Vancouver Sun Article
Front page in the Business section in today’s Vancouver Sun.
Complaint spurs review of poll cited in union disclosure fight
Polling standards group investigates results widely referenced by tory MP
BY PETER O’NEIL, VANCOUVER SUN JULY 5, 2013 6:20 AM
A poll used by B.C. Conservative MP Russ Hiebert to argue there is wide public support for his controversial union disclosure bill is the subject of a disciplinary review by the organization that sets polling standards, The Vancouver Sun has learned.
The 2011 poll was paid for by the Canadian Labour-Watch Association, a Vancouver-based organization that is partly funded by non-unionized construction firms and provides, among other things, information on how to decertify unions.
Two Saskatchewan academics have argued since late 2011 that the poll by Nanos Research doesn’t give an accurate picture about public attitudes on disclosure of union spending practices.
The Canadian Labour Congress, using information compiled by the academics, filed a formal complaint with Canada’s Marketing Research and Intelligence Association.
“We’re in the process of looking into it,” said MRIA interim executive director John Ball.
He said the association expects to make the MRIA panel’s findings public this month. Veteran Canadian pollster Allan Gregg, asked by The Sun to review the Nanos poll, said he’s glad the association is taking the complaints seriously. He said that, in his opinion, the two key questions were crafted in a “horrendously biased” way to get the results LabourWatch wanted to promote tougher disclosure laws for unions.
Gregg, chairman of the Harris/Decima polling firm, told The Sun that, in his view, “this is not the kind of polling that people in our discipline should be doing. Clearly it’s being done by an advocacy group that’s got a particular perspective on the world and an axe to grind, and they’re using the poll not to illuminate their understanding of public opinion but as a PR (public relations) tool.”
Hiebert’s disclosure bill was blocked by the Senate last week. But the Harper government has indicated it will make a new bid to pass a law forcing unions to publicly disclose an array of financial details.
Hiebert has repeatedly cited, as recently as this week, the Nanos poll in defending his call for unions to make public all transactions over $5,000 and all remuneration to employees totalling more than $100,000.
Hiebert has refused requests from The Vancouver Sun for an interview but issued a statement saying he stands by his comments.
The MRIA’s Code of Conduct requires members to do research in a way that boosts public confidence by not producing polls that are “inaccurate or misleading.” An association member that violates standards could face sanctions ranging from a warning to revocation of their membership in the organization, said Ball.
The MRIA doesn’t disclose details of its panel investigations or the identity of complainants, said Ball.
But CLC president Ken Georgetti confirmed, when asked, that the CLC filed the complaint. The CLC used information assembled by two professors in the University of Regina’s business administration faculty, Sean Tucker and Andrew Stevens, who began sending emails in October 2011, to pollster Nik Nanos, LabourWatch President John Mortimer, and eventually the MRIA, raising questions about the poll.
“New results from a Nanos Research survey … shows the majority of working Canadians (whether unionized or not) say they want union financial transparency,” a Sept. 5, 2011 LabourWatch news release began, adding that 83 per cent of “working Canadians believe that Canadian law should require both public and private sector unions to be financially transparent with the public.”
The CLC and the two academics say there are two key flaws with the random telephone poll of 1,001 Canadians, conducted in the summer of 2011. A poll of that size is accurate to within 3.1 percentage points, 19 times out of 20, according to Nanos.
They argue that respondents, when asked the broad question about whether unions should publicly disclose spending, were “primed” by a preamble designed to sway them.
The preamble began: “As you may know, public and private sector unions do not pay taxes, the union dues of unionized employees of the private and public sectors are tax deductible, and their strike pay is not taxable. In addition, taxpayers pay the wages of civil servants and, therefore, fund their union dues.”
Respondents were then asked if they agree or disagree with the assertion that it “should be mandatory for unions from both the private and public sectors to publicly disclose detailed financial information on a regular basis.” Respondents 83 per cent either “completely” or “somewhat” agreed; just 14 per cent “somewhat” or “completely” disagreed. The rest were unsure.
“We’re concerned about how Labour-Watch is using misleading and inaccurate information,” Georgetti told The Sun. “The number they quote, the 83 per cent, they primed the question to get that number.”
The academics, in a joint statement sent this week to The Sun, shared Georgetti’s concern about the preamble.
Tucker and Stevens said that in their opinion, the LabourWatch-Nanos objective was “to use one-sided information to plant a seed in the minds of respondents that public disclosure of union financials is needed.”
“From a professional and scholarly standpoint, this type of question is highly inappropriate as it uses a methodological approach that primes respondents.”
Nanos noted that his poll’s findings were backed up by a Leger Marketing survey done in Alberta and financed by Merit Contractors Association, which is affiliated with LabourWatch.
That online survey of 501 employed Albertans in 2012 found that 86 per cent either “somewhat” or “completely” agreed with the proposition that it should be mandatory for unions to regularly make public financial information.
The second complaint relates to another question.
Nanos removed any reference to the question in its poll results, yet curiously included in the very first sentence following the poll report’s executive summary a statement related to the results of the question.
“Canadians were divided on whether the Canadian public or just union members/unionized employees should have access to unions’ financial information,” Nanos wrote in a statement that appears to contradict LabourWatch’s 2011 news release.
Nanos, after getting several emails from the academics, added a comment to the report on Oct. 28, 2011, explaining that he had erred in designing that question by giving respondents the unclear option of three choices of who should receive “access” to financial information – “unionized employees,” “actual union members only,” and “the Canadian public (i.e. everyone).”
The potential for someone to “fall into all three categories” could lead to “confusion” among respondents, he explained. Nanos also pointed out the term “access” was too vague.
The two academics, in a statement sent to The Sun this week, said they think Nanos’s explanation for not releasing the results was “weak.”
“We suspect Nanos refused to share the result because his client, Labour-Watch, would not want this information disclosed as it would cast doubt on the level of public support for Hiebert’s controversial private member’s bill, Bill C-377,” the academics told The Sun.
Nanos, a former MRIA executive director, defended his poll’s integrity.
“It is unfortunate that our firm designed one question with faulty answer options presented to Canadians,” he said in an email.
“We accept responsibility for this mistake and advised the client immediately of our concerns about the reliability of the data for that one question. Out of a 26-page report, there is one problem sentence related to that question. We are sorry.
“The results for the study stand as is and one flawed set of answer options should not detract from the overall study findings. Any interpretation beyond this is complete conjecture.”
The “Canadians were divided” line was still contained in the reports posted on the LabourWatch and Nanos Research websites as of midday Thursday.
LabourWatch’s Mortimer defended his organization’s decision to not make the data public.
Nanos “advised us that the response data was … unreliable and per industry standards should not be released,” he said in an email. “Therefore, we have not and will not release data that should not be used, let alone released. We stand by all of their work including their approach to handling the flawed question and data.”
Asked on Wednesday why the “Canadians were divided” statement was still on the report posted on his website, he immediately replied: “Great point. I have emailed Nik Nanos to ask for a corrected report from them, too, that we can replace the current one with that has the incorrect sentence.”