The Star Tribune reports in this morning's edition that its Minnesota Poll shows Amy Klobuchar with a 19-point lead in the Senate race against current Republican Congressman Mark Kennedy. The Strib describes this as "surprising", but for those of us with experience in the atrocious polling done by the Strib, it comes as no surprise whatsoever:

DFL Senate candidate Amy Klobuchar has opened up a strong early lead over GOP rival Mark Kennedy in a Minnesota Poll that shows Klobuchar with 50 percent of likely voters' support, compared with 31 percent for Kennedy.

Much can change between now and November. But in what had widely been considered a close race, Klobuchar in midsummer has more support than Kennedy in nearly every demographic category: men, women, liberals, independents, lower- and upper-income Minnesotans, seniors, urban dwellers, suburbanites and outstaters.

Kennedy is most popular with younger Minnesotans -- he leads Klobuchar 63 to 16 percent among those under age 25 -- and with Republicans, but he is behind in nearly every other category the July 6-11 poll measured.

It's helpful at this juncture to recall the MinnPoll's history in predicting elections. Over the last twenty years, the poll has miscalculated Republican support every election cycle, getting increasingly worse as time goes on while overestimating Democratic support. When Al Gore narrowly won the state in 2000 (by 2.4%), the MinnPoll estimated a ten-point gap in its final result days before the election, claiming that Bush would only take 37%. He wound up with 45.5%. In the 2002 Senate race, the MinnPoll predicted Walter Mondale would beat Norm Coleman by five points; he lost by two, as MinnPoll underestimated Coleman's support by over eight points.

Their incompetence doesn't just apply to two-way races. In 1998, they predicted that Skip Humphrey would win the governor's race and draw 35% of the vote, with Jesse Ventura coming in last place with 27%. Those positions were reversed on Election Day, with Humphrey only able to attract 28%. They did manage to predict that Tim Pawlenty would win the 2002 three-way race to replace Jesse, but they underestimated his support by nine percentage points.

If you get the notion that the Strib skews its polls against the GOP, you'd be correct. In the last twenty years of elections for President, Senate, and Governor, the Strib has underestimated Republican support by an average of over seven points. The only race in which they did not underestimate Republican support was in 1990, when they picked Rudy Boschwitz to beat Paul Wellstone for the Senate.

It's more helpful to look at other polling and to skip the worthless MinnPoll. Rasmussen, for instance, just finished polling for this race (June 26th), and had Klobuchar leading -- but only 47-44, within the margin of error. Zogby had it 49-41 Klobuchar on June 21. The MinnPoll has a gap more than double that of Zogby -- who has his own accuracy problems -- a "surprising" result, indeed.

The Star Tribune showcases its usual hackery in these results. No one who lives here is fooled by the MinnPoll any longer.

Trackback Pings

TrackBack URL for this entry is