Terry Branstad leads Governor Chet Culver by 50 percent to 34 percent, according to a new survey by the Republican firm Magellan Data and Mapping Strategies. Culver polled 40 percent against Bob Vander Plaats, who was nearly tied with the governor at 39 percent. Culver led Rod Roberts 38 percent to 32 percent in this poll. Magellan surveyed 1,353 “likely Iowa general election voters” on March 30, and the poll has a margin of error of 2.7 percent. Several pollsters have now found Branstad at or above 50 percent against Culver, but this statistic is even more worrying:
Among the key voter subgroup of Iowa independent voters, which Magellan projects to constitute 25% of voter turnout in November, Chet Culver trails Terry Branstad by 34 points, 55% to 21%, trails Bob Vander Plaats by 7 points 36% to 29%, and is statistically tied with Rod Roberts 28% to 29%.
Recent polls by Selzer and Rasmussen have also found Branstad way ahead of Culver among no-party voters. The governor has work to do with this group. There’s no guarantee that Magellan’s likely voter screen is accurate, but no-party voters did constitute about 26 percent of the electorate in the 2006 general election.
Magellan’s numbers for Culver’s and President Barack Obama’s approval ratings were lower than I’ve seen in any other Iowa poll. Culver was at 30 percent approval/56 percent disapproval, and his favorability ratings were 33 percent favorable/58 percent unfavorable. About 48 percent of respondents had a favorable opinion of Branstad, while 36 percent had an unfavorable opinion. President Obama was at 42 percent approval/50 percent disapproval.
The Congressional generic ballot numbers also leaned Republican. Magellan asked, “If the election for Congress was being held today, and all you knew about the two candidates was that one was a Democrat and the other was a Republican, for whom would you vote?” Statewide, 40 percent of respondents said Republican, 33 percent said Democrat and 27 percent were undecided. Republicans led the generic ballot in all three Democratic-held House districts, including a difficult-to-believe six-point edge in Dave Loebsack’s district (IA-02), which traditionally has the strongest Democratic voting performance. (Keep in mind that the margin of error for subgroups in a poll is larger than the margin of error for results including the whole sample.) In this thread at The Iowa Republican blog, commenter dblhelix noted,
The [likely voter] model is extremely tight, w/ 86% responding “extremely likely” / 9% “very likely” on voter participation. I can believe R +1/large nr of undecideds in CD3 at this time, but R+6 in CD2? The “less reliable” Dem voters will fill that in, easily. Throw some points back to Dems, but the ballots in CD1-CD3 remain competitive, and again, it’s [voters aged] 35-44 driving this.
As a general rule, the tighter the likely voter screen, the more a poll will favor Republican candidates. From that perspective, it’s surprising that Branstad’s lead over Culver is “only” 16 percent. I doubt that Iowa’s first or second Congressional districts will turn out to be competitive races this fall, but no Democrat should be complacent. Our success in the Congressional races and especially in the battleground statehouse districts will depend on turning out people who wouldn’t tell a pollster in March that they are “extremely likely” to vote.
Click here to download files containing topline results, crosstabs and a presentation summarizing the results of the Magellan poll. The survey also included issue questions on health care reform and the federal stimulus bill, among other things. Share any relevant thoughts in this thread.