Mention the word statistics and the eyes of many simply glaze over. Yet while we may not appreciate it, statistics profoundly affect our daily lives.
Statistics are essential for a democratic society — we need good information on how society is performing. For example what is the state of the economy, where are the new jobs, are living standards growing or stagnating, are we healthier, what’s happening to the environment, what’s happening to our productivity, how fast is our society aging, is energy use per capita growing or shrinking, what is the health of our rural communities, or how many Canadians have access to the Internet?
These are just a few examples of statistics published routinely by Statistics Canada. They constitute an ongoing report card on how well our country is doing and are statistics that can be trusted. The minister of Finance cannot boast that the economy is doing well if statistics show it isn’t.
Yet Statistics Canada is at risk, as Ivan Fellegi, Canada’s chief statistician from 1985 to 2008, spelled out in lucid terms at this year’s Symons lecture in Charlottetown. For the first time in its 94-year history, a government interfered with the way in which Statistics Canada carried out its responsibility to provide Canadians with statistics collected in a professional manner and hence trusted to be reliable. This is the mandatory long-form census affair.
While the cabinet approves the actual questions that appear on the census form, the methods used to collect the census have been determined by the chief statistician, based on professional standards and techniques that ensure that the information collected is accurate and representative of the country.
By turning the mandatory long-form of the census into a voluntary survey, the federal government took the unprecedented step of telling Statistics Canada how to do its professional job, contrary to the Fundamental Principles of Official Statistics, as set out by the United Nations and which Canada had signed.
The federal government chose to announce its decision late on a Friday afternoon, before a summer long weekend, presumably hoping no one would notice. To control the official line, Statistics Canada was forbidden from answering media questions. That was taken over by Industry Minister Tony Clement, the minister responsible for Statistics Canada.
Two shabby actions followed. Clement, currently under investigation for G-8 pork barrelling, took what can only be described as the extraordinarily arrogant stance that Canada’s statisticians “all work for me.” In other words, he was entitled to boss them around any way he wanted.
Second, he implied that Statistics Canada had agreed that making the mandatory long-form voluntary would work. This was simply not true and it was on this issue that Fellegi’s successor, Munir Sheikh, felt he had no choice but to resign if the professional integrity of Statistics Canada was to be preserved.
This interference in our statistics agency matters in two ways, as Fellegi argued.
First, it will affect the census results because the data from the voluntary survey forms will not be statistically reliable.
Second, he said, “it could have deeply harmful long-term impacts on Statistics Canada itself.”
The core point, though, is that Canadians must be able to trust the reliability and impartiality of Statistics Canada, which should mean that politicians must not be allowed to meddle with the way Statistics Canada performs its professional role.
Public confidence is important for democratic accountability. But, as Fellegi says, it is also important for evidence-based decision-making.
Statistics are also important for research.
And they are also important for what Fellegi calls information brokering — the reality that indexing of pensions, equalization payments, allocation of seats in Parliament and many other decisions depend on reliable data.
This is why Fellegi wants the Statistics Act to be amended to incorporate the UN’s Fundamental Principles of Statistics and at the same time he proposes that the chief statistician be chosen through a transparent committee of wise Canadians, such as former governors of the Bank of Canada.
Independent and trusted institutions are a public good. Statistics Canada, clearly, is a public good.
A repeat of the shameful behaviour of the federal government on the census must be prevented, and with the next census planned for 2016, now is the time to act on Fellegi’s recommendations.
Economist David Crane is a syndicated Toronto Star columnist. He can be reached at firstname.lastname@example.org