The UNH poll I've already referenced twice (see here and here) is a tracking poll, and the new numbers out this morning show Democrat Paul Hodes extending his lead over incumbent Republican Charlie Bass to 13 points (49 to 36) - up five points overnight (the numbers from NH-1 were unchanged).

The Bass campiagn questioned the accuracy of the poll, and the Union-Leader reports on the response of UNH poll director Andrew Smith:

Smith said that statewide, 42 percent of those who first said they were registered undeclared voters later identified themselves as Democrats, while 33 percent called themselves independents and 25 percent Republicans.

Yesterday's 2nd District poll included 102 people, or 35 percent of the sample, who said they were registered Democratic likely voters; 73, or 25 percent, registered Republican likely voters; and 115, or nearly 40 percent, registered undeclared likely voters.

Those numbers differ from the official voter registration figures for the entire state. According to the Sept. 12 primary voter checklist on the Secretary of State's Web site, 26 percent of New Hampshire voters are Democrats, 31 percent Republican and 43 percent undeclared.

Smith said many more Republicans than Democrats were eliminated from his poll because they said they will not or probably will not vote.

Smith said he has used this methodology since about 1985 at the University of Wisconsin, the University of Cincinnati and, since 1999, at UNH.

"I don't weigh this toward any particular party," he said. He also said that unlike the method used by pollsters who try to reflect the voter affiliation makeup of the state in their samples, "I don't try to assume who is going to show up to vote."

He said that many campaigns use lists of people who voted in previous elections as their samples, but he said, "That makes an assumption about what is likely to happen this time around."