Counting on long-form census data


By Roderic Beaujot and Zenaida Ravanera
Thursday, September 23, 2010
A voluntary National Household Survey cannot take the place of the long-form census.
This was the message of Munir Sheikh when he resigned as Statistics Canada chief statistician over the government’s decision to scrap the long-form.

On July 21, on behalf of The University of Western Ontario, Ted Hewitt, Research and International Relations vice-president, wrote to Tony Clement, Minister of Industry, to urge the reinstatement of the use of the long-form for the 2011 Census.

Western is in good company, with now more than 350 groups and organizations having written similar letters. Speaking with the editorial board of The Globe and Mail on Sept. 16, Mark Carney, Bank of Canada governor, expressed concern that the changes to the census could undermine the data that the Bank of Canada uses to analyze productivity, labour and households in assessing Canada’s economy and steering it in the right direction.

When other surveys are undertaken, whether by Statistics Canada or by other agencies, procedures are made to properly represent the underlying population and to calibrate against missing information. This will be more difficult to do, if not impossible to do properly, without the data gathered through the long-form census. This is one of the most serious repercussions of the change to a voluntary National Household Survey. 

At a local level, sociology professor Paul Whitehead, a long-time member of the London and District Catholic School Board, observes the change will affect the quality of the data used to administer the school system. In particular, resource allocations both from the Ministry to school boards and from school boards to individual schools are based not only on population counts (available in the census short form), but also on indicators of proportions of the population with low income status, and second language learning needs, measured by indicators of ethnicity, language and immigration status.

Ontario and most other provinces have also advocated for reinstatement of the long-form. The Quebec Government’s Institut de la statistique Québec made the succinct case that the change to a voluntary survey will have impact on three dimensions of data quality: reliability, comparability and coherence. The lack of testing and the administration of the National Household Survey as voluntary will undermine the reliability of data, especially for small administrative areas. Comparability to earlier censuses will be affected, and thus the analysis of change, and the extent of effectiveness of policies and programs at the national, provincial and local levels. A further problem could be a serious lack of coherence between the mandatory Short Form Census and the voluntary National Household Survey.

There is strong public recognition of the importance of the census, one indication of which is media attention associated with the release of census results. Far from seeing the census as an onerous imposition, the vast majority of Canadians take pride in “counting themselves in,” in being part of the diversity of the country, and how one’s specific circumstances contribute to this diversity.

It is important that we be protected from government intrusion in our lives, but this right needs to be balanced with the responsibility to provide information that will help administer the policies, programs and businesses of the public and private sectors. Statements coming from the Minister of Industry, who is responsible for Statistics Canada, indicate a lack of recognition of the value of social science research toward the proper functioning of our country. Data and research are not just for the idle curiosity of academics; they are needed to make our country work better at all levels. 
No one should go to jail for failure to answer the census, be it the short form or the long-form. But, while Statistics Canada has world class protocol and procedures to ensure the maximum utility of the data and to guarantee against the release of individual data, even to other parts of government like Revenue Canada, the threat of jail needs to remain for anyone who would release census data pertaining to any individual.

There is much interest in the comparability of data across countries. A change to a voluntary survey would affect the reliability of the comparisons. Some countries have made major changes in census gathering. The United States has gone down the road of a continuous American Community Survey (ACS), but this was a decade in the making, and the initial decision to make the ACS voluntary was reversed. Others, like the Scandinavian countries, have gone down the route of using administrative data. However, such type of data gathering would require the linking of various data sources and registering one’s change of address with the government. In Canada, this would be considered much more intrusive than the use of the long-form census.
The authors are professors in the Department of Sociology and are members of the executive committee for the Cluster on Population Change and Lifecourse.

Also from this web page:


8:30 a.m. to 5 p.m.
(holidays excluded)


Helen Connell (

Jason Winders (

Paul Mayne (

Heather Travis (

Advertising Coordinator:

Denise Jones (

Off-Campus Advertising Sales:
Chris Amyot, Campus Ad (

National Advertising Representative:
Campus Plus



Western News, Suite 360
Westminster Hall
The University of Western Ontario, London N6A 3K7














Western provides the best student experience among Canada's leading research-intensive universities.