AK-Sen: Murkowski Drawing Plenty of Votes, But…

Opinion Research for CNN/Time (9/24-28, likely voters, no trend lines):

Scott McAdams (D): 22

Joe Miller (R): 38

Lisa Murkowski (I): 36

(MoE: ±3%)

Craciun Research for the Alaska Native Tribal Health Consortium (9/24-25, likely voters, no trend lines):

Scott McAdams (D): 19

Joe Miller (R): 30

Lisa Murkowski (I): 41

(MoE: ±5.7%)

Nate Silver, however, hits on some problems with the methodologies of both polls:

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.

Indeed, and as Nate later tweeted, perhaps the best approach that pollsters should take would be to ask voters if they’re voting for McAdams, Miller, or a write-in. That would then be followed by a prompt to ask who they’re writing in (rather than offering Murkowski as a choice). However, this methodology would be difficult for a robo-pollster to accommodate (though it shouldn’t be too hard for Opinion Research or Craciun, who use live interviewers). After all, the way these pollsters are framing the choice doesn’t reflect the realities of the ballot.

It’s also worth noting that Scott McAdams only went up on the air last night for the first time, so add that as another reason why these polls may represent something of a high-water mark for Murkowski that she is unlikely to obtain in November.

Meanwhile, CNN/Time also polled the gubernatorial race…

Ethan Berkowitz (D): 38

Sean Parnell (R-inc): 57

(MoE: ±3%)

Interestingly, Berkowitz’s margin is actually slightly worse among registered voters (at 58-36). The guy is tenacious, but this looks pretty tough.

33 thoughts on “AK-Sen: Murkowski Drawing Plenty of Votes, But…”

  1. He’d probably have a better chance there than now (although to be fair, running for governor was still probably smarter than running for senator when he announced given that Murkowski was pretty much seen as invulnerable at the time).

  2. Based on http://www.swingstateproject.c

    Assuming the apples to oranges comparison is valid

    Murkowski is taking more from McAdams than Miller — roughly 2x as much.

    IF Murkowski’s numbers begin to fade, will her support go more towards McAdams?

    If her support is distributed on a 2/1 basis towards McAdams, that would put him at least in striking distance.

    And what of the Libertarian? Don’t they typically get a significant share of the vote in statewide Alaska elections? Or is all of that support magically going to Miller this time around?

  3. Q1. If the Senate election were today, would you vote for….Democrat Scott McAdams? Republican Joe Miller? Libertarian David Haase? Or, would you write in someone else’s name?

    Q2. (If Write-In) Whose name would you write in for US Senator…..Sarah Palin? Tony Knowles? Lisa Murkowski? Mike Gravel? Or, someone else?

    This forces Murkowski voters to pick write-in instead of getting her as a ballot option, but also forces them to explicitly pick Murkowski from all the names you could write in. Palin, Knowles, and Gravel, are just placeholders–you could throw in the name of any Alaskans there. But I think this would give more accurate results than, “Are you going to write in Lisa Murkowski?”

  4. As Nate implies, these polls are next to worthless… BUT in the absence of more reliable information, could create a self-fulfilling prophecy that McAdams can’t win, while Murkowski can.  DSCC needs to get on the horn to Ivan Moore, stat.

  5. Maybe Murkowski can win as a write-in? Given, it’s not probable… But possible, especially if she can deny Miller any help from the GOP establishment.

    But more likely, she may just end up going the way of Charlie Crist (who’s actually on the ballot!) and fade away as more voters pick McAdams or Miller.

    Now I can see the reasoning behind McAdams’ first ad.

Comments are closed.