A week before Iowa’s primary election, Public Policy Polling released a poll showing former Governor Terry Branstad leading Bob Vander Plaats 46 percent to 31 percent, with State Representative Rod Roberts well behind at 13 percent. The firm surveyed 474 “likely GOP primary voters” between May 25 and 27, and the margin of error is plus or minus 4.5 percent.
The polling memo by Tom Jensen notes, “Branstad gets 42-68% of the vote across the ideological spectrum, but does worst against the 74% conservative majority, edging Tea Party favorite Vander Plaats by just 41-35.”
This poll supports what I’ve been thinking for months about Roberts. He is the best surrogate Branstad could have in this primary, diluting the votes of the social conservative base that doesn’t trust the former governor. If one candidate consolidated the “not Branstad” vote, the topline result would be nearly a dead heat.
If PPP’s survey is accurate, Branstad will win next Tuesday’s primary, but with the advantages he took into this race he should be getting 60 to 70 percent of the Republican vote. He’s done the job before, he will have spent more than $2 million before the primary (more than his opponents combined), and he has been advertising statewide on television and radio since the beginning of April. Roberts and Vander Plaats could manage only limited ad buys, and Vander Plaats just went up on television the day before PPP’s poll was in the field.
Iowa Independent highlighted a notable passage from PPP’s polling memo:
Among voters that actually know who Vander Plaats is – whether they see him favorably or unfavorably – he leads Branstad 42-37. The question is if there’s enough time left for Vander Plaats to completely make up the huge gap in name recognition he began the campaign with. Seventy-nine percent of voters have an opinion of Branstad, with it breaking down positively by a 59/20 margin. Meanwhile only 66 percent have an opinion of Vander Plaats with 47 percent seeing him favorably to 19 percent unfavorably.
There are very clear age divisions in the race. It’s tied among voters under 45, who may not even remember Branstad’s time as Governor. But he’s up 55-20 with senior citizens, who are certainly likely to remember his tenure, and that’s fueling most of his overall victory.
Vander Plaats was never going to be able to match Branstad’s spending dollar for dollar with the huge support for Branstad among Iowa’s business Republican elite. But if Vander Plaats had saved more of what he raised in 2009, he might have been able to raise his name recognition much more this spring.
As for the differences between younger and older respondents, I would think almost any Iowan over 30 remembers Branstad as governor. I suspect that this discrepancy tells us there are a lot more moderate Republicans over age 45 than under age 45. Branstad leads Vander Plaats among moderates by a huge margin in the poll. The Republican Party has grown much more conservative in the last decade or two, so younger moderates might naturally identify more with Democrats or no-party voters.
Incredibly, this is the first public poll of the Republican primary since last July, when The Iowa Republican blog commissioned a survey by Voter/Consumer research. That poll found Vander Plaats way ahead of the rest of the declared Republican candidates, with only Branstad hypothetically able to make the primary competitive.
Branstad created an exploratory committee to run for governor last October. Since then, Selzer has done two Iowa polls for the Des Moines Register, Research 2000 has done three polls for KCCI-TV, The Iowa Republican commissioned another poll in January, not to mention several Iowa polls by Rasmussen. All of those surveys tested Governor Chet Culver against his Republican challengers but not the Republican primary. The lack of polling on Branstad against Vander Plaats and Roberts is a continuing mystery to me, given how many polls have been conducted on Democratic or Republican primaries in other states. You would think that at the very least The Iowa Republican blog would want to poll the GOP primary. The fact that they haven’t suggests that last summer’s primary poll may have been intended primarily to help the people recruiting Branstad to run for governor again. Rasmussen is the most prolific pollster in the country, and has polled Republican primaries in many other states. Maybe Rasmussen really is just interested in setting a narrative rather than polling the most newsworthy races.
Share any relevant thoughts in this thread.
P.S. PPP also polled the Iowa Democratic primary for U.S. Senate and found Roxanne Conlin way ahead with 48 percent, compared to 13 percent for Bob Krause and 8 percent for Tom Fiegen. I posted a big linkfest on the Democratic U.S. Senate primary at Bleeding Heartland.
UPDATE: PPP director Tom Jensen told me today that no one commissioned the Iowa poll PPP conducted last week. He was responding to speculation on an Iowa republican blog that a Democratic 527 group may have commissioned the PPP poll to see if their direct-mail and tv commercial attacks on Branstad are working.
For a long time I accepted that robopolls were as good as live call polls, because the evidence was showing just that.
But both Rasmussen and PPP have had some crazy poll results this cycle, and it’s making me doubt them a little more. Don’t get me wrong, I’m not treating them equally at all, and PPP is FAR more credible. But there are plenty of indications that robopolls have a bias toward detecting hyperengaged voters who are not close to representative of the electorate, and also that respondents tend to give unreliable responses to robopoll questioning. On the latter point, the classic examples are PPP’s poll-testing of crazy questions like whether respondents believe Obama is the anti-Christ. The numbers on that and other crazy questions are so high that it raises doubt about polling and has led me to conclude that respondents are much more inclined to respond impulsively and unreliably, rather than thoughtfully and honestly, to robopoll questions.