Polls apart

| | Comments (2)

Earlier this week, in three installments, the Tulsa Whirled and KOTV released the results of a poll of 500 likely voters about Tuesday's bond issue, satisfaction with the progress of Vision 2025, and the popularity of the Mayor, the City Council as a body, and each individual councilor. The poll showed the bond issues passing, a majority satisfied with progress on Vision 2025, the Mayor with a 60% approval rating, and the Council with a 34% approval rating. On the individual councilors the undecideds were around 50% on each. The decided voters broke 50-50 for and against Chris Medlock, Jim Mautino, and Roscoe Turner; the other councilors had a majority approving. The poll has a margin of error of +/-4%.

The supplied answers to one of the questions about the Council were just bizarre:

We wanted to know if voters think the current conflict among members of city council is helping or hurting the city. When asked if the conflict is helping or hurting the city's growth and development, 60 percent say the conflict is hurting the city by preventing important growth and development. 22 percent say the conflict helps the city by raising questions about the city's growth and development. 9 percent say it's just business as usual. Another 9 percent had no opinion.

It would make sense to ask the "helping/hurting" question without qualification, but adding the reasons is more typical of a "push poll," in which the poll is really a vehicle for delivering a partisan message to the voter. What answer would you give to that question if you feel that the conflict is helping the city by exposing long-hidden problems in city government? Interesting, too, is that they didn't give the voter a chance to say which side he blamed for the conflict.

It's possible that this is the one poll in twenty that is outside the margin of error -- the odds are 19-1 against drawing such an unrepresentative sample. How the pollster qualified likely voters is another factor that can affect the result of a poll. Did they simply ask the respondent, "Are you likely to vote?" Pollster Ed Goeas has said that older voters tend to underestimate their likelihood of turning up at the polls, while younger voters tend to overestimate. A more sophisticated approach uses a voter's actual history of showing up at the polls, as recorded in the state's voter database. If the sample of voters for this poll was drawn from the general voting population, I would expect different results if the sample were taken from those who regularly vote in municipal elections.

That said, I'm not surprised by the results. Although I would have been happy to know that most voters in the city agree with me, the reality is that most people aren't paying close attention, and they will tend to have a positive view of local officials unless they're given some specific reason to believe otherwise. Many people still depend on the Whirled for their understanding of local events, and Councilors Mautino, Medlock, and Turner have been hammered in the news pages and on the editorial pages since they took office. It is surprising that Vision 2025 only has a bare majority of support, lower than its margin of victory in 2003.


Shadow6 said:

Or, the poll could be as accurate as the 2004 Carson/Coburn poll. They called THAT as a horse race, and Doctor Tom drubbed Little Brad.

Shadow6 said:

Correction: The World had Carson up by 7 points nine days before the election:
"The survey of 753 registered voters statewide was taken Oct. 14-19 by Tulsa-based Consumer Logic, which said the poll had a margin of error of plus or minus 3.5 percent.

The polling firm is independent of the sponsors of the Oklahoma Poll, the Tulsa World and KOTV-channel 6.

Those surveyed were asked which Senate candidate they would vote for if the election were held today.

Forty-seven percent favored Carson, 40 percent chose Coburn, and 13 percent were undecided, the polling firm said."

So I would not trust those polls very much. It's just chest-thumping, and an attempt to influence public opinion.

About this Entry

This page contains a single entry by Michael Bates published on April 2, 2005 9:50 PM.

A Schiavo parallel was the previous entry in this blog.

GO Bonds, GO is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.



Subscribe to feed Subscribe to this blog's feed:
[What is this?]