The decision of the last federal government to proceed with the elimination of the mandatory long-form census brings to mind the words of former New York Yankees catcher Yogi Berra: “It’s like dèja vu, all over again.” That’s because Canadians have been down this road before. This month, the census forms went out in the mail. For the first time, 4.5 million households that have been selected on a stratified and sectorially random basis to receive the long-form census were asked to fill out the form on a voluntary basis. From past experience, we know that a government commitment to promote voluntary compliance will not work. Here’s how we know.
Starting with the 2006 census, for the first time all participants were asked to indicate (by checking a box) whether their responses on the short form could be made public after 92 years. If respondents said no, the form wasn’t destroyed, but access to it in its name-specific format was forever prohibited. In cases where respondents didn’t answer the question, the default position was no.
Before 2006, there was no opt-in, informed consent clause. Rather, all pre-1911 Canadian censuses, with their name-specific personal information, have been made publicly available – and that without a single complaint to the Privacy Commissioner. But, because of amendments made to the Statistics Act in 2005, Canadians filling out the 2006 census now had to agree to have their responses made public in the distant future.
Historians, archivists, and genealogists were strenuously opposed to any consent clause, but if forced to accept one, they preferred an “opt-out” clause. Even if only a small percentage of Canadians chose to keep their census information permanently inaccessible, then the integrity of the national census as a source of genealogical and historical information would be forever compromised. Moreover, it is impossible today to know what might be historically important tomorrow, and descendants of census participants might come to regret that their ancestors long before them had said no to the opt-in question.
During the committee hearings on the Statistics Act amendments, representatives of both Statistics Canada and Archives and Library Canada vowed to promote and encourage a “yes” answer to the informed consent question. An Industry Canada press release (in November 2004) even promised a public communications campaign “to encourage Canadians to allow future access to their census records to preserve Canada’s history for future generations.”
So, what happened? Over the next year, in preparation for the 2006 census, Statistics Canada added a page to its website, entitled, “The 92-year question – say Yes!” It provided exactly the kind of information that respondents needed to understand and appreciate the significance of census records to future historical and genealogical research. But this explanation was available nowhere else than the website page, nor was it strongly advocated in a public-relations blitz.
The actual wording of the question on the 2006 census form also did not facilitate a “yes” response. It began with the rather ominous warning – “The Statistics Act guarantees the confidentiality of your census information”– and then asked respondents whether they wanted “to make your census information available in 92 years for important historical and genealogical research.”
By March 2007, when the aggregate results of the 2006 census first were made available, we learned that only 56 percent of respondents had consented to do this. Forty-four percent of Canadians had decided that their privacy was more important than access in 92 years. Consequently, their personal census information would be permanently inaccessible for research, including to their own grandchildren.
This outcome was disappointing, if not distressing. The value and usefulness of census data as a future source on everyday Canadian lives had effectively been undermined by the opt-in clause. Or, as Yogi put it, “The future ain’t what it used to be.”
Will the handling of the now-voluntary long-form census be different in 2011? The former Conservative government publicly pledged to get more respondents to participate. But the 2005 amendments to the Statistics Act, which included the opt-in clause, were a hard-fought, brokered deal, where different interest groups reached a compromise and settled on a course of action.
There was no such agreement on the fate of the long-form census. The Harper government decided unilaterally that it would no longer be mandatory. And as the 2006 census experience sadly demonstrated, “opt-in” and “voluntary” do not produce reliable, comprehensive statistics that are essential to informed public policy decisions. “If you don’t know where you’re going,” said Yogi, “you might wind up some place else.” When it comes to the Canadian census, it seems he had a point.
Bill Waiser is a professor of history at the University of Saskatchewan and author of Saskatchewan: A New History.