- AP Photo/Stacy Thacker
- Mayor Emanuel last Wednesday with Hillary Clinton, who was in town promoting her new book.
“Rahm re-election on ropes” a Sun-Times headline proclaimed May 10.
That conclusion was based on how Mayor Emanuel had fared in a poll the paper commissioned. In a telephone survey, 511 registered voters who said they were “very likely” to vote in February’s mayoral election were presented with a field of five candidates. Emanuel finished first, but just barely, and only 29 percent of the respondents said they’d vote for him. Cook County Board president Toni Preckwinkle came in second with 26 percent. The other three candidates were Chicago Teachers Union president Karen Lewis, who got 10 percent; 2nd Ward alderman Bob Fioretti, who got 5 percent; and former 9th Ward alderman Robert Shaw, the only declared challenger, who got 3 percent. Twenty-seven percent said they were undecided.
The poll continues to be cited as evidence for the mayor’s weakness heading into his reelection campaign, so I thought I’d take a closer look at it.
Emanuel got most of the white vote, but his support from minorities was dismal, the Sun-Times said. Only 8 percent of African-American respondents backed the mayor, while 35 percent favored Preckwinkle. And Preckwinkle did even better and the mayor much worse among Hispanics: Preckwinkle got 40 percent, Emanuel just 2 percent.
“Right now, Rahm is not connecting” with voters, Michael McKeon, the president of the firm that conducted the poll, McKeon & Associates, told the Sun-Times. “If he doesn’t do that, he’s gonna lose.”
Emanuel’s critics gleefully circulated the story on Facebook and Twitter. Major news organizations spread the news as well. “Is Rahm Re-electable?” a Wall Street Journal headline asked on May 19. The mayor “risks losing re-election in February if he doesn’t turn things around quickly,” the WSJ essay maintained, citing the Sun-Times poll as evidence. The essay’s author, Allysia Finley, highlighted Emanuel’s bad showing among minorities. Valerie Strauss, who covers education for the Washington Post, pointed to the poll on May 21 while asserting that Emanuel’s support “has fallen significantly in part because of his school reform policy.” Closer to home, my colleague Ben Joravsky wrote that the mayor “appears to be in a heap of trouble,” as indicated by the poll results.
Emanuel may in fact be vulnerable, but the poll hardly showed his reelection to be on the ropes. The only opponent who rivaled his total support—Preckwinkle—hasn’t said she’ll run.
And some of the poll’s other results make me skeptical about the whole thing. The Sun-Times labeled Emanuel’s “meager” 2 percent from Hispanics a “warning sign” for the mayor. I think it’s a warning sign too—about the poll.
The mayor’s 8 percent African-American vote could be partly attributed to the fact that three of the four other candidates (Preckwinkle, Lewis, and Shaw) were African-American. But there were no Hispanic candidates in the proffered field. And even if there had been—only 2 percent for an incumbent mayor? Wouldn’t he get more than that by accident, and from name recognition?
The Sun-Times article about the poll provides a link from which “the complete poll results can be downloaded.” That link is to the website for McKeon & Associates, the polling firm. The website has a list of “media hits” from recent surveys the firm has conducted—but no information about the Chicago mayoral poll.
I called McKeon last week and asked him for the complete poll results, and he e-mailed them to me.
Those results showed that the Hispanic “warning sign” for the mayor was based on a microscopic sample. Only 60 Hispanics were polled (compared with 245 whites and 171 African-Americans). Half of the Hispanic respondents were undecided. Of the 30 decided Hispanics, a single respondent said he or she would vote for Emanuel. If not for that lone respondent, the Sun-Times story might have described the mayor’s troubling zero percent Hispanic support.
Among the decided Hispanic voters, the results were also curious. Preckwinkle would have done only slightly better if the poll had been taken at a Toni For Mayor rally. She got 24 of 30 votes from the decided Hispanics, or 80 percent. Fioretti and Shaw got two votes each, Emanuel and Lewis one.
McKeon attributed the Hispanic results to the small sample size. “You’re talking subcategories there, OK?” he said. A larger poll would have provided more reliable subcategories, but “that’s a lot of money.” The sample of Hispanics was so small that not much weight should be put on those results, he said.
But people were putting weight on the Hispanic results, I reminded him. McKeon said he had no control over that. Then he himself put weight on those results. Emanuel’s support among Hispanics may really be higher than the poll indicated, he said, but the 2 percent he got “is a very negative trend” for the mayor.
The Sun-Times story about the poll had noted that its margin of error was plus or minus 4.2 percentage points, “higher when the results are broken down by demographic factors.” I asked McKeon how high the error margin was for the demographic results. He estimated 20 percent.
Which is indeed a bit higher than 4.2 percent.
A couple of other peculiarities I noticed in the results McKeon sent me:
—There were 27 respondents between ages 18 and 30. All 27 said they were undecided.
—Although the number of women and men respondents was about equal (259 women and 252 men), Alderman Fioretti got 28 votes from men and none from women.
I called Fioretti to ask him about this warning sign concerning his support from women, but I haven’t reached him. Former alderman Shaw seems to have the same problem: he got 14 votes from men and none from women.
Odd results like these only make the poll more trustworthy, McKeon told me: “Anytime you look at a poll and everything’s perfect in it, it’s phony.”