Survey accurate within, well, nothing at all


Advertise with us

DUE to the reduced accuracy of the nation’s new census methodology, the proportion of Canadians who are visible minorities is either “one in five” or “totally awesome.”

Read this article for free:


Already have an account? Log in here »

To continue reading, please subscribe with this special offer:

All-Access Digital Subscription

$4.75 per week*

  • Enjoy unlimited reading on
  • Read the E-Edition, our digital replica newspaper
  • Access News Break, our award-winning app
  • Play interactive puzzles

*Pay $19.00 every four weeks. GST will be added to each payment. Subscription can be cancelled anytime.


Hey there, time traveller!
This article was published 09/05/2013 (3559 days ago), so information in it may no longer be current.

DUE to the reduced accuracy of the nation’s new census methodology, the proportion of Canadians who are visible minorities is either “one in five” or “totally awesome.”

The face of the average Manitoban is either a 21-year-old aboriginal male or an 80-year-old Klingon warrior from the upcoming J.J. Abrams movie Star Trek: Into Darkness.

The number of Canadians who declare themselves as having no religion is either 23 per cent or Michael Stipe from R.E.M.

OK, so I’m being facetious. Unfortunately, no one really can say for certain what’s happening in this great and apparently growing nation, thanks to the federal government’s perplexing decision to do away with the mandatory, long-form census.

In 2011, it was no longer compulsory for Canadians to respond to what’s now known as the National Household Survey. Instead, you merely had to respond to the questionnaire if, you know, you were in the mood or didn’t have anything more exciting to do at the time.

As a result, in its big release on Wednesday, Statistics Canada was only able to present data for about three-quarters of the nation’s census subdivisions. In the other 25 per cent, too many people — more than half the population, in fact — did not bother to fill out the form.

What that means is the 2011 census suffers from a flaw statisticians call “non-response bias,” which is a neutral, scientific way of saying “not enough bloody data to see the whole picture.”

Entire towns and rural municipalities were left out of the survey report, which is bad news for any government effort to tailor policies to the needs of its population. Less obviously, there’s no assurance of the accuracy of the census data for the subdivisions that were in the report.

The overall non-response rate was about 31 per cent in 2011. The last time filling out the census was mandatory, the non-response rate was closer to six per cent. The latter number was small enough for statisticians to fill in the data gaps with some measure of accuracy.

Efforts to adjust the data now are less reliable — and way more annoying to the policy-makers who rely on accurate information to paint a picture of any segment of the population, whether it’s Winnipeggers or Roman Catholics or single people caring for their children.

It will come as no shock to learn people who work in the field of statistics are unhappy with the state of the data.

“It’s not ideal,” said Curtis Brown, an associate with Winnipeg’s Probe Research, using very strong words for his famously dispassionate profession.

“It’s potentially not going to be as reliable and there will be validity concerns, just because of the way it was done. Time will tell in the analysis whether it holds up to scrutiny.”

Before the feds rolled out the new voluntary survey, they conducted simulations to see how reliable it would be. According to the Globe and Mail, those tests revealed sampling errors for people with low incomes and First Nations people — two very important components of Winnipeg’s population.

Without accurate data, the city, province and Ottawa run the risk of both failing to meet the needs — and spending way too much on any particular program aimed at any particular group. In Winnipeg, anti-poverty efforts and aboriginal community-development programs are particularly important.

While there is no reason to doubt the 2011 survey’s assertion Winnipeg has Canada’s largest First Nations population, there’s very good reason to question the actual count.

Opponents of the mandatory census argued the questionnaire would say that’s none of the government’s business. To this particular brand of libertarian, the mandatory census constituted an intrusion into the private lives of citizens.

But there are two kinds of liberty — the freedom to do stuff and the freedom from harm. A libertarian case can be made it is wrong to subject anyone else to harm through our own inaction.

In other words, filling out your census form may just be one of the most valuable duties any citizen can perform, because the data offer academics, policy-makers and politicians the tools they need to improve lives, not to mention create accurate budgets and spend public money more wisely.

And to think it only took a few minutes of our time.

Report Error Submit a Tip


Advertise With Us