12 Points Up in Monday Polls, Bloomberg Wins by 5. Was It All Just a Con?
Hang on a second.
A week ago Quinnipiac had Bloomberg up 18 points. Yesterday they had him up 12 points.
If the polls were anywhere near right, and the shift in their numbers reflected the rate of Thompson's gain, he should have lost by about 10 points.
What happened?...
Even the stalwart Times admits Bloomberg's "margin was far smaller than expected." Last night Liz Benjamin looked at the numbers and said they were "not decisive. Not at all," and noticed NBC, who counsel we had drunkenly accepted earlier, had withdrawn their early victory projection.
Time magazine calls Bloomberg's victory, long portrayed as an impending landslide on the magnitude of his 2005 58-39 crushing of Freddy Ferrer, a "squeaker."
Did Bloomberg lose a bunch of votes? Our guess is that he never had them to begin with.
On October 22 a Marist poll showed that "likely non-enrolled voters have moved toward Thompson." For one thing, this was a remarkable shift -- a month earlier Marist had seen the unenrolled breaking 65-31 for Bloomberg. On the 22nd they found them going 48-41 for Thompson.
Is that believeable? We didn't catch this at the time -- and we admit, we got played as bad as everyone else -- because we figured if those "voters" weren't enrolled at the time of the survey, which came after the voter registration deadline, their preference hardly mattered. In hindsight, aside from the question of whether their respondents understood what "enrolled" meant, we wonder if Marist or any other polling organization ever talked to certain kinds of people who were equipped with the franchise but might not respond to phone polls, even if they were "conducted in both English and Spanish."
A lot of citizens speak Farsi, Urdu, Creole, Mandarin, Russian, etc. A lot of citizens don't take calls from pollsters, or any strangers, whatever their language.
A lot of citizens are invisible, at least to the people who get paid to "reach out" to them, and not always because of the language they speak.
We respect professionals, including polling professionals, but the idea most large organizations have of who we are -- as revealed by the marketing that is thrown at us everywhere we go -- is often woefully inaccurate. This year the pollsters seemed to figure we were just like the people in the Bloomberg ads -- "multicultural" and "diverse" in a TV-friendly way, walking with the Mayor down clean sidewalks, shaking hands with him and optimistically looking toward the future.
But to a shocking extent, they miscalculated -- and so did we.
Write Comment
Email to Friend
Print Article
