A poll released last week showed significant voter opposition among Michiganders to giving health benefits to unmarried partners of public employees. But poll experts and LGBT activists in the state are accusing the pollsters of forcing out the finding with leading questions and unrepresentative demographics. Meanwhile, those who funded the poll are standing by it.
The poll of 600 likely 2012 voters, conducted by Lambert, Edwards & Associates out of Grand Rapids and paid for by Denno Research of Lansing, was done Jan. 19-21. It found 54 percent opposed to providing health benefits to gay and lesbian couples or unmarried couples, while 32 percent supported the benefits.
The poll comes less than a month after Gov. Rick Snyder signed a bill prohibiting public employers from offering health benefits to public employees.
But the poll itself is coming under scrutiny because of the way questions were structured. The pollsters asked:
Should the State of Michigan and local municipalities use taxpayer money to offer health care benefits to domestic partners of employees--in other words, offering benefits to gay and lesbian couples and couples in common law marriages? Do you strongly support, support, oppose, or strongly oppose using taxpayer money for this purpose?
"I think this question is asked in a leading way because it emphasizes the term 'taxpayer money,'" says Bob Witeck of Witeck Communications in Washington, D.C. While Witeck is not a polling firm, it has partnered with Harris Interactive, which conducts the Harris Poll, for a dozen years in polling American opinions on LGBT issues. "For instance, if you asked about almost anything state and local governments do, and early in the question, focus on 'taxpayer money,' it is far more likely to inflame popular biases that public funds are either wasted or corrupted."
Witeck says the poll is an "outlier" poll because of the question wording, and the findings. Most national polls have consistently found that Americans support providing health care and other benefits to same-sex couples and unmarried heterosexual couples.
"Questions about same-sex partner health benefits and about pay and benefits parity are frequently asked in national and local poll samples. When asking these kinds of questions, respondents often are informed that the dependents of other employees, in married heterosexual couples, are eligible for health benefits, while their co-workers’ dependents or partners are not," Witeck said, explaining the outcomes of other polls on the issue. "When the facts about differential treatment or unequal rewards is clarified, then in poll after poll most Americans consistently believe that equal pay (and benefits) should be the norm – and that whether one is gay or straight, is irrelevant. In the framing of this question, this also might sound to some respondents as 'special' or 'unequal' status for gay couples which is not the case."
"It was intended to find this result," says Witeck. "It doesn't ask it as a matter of parity."
Witeck's view on the poll doesn't phase Dennis Denno, the political operative who commissioned it.
"I stand behind the validity of the question and the results," Denno said in an email statement. "The question is clear-cut and not confusing."
Denno, who now runs Denno Research, was formerly chief of staff to former Democratic state Sen. Buzz Thomas of Detroit.
Questions have also been raised about the demographics of the poll participants.
More than 400 of the poll respondents -- about two-thirds -- were above the age of 50. The U.S. Census puts the adult population of 45 and older at about 40 percent. In addition, nearly one third -- a total of 168 -- of the poll respondents hailed from West Michigan. This was the largest geographic demographic in the sample and West Michigan is considered a very conservative part of Michigan.
"And you will notice that we used cell phone lists of voters to successfully include the opinions of younger people," Denno wrote about the age question. "Finally, the Western portion of the state is defined as the following counties: Allegan, Barry, Berrien, Branch, Cass, Grand Traverse, Ionia, Kalamazoo, Kent, Lake, Leelanau, Mason, Mecosta, Montcalm, Muskegon, Newaygo, Oceana, Osceola, Ottawa, St. Joseph, and Van Buren."
"Simply put, this kind of poll result seems intended to distort the public policy issues and public understanding," Witeck says. "It is framed in a leading way, does not offer enough information to individuals about the topic, and as a matter of polling ethics, when released to the media, it must include its methodology which assures readers the survey was conducted according to industry standards, and that the sampling was properly weighted to reflect a demographic cross-section of Michigan citizens."
Witeck is not alone in challenging the question -- and the results. Emily Dievendorf, policy director of Equality Michigan, also slammed the poll.
"With questions worded that way, this result is, unfortunately, predictable. It's totally inconsistent with any reputable poll done in recent years, but I would guess the firm's goal was earned media - not necessarily survey integrity," said Dievendorf in an emailed statement. "This is not an appropriate time to risk misrepresenting the will of Michigan residents when the call for equality has been so consistently strong. Discussion surrounding the needs of the gay community should not be treated as just an opportunity to be visible. We are talking about Michiganders and their families who are being harmed in this political climate. The social climate needs to be portrayed as it is - fed up with intolerance and ready to include everybody."
David Holtz, executive director of Progress Michigan, expressed anger at the poll.
"My God, I guess the corporate press will report the results of any old poll regardless how flawed it is," Holtz said in an email statement. "The wording of the question--which hypes the "taxpayer funded" language combined with skewed sampling--tells you almost nothing about how most voters really feel about the issue of domestic partner benefits. This is a question designed to achieve a certain result."