SurveyUSA uses a mulligan on Measure 85, and (surprise!) it's passing.
Kari Chisholm
You might remember that a month ago, SurveyUSA did a poll on behalf of KATU-2 that included the Worst Poll Question of All Time™. That question was:
On Measure 85, which is about a corporate tax "kicker," are you ...? Certain to vote yes? Certain to vote no? Or not certain?
As I noted at the time, the question failed to tell poll respondents what Measure 85 would actually DO with the kicker - raise it? lower it? spend it all on margaritas at Senor Frog's in Cancun?
A month ago, that stupid question got 14% in favor, 21% opposed, and 65% undecided. And the tell was that Democrats and liberals had the lowest numbers in favor.
Well, in a poll conducted this week, they asked the question again. And again, they got stupid numbers: 15% in favor, 25% opposed, and 60% undecided.
But after defending the idiotic poll question a month ago, this time SurveyUSA asked a detailed follow-up that matches what voters will actually see on their ballot. Here's the question:
Let me read you a more detailed description of Measure 85: The title of the ballot measure says: Amends Constitution: Allocates corporate income/excise tax "kicker" refund to additionally fund K through 12 public education. A Yes vote allocates the corporate income and excise tax "kicker" refund to the General Fund to provide additional funding for K through 12 public education. A NO vote retains existing corporate income and excise tax "kicker" that requires refund to corporations when revenue exceeds estimated collections by two percent or more. Based on this description, will you vote yes? Or vote No? on Measure 85?
And this time, we got some real numbers: 53% in favor, 26% opposed, and 21% undecided. And as you might expect, liberals were 72-7 in favor, moderates 55-26 in favor, and conservatives just 36-41. (But hey! More than a third of self-identified conservatives are in favor of sending the corporate kicker to public schools!)
So, props to SurveyUSA for recognizing their stupidity the first time around AND for asking it two ways this time, giving us an object lesson in the critical importance of survey design. See the full SurveyUSA crosstabs here.
Oh, and in other news, President Obama leads Mitt Romney 49 to 42%. That's just 2% closer than a month ago. (And the poll was almost entirely taken before the second debate.)
Measure 80, the marijuana measure, is at 36% in favor and 43% opposed. Basically no change in a month.
And the two casino measures, 82 and 83, are going down in flames - with 28% and 26% in favor, respectively - which represents zero change from a month ago. And now we know why they pulled the plug.
 |
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
10:05 a.m.
Oct 19, '12
... can we draft a Measure that dedicates a portion of the kicker toward margaritas as Senor Frog's?
10:00 p.m.
Oct 19, '12
Surveys have shown that there is a growing number of cell phone only households, about 35%, and another 15% that are mostly cell phone. Notably pollsters either don't poll them at all or under poll them. Survey USA in its internals includes a 23% sample of cell only voters. Good, but insufficient, since we know that cell users are heavily Democratic and pro-Obama. For that reason I believe it is likely that President Obama will go into double digits here in Oregon and Dems will fare even better. In addition considering that the LV voter screen is heavily weighted against younger voters, first or second time voters, or minorities, their findings short change the Democratic side.
6:28 a.m.
Oct 20, '12
Bill i share your concerns about under polling due to a measurable number of cell phone only households. It brings to mind the infamous 'Dewey Defeats Truman' headlines.
However, where do you base your assertion that statewide, cell users are both heavily democrat and pro-Obama?
1:32 p.m.
Oct 20, '12
That assertion is born out by every polling company nationwide, and it holds true in this particular sample. The demographic is younger and more Dem. and pro-Obama.
2:06 p.m.
Oct 20, '12
I believe my assertion will prove born out in election results. President Obama will win by double digits in Oregon, more than +7.
10:59 a.m.
Oct 20, '12
While Survey USA may only use 23% cell, Gallup announced recently that it was switching to 50% cell phone samples and they have Romney up 6% nationally. Not saying there are not other issues with their survey screening, but don't count on this aspect to do better than the polling averages show.
1:36 p.m.
Oct 20, '12
If Gallup consistently structures their sample so that it represents what they think the electorate Voter ID is, that is R++ , and then construct a "likely voter" screen that eliminates first or second time voters, or minorities based on their assumptions about turnout, then they will come up with the same results regardless of whether they sample cell phones. I anticipate that as Election day comes closer they will modify their LV and voter ID screen so as not to be the outlier on election day.
2:05 p.m.
Oct 20, '12
Some good analysis of the flaws in Gallup. Basically Gallup overweights their sample with white old men. The voting universe has changed:
http://www.dailykos.com/story/2012/10/20/1147573/-Gallup-Internals-Corrected-for-predicted-racial-turnout-Romney-at-49-3
And this: http://www.huffingtonpost.com/alan-abramowitz/election-polls-gallup_b_1989865.html?utm_hp_ref=@pollster
7:05 a.m.
Oct 21, '12
One more "trick" Gallup uses in the LV voter screen. They project an 80% turnout rate for the white (non-hispanic) vote. Totally implausible.
http://www.dailykos.com/story/2012/10/21/1147769/-Gallup-s-LV-model-has-Non-Hispanic-white-vote-at-80