Just one day after taking office, the Liberal government moved to fulfill one of its election promises and reinstated the mandatory long-form census in time for the 2016 head count, a move that was cheered by university researchers and others who depend on the data for their work. The announcement reversed a controversial decision by the previous Conservative government to abolish the 2011 mandatory long-form census in favour of the voluntary National Household Survey (NHS).
“We are committed to making evidence-based decisions on programs and policies and to providing better and more timely services to Canadians,” said Navdeep Bains, the newly-appointed Minister of Innovation, Science and Economic Development, in announcing the change. “Today Canadians are reclaiming their right to accurate and reliable information.” Mr. Bains made the announcement in Ottawa on Nov. 5 alongside Jean-Yves Duclos, Minister of Families, Children and Social Development and an economics professor at Université Laval.
“It has been 10 years since we last had a good hard look at ourselves as Canadians,” said Stephen Toope, president of the Federation for the Humanities and Social Sciences and former president of the University of British Columbia. “We are now back on track to knowing who we are, in all our diversity. This is essential to building a more prosperous, inclusive future for Canada.”
The census, administered every five years, gathers information on a wide range of factors including education, labour and income that, because of the large sample size, can be used to analyze data for small geographic areas and population subgroups. It is widely used by researchers, business groups and other levels of government to conduct research and make policy decisions. The voluntary survey used in 2011 included the same questionnaire as previous censuses but without the threat of penalties for those who failed to complete it. Previously, those who didn’t complete a mandatory census could face fines of up to $500 and a jail term of up to three months, although the penalties were rarely enforced.
Still, the change resulted in a steep drop in the response rate to 69 percent for the 2011 NHS from 94 percent for the 2006 census. The voluntary survey also cost $22 million more to administer because Statistics Canada sent out more questionnaires (4.5 million in 2011 versus 3 million in 2006) in an effort to offset the anticipated drop in responses.
“The change meant we had worse data and more expensive data,” said Alain Bélanger, president of the Canadian Population Society and a researcher at the Institut national de la recherche scientifique, part of the Université du Québec network. It also made the data not directly comparable to that of previous census years. And for about 25 percent of municipalities and small geographic areas, the response rate was so low that the agency didn’t release the results at all because of quality concerns, said Dr. Bélanger, who worked at Statistics Canada for 17 years, including two in the census division.
Despite the drawbacks, the overall quality of 2011 NHS federal and provincial-level data was good. But, at the neighbourhood level, it “was clearly problematic,” said Don Kerr, a demographer at King’s University College at Western University. “There’s a black hole there for many municipalities, particularly in rural Canada,” said Dr. Kerr, who conducts research for the London Poverty Research Centre at King’s and relies heavily on neighbourhood data for his work.
Dr. Kerr, who has also spent time at Statistics Canada, urged the federal government to go further and restore the autonomy that Statistics Canada and the chief statistician had traditionally enjoyed. Many researchers saw the shift to a voluntary NHS as a politically motivated move. “We can’t have this political interference continuing,” said Dr. Kerr.
In an op-ed article in the Globe and Mail, Munir Sheikh, former chief statistician of Canada and now an executive fellow at the University of Calgary’s School of Public Policy, called on the federal government to change the Statistics Act to make the agency truly independent. The act gives the responsible minister the final authority in deciding technical statistical matters and gives cabinet the authority to determine questions that go into a census, he noted. “This is simply not right,” wrote Dr. Sheikh, who resigned his post at Statistics Canada to protest the cancellation of the mandatory long-form census. “I believe the contents of a census should be a decision purely based on a country’s data needs and not on the politics of the day.”
Despite the widespread criticism of the voluntary NHS, the change had spurred “a lot of creative thinking” at Statistics Canada about how to conduct a census, said Miles Corak, economics professor at the University of Ottawa’s graduate school of public and international affairs, something he hopes will continue. Among other things, the agency had started to explore alternate methods of tallying population counts through the large troves of information contained in government databases known as “administrative data,” such as income tax rolls and voter’s lists. “This is extremely innovative and cost-effective,” he said.
A number of other jurisdictions are also moving in this direction. Earlier this year, the British government officially adopted the Census Transformation Programme, a plan to consider using less-costly methods of running a census, including the use of administrative data, by 2021.
Dr. Corak also cautioned the government against reverting to the use of “a big stick type of attitude,” towards those who don’t complete the questionnaire. For some groups, the census is seen as an intrusion of privacy and they resented the notion that it was mandatory, Dr. Corak said. Legal threats are not going to win them back, he added. In the future, he’d like to see the government promote the completion of the census as a duty that is part of good, responsible citizenship.
Although buoyed by the return of the long-form census, Statistics Canada still faces significant funding challenges. The agency receives base funding from the federal government to gather certain data such as the census, as well as inflation, GDP and labour force figures that are crucial to setting macroeconomic policy. It also gathers data for other federal departments and levels of government on a cost-recovery basis. In recent years, Statistics Canada, along with other ministries, had its base funding reduced. At the same time, other departments and levels of government, faced with budget restrictions of their own and shifting priorities, reduced the amount they spent on data collection. As a result, Statistics Canada took “a double hit,” resulting in the termination of several surveys, said Dr. Corak. Still, the agency remains “a fundamentally well-managed place,” said Dr. Corak, who spent a number of years working in its research division.
The federal government followed up its announcement on the return of the long-form census with another days later lifting restrictions on federal government scientists that had prohibited them from speaking publicly about their research.
The policy changes are “highly encouraging,” said Scott Findlay, an associate professor of biology at the University of Ottawa and a co-founder of Evidence for Democracy, an advocacy group that had protested the muzzling of federal scientists and the elimination of the long-form census under the previous Conservative government.
Dr. Findlay said the Liberal’s election pledge to create a federal Chief Science Officer will be more difficult to implement quickly because of the financial resources required. He said it isn’t clear if the government intends to reinstate the post of a National Science Adviser – a position held from 2004 to 2008 by Arthur Carty, former president of the National Research Council, until it was eliminated by the Conservative government – or have it take another form.