OTTAWA — Clients of Statistics Canada are concerned about “gaps” in the agency’s latest analysis of Canadian data published in the past week, casting doubt about whether they have enough information about trends to make good decisions and policies.
Staff — within the federal organization that measures and analyzes Canadian trends and statistics — say budget cuts took their toll on its latest product, an analysis of results from the 2011 National Household Survey on the Canadian work force, its education and commuting habits.
One senior analyst who managed a report on commuting habits was blunt when asked about the differences between the agency’s 2006 analysis of commuting habits from the mandatory long-form census and its latest study, based on data from the 2011 survey’s voluntary questionnaire.
Both questionnaires asked detailed questions about personal habits, housing and other related issues in order to provide a portrait of the population, but the analyst, Martin Turcotte, suggested the agency had limited resources to analyze the latest numbers from 2011.
“The report in 2006 was 40 pages long and this one is shorter. So a choice had to be made,” said Turcotte. “We decided, based on our resources and the time available and the staff, to have a shorter version that specifically focused and actually (gave) more general analysis than what was done before.”
Turcotte said some of the choices it made included offering new information about average commuting times to work, which prevented the agency from offering analysis of other statistics such as the habits of different age groups.
The federal government announced in the summer of 2010 that it would eliminate the long-form census, a mandatory survey sent to one out of five Canadian households, with a voluntary survey, suggesting that the detailed questions were intrusive and that Canadians should not be legally forced to answer them. The short census questionnaire remains mandatory, once every five years, and is sent to all Canadian households.
The main lobby group representing Canadian municipalities said it had concerns that some “gaps” might hamper planning in cities for effective decisions on infrastructure, affordable housing, bus routes as well as programs for new Canadians..
“We are calling on the government of Canada to work with all orders of governments to figure out how to fill the gap and provide communities of all sizes the most reliable information,” said Claude Dauphin, president of the Federation of Canadian Municipalities. “Municipalities rely on Statistics Canada information to understand the changing needs of our cities and communities.”
Statistics Canada spokesman Peter Frayne said the agency’s chief statistician, Wayne Smith, was not available for an interview Friday.
Environment Canada, which references some Statistics Canada data in its annual greenhouse gas emissions inventories – a mandatory report required under the Canadian government’s international climate change commitments – told Postmedia News it needed to review “available information and assess its potential use” in the national inventory reporting.
Although budget cuts have affected the agency’s capacity for analyzing its new data from 2011, Frayne said it spent more money on the last survey, sending out additional questionnaires to households.
The agency estimated it needed $22 million in supplementary funding to produce and mail-out the additional questionnaires, along with increased follow up in the field. Overall, it estimated costs of $652 million on the 2011 census and questionnaire program, versus a budget of $567 million in 2006 for the long-form census.
Christian Leger, the president of the Statistical Society of Canada, said the federal agency succeeded in getting a better response rate than private polling firms because of its brand name. But he said a non-response rate in the voluntary survey of 32 per cent opens the door to numerous problems when compared to the six per cent non-response rate from the mandatory long-form census in 2006.
“The analysts at Statistics Canada did the best they could,” said Leger in an interview.
But Leger, whose organization represents about 1,000 professional Canadian statisticians, said smaller towns as well as some suburbs in large metropolitan areas could be shortchanged by the quality of the reports that will lack sufficient data to help them manage some services, including schedules for public transit and infrastructure.
“At a global level or a provincial level there’s a way to ensure relatively good estimates that make sense, globally,” he said. “But the more we go to smaller samples, the more random it becomes to draw conclusions.”
