Poll shows Mathis leading in Iowa Senate district 18

Democrat Liz Mathis has banked more early votes than Republican Cindy Golding for Tuesday’s special election in Iowa Senate district 18, and a survey by Public Policy Polling puts her narrowly ahead in the final days of the campaign.

Click here for PPP’s full polling memo, including the questionnaire and cross-tabs. Tom Jensen posted highlights here. Mathis leads Golding by 52 percent to 46 percent, even though respondents split 44 percent to 44 percent on whether they would prefer Democrats or Republicans to be in control of the Iowa Senate. High name recognition from local television work and a strategy to avoid taking a stand on controversial issues are probably helping Mathis outperform the desire to maintain a Democratic-controlled Senate.

Public Policy Polling surveyed 878 “likely voters” in Iowa Senate district 18 between November 4 and November 6. The margin of error is plus or minus 3.3 percent. I inquired about the likely voter screen; PPP responded that they pulled a list of voters who had voted in at least one of the last three general elections, and the call script began by asking respondents to “hang up now” if they were not planning to vote on Tuesday.

Opponents of same-sex marriage rights slightly outnumbered supporters by 46 percent to 42 percent, but only 11 percent of respondents told PPP that gay marriage was the most important factor determining their vote.

Both President Barack Obama and Governor Terry Branstad are in net negative territory among this poll’s respondents. Obama’s approve/disapprove numbers were 37/51, even though 51 percent of those surveyed said they had voted for Obama in 2008, compared to 43 percent who said they had voted for John McCain.

Branstad registered 39 percent approval and 42 percent disapproval. I am surprised that 18 percent of respondents were unsure whether they approved or disapproved of the governor’s performance. Remember, this wasn’t a survey of adults or registered voters in Senate district 18. This poll tried to sample likely voters in a special election for the state legislature. Presumably we’re talking about a very politically engaged slice of the electorate. Yet nearly a fifth of them don’t have an opinion on Branstad’s work as governor.

I don’t know how to interpret that finding. Maybe Branstad has done a good job of pushing his political agenda below the radar, or maybe voters aren’t sure what Branstad’s been trying to do, because the 26 Iowa Senate Democrats kept many of his proposals from becoming law.

Share any relevant thoughts in this thread.

About the Author(s)


  • to put this in perspective

    Here is your GOTV. Compare:

    Party ID of electorate

    2010: 32%D 39%R 29% NP

    PPP:  39%D 34%R 26% NP

    As of Friday, ballots returned as a 2010%:

    D 113%

    NP 96%

    R: 73%

    You have to credit the huge Dem bump to the early vote effort.


    PPP: 18-29 favor Mathis 59-38

    rapidly narrows to 51-46 (30-45) and 50-48 (65+)

    Unfortunately, PPP uses different age ranges than IA reporting. Eyeballing, it looks like the PPP electorate is a bit older overall. I think what you will see happen is that it’s GOP youth that will be depressed while Dem/NP will be elevated in the Marion precincts cited in earlier diary. For example, for 18-24, the GOP dominated in 2010, casting 40% of the ballots. Not this time, due to GOTV/targeting.

    Name recc:

    The “Mathis advantage” appears to be about 2.1pts if you compare the crossovers only. If I had planned the GOP campaign, I would have targeted Republican-leaning precincts that grew substatially in the last half of the decade to reduce the name recc factor. Did not see any evidence of this in Thursday’s numbers.

    I think the “Branstad hypothesis” is purely speculative, and I’m not necessarily buying. The story in my mind is the superlative Dem GOTV effort + superior name recc.


    45-50 isn’t too bad, and I’m guessing it was a lot worse for Dems in 2010. The NP turnout in 2010 was male-dominated — think we’ll see a gender shift here as well.

    Note that Mathis is handily winning women and narrowly winning men.

    I’m not sure why a 6pt lead is “narrow,” except that it’s wise to hedge bets just in case there’s some huge Republican surge. However, I feel pretty confident Mathis will win and that it’s more about GOTV/targeting/name recc and less about Branstad and other issues.

    • I was thinking of it as narrow

      because it’s not far outside the margin of error for the poll.  

      • poll not only info we have

        truthfully, I’d given Mathis the advantage before the absentees started rolling in only because it was obvious the Dems (and associated groups) were lining up behind her while the Republicans were fighting. The list of union locals doing GOTV is as long as your arm.

        Then there’s the absentees. The updated nrs for today don’t appear to be available yet (I get nothing clicking on the link). And the 45-50 is meaningless in the context of the absentees from NP, which are probably more favorable. Now the rest of the NPs have to show up to vote.

        One thing I didn’t like about the poll is that they didn’t break down “alredy voted” vs “have not voted.” That would have been very useful. Too much on gay marriage, not enough nuts and bolts.

        I have yet to hear a single Republican predicting victory with the usual bravado in a tight race. When’s the last time you’ve seen such a glum bunch? Good acting if internal polling has it as a toss-up.

        Then there’s the fundraising.

        Today’s “full-court press” may make it closer, but unless Dems/Dem-leaners are more scared of the weather, I’m not sure that a 1-2 day mega-effort is enough. The Republican lean in the district isn’t that large.

        I looked today for good arguments against the Dem advantages from sources closer to the ground, but all I found was Dorman’s argument about the registration numbers in the district. That has nothing to do with turnout at this point.

        The poll is just another piece of information added to all vectors pointing in the same direction. Of course, I do understand caution given all the last minute GOP eruptions across the state in 2010, but that was matched by solid early voting GOTV and greater unity.

        I will totally have to recalibrate my crystal ball for GOP surges if Mathis loses.

  • according to iowapolitics.com

    Republicans and conservative groups applied a “full-court press” today. That should make it more exciting, but I still think Mathis wins,

    • forecast is for rain tomorrow

      in the Cedar Rapids area (after the driest autumn in Iowa since 1958). Democrats better be ready to get their election-day vote out.

  • Huckabee

    now here’s a robocall you didn’t get …

    getting the big guns to call like that today means that the Mathis lead is real.

    So, can Huckabee swing it for Golding? Will the ex-governor from Arkansas lead the charge of Republicans to victory?