Wednesday, September 29, 2010

Handicapping The Alaska U.S. Senate Race Is Like Trying To Eat Jell-O With Chopsticks; No Two Polls Are Alike: A Tossup Between Murkowski And Miller

Update September 30th: Have now included the results of an Ivan Moore poll discussed in the Anchorage Press. Have also done some re-formatting to make this lengthy post a bit easier to follow.

Professional pollsters are pulling their hair out over trying to handicap the Alaska U.S. Senate race. The difficulty is in how to present Lisa Murkowski, who's running a write-in campaign, in the polls. There is no recent national precedent to serve as a reference.

There is one common denominator, though. Democratic candidate Scott McAdams is a distant last in each poll. Results of each of the five polls presented after the jump

Craciun Research Poll

On September 29th, 2010, Alaska Dispatch informs us that Craciun Research, operated by Anchorage resident Jean Craciun, polled 300 voters from Sept. 24-25 and found Lisa Murkowski on top. Cracium paid for the poll herself. Here are the numbers (note that I use the term RINO merely to distinguish Murkowski from the official Republican nominee, Joe Miller; it's not intended to be pejorative):

-- Lisa Murkowski (RINO): 41 percent
-- Joe Miller (R): 30 percent
-- Scott McAdams (D): 19 percent

Other candidates in the U.S. Senate race include Fredrick Haase (Libertarian), Ted Gianoutsos (Ind.), Tim Carter (Ind.), and Sid Hill (Democrat write-in)

Craciun insists her numbers are more accurate than her competitors because instead of merely using computers that randomly dial residents and connect with people who are likely voters, her people look at historical voting records which differ markedly throughout the state and align the numbers and demographics of those surveyed to match district-by-district.

Ivan Moore Poll

Since this post was first published, the Anchorage Press has reported the results of a poll of 750 adults conducted by Ivan Moore from Sept. 23-27. Full poll results available HERE. The U.S. Senate poll was conducted in two parts:

(1). Miller vs. McAdams, with no mention of Murkowski by pollster, although if the respondent volunteered Murkowski, the response would be recorded:

-- Joe Miller (R): 43 percent
-- Scott McAdams (D): 28 percent
-- Lisa Murkowski (RINO): 18 percent

(2). The 82 percent (573 respondents) who didn't volunteer Murkowski's name were asked a follow-up question: “As you may have heard, Lisa Murkowski is running a write-in campaign for U.S. Senate. Knowing this, would your vote for U.S. Senate remain the same, or would you write in Lisa Murkowski?” The results:

-- Lisa Murkowski (RINO): 43 percent
-- Joe Miller (R): 36 percent
-- Scott McAdams (D): 14 percent

The Anchorage Daily News is a bit confused by this. They ask if it means that Murkowski get 18 percent or 43.5 percent of the vote? While Ivan Moore is considered the second most accurate pollster in Alaska, his own personal bias against Miller is revealed in the statement "A Judgment Day of sorts is coming for Mr. Miller and it should be mighty entertaining to watch."

Dittman Research Poll

Anchorage pollster Dave Dittman has also polled the race, and while he wouldn't release the details, he says Murkowski leads Miller by two percentage points in his poll. Dittman is customarily the most accurate pollster in Alaska.

CNN/Time/Opinion Research Poll

Also on September 29th, the results of a Opinion Research poll commissioned by CNN/Time was released. Taken between Sept. 24-28, it queried 1,528 adults, including 927 likely Alaskan voters, asking the following question: "If the election for U.S. Senate were held today and the candidates were Scott McAdams, the Democrat and Joe Miller, the Republican, who would you be more likely to vote for or would you write in the name of Lisa Murkowski, who is also running?" Full results available HERE:

-- Joe Miller (R): 38 percent
-- Lisa Murkowski (RINO): 36 percent
-- Scott McAdams: (D): 22 percent

Time also revealed that this poll showed a racial split developing. Murkowski appears to be polling best amongst non-Whites, 49 percent to Miller's 24 percent, which is believed to reflect the strength of her support amongst Alaska Natives, while Miller is winning the White vote, 42 percent to 33 percent. This despite the fact that unlike Murkowski, Joe Miller hasn't introduced race into the campaign. Additional discussion of this poll available on CNN Political Ticker, the Anchorage Daily News, and The Hill and Time.

Rasmussen Poll

Of course, all these polls were preceded by a Rasmussen poll which told a different story and which I considered questionable. Their survey of 500 Likely Voters in Alaska was conducted on September 19, 2010.

-- Joe Miller (R): 42 percent
-- Lisa Murkowski (RINO): 27 percent
-- Scott McAdams (D): 25 percent

Perhaps it was on the level, because since that time, the Miller campaign has been buffeted by two unfavorable revelations; first, that Joe Miller had obtained low-income hunting and fishing licenses in 1994 despite having purchased a house assessed at $88,700 at the time, and second, that Joe Miller is late in filing required financial disclosure forms with the Senate Records Office and Ethics Committee. In addition, the Murkowski campaign was likely buttressed by the fact that Senate Republicans ultimately chose not to strip her of her seniority. The combined effect of all these actions may be eroding support for Joe Miller.

However, Craciun's polling results have not always been accurate, either. So I believe the results by Dave Dittman and CNN/Time and more representative, and at this point, it's a tossup between Lisa Murkowski and Joe Miller.


  1. Thank you for this informative post! I was one of the voters polled; I’ve looked at the various poll results you linked to in an attempt to figure out which poll I participated in but am unable to verify which one it was. I’d like to share my experience with the poll. I spent about 45 minutes talking to a person on my land line phone. There were many questions, more than revealed on the poll result links. One of the first questions asked what my race was, to which I replied, “I am a mutt” – which truly I am – I have Japanese, German, some Indian and who knows what else. I believe most of us are and we need to stop classifying ourselves so narrowly in our time of prevalent mixed relations. I’m not sure what the pollster marked. I had a couple questions which asked to pick between the Republican and Democrat (no candidate names given). To these I said I would vote, “Constitutional Conservative” I also said to the effect of, “These questions skew results as voters have more than just those two choices on the ballot, and don’t you have a blank to fill in my answer? Even the ballot has a blank write in area.” They also asked me how many handguns and rifles are in my household to which I replied, “None of your business how is this relevant to the survey?” They also asked if I specifically was Hispanic, to which I reiterated the, “None of your business” answer. I am no fool, I took statistics in college, I can see how these “unbiased” surveys try to skew results to favor the sponsors’ agenda. Statistics is not hard to understand and I think we need to take these polls with a grain of salt and not let their results affect how we want to vote because we think, due to the results, we should change our stance. I believe that is part of the pollsters agenda. Just as an aside, since the poll began with a statement that it was recorded for quality purposes I made sure to say, “I am disappointed in Lisa Murkowski. I am one who voted for her in the primaries but by running her write in campaign she is not listening to the voice of the people and I will not vote for her in the general election.” If others were polled I’d love to hear their experience.

  2. Thanks for your informative and interesting input, Rebekah. One of the reasons these pollsters ask about race is because of the doctrine of multiculturalism. Diversity and multiculturalism divides us rather than unifies us. Many people make a healthy living by promoting such division, to include the ADL, the SPLC, and the NAACP. Many sensitivity trainers make thousands of dollars per year proposing "solutions" to cope with the challenges of diversity. Diversity is big bucks to a lot of people.

  3. BOTH murkowski AND miller ARE jew/mason VOTING FOR them IS THE SAME AS CUTTING YOUR OWN THROAT................