Sandra Dunham & Leslie Rosenblood
Our census overcounts religious adherence in Canada. This leads to distortions in public policy, and to private enterprises making decisions that are based on misleading data. CFIC wants to improve the census questions related to religion and needs your help. If interested, please join our committee for census accuracy. Details for signing up are below.
Canada conducts a national census every five years. Governments use the data to realign electoral districts and calculate transfer payments. The data are also used by the private sector to make decisions. Seventy-five percent of households receive the short-form questionnaire, while the remaining 25 percent receive the “long-form” version, which asks a much longer list of questions.
The question on religion is asked on the long-form census every 10 years because “a need for this information at five-year intervals has not been demonstrated.” Data on religion are used “to support formulation, evaluation, and administration of a number of federal programs.” They are used by community and religious groups as well. Since the data are used to determine whether, where, and how to deliver social programs, it is crucial that the census be accurate.
CFIC is concerned that the question about religion leads to elevated estimations of the level of religiosity in Canada. The key question reads, “What is this person’s religion? Indicate a specific denomination or religion even if this person is not currently a practicing member of that group.” (Italics ours.)
In 2021, CFIC urged non-practicing Canadians to answer “no religion.” While we trust most of our members did, we have anecdotally confirmed that many (perhaps most) Canadians who identify by birth with a religion — even those who have never attended church and/or no longer have any connection to the creed — selected the religion of their parents. This leads to an overestimation of the number of religious Canadians, and an underestimation of the number of atheist, humanist, agnostic, and otherwise non-believing Canadians. In turn, this leads to an over-representation of government policies supporting religious groups and individuals.
In 2019, Statistics Canada conducted consultations with interested parties that included the question on religion. Participants asked that the list of religions be expanded to include Canada’s religious diversity; however, no improvement was made to the wording of the question.
We cannot allow this to happen again! In nine years, Canadians will be asked once again to answer a question on religion on the long-form census. CFIC wants the 2031 census to determine the number of Canadians who DO NOT CURRENTLY identify with any religious group. CFIC will be forming a small task force to ensure this happens. If you are someone who is concerned about this issue, please consider joining our group. If you would like to know more, or sign up to be part of the committee, please email us and tell us a bit about why you are interested and any relevant background about your ability to assist.
Check out more information here.
Chronology of Canada’s Census
- 1666 — Recorded age, sex, marital status, and occupation of New France’s 3,215 inhabitants.
- 1666 to 1871 — Questions were added related to housing stock, armaments, livestock, crops, buildings, churches, and gristmills. Beginning in 1765, questions about religion, race, and ethnicity were added.
- 1867 — The Constitution Act required that a census be taken in 1871 and again every 10 years.
- 1870 — Canada (including Nova Scotia, New Brunswick, Quebec, and Ontario) passed the Census Act of May 12, 1870.
- 1871 — First census after confederation was taken, with the goal of determining appropriate parliamentary representation by population. There were 211 questions which included religion.
- 1906 — The prairie provinces began to take a separate census of agriculture every five years.
- 1956 — The Census of Population and the Census of Agriculture began to be taken at the same time every five years.
- 1951 — The first census after Newfoundland and Labrador became part of Canada, and therefore the first to include all of what today is known as Canada.
Comparability of the question is important, as is its effect on the other surveys that use the Census data as a frame or the like.
Back in the 1980s, I was the StatsCan analyst for the Census housing questions. I sat next to the analyst responsible for the religion question. In those days, I would criticize him for the wording of the religion question, telling him that, if you shift the context to “de facto”, a true response result would be as much as 80% “secular”. Almost half a century later, the status quo prevails – people are answering the Census with the preferred religion of their ancestors. Do remember that organized religions help defray the Census cost with financial donations. All this in the service of tabulating religious affiliation numbers that exaggerate their true respective followings, and justify the religion’s continuing existence. Money talks.