Statistics Canada’s Senseless Census Decision

June 30, 2010

Open letter to the Honourable Tony Clement, Ministry of Industry and Minister Responsible for Statistics Canada and

Munir Sheikh, Chief Statistician, Statistics Canada

Dear Sirs

I am concerned by the Canwest report about the decision to restrict the upcoming Census, and the path the highly-regarded Statistics Canada and the Government of Canada seems to be headed down with regard to the type of information it is interested in collecting.

This latest decision scraps the Census long-form questionnaire in favour of a one-time survey which makes responses voluntary rather than mandatory. This move will weaken the quality and availability of data that tells us what is happening to employment, immigration, housing, incomes and education – the very issues that beg for the best policy decisions possible as we inch our way through recovery.

The Census long-form questionnaire is a unique tool that affords decision-makers a rich set of facts about Canadians, facts that are as reliable at the census tract or neighbourhood level as the nation-wide level. This is because of its huge sample size and the fact that response is mandatory. The information that the Census long-form generates is invaluable for decision-makers at every level of governance. We are unlikely to get a similarly high quality of information across geographies and sub-groups of the population from the proposed survey. Though current plans envisage circulating the new survey to one in three households, the response rate will be significantly lower because it will be a voluntary questionnaire. Response rates will be particularly depressed in areas where there are active campaigns urging non-compliance, as was the case during the 2006 Census.

Without robust Census data, it is difficult for local governments, health districts and other community service providers to respond effectively to shifting patterns of need or introduce changes – including cuts – that do the least harm or provide the greatest value for money. Indeed, it is the local level that is most hampered by this federal decision. The issue raised by cutting the Census long-form questionnaire is not just about having good information; it’s about having relevant tools for democracy.

This is not the first Statistics Canada survey to be cut or compromised during the administration of the current government in areas of inquiry that help develop or assess the impact of public policy.

For example, the Workplace and Employee Survey was discontinued in 2009, in the middle of the recession. It was the only source of annual information on job vacancies, benefits and private pensions. This mandatory survey asked employers questions about the availability of  health-related benefits, pay-related benefits and pension-related benefits to their full- and part-time employees.  The only regular source of information on private pensions was eliminated just as a flurry of concern about the sustainability of retirement incomes started to escalate.

The annual Survey of Household Spending was designed to inform us of spending patterns of Canadian households. It is about to change its methodology to save costs. Because of these changes we will no longer know what is happening to the savings rates and debt levels of rich, poor and middle class families. We will no longer know how much of their incomes rich households spend on – for example – medical expenses or energy costs compared to poorer households. The new methodology provides solid results at the most aggregated levels (what proportion of Canadian household incomes went to housing, vacations, etc.) but we are about to lose information about how Canadians spend, save, and borrow according to their age, their province of residence, and their income level. Yet this is precisely the information that is needed to monitor how Canadian households are faring in the wake of what the Bank of Canada and the national press has called the Great Recession.

Finally, the Survey of Financial Security tracks the distribution of assets and debts across income groups, age groups, family types, and regions in Canada, something virtually all advanced industrialized nations do on a regular basis.  It was last undertaken by Statistics Canada in 2005. There are no plans to run it again. It has been deemed an unnecessary survey. This is a troubling development in an era in which growing inequality and concentration of incomes and wealth have become an intractable issue, in good times and bad.

These have all been political decisions. The decision to stop inquiring about the world around us is as political as the decision to ask questions. The issues that are no longer being probed by the government or Statistics Canada are not going away.

Without a foundation of reliable, consistent information, evidence-based public policy is impossible. It is troubling to think that our elected leaders think decision-based evidence-making is preferable. This may work for a time, but it is not a durable strategy.

We at the Canadian Centre for Policy Alternatives urge to the Government of Canada to stop politicizing the work of Statistics Canada and retain utilization of the Census long-form questionnaire. We urge Statistics Canada to resume and restore surveys that are able to track trends in the distribution of incomes and wealth, savings and debt, the trends that will make or break our progress towards sustainable and broad-based recovery.

Armine Yalnizyan

Senior Economist

Canadian Centre for Policy Alternatives


  • I wish I could comment, but I have already been threatened and served with a legal notice for making comments on here about stats Canada. So all I will say is this- I would have never guessed this in a million years- no information- I guess there are many ways to run a modern democracy- but in all of the space that we have labeled modernity, I have never seen steps in such directions.

