A Tale of Two Surveys
The CAQ, Bill 21 and the Making of Public Opinion (Part 2 of 3)
It was a poll done by CROP and released on November 26 that led the media to almost unanimously embrace the idea that public opinion was firmly behind the CAQ’s policy. The poll, conducted on November 14–19, found that two-thirds (65%) of Quebeckers support the CAQ’s proposed ban on “visible religious symbols” for public employees in positions of authority, with less than a third (28%) opposed — at least initially. But popular support for the CAQ’s measure dropped dramatically when respondents were asked about implementing the law, i.e. firing teachers and other public servants that refuse to comply. Just under half (49%) of Quebeckers supported the ban when reminded of its consequences, while nearly as many were opposed (42%) to the government firing employees for their religious beliefs. Roughly one-third of this swing in opinion was due to CAQ voters changing their minds.
How was it that so many people supported the CAQ ban but opposed its implementation? When questioned about these paradoxical responses in an interview on CBC Radio, CROP President Alain Giguère clearly had trouble explaining the outcome. Although he stood by the integrity of the results, Giguère said that the “very high” level of opposition to firings was the “weak point” of the survey. Reluctantly, he admitted that the consensus on the issue of religious symbols was perhaps not as strong as it seemed: “If we see more real-life people losing their jobs, I don’t know, maybe that will change.” Quebeckers’ strong opposition to firing public servants for religious motives was, in Giguère’s opinion, the poll’s “most surprising result.”
The CROP President’s surprise was completely unjustified. Polling conducted during the Charte des valeurs debate repeatedly showed shifts of opinion when it came to questions concerning the implementation of a religious symbols ban. Over the 2013–2014 period, support for the Charte oscillated between 42% to 51% and approval for banning religious symbols was generally around 60%. This spontaneous approval, however, dissolved upon contact with the realities of implementing the PQ’s far-reaching ban, with a solid majority (55% to 75%, depending on the poll) opposed to firing employees in the name of secularism. Hardly surprising, then, that Bernard Drainville, the minister responsible for the proposed Charte, studiously avoided any mention of the word “firings” in his comments to the media at the time, even if the PQ’s policy clearly entailed many people losing their jobs.
Giguere’s surprise is especially surprising when we consider that CROP’s own polling at the time had recorded similar swings of opinion. In 2014, CROP’s Vice-President Youri Rivest acknowledged that there was “a kind of tension” in public opinion concerning restrictions on religious symbols, one that manifested “when you move from the symbolic to the real … when you are going to fire a nurse, for example.”
In a 2013 column for the Journal de Montréal, former Premier Jacques Parizeau argued that Quebeckers’ support for banning religious symbols was broad yet shallow, based on the stereotyped image of Islam presented to them on television and in newspapers. “The reaction is obvious: we don’t want that here!” This “initial reflex,” however, gave way to sympathy when Quebeckers were confronted with the real-life consequences the Charte would have for Muslim women, Parizeau noted:
Quebeckers are neither mean nor vindictive. When they are told that some women could lose their jobs because they do not want to give up their headscarf for religious reasons, three-quarters (if we believe the polls) say ‘no, that wouldn’t be right.’
CAQ deputies have frequently sounded both mean and vindictive, or at the very least callous and unfeeling, when discussing the human consequences of their policy. “You’ve got to think, each time we make a law, somebody’s dream somewhere is crushed,” Iberville MNA Claire Samson told journalists in late January. Public Security Minister Geneviève Guilbeault has raised the prospect of religious symbols raids, saying police could be called upon to enforce the ban. “The law is the law,” Guilbeault said.
The risk that religiously-motivated firings could spark broad opposition appears to be the main reason the CAQ government has included a ‘grandfather clause’ in Bill 21, partially exempting current employees from the ban. From the outset, the government has vacillated on whether it would enforce its ban by firing recalcitrant workers. Ultimately, Legault came down in favour of the exemption to avoid the creation of martyrs in the public sphere. “At the helm of the government, there are fears that the protestations of a ‘madame Lazhar’ fired because of her hijab might make headlines — an allusion to the film Monsieur Lazhar, about a North African teacher newly arrived in Québec,” La Presse reported, after news of the inclusion of a ‘grandfather’ clause was leaked. With its Charte 2.0, the CAQ appears determined to not make the same mistakes as the PQ did in 2014.
Three days after the publication of the CROP poll, Vox Pop Labs (the polling firm responsible for Radio-Canada’s Boussole électorale) released its own survey on religious symbols, drawing on a much larger sample size and using an innovative methodology. The CROP poll, like most public opinion surveys on the topic, had asked for people’s opinions on banning “visible religious symbols” without defining what this term meant. But Vox Pop Labs found that support for the CAQ’s ban was much lower when respondents were asked about specific religious symbols, as opposed to religious symbols in general.
The Vox Pop Labs survey asked whether a given religious symbol should be banned for public servants in positions of coercive authority (judges, police officers and prison guards, the so-called “Bouchard-Taylor recommenda-tions”), for positions of authority plus public sector teachers (the CAQ ban, as initially proposed), or in public more broadly. In a further departure from the norm among polling firms, Vox Pop Labs opted for pictograms to represent the symbols in question rather than simply naming them.
The company explained that it was “dissatisfied with the types of questions that were being used to gauge public sentiment on this issue.”
These questions would typically conflate many different religious symbols in a single question, such as “Do you support or oppose a ban on religious symbols for public school teachers?” Questions asked in such a way miss a key piece of information: what people actually think of when they encounter the term “religious symbols.” Are they thinking about a full veil that covers the face, like a burqa? Or are they thinking about hijabs–veils that cover only the hair? How people interpret a blanket term like “religious symbols” will affect the results. We believed that there was also a strong possibility that many people were interpreting it as meaning more ostentatious religious symbols, such as a niqab or a burka, thus increasing support for the ban.
The second reason is that we wanted to stay away from actually naming religious symbols. Not everyone knows what a kirpan or a niqab is, or how a niqab may differ from a burqa. In fact, we wanted to avoid naming symbols altogether so as not to accidentally prime respondents one way or the other.
The Vox Pop Labs survey was a direct challenge to the methodology adopted by CROP and many other polling firms. And its results confirmed that popular support for restrictions is highly contingent on what kind of religious symbols respondents have in mind.
Vox Pop Labs found that Quebeckers were “unequivocal” in their support for banning the face-covering niqabs and burqas — items worn by only a handful of Muslim women in the province, none of whom work as teachers, prosecutors, judges or police officers. But they were far less convinced of the wisdom of legislating against more commonplace religious headcoverings, with support for banning hijabs, turbans and kippas falling dramatically. Overall, only 55% of respondents favoured applying the Bouchard-Taylor restrictions to these symbols, a surprisingly low level of support for what is regarded in the media as an absolute minimum consensus. And when restrictions for these religious symbols were extended to teachers, support dropped to only 41%, meaning that 59% of respondents objected to one aspect or another of the CAQ’s ban. Vox Pop Labs reported “striking” differences across age groups, with the Bouchard-Taylor restrictions receiving the support of only 43% of youth (18–34 year olds) and the CAQ ban “a measly 26%.” (Vox Pop Labs, October 10–25)
Other recent polls confirm that public opinion is highly contingent on what kind of religious symbols respondents have in mind. A cross-Canada survey conducted by Angus Reid Institute found that 65% of Quebeckers were in favour of the ban while 28% were opposed, figures identical to the CROP poll. But when asked about banning specific religious symbols for public servants, opinions varied widely depending on the religious symbol. Quebeckers were 91% in favour of banning niqabs and burqas, but only 48%-57% in support of banning hijabs, turbans and kippas. Opposition, meanwhile, rose to 41%-48% depending on the symbol.
These polling results illustrate just how dramatically people’s opinions can change when confusion between niqabs and burqas versus other religious symbols is eliminated, and suggest that a significant chunk of the popular backing behind banning religious symbols is fictitious. The consensus on banning religious symbols is a statistical mirage created by pollsters and the media, a distortion of political reality that has attained the status of unquestionable truth.