Guest Blogger, Rick Heimstra, Director, Centre for Research on Canadian Evangelicalism
For over a century, Statistics Canada (StatsCan) and its predecessor have collected religious affiliation information once every decade through the Canadian census, but not anymore. Starting in 2011, religious affiliation and other demographic information will be collected by a new National Household Survey (NHS). The NHS will replace the long form of the Canadian census that to date was sent to 20% of Canadian households. The new NHS will be sent to a little over a third of Canadian households, but unlike the census, the NHS will be completely voluntary. StatsCan says they’re “ counting on Canadians who receive this survey to recognize the importance of this information and to respond to the survey.”
A voluntary multi-page form to be completed for the government of Canada … would you complete it? I think StatsCan’s confidence may be misplaced.
Canadians may ignore the survey for a variety of reasons. Three reasons religious data may be seriously undercounted are:
- Time. People are busy, and, increasingly, pollsters of all descriptions are seeing falling response rates.
- Experience of persecution. Some religious groups, such as Mennonites, have experienced government persecution (yes, even in Canada, and most recently last year when a community religious school was ordered closed in Québec) and are not anxious to report their religious affiliation.
- Theological reasons. Some groups have theological reasons for opting out of census gathering.
Can a national survey that substantially misses or undercounts religious groups really provide an accurate national religious portrait? Of course, good statisticians always estimate and compensate for error in their data, but there is good evidence that when it comes to religion generally, and Evangelicals in particular, StatsCan has not done their homework. There have been ongoing problems with how the Canadian census has counted and reported on Canadian evangelicals, and due to changes in the 2001 religion question there were serious historical comparability issues with the 2001 census data. With the voluntary NHS, we will have larger and uneven religious undercounting errors introduced, further impairing the historical comparability of the data.
Rather than moving the religious affiliation question out of the census it should have been kept where it was and strengthened with a companion religious participation question. Unlike in the past, religious affiliation is no longer a reliable correlate of religious participation. Just because I say “church x is my church,” doesn’t mean I go there. If there is one thing sociologists have told us time and again it is that religious participation, far more than religious affiliation, shapes who people are and how they behave. Canadian census takers have not kept pace with this societal change by strengthening their religious affiliation question with one on religious participation.
Historically census numbers have been viewed as an objective count of a religious group’s numerical strength. Census numbers are one potential measure of influence, and sometimes create influence. If a religious group doesn’t show up in StatsCan’s numbers, will they have influence in the public square? Or, will their influence potentially go unnoticed? Will they be afforded the same religious liberties as other groups? Distortions in StatsCan’s data will inevitably produce distortions in the Canadian socio-political landscape while potentially compromising the formation of good public policy. This is a sad day for Canada, and it is one that could dramatically affect religious liberties.
StatsCan made the results of the last two censuses available online. This has been enormously helpful to public policy makers and to ministry leaders who want to more intelligently serve their communities. Currently General Social Survey (GSS) data isn’t made so freely available. Will the new NHS – which appears to be a very large GSS – be as openly and widely distributed as the census was? Who loses and what community ministries and charitable works get hurt when this information gets withheld?
Moving the religious affiliation question off the census and into a voluntary survey is a disappointing proposal. It is one that will hamper Canadians’ ability to understand the impact of religion on our private lives and our public square. StatsCan, in one decision that ignores both history and the future, may have removed government data from its place on the social landscape of our country.