Rick Nolan faces reporters (minnesota.publicradio.org).
by Tony Petrangelo
Oct 23, 2012, 7:00 AM

Two new polls show Michele Bachmann and Rick Nolan ahead

New polls of Minnesota’s sixth and eighth congressional district races were released by the StarTribune Monday morning. For these survey’s they are utilizing the services of Pulse Opinion Research, a subsidiary of well known pollster Rasmussen.

Pulse Opinion Research (10/22, no trend lines):

Chip Cravaack (R-inc) 43
Rick Nolan (D) 50
Undecided 7
(MoE: ±3%)

This is the largest lead we’ve seen for Rick Nolan yet and he’s hitting 50% to boot. The previous high for the former congressman was 48% in a poll conducted by Victoria Research for the Nolan campaign.

And while Rick Nolan’s level of support is the highest we’ve seen in any poll so far, Chip Cravaack’s level of support is within the range that has already been defined by the previous five polls. In those polls his lowest top line number was 42%, while his highest was 45%. The 43% he’s getting in this poll falls right in line with those numbers.

Chip Cravaack is consistently polling at or below 45%. For an incumbent congress person this is a big red flag and a sign that he’s in a lot of trouble.

Pulse Opinion Research (10/22, no trend lines):

Michele Bachmann (R-inc) 51
Jim Graves (D) 45
Undecided 4
(MoE: ±3%)

This is now the second public poll of this race and they have both had Michele Bachmann at or above 50%. In contrast to her neighbor to the north, this is a somewhat comfortable position for an incumbent to be in.

While this race showed signs of tightening at one point, those signs ceased once Michele Bachmann’s multi-million dollar ad blitz began in full force. With only two weeks to go, this race is now looking like an extremely tough climb for the challenger Jim Graves.

Since I’ve been doing race ratings for the Minnesota Legislature, I’ll take a stab at ratings for these two races.

MN-6: Likely GOP
MN-8: Lean DFL

And these are the only two seats that really need ratings, as the rest can comfortably be given Safe labels for their respective incumbents. A case could be made that MN-2 should be Likely GOP (I’m not going to make it), but other than that, the rest are off the board at this point.

Because this is a StarTribune poll, there are, here comes my shocked face, criticisms of it from pretty much the entire Minnesota Right Wing Blogosphere. And Minnesota Republican operatives to boot.

Yes, utterly bogus!

If you go and read David Brauer’s piece (and I suggest that you do) that is cited as evidence of this polls total bogusness, you will find that Brauer does not actually say the poll is bogus. He simply goes over many of the methodological problems with the poll. And there certainly are many.

On the surface, you’d think the Strib poll, conducted by Texas-based Pulse Opinion Research, might be better. POR used a relatively large 1,000-person sample — nearly twice as big as SUSA’s — reducing mathematical sampling error to plus or minus 3 percentage points, compared to SUSA’s 4.2.

However, there are some methodological red flags that give this polling junkie pause.

The methodological problems outlined include only calling on one night, as opposed to over the course of multiple nights. This means there is no attempt to recontact those who don’t respond, which leads to an even greater non-response bias than normal.

Additionally, as a robo-pollster, POR like it’s big brother Rasmussen, does not call cell phones. As time goes on that becomes more and more of an issue.

And beyond the methodological problems, as Brauer points out, POR generally sucks as a pollster.

But none of this makes the poll bogus, it just makes the poll less reliable then the other more methodologically sound polls that have been done so far.

SurveyUSA and Public Policy Polling, the two organizations who have had the most accurate results in Minnesota over the last few cycles, both have their own methodological problems.

SurveyUSA for instance, while they call cell phones, they use live operators to do it. So they are conducting one part of their poll with robots and the other part with actual people.

PPP who is a robo-pollster like SurveyUSA and POR, takes the same approach as POR and doesn’t even bother to call cell phones, instead opting to add additional weight to the younger age demographic to make up for it.

The simple reality of the modern media marketplace is that no one wants to pay top dollar for polls, this is why SurveyUSA, Public Policy Polling and Rasmussen are so successful, because they offer a low cost alternative.

And while you can point to perceived flaws in pretty much every poll, you shouldn’t be looking at individual polls anyway. You should be looking at the average of polls.

And here’s the average of the polls (all of them, public and internal).

MN-6: Michele Bachmann 49, Jim Graves 44
MN-8: Chip Cravaack 44, Rick Nolan 46

Thanks for your feedback. If we like what you have to say, it may appear in a future post of reader reactions.