Making sense of the 2011 census
There is considerable interest in Canada on census issues, as became quite obvious during the controversy over the long- form census in the summer of 2010, which led me to resign as chief statistician of Canada.
Statistics Canada has now released the first instalment of the data from the 2011 Census that took place in May of last year. These data relate to population and dwelling counts. Further instalments of data related to age and sex of the population, families, households and marital status will be made available through the year.
From 1971 to 2006, the census included two parts: the short form and the long form. The short form included questions of a tombstone nature with the main objective being a head count. The long form included the remaining questions that were focused on getting respondents’ socio- economic information in areas such as the labour market, income, transportation, education, disabilities, housing, citizenship and ethnicity.
For the 2011 Census, the federal government decided to eliminate the long- form census that had 53 questions while maintaining the short form with eight questions. Two questions on language that were previously in the long form were added to the short form. Still being a census, the shortform questionnaire remained mandatory. The quality of the short- form data being released starting Wednesday, therefore, should be broadly as good as that released from the previous censuses.
Wednesday’s release showed that, on average, the Canadian population over the past five years grew 5.9 per cent to reach 33,476,688. Alberta, British Columbia, Saskatchewan, Yukon and Nunavut posted above- average, and all other provinces and territories below- average, growth. The sources of population growth over the past decade have been two- thirds through immigration and one- third through natural increases.
The short- form questions on demography have many uses. For example, this information is essential to enumerate the population, to provide counts of population by federal election district, to redraw the electoral boundaries, and to provide a base for the federal fiscal transfers to provinces and the territories. The language questions in the short form allow the government to meet the minority language provisions of the Charter.
The census part that used to be the long form was replaced by a voluntary survey that was undertaken by Statistics Canada at the same time as the short form or soon after. Its results will begin to be released in early 2013.
It is this action of the government that generated considerable angst and concern across the country. Close to 370 groups objected to the decision. Critics included, among many others, provincial and municipal governments, academics, think- tanks, private- sector researchers, NGOS, religious organizations and the media.
What was their problem? They all understood that, while the long- form data would still be available from the new voluntary survey, they would not be reliable enough for the purposes for which they would use these data. This concern flows from a basic understanding of the theory of statistics.
It is a statistical fact that a voluntary survey cannot become a substitute for a mandatory census because of uneven response rates from different population groups and different size geographic areas. Increasing the sample size cannot offset this problem, and may make it worse. Hence, many data users, including the federal government, will lose the data quality they need.
Furthermore, to the extent that the long- form census data provide a benchmark for other Statistics Canada surveys, there is a risk that the quality of data from these other surveys would suffer.
We would need to wait a year for the survey results to see the quality of the new data. However, two observations regarding the quality of the new survey data have been made by many, based on statistical theory: first, these data would not be strictly comparable to the data from previous censuses, although comparisons would continue to be made in the absence of anything else; and, second, data for some smaller sub- groups of the population and smaller geographical areas may not be released because of unacceptable quality.
As the clock is ticking for a decision on the 2016 census, I hope the government would consider restoring the long form census.