Nonprofit, nonpartisan journalism. Supported by readers.


DFLers ‘up big’ in Minnesota, NBC/Marist poll finds

Sen. Tina Smith
MinnPost photo by Lorie Shaull
NBC finds appointed Sen. Tina Smith ahead of Republican challenger Karin Housley 54-38 percent.

NBC and its polling partner Marist are out with a Minnesota-only poll that is very good for Democrats in general (yes, they are so clueless that they don’t seem to know that our Dems are DFLers). The results are out of line with most other recent polling, and all in the same direction, so take them with a grain of salt (although it could be that the Dems are pulling further ahead in good ol’ Minnesota).

Under the headline, “Democrats Up Big in Minnesota as Trump Approval Sags,” the poll also finds DFL candidates in most of the major state races have big and widening leads.

According to NBC/Marist poll:

DFL gubernatorial nominee Tim Walz leads Republican Jeff Johnson by 55 percent to 38 percent among likely voters. Other polls have generally shown Walz ahead, but not by that much. For example, the Star Tribune/MPR Minnesota Poll, published Sept. 21, showed Walz ahead 46-36.

DFL U.S. Sen. Amy Klobuchar leads Republican challenger Jim Newberger by a whopping 63-33. No one has ever shown that to be a close race.

But in our other Senate race, NBC finds appointed Sen. Tina Smith ahead of Republican challenger Karin Housley 54-38 percent. The Strib/MPR poll had it 44-37, and while it has seldom been treated as a toss-up, some of the national pundits have believed it could be quite close. And it could be, if NBC/Marist’s poll is wrong. If it’s right, a 16-percentage-point lead is unlikely to disappear in the four weeks remaining till Election Day.

In races for the U.S. House, Minnesota has been blessed (or cursed, if you don’t like TV ads filled with scary music and ridiculous claims) with four competitive races. NBC/Marist didn’t poll those four races individually but instead asked their statewide sample of likely voters whether they planned to vote for the Democratic or Republican candidate for the U.S. House. That one came out 54-40 in favor of the generic Democrat.

Again, without polling specific House contests, the poll found a general preference for DFLers across all age groups, all income groups and all regions except southern Minnesota. That exception might be good news for Republican Jim Hagedorn, who has been locked in a tight race with DFLer Dan Feehan for the southern First District seat that Walz is vacating.

But if DFLers win the other three close congressional races, we would end up with a House delegation of six DFLers and two Republicans, compared to the current 5-3 split. (And, of course, I don’t need to remind you that polls are not always successful at predicting outcomes and there are still almost four weeks for things to change.)

Now then, as to President Trump — who will not be on the ballot, but who shocked the world two years ago by almost carrying our dear old state, which has gone for the Democrat in every presidential election since 1972 — 56 percent of likely voters told NBC/Marist that they disapproved of the job he is doing as president (44 percent of them said they disapproved strongly). That compares to 38 percent who approved of his work so far from the Oval Office, just 26 percent strongly.

I’ll say again that polls should not be taken as reliable predictors of outcomes, as 2016 showed.

Another word of caution: Reporting on the same poll this morning in his daily email, political junkie Blois Olson notes, “This same poll reported a poll on July 27, 2018 that showed Jeff Johnson trailing Tim Pawlenty by 19 points. That poll turned out to be off by 28 points with Johnson winning the election by 9. … The July poll also had Lori Swanson winning the DFL primary.”

NBC’s full writeup of the poll can viewed here.

The question wording and detailed breakdowns of their findings are here.

Comments (24)

  1. Submitted by RB Holbrook on 10/11/2018 - 09:15 am.

    Counting down until the inevitable “Who can trust the polls, remember how they predicted Hilary would win?” comments in 10, 9, 8 . . .

  2. Submitted by Hiram Foster on 10/11/2018 - 09:23 am.

    What’s the track record of the poll?

  3. Submitted by Pat Terry on 10/11/2018 - 09:24 am.

    Wow, disappointing to see even Eric Black misstate what happened with polls in 2016. The final polls got the result (a 2-3 point Clinton advantage) exactly right.

    • Submitted by John Evans on 10/11/2018 - 09:43 am.

      Or, as Charlie Pierce said, “I totally got fooled. I thought he’d lose by about three and a half million votes!”

