Data deficit: How are Canadians coping sans long-form census?
Canada’s long-dead long-form census is in the news again.
Liberal MP Ted Hsu’s private members bill, which proposes to bring it back but eliminate the threat of jail time for those who don’t fill out the mandatory long-form census, has brought the issue back to the fore – even though the bill has scant chance of passing when it’s put to a vote in a majority Conservative House of Commons next week.
But if Canada’s gold standard of population data’s gone for good, what does that mean for the individuals, governments, businesses, planners, health authorities (essentially, everyone) who depended on it?
Saskatoon Health Region has spent months trying to find a way to use household survey data – with little success.
“We’ve just found too many concerns with it. We’ve been trying,” said Cory Neudorf, Chief Medical Officer of Health.
“There’s much higher non-response rates for that survey in the vulnerable parts of our city and in certain rural areas. …
“We’re ending up, in some cases, with having to suppress data for the very areas we need to study.”
Many, including the regional planning council Neudorf sits on, are still using 2006 census data, which is the most recent reliable information. But that misses not only the recession but also the huge economic changes facing places like Saskatoon in the past decade.
Starting later this year, hospitals in the Saskatoon health authority will ask patients questions such as their employment status, housing and income.
“Not everyone will want to do that, but we’re trying to explain how important this is to health planning but also, potentially, to impact your care at this visit if a doctor or a social worker working with us knows your social condition it might affect your discharge plan,” Neudorf said.
A similar project was a success in several Toronto hospitals, which are now rolling it out to other health clinics in the area.
And, contemplating the likely defeat of Hsu’s bill, Neudorf argues “this isn’t a partisan issue.
“It makes good sense from a policy perspective. and even from an efficiency perspective, to get the necessary information for appropriate planning and make sure the tax dollars we’re spending on government programs do what they’re intended to do,” he said.
Now, “it’s more like shooting in the dark.”
Meanwhile, Toronto Public Health can’t check the accuracy of its own small-scale surveys: It usually uses the census as a benchmark, said Paul Fleiszer, manager of surveillance and epidemiology. They’ve shared the cost of other income-related data with other agencies, but are left high and dry when it comes to other determinants of health.
“As time passes, our understanding of Toronto’s population and how it has changed over time will become blurred,” he said. This will hurt services including “initiatives for preventing chronic disease, ensuring healthy pregnancies, children and families, dental health and controlling communicable diseases. …
“Other than reinstating the long-form, I’m not sure there’s a viable alternative, which makes me very concerned about what we’ll do going forward.”
As Global News reported last year, some health authorities who rely on neighbourhood-level data for everything from immigrant services to epidemic planning collected info on their own.
But doing so can be prohibitively expensive.
And Global News analysis has shown the survey missed some of Canada’s key communities.
So what other options are there?
There’s also administrative data – personal information collected by various branches of government. Income tax data can be helpful, if you have the time and money to request special breakdowns from Statistics Canada.
Much more valuable is the longitudinal administrative data bank, which combines statistics on both income and demographics. While this hasn’t been accessible in the past, a pilot project to make it available began in 2013.
Because administrative data tracks real people, it can be both more precise and more of a privacy minefield: It would require meticulous anonymizing to make it releasable.
That said, University of British Columbia economist Kevin Milligan thinks this kind of administrative data will have to form the backbone of reliable Canadian info.
Losing the census “significantly hurt the quality of data we have,” he said.
But with administrative data, “you can answer questions a lot more precisely. … It allows you to drill down a lot more,” he said.
“I get excited, as a researcher, about those possibilities.”
Frances Woolley, associate dean of Carleton University’s economics department, likens it to a radio station that grows weaker and more obscured by static the farther you drive. You could put up with the static and keep using 2006 census data as it grows weaker over time; you could switch to a U.S. station; you could give up and turn off the radio altogether.
“As you get further and further away from having a nationally representative sample of the population, you become less and less aware of what potential problems you’ve got.”