In OR-5, inexplicable polling: Is Kurt Schrader down by 10? Or up by 12?
Kari Chisholm
Things seem extraordinarily volatile in the race for Oregon's 5th congressional district.
Late last week, SurveyUSA released a poll done for KATU that showed Scott Bruun leading 51 to 41 percent. The poll was done October 17-19, with a 4.1% margin of error.
And yet today, the Oregonian released a poll that shows Kurt Schrader leading 50 to 38 percent. The poll was done October 18-21, with a 5% margin of error.
It almost makes you wonder if one of them accidentally swapped the names in publication. Assuming that there's not some kind of absurd error like that, how do we make sense of the polling?
As the O's Jeff Mapes noted, SurveyUSA seems to have made a curious choice with regard to makeup of the electorate:
I wonder about SurveyUSA's estimate of who will turn out in the election. The latest voter registration figures I've seen by congressional district - from September - shows Democrats with a five-point registration edge, 40.1 percent to 35.4 percent for Republicans.
Yet the SurveyUSA poll estimated turnout this way: 44 percent Republican, 37 percent Democrat and 18 percent other. Even though Republican turnout is likely to be higher than Democrats this year, that doesn't seem right to me.
That may be true. It may also be false.
Here's what we do know: The Schrader/Bruun race is shaping up to be exactly what the 5th District has often been -- a tossup.
Here's what you can do. Go make get-out-the-vote calls in the Democratic HQ in Oregon City or make a donation to Kurt Schrader. In a race this close, everything matters.
 |
More Recent Posts | |
Albert Kaufman |
|
Guest Column |
|
Kari Chisholm |
|
Kari Chisholm |
Final pre-census estimate: Oregon's getting a sixth congressional seat |
Albert Kaufman |
Polluted by Money - How corporate cash corrupted one of the greenest states in America |
Guest Column |
|
Albert Kaufman |
Our Democrat Representatives in Action - What's on your wish list? |
Kari Chisholm |
|
Guest Column |
|
Kari Chisholm |
|
connect with blueoregon
12:14 p.m.
Oct 25, '10
Full disclosure: My firm built Kurt Schrader's campaign website. I speak only for myself.
12:15 p.m.
Oct 25, '10
Nate Silver recently charted the house effects of SurveyUSA and Elway in Washington. SurveyUSA exaggerates the GOP strength by 4.5%, while Elway exaggerates the Dem's by 3.2%. We haven't seen that same effect to that degree in Oregon--but we also haven't seen a lot of Elway polling here.
I'd say you're right--house effects exaggerated by a crazy year when models are getting blown up right and left.
12:54 p.m.
Oct 25, '10
Survey's voter screen has been totally skewed this election season making it an outlier in most polling I've seen. I'll go with the Northwest pollster, Elway. I think that national pollsters also give our vote-by-mail the same turnout projection that other states have. Even Survey with their skewed numbers in votes already cast had Schrader ahead by one.
1:05 p.m.
Oct 25, '10
I think Schrader will pull it off. He's a moderate democrat who's supported by the NRA, and I think he was the clear winner in the KGW debate over Bruun.
2:55 p.m.
Oct 25, '10
I would really hate to see Bruun be my representative. I am doing GOTV and hope everyone else does, too.
6:29 p.m.
Oct 25, '10
Given the amount of TV ads I see, the DLCC must be scared to death. Not my district but by far the most interesting congressional race in Oregon. Funny to watch Schrader attempt to gingerly distance himself from Obama while so many ardent progressives are disillusioned with what they perceive as Obama's centrist leanings. If nothing else, these elections make for great political theater.
12:24 a.m.
Oct 26, '10
I going down to the phone bank at Marion Democratic HQ in Salem to make some GOTV calls with Kurt Schrader at 5 PM Wednesday. I think he will phone with GOTV callers in Oregon City on Tuesday about the same time.
12:40 p.m.
Oct 26, '10
One fact often forgotten when reviewing polls, is that the "+/-" "margin of error" is a margin of error within a certain confidence rating. So, if the "margin of error" is 5% at the 95% level of confidence, then you are saying that 95% of the time your number will be within 5% of the actual number for the population represented by the sample. Conversely, 5% of the time your number is outside that range, and so could be completely whacko. Reviewing a series of polls can often give you an accurate idea of where the population is, and if one poll is entirely different from the others, it is very likely that that one poll falls in the 5%.