Update: Cell-phone-excluding polls — still hurting Dems, even if GOP still leads

Update: FiveThirtyEight.com’s Nate Silver, who studies and aggregates polls, has some thoughts here.

With SurveyUSA/KSTP set to release a new Minnesota gubernatorial poll today that excludes voters without landlines, I thought I’d pass along the Pew Research Center’s fresh update on the matter.

Good news for Dems: Landline-only polls are biased against you by several percentage points.

Good news for Republicans: You’re still going to stomp the Democrats.

Pew — which shares a pollster (Princeton Survey Research Associates) with the Star Tribune — surveyed 2,816 registered voters, a very big sample. When just landliners were polled, 46 percent favored Republicans, and 45 percent favored Democrats. Add in cell-phone-only voters (CPOs) and the Dems take the lead, 47-44.

Pew/PRSA also screened for likely voters, and here’s where the numbers get good for the GOP. The landliners go 53-41 for John Boehner’s crew, but when CPOs are added, it’s 50-43 GOP.

Important notes:

1. This is a generic Congressional ballot nationwide — it’s not specific to Minnesota.

2. This isn’t taking anybody’s temperature on a governor’s race.

3. A reminder: The Star Tribune does include CPOs; SurveyUSA, Rasmussen and MPR/Humphrey Institute polls don’t. Rasmussen released a poll last week showing DFLer Mark Dayton leading by 2 percentage points. Both the Strib and MPR/HHH have September polls showing Dayton with a 9- and 11-point lead, respectively. SurveyUSA’s last poll, on Sept. 12, showed Dayton up by two. (All results here.)

4. Pew’s number has bounced around a bit: a July-August survey showed no difference when CPOs were added in. But three of four surveys, including the most recent one, have.

5. Pew, curiously, doesn’t list a margin of sampling error for its latest survey.

You can also learn about all our free newsletter options.

Comments (1)

  1. Submitted by Greg Kapphahn on 10/15/2010 - 09:33 am.

    With the economic turmoil we’re currently experiencing, the possibility that cell phones offer to be instantly available to employers who might be trying to reach you, and the doubling of expense necessary to maintain lines in both systems, I strongly suspect that the surveys on which the ideas regarding WHO the CPO folks are and how MANY of them there are likely to be considerably outdated.

    If current poll results are based on outdated assumptions regarding CPOs it’s quite likely that, in reflecting actual and current reality the results on the ground on November 2nd, will turn out enough more in the Democrat’s favor than currently being predicted as to appear to be a “stunning upset” of the Republicans (if of course the MSM, being owned by those same Republicans, can gain the permission to, and has the courage to report it that way, of course).

Leave a Reply