Minnesota Public Radio/Humphrey Institute (8/25-29, likely voters, 5/13-16 in parens):
Mark Dayton (D): 34 (35)
Tom Emmer (R): 34 (31)
Tom Horner (IP): 13 (9)
Undecided/Other: 19
(MoE: ±3.6%)
One red flag from this poll is its sample composition — it pegs the Minnesota midterm electorate at 46% Republican, 41% Democratic, and 13% independent. An early August poll by SurveyUSA had a partisan composition of 36D-32R-28I, and showed Dayton leading by 46-32. Indeed, Dayton has been leading in the last five polls of the race, the last three of which by margins varying between 9 and 14 points — and that includes Rasmussen. So I’m not convinced by this poll’s top lines, especially considering the pounding that Emmer’s been taking on the airwaves in the past two weeks.
That’s not to say that Dayton is incapable of fumbling this one — it’s just that most of the evidence suggests that he has yet to do so.
That is the key question. It is plausible, but not likely that people would ID themselves with such a ratio this cycle. I still rank this lean Democratic until I see more polling.
FWIW here is the Party breakdown from the exit polls of the last 3 elections.
2008: D-40 R-36 I-25
2006: D-40 R-36 I-24
2004: D-38 R-35 I=27
In a tough cycle I could buy an even Party breakdown but I am not buying a 6 point Rep lead. Plus the Ind sample is about half of what it should be.
I strongly doubt it looks like this. If it does really does, than Dayton is in trouble. The only reason I could think of this race being tied now is the ads that Republican group is running against Dayton, and I doubt they have played often enough or would be effective enough yet to move this race so much in a short period of time.
Minnesota is just a hard state to poll. SurveyUSA was noticeably off in their polls for Minnesota, and this sample seems very much GOP-skewed also.
Are there any pollsters who have a good track record polling Minnesotans?