Census flap started with broken 'promise'

Once again Canada is in the throes of a debate about the census. Prime Minister Stephen Harper is being criticized for his government's plan to change the way the census is administered. As it stands, one-quarter of households are required by law to complete the census long form, but the Harper government has proposed moving to a system in which one-third of households are asked to voluntarily complete the long form.

Critics of the proposed change include members of the scientific and academic communities, the professional research community (including Munir Sheikh, the head of Statistics Canada, who resigned over the issue), some business groups and social activist organizations that worry the shift to a voluntary form will skew future census data by under-counting some marginalized Canadian groups.

So why drop the mandatory long form? Is the government, in the words of NDP heritage critic Charlie Angus, turning the census into a "manufactured crisis to play to a partisan base?"

Both Harper and Industry Minister Tony Clement have argued the change would address the concerns of Canadians who find the questions on the long form to be highly sensitive and personal (income, religious practices, sexual orientation and ethnic heritage). But in seeking to address the privacy issue, critics argue the government is sacrificing statistical accuracy.

There may, however, be a way of satisfying both groups.

For decades, Statistics Canada has gathered this highly sensitive personal information -- under threat of legal sanction for non-compliance -- in return for the promise that the original, individually identifiable census forms would never be made public.

Commonly referred to as the "Laurier Promise" (after Liberal prime minister Wilfrid Laurier), this confidentiality pact goes back to the time of the Special Western Census of 1906.

This changed in 2005 when Paul Martin's Liberal government introduced a bill designed to make the original census documents public by transferring them to the National Archives.

For Canadians researching their family histories, this was good news. But for those worried about privacy issues, this move represented a betrayal of trust -- personal information gathered under the authority of the census would no longer be held in trust forever.

There are, of course, those who argue that the Laurier Promise was never intended to remain in perpetuity. But that was certainly not the position of either StatsCan or the privacy commissioner.

Historically, StatsCan strongly opposed the release of individually identifiable census documents and fought earlier efforts to transfer census documents to the National Archives. Legal opinions written by the Department of Justice supported the notion that the release of individually identifiable census forms was prohibited. And Canada's privacy commissioner also spoke out against the Martin government's proposed changes to the Statistics Act.

To address these concerns, the Martin government's new rules included a provision whereby individually identifiable census forms would only be released to the public after a period of 92 years.

But it is hard to have confidence in this promise. If the previous policy of absolute confidentiality can retroactively be reversed to allow publication of individually identifiable information in 92 years, why could the policy not be changed some day in such a way that individually identifiable census records are released in 62 years, or 32 years, or less?

It is little wonder then that many Canadians chafe at the idea of having to provide highly sensitive personal information to the government knowing that this information will one day be made public for all to see.

Perhaps the way out of the current impasse would be to return to the status quo of the Laurier Promise. By restoring the Laurier Promise, those Canadians who now object to the long form as a potential invasion of privacy might be convinced to support the idea of mandatory compliance knowing that their individually identifiable census forms will never be made public.

And businesses, researchers, social activists and other groups concerned by the move to a voluntary long form could be assured that they would have access to the aggregate data they require.

Personally, I find the justification for the abrogation of the Laurier Promise to be especially disturbing. According to the preamble of the 2006 census form, the release of individually identifiable information was intended to assist historians and genealogists in their research endeavours.

Naturally, these groups would not be satisfied by a return of the Laurier Promise. But all is not lost.

If some distant cousin, nephew or other member of what I like to call the "Genealogical Industrial Complex" (e.g., for-profit firms like Ancestry.com) wants to find out about my life history, they can do what distant relations have had to do in the past -- come visit me in an old-age home and listen while I ramble on about the "good old days." And if they bring me some doughnuts -- preferably a walnut crunch -- I promise to reveal all.

Mark Yaniszewski is adjunct professor, department of political science, University of Toronto, Mississauga.

Republished from the Winnipeg Free Press print edition July 28, 2010 A12