Is the abandonment of religion greatly exaggerated?

The headlines around the release of the National Household Survey have proclaimed "Many Canadians are losing their religion". Yet it is not obvious this claim stands up to serious scrutiny.

Canada's religious profile evolves in two fundamentally different ways. First, people immigrate or emigrate, are born or die. Every new Canadian has their own religious beliefs, and contributes these Canada's religious profile. 

Second, a person's religious beliefs can change over time - they find God or lose their faith, or convert from one religion to another. 

In aggregate, Canada-level data, these two effects are compounded. According to the results of the National Household Survey, "About 7,850,600 people, nearly one-quarter of the population (23.9%), had no religious affiliation. This was up from 16.5% a decade earlier, as recorded in the 2001 Census. But does this mean that people are losing their religion? Or that older, more religious individuals are dying, and being replaced by young Canadians and new immigrants who are less likely to have faith? Or - given that the NHS is a voluntary survey - does it reflect sampling bias?

To separate out these various effects, I did a "pseudo-cohort" analysis. In other words, I took a clearly identified group of people - those born in Canada between around 1947 and 1956 - and tracked their religious beliefs from 1991 to 2011. I chose that cohort because it was old enough to have clearly-defined beliefs in 1991, but not so old as to have begun to experience significant mortality by 2011. The analysis is called a "pseudo-cohort" because different people are sampled each time, but since the sample is taken from the same group of people, the results should be similar to those obtained by a longitudinal study.

One of the striking things about this table is the tiny numbers belonging to non-Christian religious traditions. I think it may be hard for people who have grown up in today's multi-cultural nation to fully grasp just how much Canada's population was dominated by people of European descent in the 1940s, 1950s and 1960s - the result of decades of explicitly racist and anti-Semitic immigration policy. 

Another noteworthy phenomenon is the lack of change in the percentage of the population identifing as Protestant/other Christian between 2001 and 2011. This could reflect robustness of the faith of Canadian Protestants, or simply a correlation between Protestantism and willingness to fill out the voluntary National Household Survey.

Catholicism, on contrast, has declined slightly. This may reflect waning faith or response bias. It should be noted that a lower response rate in one group, say Catholics, will tend to inflate that size of other groups, such as Protestants and those without a religious faith. 

So is the decline in religious faith greatly exaggerated? My take-away from this limited analysis is changes in Canada's religious make-up are come more from the emergence of a new generation of Canadians with their own spiritual beliefs, than from a decline in the religiosity of Canadian adults - although I am definitely worried about the potential for a correlation between religiosity and willingness to fill out surveys.

Comments

[Bob Smith]

I think part of the "decline in religion" story (albeit a part that isn't so much reflected in the census, but probably informs conclusions people draw about the census) reflects (i) a decline in intensity of beliefs and (ii) a shift in affiliations between traditional religious groups and newer ones.

On the first point, it doesn't surprise me that the percentage of Canadians who describe themselves as Catholic or Protestant hasn't changed much. But you wonder how many of those Prostestants or Catholics attend church other than for weddings or funerals? If you asked me, I'd probably identify myself as a Protestant, but I can't say that I've ever voluntarily darkened the door of a church. So religious identification may not have changed, but religiosity sure has.

A related point is that there has probably been a shift in religiousity betwen different groups, resulting in a perceived decline in religiosity for many people. So, for example, if you looked at Quebec Catholics, unquestionably there has been a decline in "religiosity" since, say, the 1960s. On the other hand, increasingly Canada's Catholics consist of, say, Phillipinos, who are often intensily religious. Ditto between say, WASPs in Toronto, and the often fiercely Protestant Toronto Korean or Caribean communities. If you asked a Quebec Catholic or WASP Torontonian if religiousity was declining, they'd tell you "yes", and in their community, it certainly is. But that's because Quebec Catholics, or Torontonian WASPs don't attend the same churches as their Philipino, Carribean or Korean countrymen.

On the second, but related, point, I think there has been a shift between traditional Christian churches (the United Church being the most telling example) and new, often evangelical churches. That isn't neccesarily consistent with the "decline in religiosity" story since the new evangelical church are often noted for their intense "religiosity", while the United Church has credibly been accused of being the spiritual branch of the NDP. So the largely white (and old) United Church has seen its membership plummet, while new evangelical churches (often packed to the rafters with new Canadians) are doing a booming business. But for Canadians familiar with the established churches, their decline could well be seen as a decline of religiosity in general.

[Frances Woolley]

Bob - unfortunately the 1991 PUMF doesn't allow me to produce a particularly fine breakdown of Protestant religions, so I can't analyze changes within Protestant groups over time.

I agree with you that many of the changes we're seeing at the national level are being driven by young Canadians and immigrants. But I know from yesterday's post that Filipino #s are a bit on the high side in the NHS, and probably some other groups are somewhat on the low side. This makes me somewhat sceptical of the national-level religious trends.

