With the news that Tom Emmer has pulled to a tiny, inside-the-margin-of-error lead in the latest Rasmussen poll, I’ve joked that it’s about time for a Star/Tribune “Minnesota Poll” showing Mark Dayton leading by an improbably huge margin.
And sure enough, here it is. It shows Dayton leading Emmer 39-30, with Horner eating up 18 points.
The key, as with all Minnesota Polls, is in the sampling; the methodology seems to be “poll Democrats until we get the result that’ll make Democrats feel better and want to come to the polls”. And if you look at the “methodology” page (and the Strib has got to be assuming you won’t), it’s right there:
The self-identified party affiliation of the random sample is: 35 percent Democrat, 28 percent Republican and 28 percent independent. The remaining 9 percent said they were members of another party, no party or declined to answer.
So the Strib claims to believe, in this electoral season, that there will be five Democrat voters for every four GOP voters, and (apparently) that independents will break the same way they did two years ago.
The Washington post carried this oddly-constructed PDF showing the Minnesota Poll’s statistical history over the past five decades or so. For MN Polls conducted since Rob Davies took over the Poll (I’ll add emphasis)…:
- The final GOP poll number was on average 5.20 percent points under the actual GOP result in the election. -5.20 percentage points is outside the margin of error in the Minnesota Poll.
- The final DFL poll number was on average 2.06 percent points under the actual DFL result in the election. -2.06 percentage points is inside the margin of error in the Minnesota Poll.
- Since 1998, the Minnesota Poll has underestimated the GOP result in elections by an average of 7.26 percent but underestimated the DFL result by only .054 percent.
I’ll remind you that if the Minnesota poll were accurate, we’d be referring to Governor Humphrey (the poll showed Skip with a strong lead over Coleman, with Ventura well out of the running), Senator Mondale (who had a five point lead in the MN Poll on the eve of the ’02 election), Governor Moe (to whom the MNPoll gave a slim lead, while significantly overpolling IP candidate Tim Penny in ’02), Governor Hatch (yep, slated to win in ’06)…
…indeed, the only year they’ve been genuinely accurate was ’08. The year of the great Democrat/DFL blowout. That’s because, in effect, the Minnesota Poll always predicts DFL blowouts. They were finally right in ’08.
This poll, like all Minnesota Polls, has only one purpose; to help revive the DFL’s flagging spirits. Mark Dayton has run a perfunctory, frankly terrible campaign, notable for the nastiness of the “third-party” attack ad campaign largely paid for by Dayton, his ex-wife and family. His support is slipping in reputable polls.
And that’s why we have the Minnesota Poll.
The serial dishonesty and, let’s be honest, in-the-bag-ness of the Minnesota Poll was the straw the broke the camel’s back when I last unsubscribed to the Strib back in 2004.
If you are a Minnesota conservative who is hoping for a sane governor next year, this is not a reason to jump off the ledge – but it is a reason to remember this is a tough campaign. Emmer’s doing well, but he still faces a full-court press – the bought-off, in-the-bag media, a ruthless pack of well-heeled institutional hyenas, and the DFL machine. If you can peel off a few bucks, or spend some time calling or door-knocking, it’s needed.
Bring on the real polling – in November!
For many clients, PDS provides a cleaned, unweighted dataset. But for the client who does not want to weigh the data themselves, PDS can provide weighting services to take into account known probabilities involving the sampling process or known variations in the non-response among groups.
In other words, these non-response groups [the large percentage of “no-response/different party” respondents – a high nine percent in the MN Poll – Ed] have been manipulated. By manipulated, I mean, they made it up. In their vast experience as pollsters, they have discovered that non-response means, this, whatever this may be. Since they are providers to Pew Research, a group with notoriously liberal findings, we can conclude that non-response to Princeton means they’re probably liberals.
So, to recap. We have a poll with a huge margin of error for the population size, a presumption of liberal bias when weighing cell phone sampling, a large number of non-independent, non-responsive people, who are presumed liberal by a liberal polling firm who works exclusively with liberal groups.
More as the story warrants.