When the federal cabinet met on the morning of Jan. 6, the first item on the agenda was the Queen's upcoming visit to Canada. But before long, talk turned to the census.
There was concern in the room that some of the questions simply went too far, particularly one about ethnic background.
According to a copy of the minutes, the Prime Minister "added that, personally, he would prefer to see the question omitted. He found it hard to see how the bonds of unity and the development of the concept of one nation would be furthered by the perpetuation of a system which emphasized the racial origin of the groups that made up the population."
Others in the room noted that removing the question would meet with stiff resistance from French-speaking groups, so cabinet decided, on that day, to keep the census as is.
The Prime Minister in this case was not Stephen Harper, but John Diefenbaker. The meeting was in 1959. And yes, Canada still has the same Queen.
Records of census discussions around the cabinet table in the National Archives add some context to the current debate unfolding over the Harper government's decision to make the long-form census voluntary.
History shows that the census isn't carved in stone: Its questions and methodology have been subject to constant change.
Initially taken once a decade, the survey was at one time a "census of population and agriculture," which in turn replaced earlier tallies of household muskets and swords. Intermediate surveys were added in 1956 so that a census took place every five years.
Liberal Prime Minister Louis St. Laurent's cabinet debated whether the provinces were freeloading off of Ottawa's research, a suspicion voiced by Industry Minister Tony Clement during the current controversy.
"An extension of the census [to every five years]would be of no great direct benefit to the federal government," state the minutes of a November, 1954, cabinet meeting. "Perhaps the provinces urging it might be asked to make a contribution towards the added cost."
Fast forward 30 years, and the census debate flared up again.
When Brian Mulroney's Tory government came to power in 1984, it announced Ottawa would stop prosecuting a B.C. woman who refused to answer some of the census questions, particularly one about the number of bathrooms in her house.
That November, the Mulroney government cancelled the 1986 national census outright in a bid to save money. Compared with the current outcry, the reaction was muted. Trade groups and academics wrote letters, but it was nothing like this summer's flood of complaints.
"What I have found surprising is the utter silence that has greeted [then-finance minister Michael Wilson's]proposal to cancel the 1986 Census of Canada," reads a letter to The Globe and Mail from a University of Toronto sociology professor published on Nov. 24, 1984.
And yet the government retreated less than a month later, reinstating the census.
Harvie Andre, who was the minister responsible for Statistics Canada at the time, recalls the strongest pushback came from other government departments that needed the basic census data to calculate programs like equalization.
In a phone interview Monday, Mr. Andre said he would occasionally hear privacy concerns about the census from constituents, but that was never the main concern of the Mulroney government. He noted that the 1986 census was reinstated with new fees for business to address the government's concern about cost.
As for the current debate, Mr. Andre said the private sector now has other ways of collecting the consumer data that comes from the long-form census.
"The need is not proven," he said, predicting the current controversy will blow over. "That's our big issue? Come on folks, shake your head."
Ivan Fellegi, Canada's chief statistician from 1985 to 2008, strongly disagrees. He insists this latest cabinet-approved census change is completely different, threatening the reputation and very existence of Statistics Canada. Having sat in on many of those cabinet debates, Mr. Fellegi said this is the first time politicians have changed the methodology of the census without testing to ensure the move is scientifically sound.
"There were lots of debates, but in the end, scientific principles always prevailed, and it was never done without testing," he said. "That's a huge difference."
Bill Curry