[Determinant]

On the second, but related, point, I think there has been a shift between traditional Christian churches (the United Church being the most telling example) and new, often evangelical churches.

Numbers, please. As a member of the United Church and an NDP member, I'll respond to both points, but when you say "plummet", the United Church's decline has been less than that of the Presbyterians. And your very Protestant Korean immigrants are also very likely to affiliate with the United Church, the United Church and the Presbyterian Church in Korea have a very, very long history together.

Further, before 1945 the United Church's membership numbers were very choppy. The great post-war "boom" lasted from 1945 to 1968, the decline was so big because the boom was so big. We're reverting to historical patterns seen pre-war.

Second, there is the ever-problematic issue of culture and nose-counting. The United Church is a confessional church; Members, properly speaking, are those confirmed and placed on the Roll of Membership by Session. Your normal regular goer who isn't a Member is an Adherent, even if baptized. In the Roman Catholic Church, if you're baptized Catholic, you're Catholic, or as Monty Python so eloquently put it "you're Catholic the moment Dad c**e". Trying to square this difference in self-identification is very difficult.

"...the United Church has credibly been accused of being the spiritual branch of the NDP."

Other way around, the NDP is the political branch of the United Church (the Convention Baptists get mention here too). The United Church came first with the Social Gopsel and a history of Methodist social activism and Congregationalist free-thinking; and that begat the NDP. Both the NDP and the United Church are structured on very similar lines: they are both bottom-up organizations based on a hierarchy of deliberative assemblies. Getting a policy resolution or a Proposal through either entails nearly identical steps.

Lastly, I would draw your attention to how big the United Church (and Anglicans, and Catholics) are in terms of number of buildings. The United Church has way, way too many buildings. Your average new Evo place has one location in a town, the United Church will have three or four. Some were redundant since 1925, others were built big with the expectation of growth; many, many small churches were built in the country when people had large farm families and no automobiles. Now those churches are redundant.

[Bob Smith]

"unfortunately the 1991 PUMF doesn't allow me to produce a particularly fine breakdown of Protestant religions, so I can't analyze changes within Protestant groups over time."

No, and I'd expect that it would be a hard distinction to capture statistically, so that wasn't intended as a criticism, but I think it's a reality that shapes the narrative.

People look at the traditional churches that they're familiar with and they're empty (or closed) and they tell themselves a story about a decline of religiosity. If I went to my mother's United Church, it would be filled with White geriatrics and on the edge of survival (if it hasn't closed already, I don't think my mother's been there in years, and she was the one who kept it financially alive). On the other hand, I look at the big Baptist church out in my neighbourhood and it's got a parking lot full of school buses that it uses to transport the faithful (mostly young families, of a kaleidoscope of colours and cultures) to church from around the GTA. From my mother's perspective (and she's probably representative of the sort of white, educated, upper-middle class Canadians who drive public discourse) religiousity is clearly declining. From the perspective of the new (or maybe not so new) Canadian jamming that Baptist church every Sunday, that's not so obvious.

Determinant: I knew the United Church reference would get you going. The numbers on declining the United Church membership are readily available from United Church publications, but for a nice summary see http://www.spiritrenewalministries.com/United%20Church.html (full caveat, I know these guys have their own agenda, but the numbers are right). And UCC membership has hardly reverted to the historical trend - current membership is significantly below what it was in 1949, when Canada's population was almost a third of what it is now. (http://www.davidewart.ca/2009/02/united-church-of-canada-people-trends.html). The post-war boom was a function of population growth, the post-1960 decline, despite rapid population growth, is a symptom of something far more troubling for the United Church. The United Churche's problem isn't an excess of buildings, it's a deficit of people.

[Determinant]

I said Pre-1945, Bob. I have a nice Observer article showing how choppy the trends were before 1939, 1949 was a boom year. You have to take the average from 1925-1939 and also look at pre-Union churches before 1925.

I live in the Bay of Quinte Methodist Tract where United (ex-Methodist) churches are so frequent you can't throw a baseball without hitting one. Everybody knows the density is silly but doing something about it is another thing.

On you Baptist example, the United Church doesn't bus people in from all over the city, we have a church in every neighbourhood. And that makes a difference in how full the buildings are.

And the United Church hasn't benefited as much as the Catholic Church from Canada's immigration patterns. The Koreans help and so do some of the Caribbean immigrants, but few other groups boost our numbers. Most of the time immigrants are from historically Catholic or non-Christian countries.

[Shangwen]

Well, one thing is for sure: however bad the StatsCan data gets from the boneheaded census decision, it's not as if media numeracy has been made worse.

