If the polling commissioned by the Tampa Bay Times was always right, Mitt Romney would likely be President of the United States. After all, the Times had Romney winning Florida by six points. And if Romney had won Florida and we had ham, we could have had ham & eggs. If we had eggs.
The Times also had Bob Buckhorn finishing third in Tampa’s 2011 mayoral race and Deveron Gibbons making the run-off in St. Pete’s 2009 municipal election. Mind you, the Times polls are more often right than wrong, but it’s their poll which most often shapes the coverage of local races, so the veracity of their polling results is very important.
With that in mind, a prominent local political consultant (who wishes to remain anonymous) who is supporting Foster and is helping to direct some pro-Foster paid media, is taking issue with the Times polling methodology. Their internal polling, while having never shown Foster in the lead, indicates that the race is a lot closer than six points; in fact, in their last poll, they say Foster is within a point. Which makes sense considering outside interest groups have spent well over $100,000 during the last two weeks attacking Kriseman via the mailbox and paid television commercials.
This consultant prefaces their comments by saying:
“I don’t think there is necessarily a conspiracy (I want that to be clear) and I don’t have an overly strong opinion on where the race currently stands based on public polls. I just want to bring some attention to what the story claims to be their methodology. Since I don’t have access to it I can only infer based on the information in the article.”
“Given that your own poll suggested the same spread its possible the writer (Puente) wrote that line not realizing what it means.”
“As aside, I think Puente has done a very good job covering this race and should be commended.”
Here is this contrarians’s take on the Times poll…
He notes that in the story about the poll, there is this line: The poll “was weighted to reflect the age, gender, race and political party of St. Petersburg registered voters.”
This would underweight Rs by 10% and overweight Os by 10%. Weighted by current registration via the Times poll: D 48% R 28% O 24%.
The contrarian notes that current Early Vote turnout is: D 48%, R 35%, O 15%.
The contrarian then reviews past turnout:
Actual 2009 Turnout: D 48% R 38% O 14%
Actual 2005 Turnout: D 52% R 34% O 14%.
The bottom line here is that the Times poll is saying that Republican turnout will be ten points less than what it was in 2009, despite early vote turnout running just three points off from where it was in 2009.
Keep in mind that registration for 2005, 2009, and 2013 has been consistent within a .5 or so : Democrats 48%, Republicans 28%, Others 24%.
The contrarian then examines the recent poll commissioned by my blog:
“I like the way you guys set-up your polls. What I take away from the “Already Voted” Vs “Likely to Vote” is that the major groups of “undecideds” will break disproportionately to Mayor Foster. Given access to my own polls and public polling this seems to be a pretty obvious conclusion. This DOES NOT mean Mayor Foster will win, only that I believe, as I always have that the race will be very close 1-3 points separating the two. “
“Your poll correctly weights both Voted and Likely, this means the closeness in “already voted” cannot be attributed to higher rates of returns by Republicans.”
“Final Note: How can both polls be right with such different methodologies?”
Good question, Mr. Contrarian. Good question.