  • I wrote the following letter to the editor of the G&M which I guess will not be printed there:
    I and many others have used data from the Census long form to analyze what has been happening in Aboriginal education since 1991. We have found that only about 40 percent of First Nations students on reserve are completing high school and the results are only a little better off reserve. High school completion rates do not seem to be improving much over the last decade. The Census long form is the only national source of information on Aboriginal educational achievement. Without the Census long form there will be no information about whether Aboriginal education results are improving and no data with which objectively to assess policy initiatives. Aboriginal education is but one example among many of the singular importance of the Census long form to social and economic issues vital to Canada’s future. The cancellation of the mandatory long form Census survey is astonishingly short-sighted. The Harper government should immediately reverse this decision.

  • In the case of transportation in Canada, many data series have been discontinued over the last couple of decades -these are data on the “supply side” of transportation, which is a sink for quite large infrastructure expenditures which need to be “got right” a priori. Now, StatCan, through this degrading of statistical reliability of the “census long form” results, will additionally degrade the reliability of the demand-side data. While analysts can interpolate one missing datum point, as time rolls on, the effects of this decision on the provision of transportation infrastructure are bound to become more and more adverse. “In the land of the blind, the one-eyed man is king.” -but will that one-eyed man make the best decisions? Surely the collectivity of many researchers’ capabilities would give a superior result.

  • the census long form is a key to many others surveys at Statcan. It is used as a benchmark for income, labourforce and several other surveys. This benchmarking ensures that the sampling process adheres to some notion of the population. For example, the census long form is the only source of reliable data for, Occupation. We have no other source, this at the national and the local level. How can one run a modern democracy without knowledge over fundemental concepts such as labour force information.

    The other major component that is lost, is the local level. No other source of data could provide the information needed to sketch the portraits policy makers and the general public require over small communities, settlements and for that matter neighbourhoods in the bigger cities. It truly is one of the worst decisions I have ever heard of in terms of information gaps. And to think all the money spent on collecting business statistics, yet when it comes to people, well we all know Mr. Harper’s viewpoint- businesses are much more important.

    Truly, this proves what some have been saying on this blog all along, there is a new danger lurking within Statistics Canada, and the neutrality that was honed and protected for decades is now being thrown out and a bias has been deposited at the top of the organization.

    With decisions such as this it is more proof that people such as Mr Sheikh who once served as a lap dog for Tom D’aquino from various business lobby groups.

    Democracy within the Harper governement is nothing but a target to destroy.


  • After having had many transportation “supply-side” data series terminated over the last couple of decades, we are now faced with losing the “demand side” data, too. As transportation uses expensive infrastructure (often provided mainly by the public purse), it is important to get it right the first time. Such will be made all the more difficult with the “long form” reduced reliability

  • I started an online petition to oppose scrapping the census long form. You can sign it here:

  • Without the data from the long form census process, all Canadian surveys, including the 30% survey purported to replace the current process, will be biased on important dimensions such as income, immigration status, education, housing tenure and many other. NO RESEARCHERS WILL HAVE DATA WITH WHICH TO CORRECT FOR THESE BIASES. Even Statistics Canada, analyzing the response from the

    Statistics Canada, currently the most highly rated central statistical organization in the world, upholds the highest possible ethical standards and degree of confidentiality. Let it do its necessary work.

  • O. Francis G. Sitwell

    The long for census should be kept. I would be willing to complete it. Though at least one person has refused to do, he was NOT sent to Prison. The question about cracked floor tiles seems unnecessary

  • The issue is not the census, but the intimidation used to create compliance among vulnerable segments of the population, particularly the poor and aboriginals. This is citizen abuse, and the fact that we’ve been doing it forever is no excuse. I was quite astonished when I reviewed the questions on the 2001 and 2006 long-forms. Asking people if they’re in a same-sex common-law relationship, or which of the children in the household were born out of wedlock, seems pretty personal. If the government must ask these sorts of questions fair enough, but people who feel they’re too intrusive must be free to decline to answer without any fear of retaliation.

Leave a Reply

Your email address will not be published. Required fields are marked *