Posts tagged ‘Census’

July 20, 2011

Census 2011 looms as a failure

I couldn’t convince my own mother to do the long form census from Statistics Canada. Oh sure, she, the mother of a researcher, meant well but it is v-e-r-y long. At least, I console myself,  she recycled the form, meeting that civic participation bar.

A colleague explained that he had tried to complete the on-line the National Household Survey (the long form census) more than three times but he finally gave the task to another household member because each question required another hunt through the family files: housing costs, education levels, etc. Why couldn’t they make it easier, he asked?

This is a small sample, to be sure, but, like doing our taxes, filling out long government forms is no fun. So the response rate to the survey has got to be poor or very uneven.

Jennifer Ditchburn from Canadian Press quoted, “a lot of people shut down the conversation quickly when they find out it’s not mandatory.” She also reported Statistics Canada is not following up on incomplete Household Surveys and is settling for incomplete ones.

What will happen to the quality of the data, then? I recently asked a medical officer of health. He too had heard about the low response rates and despaired how this was what had been predicted during the discussions before parliament when the census revisions emerged. We’ll have to do estimations, he explained, to test how reliable the data is and then we will muddle through.

And that epitomizes the very issue with the census – a census is, by definition, a count of every household (or in the case of the long-form census, one in five random households). This census is no longer a census, but simply a sample of the Canadian population.

While the progress of the data collection is not public, other hints of the (non-)returns tell a worsening story.

The short-form census, the one that basically asked us to re-type what was on the mail label and add family language, must have a poor response rate too.

At the community agency where I work, clients were confused by government messaging. Hadn’t the whole debate about the census concluded that it was voluntary? Staff have had to explain that the short-form census is still required, while the long-form one, no longer called the census, is not.

Confused? I’ll say. So are others, as documented in the Hamilton Spectator and the Halifax Chronicle Herald this week (see the comments section for further evidence even).

I feel particularly mournful about this because, as governments move towards integrated policy responses, such as poverty-reduction strategies, and as our computational capacities increase, the census has become integral to good evidence-based decision-making. So just as the need and our capacity to explore better policy-making are emerging, our ability to do so has been undercut.

What a crying shame.

April 11, 2011

Statistics Canada 2011’s long form census questionnaire will play out neighbourhood by neighbourhood

Within less than a month, Canadians will be filling in the new census forms delivered to our front doors, which we all have to answer. One month later,  a third of us will be given the voluntary long form, now called the National Household Survey.

People smarter than me have pointed out how this new format will hurt the reliability of the census. We know that low-income people and others who are not included in full civic  participation are less likely to participate. And, frankly, if they are not counted, then the government will look good.

“Look, fewer poor people in Canada!” And then, because dollars follow the evidence presented, “We can cut some of those costly support programs.”

That exact logic has some of us in the community sector worried. If people in our neighbourhoods are not counted, we will not be able to make the case for the need.

Toronto had a more small scaled rehearsal of this census “undercount” problem in 2006. Key Toronto organizations, City of Toronto staff and local academic researchers all raised concerns about undercounting in some key Toronto neighbourhoods. As a result, Statistics Canada went out and re-sampled the target areas.

In fact, when the Inner City Advisory Committee at the Toronto District School Board looked at the last census, they also worried about the undercounting problem and moved a motion to encourage local schools to set up form-filling clinics to help parents to complete the census.

Schools and community agencies are close enough on the ground to reach people who live in basement apartments, or who speak one of the official languages as a third or fourth language, or who have limited literacy skills. These are the people who are less likely to fill in the census form — especially if it is voluntary — so helping them to do so, helps build a more accurate picture of the neighbourhood.

On the other hand, some are arguing that we should boycott the voluntary long census form. The data, by most measures, will be unusable because the methodology has changed so much. Any data collected this way cannot be compared with earlier censuses. “Why participate?” they ask.

So, in the end, what community agencies and local schools are left with the prisoner’s dilemma.

  • If some of us, working for the benefit of our local community, support a higher response rate, our neighbourhoods will be helped,  but others, who didn’t do the additional outreach, will be hurt in the comparison.
  • If none of us work to support a higher response rate, then the resultant undercounts will hurt our clients.
  • And the final option, that we will all work to improve the census, seems the most unlikely scenario of all.

What we choose, and what others choose, will have consequences for all of us.

read more »

%d bloggers like this: