Blogs & Comment

New national survey provides incomplete picture of Canada’s changing mosaic: Mark Brown

We're changing but how?

If you’ve owned a home, you know all about the concept of deferred maintenance: you put off work like roof repairs to a later date in order to save some money now. Homeowners take comfort in these choices because there is no immediate impact on their lives. The trouble is, these decisions often boomerang back to you later and often at a higher cost.

This is precisely the predicament Statistics Canada finds itself in after Ottawa nixed the mandatory long form census in 2010 in favour of the voluntary National Household Survey. The data may look fine now, but the quality of the data going forward is an issue.

And there is reason to be concerned. Statistics Canada is suppressing data on more communities than ever before due to quality issues. Data for more than 1,100 census subdivisions (CSD) was suppressed from the latest release of the National Household Survey, says Marc Hamel, a census manager with Statistics Canada. That’s up sharply from the 245 CSDs that were suppressed following the release of data from 2006 census.

While the agency often suppresses info for privacy reasons when there is a small sample size, this time around there has been a significant increase in the number of communities where Statistics Canada suppressed data due to a lack of response. These aren’t just little clusters of houses along the road. Some CSDs are home to more than 10,000 people and include places like Prescott, Ont. and Tofino, B.C. (Click here to see Statistics Canada’s full CSD suppression list).

Some wonder if this list should be even longer. Jan Kestle, founder and president of Environics Analytics, wrote in a note to clients immediately following the survey’s release that the method applied by Statistics Canada in the latest survey was “somewhat less stringent than they have used in the past in an attempt to release as much data as possible while staying within the quality limits.”

At the macro level there is little concern about the quality of the data, but it’s the smaller CSDs that pose a challenge for firms like Environics. These small data areas may not seem important but they’re cortical to marketers and social researchers. This data helps decide everything from where growing families need childcare services and whether car buyers are more likely to go with a pickup truck over a minivan, explains Kestle. The data is just as critical to the public sector to guide policy decisions on everything from education to health care.

“Data about subsets of the population are extremely important to researchers,” notes Kestle. But in the latest release, which focuses on immigrants, aboriginals and different socioeconomic or age groups, the concern is that these groups didn’t respond in a representative way and might not be properly counted.

Kestle warns the shift to a voluntary survey comes at a particularly bad time. “We are in a time of great demographic change in Canada and the past cannot effectively predict the future for a number of variables with new data,” she writes.

The National Household Survey shows that Canada is indeed in the midst of a major demographic shift. Immigrants now make up almost 21% of Canada’s population, which is highest level of any G8 country. At the same time Canada’s Aboriginal population grew by 20%, which is well above the 5% growth rate for the rest of the population.