MN: Pawlenty (R) leads Hatch (DFL) by 2

[Tender Branson]

New Poll: Minnesota Governor

Summary:

Pawlenty (R): 43%
Hatch (DFL): 41%
Peter Hutchinson (I): 5%
Ken Pentel (G): 2%

Poll Source

[Sam Spade]

FYI, this is the infamous Star-Tribune poll which usually overstates Democratic support by anywhere between 5-10 points depending on year.  This makes me wonder even more about Rasmussen polling in the state, but I'll wait to make any changes until M-D comes out with one.

Link to poll's internals are here.

http://www.startribune.com/587/story/554728.html

[I spent the winter writing songs about getting better]

I haven't seen the usual Dem bias in Star Trib polls since 2004, so it's likely they change their polling methods. Whether or not it's any good now remains to be seen. I'll admit these numbers appear more accurate than the Rasmussen ones, but I am still quite happy that Pawlenty is well under 50%.

Of course, what makes me doubt it is the third parties getting 7%, considering how anti-third party the current poliitical climate is.

[Sam Spade]

I haven't seen the usual Dem bias in Star Trib polls since 2004, so it's likely they change their polling methods. Whether or not it's any good now remains to be seen. I'll admit these numbers appear more accurate than the Rasmussen ones, but I am still quite happy that Pawlenty is well under 50%.

Of course, what makes me doubt it is the third parties getting 7%, considering how anti-third party the current poliitical climate is.

BRTD, they've had the same polling sample since the mid-1970s, the same sampling of DFL vs. Rep., even though the state has changed a lot since then, as I'm sure you'd admit. 

They've been totally wrong on every Minnesota election since 1994 that I can remember to the Democrats by the margin I listed above (and maybe slightly before that), so why would they change it now, if they haven't changed it before.

And we haven't had any election to measure them up against since 2004, so how would we notice that their sampling has changed?  They were 5 points too far to the left of Kerry's result in 2004 in Minnesota, even with an above-average turnout for the DFL (even for them in Minnesota).  They were even worse in 2002 and 2000, proving my point.  Let's see them actually get at least one general election within MOE (more like two for me) before we can say they've "changed their model".

Look, I normally say that all of this "media bias" in polls is total BS, especially when touted on the "right" or "left" (more often it's the former), but when you have records like this, it's hard to ignore the facts and the history in making a judgment.

The one thing where I'll say that you might be quite right on is the power of the third-parties in this race.  However, unlike some other states (Texas comes to mind), Minnesota has a fairly strong tradition of third-parties competing and pulling strong percentages (for third parties), so I'm going to simply remain cautiously agnostic on that issue for the time being.

[I spent the winter writing songs about getting better]

I haven't seen the usual Dem bias in Star Trib polls since 2004, so it's likely they change their polling methods. Whether or not it's any good now remains to be seen. I'll admit these numbers appear more accurate than the Rasmussen ones, but I am still quite happy that Pawlenty is well under 50%.

Of course, what makes me doubt it is the third parties getting 7%, considering how anti-third party the current poliitical climate is.

BRTD, they've had the same polling sample since the mid-1970s, the same sampling of DFL vs. Rep., even though the state has changed a lot since then, as I'm sure you'd admit. 

They've been totally wrong on every Minnesota election since 1994 that I can remember to the Democrats by the margin I listed above (and maybe slightly before that), so why would they change it now, if they haven't changed it before.

And we haven't had any election to measure them up against since 2004, so how would we notice that their sampling has changed?  They were 5 points too far to the left of Kerry's result in 2004 in Minnesota, even with an above-average turnout for the DFL (even for them in Minnesota).  They were even worse in 2002 and 2000, proving my point.  Let's see them actually get at least one general election within MOE (more like two for me) before we can say they've "changed their model".

Look, I normally say that all of this "media bias" in polls is total BS, especially when touted on the "right" or "left" (more often it's the former), but when you have records like this, it's hard to ignore the facts and the history in making a judgment.

I have no clue what would provoke such a change, but one would have to be pretty naive to argue there's a 5 point slant in most of the post-2004 polls. That'd give Republicans a 9-point identification advantage for example based on one poll and it'd mean Rasmussen is 12 points off here.

We can't see for sure until any actual election happens, but clearly something is up here.

There was also one StarTrib poll that had Kerry+5, which was more accurate than any poll with Bush winning Minnesota.

[Sam Spade]

I haven't seen the usual Dem bias in Star Trib polls since 2004, so it's likely they change their polling methods. Whether or not it's any good now remains to be seen. I'll admit these numbers appear more accurate than the Rasmussen ones, but I am still quite happy that Pawlenty is well under 50%.

Of course, what makes me doubt it is the third parties getting 7%, considering how anti-third party the current poliitical climate is.

BRTD, they've had the same polling sample since the mid-1970s, the same sampling of DFL vs. Rep., even though the state has changed a lot since then, as I'm sure you'd admit. 

They've been totally wrong on every Minnesota election since 1994 that I can remember to the Democrats by the margin I listed above (and maybe slightly before that), so why would they change it now, if they haven't changed it before.

And we haven't had any election to measure them up against since 2004, so how would we notice that their sampling has changed?  They were 5 points too far to the left of Kerry's result in 2004 in Minnesota, even with an above-average turnout for the DFL (even for them in Minnesota).  They were even worse in 2002 and 2000, proving my point.  Let's see them actually get at least one general election within MOE (more like two for me) before we can say they've "changed their model".

Look, I normally say that all of this "media bias" in polls is total BS, especially when touted on the "right" or "left" (more often it's the former), but when you have records like this, it's hard to ignore the facts and the history in making a judgment.

I have no clue what would provoke such a change, but one would have to be pretty naive to argue there's a 5 point slant in most of the post-2004 polls. That'd give Republicans a 9-point identification advantage for example based on one poll and it'd mean Rasmussen is 12 points off here.

We can't see for sure until any actual election happens, but clearly something is up here.

There was also one StarTrib poll that had Kerry+5, which was more accurate than any poll with Bush winning Minnesota.

There were 4 StarTribune polls in 2004, giving these respective results:

4/2/2004:  Kerry 50%, Bush 38%, Kerry +12%
9/13/2004:  Kerry 50%, Bush 41%, Kerry +9%
10/11/2004:  Kerry 48%, Bush 43%, Kerry +5%
10/29/2004:  Kerry 49%, Bush 41%, Kerry +8%.

Final result was Kerry +3.48%.  Ironically, Zogby had the best result in the state per MOE @ Kerry +6%.  Not a good year for Minnesota polling.

Interestingly, out of the 28 public polls done in Minnesota that were not partisan polls (I'm including Strategic Vision in this) through the year of 2004, 8 public polls showed a lead greater than Kerry +6% or above, 3 were Star Tribune, 2 were Rasmussen summer polls (+7%,+9%), 1 was a uni poll at the end of the campaign(+7%), 1 was Gallup's last poll (8%) and 1 was Zogby's last poll (+6%)

[AuH2O]

The StarTrib is a garbage newspaper. Not wise to trust their polling, even if a particular poll is not heavily biased it still is probably a poor overall sample.