The decline-of-religion story is really two things: (1) changes in religious affiliation and intensity, (2)substitution of non-religious social institutions for older religious ones. You could call this a 50-year change or a 500-year change, depending on your analysis. If you compared the change in people who proclaim a religion and affiliation with religious social institutions (schools, business associations, fraternal orders, etc) to the change in people who simply proclaimed a religious affiliation, I suspect you would see a lot of the stability in overall religiosity coming from growth in populations that support religion-affiliated social institutions. That is, there would be a straight-line decline in your typical C+E Christians, but significant growth in the "socially religious" (primarily Hindus, Filipinos, and Koreans). This is just a conjecture of course, but many immigrants are essentially Jews--cultural groups where even the overtly non-religious can comfortably take part in religious life because the social bonds are so strong. This is certainly my own experience with my own group (Buddhists, and let's drop that business about Buddhism not being a religion, by the way).

You certainly don't see this kind of secular comfort in Catholic and Protestant communities (except of course the leadership of the UCC).

[Andrew F]

I think this is an issue that is often poorly studied. People identify with religious traditions even if they don't worship, don't believe in god, and disagree with the tenets of the religion. They are religious in name only. In the UK, one study showed that many people claimed to be christian because they considered themselves to be a good person (ie, being good is synonymous with being Christian).

So, the decline in religiosity across the West is being masked somewhat by the fuzzy, unclear questions about self-identification, and not questions like "Do you believe in the resurrection of Jesus Christ". If you don't, you should have a hard time believing you are Christian, but a significant number of people manage this feat.

[Frances Woolley]

Shangwen: "(2)substitution of non-religious social institutions for older religious ones."

Yup, if "yoga" was a possible answer to the religion question, religiosity would be soaring! And indeed yoga does have a lot in common with traditional religious practice e.g. emphasis on meditation, spirituality.

"even the overtly non-religious can comfortably take part in religious life because the social bonds are so strong" - one thing happens, though, is that 2nd and 3rd generation Cdn families can become very religiously diverse over time. My own extended family has gone from being entirely homogeneous to quite mixed - some people converted, others married people of different religions and ethnicities. It's not so easy to take part in religious life when different family members go to different churches. Or else one becomes a consumer of religion "no, not St X, I didn't like the priest, the music's better at St Y, on the other hand, St Z always ends on time".

Andrew F - thanks, that's an interesting point.

Determinant - this raises interesting questions about the structure of protestant religions, i.e. the tendency to be more nationally or locally defined, less universal.

[nottrampis]

keep it up Frances.

Really interesting even though it is on Canada.

We love you in OZ!

[Determinant]

Determinant - this raises interesting questions about the structure of protestant religions, i.e. the tendency to be more nationally or locally defined, less universal.

Reformations were generally national; they had to pass with the approval or disapproval of the King, or in a few places the republic. The Reformation also engendered the Counter-Reformation, which impelled the Roman Catholic Church to a greater level of uniformity than had been the case before. Each church has its own culture and organization which is a specific products of its own history.

Or in short, in England there are a thousand religions and four sauces; in France there are a thousand sauces and four religions.

The United Church, and I'd say most mainline Protestant Churches have shot themselves of the "social religionists" that Shangwen mentions. After the 1960's people (especially in the core Protestant demographic) felt no need to go to church to maintain social standing, so what's left are the people who actually believe this stuff. Except for the Very Rev. Bill Phipps, but he's a heretic.

You certainly don't see this kind of secular comfort in Catholic and Protestant communities (except of course the leadership of the UCC).

What? The United Church does entail a commitment (explicitly to the Trinity for Members, to the Basis of Union for ministers, though some have to have crossed their fingers when they said they were in "Essential Agreement" with it). There has been a strong backlash against "splash and run" baptisms in the last 30 years. In UCCan theology, infant baptism means a commitment to raise the child in the Christian faith and a vow to do so is taken by the congregation which holds the baptism (this is plain old Presbyterianism). Since that didn't work in so many cases, baptism isn't a matter of course for infants anymore in many congregations. We want to see some commitment from the parents.

And finally, somebody mentioned Oz. The Uniting Church of Australia is the United Church of Canada's sibling, the former came together in 1977 and we did in 1925; both had exactly the same mix, including the same number of Dissenting Presbyterians. There is so much to like about Australia for a Canadian. Such a wonderful sibling we have!

[Patrick]

Just a seconding of Andrew F's comment. If you were to ask more pointed questions that got to the heart of being e.g. Christian (immaculate conception, resurrection, the holy trinity, etc ...), I suspect you'd find very few people even had any idea what you were talking about. Except maybe atheists.

[Shangwen]

Frances, I'm interested in your comments on religiosity and willingness to fill out surveys; I did a quick PubMed search and found several articles strongly correlating religiosity with adherence to medical treatment as well (included an unexpected one about African Christians being more adherent to AIDS therapy than the non-religious). I can imagine the ideological leaps some people would make with this ("They all submit to authority!") but the underlying dimension could also be higher levels of trust, or confidence, or indeed more comfort with authority.

Again on the measurement issue, there are other people more prone to filling out surveys, and people unlikely to do it. So, if the voluntary census continues, we will eventually have a picture of Canada as a nation populated entirely by religious people, retirees, and the chronically unemployed. Awesome!