We Know What You Did Last Sunday (…We read your diary)

Sociologist Philip Brenner has produced an excellent series of journal articles on the measurement of church attendance in the U.S. and cross-nationally. Some of his work utilizes time diary studies. These are unique and well suited for measuring attendance. In standard surveys, when an interviewer asks a respondent if they attended religious services this week or how often they do so generally there are social desirability pressures that often lead some non-attenders or infrequent attenders to “over-report” their attendance. Time diaries remove this pressure as the participant is just filling out a table of time of what they did in a day. They are not “primed” in any way to report any specific activity (…these are the data often used for estimating how many hours Americans spend watching television or commuting to work per day).

Brenner’s most recent article in The Sociology of Religion looks at cross-national attendance data. In his review of the literature he writes an excellent short description of “the effect of life-stage” on religion, which I discussed in a previous post: “After being baptized in, named by, or introduced to the religious community as an infant, the young individual attends regularly until s/he leaves the parental home. The stage of ‘prodigality’ characterized by a lack of attendance, ends after the young adult marries and returns to a pattern of regular attendance, perhaps hastened by the arrival of a child. Attendance remains relatively stable until fluctuations in midlife, after the departure of adult children” (2012, p. 366). Following this life-stage model, one of the reasons we might see declining church attendance is a growing adultolescence where people wait longer to marry or have children or choose not to do so at all. Time diaries present a unique and very accurate method to study this model of religiosity.

The U.S. government conducts a variety of large scientific surveys (calculating unemployment, income data, etc.) but typically does not ask anything related to religion in these studies. There is one exception and it is a time diary study called the American Time Use Study (ATUS). This is conducted by the U.S. Census Bureau and sponsored by the Bureau of Labor Statistics. If a respondent attends church on Sunday and records this in their time diary this is captured in the study. ATUS is one of many other time diary studies that have been conducted in recent decades in the U.S. which have been collected and formatted for combined use by the Centre for Time Use Research (CTUR) as the American Heritage Time Use Study (AHTUS). This allows researchers to study how Americans have used their time from the 1965 to 2011—including the hours devoted to religious worship.

So let’s read thousands of Americans diaries…  Below is a figure summarizing the percentages of American adults who report time spent on religious worship by day of the week in each decade since the 1960s. Attendance declined in the 1970s and again in the 1980s but has since stabilized at a level where just more than one in four Americans record attendance at religious services on Sundays in time diary studies. There are also noticeable upticks in attendance on Wednesdays and Saturdays. Note that some of the people reporting worship on these other days are also doing so on Sunday so the percentages are not necessarily additive. 

Currently, America’s Millennial Generation (born 1982 or later) includes those ages 31 and younger. How do the diaries of people in this age group compare across time? Just as the life-cycle model would predict, the 18 to 31 cohorts always have lower worship levels than the overall general population. However, there is no evidence of any recent decline in their attendance. If anything, there is a slight uptick in worship time among this demographic since 2000—including in the first two years of the 2010s which measures the Millennials specifically.

Disaggregating the data from 2003 for Sundays specifically we can see a bit more of this pattern of “revival” among Millennials. In 2011, their diaries began to look a lot like those of the previous generation (born 1961 to 1981; Generation X) yet still a bit below those of the Baby Boomers (born 1943 to 1960) and the Silent/G.I. generations (born before 1943).

In a recent post I spoke of a desire for more data that would reveal what the “competitors” to Mass attendance are for Catholics (i.e., what else are they doing on Sundays that might lead them to miss Mass?). Unfortunately, the time diary studies rarely ask for the religious affiliation of participants so we can’t break this out. But we can draw broader conclusions form these data. If you are a Millennial we can guess you might be sleeping through church. Forty-three percent of the 24 hours of Sunday are spent on sleep by those ages 31 and younger—more than the members of any other generation. Now it is true that younger people need more sleep and getting older may often bring on mild insomnia so perhaps we should not be too surprised. Sunday is after all a day of rest is it not?

Millennials are also ever so slightly more likely to be working on Sunday—perhaps a product of our modern economy where everything seems open for business on Sunday a.m. The youngest Millennials may be more likely than older Americans to have weekend retail or service jobs preventing their church attendance. Television may not be keeping them away from church. Millennials spend the least amount of time with the tube whereas the retired set are more avid watchers. With that said hours of television in each generation still greatly outnumber average time spent on worship (note that these averages include “0” time observations for those who do not report the activity at all. If one looks at worship time only for those who report worship there is almost no difference in time spent on this by generation or any differences between decades. Services have not become shorter or longer over time).

So how can we know more specifically about what Catholic Millennials are doing? To see a more detailed portrait we can look to recent CARA Catholic Polls (CCP). In the table below (click the table to enlarge) we show the many aspects of life reported by adult Millennials—who currently number approximately 11.5 million people—in CCP surveys. 

