Another new poll (PDF) on Minnesota’s U.S. Senate race, this one mostly good news for Norm Coleman.
A SurveyUSA poll taken last week, after the flap about Al Franken’s Playboy articles and after the DFL convention endorsed Franken, finds Coleman leading Franken by 52 to 40 percent — the biggest lead Coleman has held in any recent poll. The poll aired Monday night by KSTP-TV, which sponsored the poll.
If attorney Mike Ciresi were the DFL candidate, SurveyUSA finds Coleman leading by 50 to 40 — probably not enough of a difference for him to use as justification for jumping into a primary against Franken.
Then there’s the new Jesse factor. When SurveyUSA offered registered voters a choice of Coleman-Franken-Jesse Ventura, the result was 41-31-23, respectively. (A recent Rasmussen poll found that a Ventura candidacy widened Coleman’s lead over Franken. This one shows a slight narrowing of the lead.)
SurveyUSA asked about a Coleman-Ciresi-Ventura matchup and got 41-28-26.
They tried some matchups in which attorney Dean Barkley was the Independence Party candidate instead of Ventura. Barkley didn’t do anywhere near as well Ventura, and has recently taken a new job that he says rules out any chance of his running for office this year. But Coleman continued to win, by healthy margins, against any combination of DFLers and Independence-ites.
The Rasmussen poll, taken at about the same time, had shown Coleman with only a 48 to 45 percent lead over Franken in a two-way race, and leading by 39-32-24 in a Coleman, Franken, Ventura race.
Franken supporters were more heartened by the Rasmussen Poll, which seemed to show that Franken was still close to Coleman after the Playboy flap.
Actually, both the Rasmussen and SurveyUSA polls show little change between May and June. In May, Rasmussen had shown Coleman leading Franken by 2 percentage points in a two-way race and SurveyUSA had shown him ahead by 10. There’s no reliable way to assess whether Coleman is as far ahead as SurveyUSA finds or whether Franken is as close as Rasmussen reports.
Both polling companies rely on robotic phone interviews. Rasmussen polled 500 and had a margin of error of plus or minus 4.5 percentage points. SurveyUSA polled 609 and had a margin of error of plus or minus 4 points.
The Star Tribune did the most recent public poll using the old-fashioned (and deemed by some to be more reliable) human-to-human interviews, also with registered voters, in mid-May and found that Coleman led Franken 51 to 44 percent and leading Ciresi 51 to 43.
What think?
Until the robopolls poll cell phones they won’t mean squat.
I suspect that ALL the polls have this problem to some extent. Younger voters in particular do not use land lines, and are thus transparent to most polling. Unless a poll targets a specific sample of individuals and does whatever is necessary to reach them it’s not going to produce a representative sample.
As far as the race is concerned:
It’s early, Coleman is the incumbent, and Franken has just begun to spend his money.
If I were Coleman I’d be worried.
And how about Franken v. Jack Nelson-Pallmeyer — who would probably have won endorsement on the second ballot at the DFL convention?