Skip to main content

Former Statistics Canada chief statistician, Munir Sheikh, waits for the start of the Commons Industry committee on the long form census on Parliament Hill in Ottawa, July 27, 2010.STR/Reuters

The House of Commons Industry Committee's hearings on the abolition of the long-form census confirmed that the policy is misguided, and exposed the flimsiness of the government's case. The Conservatives should give fair consideration to the unanimous opposition, from such a wide variety of groups of differing (or no) political agendas, make two sensible changes, and restore the long-form census.

Here are the outside experts, on the functioning of the current census: Ernie Boyko, deputy director of the Carleton University Library Data Centre, said, "We use the census to adjust and weight our voluntary surveys." Paul Hébert of the Canadian Medical Association Journal, said, "It's been studied for so many years that we understand its error rate." Inuit leader Elisapee Sheutiapik, the Mayor of Iqaluit, said, "Housing data collected by Statistics Canada as part of the long-form census is some of the most valuable information we have to gauge the rate of overcrowding in Inuit communities."

So the current census is reliable, is used by policy-makers and forms the basis for other Canadian statistical analysis.

What did those tasked with data collection and interpretation say about the alternative proposal? Munir Sheikh believed he had to resign as chief statistician because of a public impression that Statistics Canada was "supporting a decision that no statistician would." Ivan Fellegi, Mr. Sheikh's predecessor, said a voluntary survey would be "unusable for the purpose of making comparisons." Don McLeish, the head of the Statistics Society of Canada, said that "we just don't know" how much bias would be introduced with a voluntary survey.

So the prospective voluntary survey is, by comparison, unsupported by professionals and unusable.

Nineteen per cent of respondents in a recent poll said they would not complete it, and it will be difficult to adjust the results accordingly. It is not a census.

The government's stated objections are not trivial. "Democratic freedom" - a respect for the personal privacy of Canadians; a desire not to intimidate them with the threat of prosecution - is an important principle, and the government has the right to set the policy in accordance with it, as Mr. Sheikh maintained even after his resignation.

But the government also has the duty to be proportional in its application of principle.

The policy, as it stands now, will result in lower-quality information and compromise Statistics Canada's authoritative data source, will cost $30-million more, has already resulted in the resignation of the implementing agency's chief executive, and has become an unnecessary political distraction. With two simple actions - another revision of the list of questions in the long-form questionnaire, and a legislative amendment (which would likely be supported by all parties) to remove the jail provision for non-compliance - the government would uphold its principles without undermining the census work.

The census debate seems insignificant to many, and comical to some. What is especially sad about it, though, is the extent to which defenders of the changes fail to heed the evidence and arguments in opposition. It is hard to deliver good policy under such conditions.