New York Times: Alaska Polls Suggest Murkowski Has Upside as Write-In
By NATE SILVER
Two new polls out Wednesday find that Senator Lisa Murkowski, who is running a write-in-bid for the U.S. Senate, is still an appealing figure to some Alaskans.
First, a CNN poll showed Ms. Murkowski running just 2 points behind the Republican Joe Miller among likely voters. The Democratic candidate, Scott McAdams, was in third place with 22 percent of the vote.
In addition, a poll provided to the New York Times by Craciun Research shows Ms. Murkowski with an apparent lead. The poll, which was conducted on behalf of the Alaska Native Tribal Health Consortium, gave Ms. Murkowski 41 percent of the vote, Mr. Miller 30 percent, and Mr. McAdams, 19 percent. It was conducted Sept. 24 and Sept. 25 among 300 Alaskans who have voted in three of the last four statewide elections.
The issue with each of these polls, however, is that they made no particular accommodation to account for the fact that Ms. Murkowski will be a write-in candidate, as her name was mentioned alongside Mr. Miller and Mr. McAdams. A Rasmussen Reports poll, by contrast, which had shown Ms. Murkowski running 15 points behind Mr. Miller, had tried to account for this by withholding Ms. Murkowski’s name when they initially read the names of the candidates, but then including her as an option when respondents were asked to press a button to designate which candidate they supported.
There are other differences between the Rasmussen poll and the two that were released today: Rasmussen uses an automated response system, for instance, whereas the CNN and Craciun polls used live callers. And the CNN and Craciun polls are more recent, during which time awareness of Ms. Murkowski’s write-in bid may have grown.
Still, given the unusual circumstances of the race, it is probably best to look at the CNN and Craciun polls as reflecting Ms. Murkowski’s upside, rather than necessarily being the numbers she is most likely to achieve. As we had noted earlier, it is probably the case that polls that treat write-in bids as though they are conventional ones overestimate their support — but it is very difficult to say how large the effect is.