  4. Submitted by Curtis Senker on 10/11/2018 - 09:33 am.

    Isn’t this the same poll that had Lori Swanson winning the DFL primary and Tim Pawlenty topping Jeff Johnson?

    Why yes, yes it is. Carry on.

    • Submitted by Mark Iezek on 10/11/2018 - 10:01 am.

      I rarely agree with you, but this polling group had Pawlenty winning by 15 points when he ended up losing by 9. They had Erin Murphy in third with 13 percent of the vote when she ended up second with 32 percent. They were so bad polling the primaries that they quite literally would been closer if they had rolled dice and reported those results.

      Not all polls are created equal. I would wait for some poll with a better track record than this one.

      • Submitted by Pat Terry on 10/11/2018 - 10:22 am.

        Polls are only a snapshot in time, and the last Marist poll was a month before the election. That last month saw Swanson embroiled in scandal and Murphy – who had almost zero name recognition – finally get on TV. The poll also had a large margin of error and many undecideds.

        Critics of 2016 polling point to big leads by Clinton. But the final polling, which occurred after the Comey letter, etc., was very accurate.

    • Submitted by Frank Phelan on 10/11/2018 - 10:06 am.

      Yup, you can’t trust the polls. Except the ones the ones for the North Dakota Senate race that you touted yesterday. Those polls are spot on.

      All rightie then…

      • Submitted by Curtis Senker on 10/11/2018 - 02:40 pm.

        That wasn’t a poll with a history of being laughably wrong.

        But you’re probably right, Heitkamp is golden.

        • Submitted by Frank Phelan on 10/12/2018 - 02:07 pm.

          Mr. Senker, with all due respect, where did I speculate that Heitkamp is “golden”?

          I merely pointed out that you often say that 2016 proved polls are not reliable. Yet for the 2018 ND Senate race, you find polls favoring the GOP candidate credible. And for the NBC-Marist poll, favoring Democratic candidates, you are back to questioning polls. Why the seeming disparity in your comments?

          For what it’s worth, fivethirtyeight rates Marist and “A”.

          • Submitted by Curtis Senker on 10/15/2018 - 07:36 am.

            “I merely pointed out that you often say that 2016 proved polls are not reliable.”

            ? When have I ever said that? I normally don’t comment on polls at all, but this one was very recently 100% wrong.

    • Submitted by William Mattessich on 10/12/2018 - 09:20 am.

      Polling for primaries is very different from polling in the general election because turnout looks different and you can’t differentiate by party ID. This poll is unlikely to miss as much as the primary polls did.

  5. Submitted by Hiram Foster on 10/11/2018 - 10:26 am.

    I tend to be much more focused on polling than the results. As pollsters themselves say, polls aren’t intended to be predictive. One of the things that means is that they aren’t of much practical use.,

  6. Submitted by Dave Carlson on 10/11/2018 - 11:25 am.

    Could these polls actually be detrimental to DFL candidates as voters may think their vote isn’t as important or needed?
    And will CD 4 and CD 5 voters tend to stay home (again) because their congressional candidates are pretty assured of victory? If they do, I suspect the statewide races will be much closer than the polls indicate and that is worrisome.

  7. Submitted by Mike Cole on 10/11/2018 - 12:19 pm.

    Why would you even print this if the primary poll was off by that much. i know because you can put up the silly little headline about the DFL being so far ahead.

    • Submitted by Frank Phelan on 10/12/2018 - 02:11 pm.

      Are you referring to the primary poll, that was taken a month before the election? A lot happened between the polling and the election. Also, the dynamics of a low turnout three way race are very different from a two party general election.

      So no, that primary poll really has no bearing here. It may well have been spot on, as a snap shot in time. No one did, or should have, predicted the results of that election based on that poll.

  8. Submitted by Paul Brandon on 10/11/2018 - 03:04 pm.

    Polls are obviously far from perfect predictors; they’re also a lot better than a coin flip. We tend to pay attention to the failures rather than the accurate predictions.

  9. Submitted by Dennis Litfin on 10/11/2018 - 04:29 pm.

    Do not give attention to pre-election polls. Just work at getting every progressive you know in to vote. Then, after the election look at the final poll…that is the one that counts.

  10. Submitted by Paul Udstrand on 10/12/2018 - 08:21 am.

    I think Smith is a toss up, and Klobuchar should be campaigning with her. If she loses I will be back to say I told you so.

Leave a Reply