The Republican polling firm Rasmussen Reports surveyed 500 “likely voters” in Iowa on September 22 and came up with bad numbers for Governor Chet Culver. Former Governor Terry Branstad leads Culver by 54 percent to 34 percent, and Bob Vander Plaats leads Culver by 43 percent to 39 percent. Culver’s approval rating is 43 percent, with 53 percent of respondents disapproving of the job he is doing.

Topline results and favorability ratings are here. Culver was viewed very or somewhat favorably by 43 percent of respondents and viewed very or somewhat unfavorably by 50 percent. Branstad’s favorability was 64 percent, and his unfavorable numbers were just 29 percent. Vander Plaats was viewed favorably by 45 percent and unfavorably by 30 percent.

These numbers will encourage Branstad, who appears likely to seek his old job again. He has said he’ll decide by October, and I’ve heard rumors that Branstad will announce his candidacy very soon (September 28 according to one person, October 3 according to someone else). I believe that the numbers we see for Branstad this month will be his high water mark, since no one has campaigned against him for 15 years.

Vander Plaats will surely cite the Rasmussen poll as proof that he can beat Culver. The whole “draft Branstad” movement grew out of fears that Vander Plaats could not win a general election.

As a rule, Rasmussen polls tend to come in with somewhat better numbers for Republican candidates and worse numbers for Democrats. Go to Pollster.com and click on almost any national or state-level race to compare recent results from different pollsters.

The recent Selzer Iowa poll for the Des Moines Register found much better numbers for Culver (50 percent approve, 39 percent disapprove). Selzer polled 803 Iowans over a three-day period (3.5 percent margin of error), while Rasmussen polled 500 “likely voters” on a single day (4.5 percent margin of error). Selzer did not poll Culver against Branstad or any other Republican.

I am seeking further information about the likely voter screen Rasmussen used, as well as the proportion of Democrats, Republicans and no-party voters in the sample. I will update this post if I receive more details. If any Rasmussen premium subscriber is reading, feel free to post a comment here or e-mail me at desmoinesdem AT yahoo.com.

The same Rasmussen poll shows Senator Chuck Grassley leading Democrat Bob Krause 56 percent to 30 percent. Chase Martyn looks at the trendlines and concludes that Grassley could become vulnerable next year. In my opinion, Grassley is still well outside the danger zone for an incumbent despite his falling approval numbers.

Click here for Rasmussen’s results on how Iowans view President Obama, the economy and health care reform proposals.

UPDATE: The commenters at Swing State Project trust Selzer a lot more than Rasmussen. One person pointed out that in late July, Rasmussen found Senator Barbara Boxer of California leading Republican Carly Fiorina by just four points (45-41), while a few weeks later Research 2000 found Boxer leading Fiorina 52-31. It appears that Rasmussen’s likely voter screen produces a sample skewed a bit toward Republicans.

It would have been helpful if the Des Moines Register’s recent poll had asked respondents about Culver and Branstad and Vander Plaats. Craig Robinson is wrong to imply that the Register might have asked those questions and decided to cover up the results. The Register published the full questionnaire from its recent poll. Some pollsters don’t think head to head matchups are useful this far out from an election.