In his letter announcing his resignation as Canada's chief statistician, Munir Sheikh said he wanted to take the opportunity to comment on "a technical statistical issue" that had become a subject of media discussion. The question was whether a voluntary survey can become a substitute for a mandatory census. The answer: "It can not."
Five days earlier, Mr. Sheikh's minister, Tony Clement, had told the press that he asked Statistics Canada if they were confident they could do their job with a voluntary survey, and their answer had been that, provided extra steps were taken in advertising and enlarging the sample size, "Yes, we can do our job."
These two statements fundamentally contradict each other. There are only two possibilities for explaining the difference: Either someone is not telling the truth, or Mr. Sheikh's and Statistics Canada's views on the usefulness of a voluntary survey have changed.
Mr. Sheikh's and Statistics Canada's views and advice to the government on the usefulness of a voluntary survey did not change. The Harper government had been contemplating the change since at least December, and it is inconceivable that Mr. Clement did not discuss the option with Statistics Canada and its head many times during that period. Canada's statisticians are highly qualified professionals with a well-deserved reputation throughout the world. They knew in December that a voluntary survey would be an expensive, biased and unsatisfactory substitute for the mandatory census, and their advice would have been that they could certainly implement a voluntary survey but that it could not be a substitute for a mandatory census.
The government could legitimately have ignored Mr. Sheikh's advice, then proceeded without the compulsory census. Statistics Canada would then have implemented the government's decision to the best of its ability. That is the role of the public service. But that is not what happened.
The problem Mr. Sheikh faced was not the choice between voluntary survey or mandatory census, but that the minister went public with inaccurate claims about the advice he had received. Mr. Clement was selective to the point that his public claims have not accurately reflected the advice given by Statistics Canada and its public service head. This misrepresentation called into question the integrity of Statistics Canada, the office of the chief statistician and Mr. Sheikh himself.
In his letter, Mr. Sheikh said he cannot reveal the advice he and Statistics Canada gave the minister "because this information is protected under the law. However, the government can make this information public if it so wishes." So far, the government has not seen fit to divulge what went on in these discussions in the lead-up to the brouhaha.
What a senior public servant does in situations like this is talk with the Clerk of the Privy Council. This Mr. Sheikh apparently did. The clerk wears two hats: He is deputy minister to the prime minister, and he is head of the public service. As head of the public service, he would be expected to be on the side of the senior public servant. In Britain, it has not been not unknown for the head of the public service to take a minister to task for such matters as misrepresenting the advice given to him by the public service. This apparently does not happen in Canada. As deputy minister to the prime minister, the clerk's duty is to protect the prime minister and government. This does happen in Canada.
The issue of a voluntary survey rather than a mandatory census is far more than the "technical statistical issue" it was described as by Mr. Sheikh. The voluntary survey will fundamentally weaken the data on which many of Canada's government and business policies are based. But describing it as a technical matter allowed Mr. Sheikh to make public his and Statistics Canada's views on the matter. If he had described it as a policy issue, he would have been in breach of his duties.
Quite likely, Mr. Sheikh discussed his resignation and the contents of his public letter with the clerk. Perhaps they agreed on his description of a major policy change as a technical issue in order to allow Statistics Canada's views to be made public. But if a deputy minister is compelled to resign every time a minister misrepresents the advice given by the public service, Canada would soon run out of competent and independent-thinking senior public servants. Something stinks in this entire affair, and it is not Mr. Sheikh or Statistics Canada.
C.E.S. Franks is professor emeritus of political studies at Queen's University and author of The Parliament of Canada.