Half of adult Millennial Catholics (50%) are in a household registered with a parish and 37% of Millennials (or someone in their household) regularly contributes to a parish weekly offertory collection. A third (33%) says they regularly read their parish bulletin. Seventeen percent attends Mass every week and 19% do so less than weekly, but at least once a month (...some of those attending less than weekly are in pews on any given Sunday meaning that the weekly numbers of Millennials at Mass exceeds 22%). Fifty percent say they receive ashes on Ash Wednesday. As we’ve noted before here, Millennials are even more active during Lent than most other Catholics. If the Church wants to say something to Millennials they should circle February 13, 2013 on their calendars as a critical mass will be passing through the doors of Catholic parishes on this day. Three in ten (30%) wear or carry a cross or crucifix and more than a quarter (27%) go to confession at least once a year.

Eighty-four percent use YouTube but only 2% of these users say they watch anything about religion or spirituality on this site. Seventy-two percent say they watch music videos, 53% watch comedy content, and 46% watch viral videos on YouTube. Note that only 7% of older Catholic YouTube users say they watch any religious or spiritual content on this site. About two-thirds (67 percent) of Millennials say they have not accessed any religious or spiritual content in any medium (e.g., television, internet, print, radio) in the three months before they were interviewed. Twenty-two percent of Millennials say they have posted a video to YouTube at some point.

Eighty-two percent of adult Millennial Catholics are on Facebook and 24% have a Twitter account. One in ten (10%) has a Catholic-related app on a mobile or gaming device. Thirty-seven percent say they primarily rely on the internet for their news. Yet three in ten (30%) say they pay “very little” or no attention at all to national news. If you want to get their attention don’t put an ad in a secular newspaper. Just 11% subscribes to a daily newspaper and only 5% say they pay for online content.

Two-thirds (65%) of adult Millennial Catholics are employed and 26% are married. If digital content is not pulling them away from Mass, work and family obligations may be doing so. Friendships are unlikely to draw many of them to Mass as only 8% say they have “many” friends in their parish. Just 6% say they are very active in their parish outside of attending Mass.

At the same time it would be a mistake to think things were much different for Catholics young and old in the past. A previous post here showed that many did not attend Mass as frequently as we “remember” in the 1950s. Part of that post referenced research by Joseph H. Fichter, S.J. and more of this is shown below with a contemporary comparison (thanks to CARA Executive Director Thomas P. Gaunt, S.J., Ph.D.). In Social Relations in the Urban Parish (University of Chicago Press, 1954), Father Fichter classified Catholics into four groups based on their connection to a parish and their activity within this in 1951:

• Dormant: “have in practice ‘given up’ Catholicism but have not joined another religious denomination.”
• Marginal: “are conforming to a bare, arbitrary minimum of the patterns expected in the religious institution.”
• Modal: “are normal ‘practicing’ Catholics constituting the great mass of identifiable Catholic laymen.”
• Nuclear: “are the most active participants and the most faithful believers.”

For comparison we show approximations of these same groups now in CARA surveys based on dormant Catholics being those who currently self-identify as Catholic but do not attend Mass annually. Marginal Catholics attend Mass on Christmas and Easter only, modal Catholics attend Mass at least once a month and nuclear Catholics are those who say they “very” involved in parish outside of Mass.

The number of dormant Catholics has increased by 9 percentage points since 1951 and the number of modal Catholics has declined by 12 percentage points. These trends are consistent with what we see among the broader population in the worship data from the time diaries. Too many have forgotten about all the “dormant” religious in the 1950s and have unrealistically embellished the size of the “modal” and “nuclear” groups.

Perhaps our “memory” of religiosity further in the past is quite unrealistic as well. Historians and social scientists for example have noted that churches were far from full in earlier times. As sociologist Rodney Stark notes, using original source material, “As for the ordinary people, during the middle ages and during the Renaissance, the masses rarely attended church. … In further support of these reports, an extensive survey of surviving parish churches in various parts of Europe reveals them to be too small to have held even a tiny fraction of local inhabitants” (pgs. 255-56).

So perhaps Millennials are just repeating the patterns of attendance we have known for more than a millennium? Even if they are it is still the case that there are a lot more potential distractions now than in 16th century Europe. Imagine how the Protestant Reformation might have unfolded if the Ninety-Five Theses were posted in a video on YouTube rather than on a church door. Luther didn’t have to compete for attention with funny cat videos.

Photo above courtesy of incurable_hippie from Flickr Commons.

Secularization, R.I.P.? Not even on Dec. 21, 2012

In his last novel F. Scott Fitzgerald wrote, “There are no second acts in American lives.” Not true. “Reborn” like a zombie this week, the theory of secularization is seemingly back from the dead (at least in the American popular imagination …see the image above). Social scientist Rodney Stark has lamented, “For nearly three centuries, social scientists and assorted western intellectuals have been promising the end of religion. Each generation has been confident that within another few decades, or possibly a bit longer, humans will ‘outgrow’ belief in the supernatural (“Secularization, R.I.P.” p. 249). This is a bit odd as people continue to speak of the theory “despite the fact that it never was consistent with empirical reality,” Stark notes (p. 249). Ironically, if you believe in the theory of secularization you are probably doing this mostly on faith. Great new data from Pew on global religious affiliation indicate this theory continues to be empirically challenged (my previous post explores some of the domestic implications of this study along with Gallup’s recent research on the topic). 

The theory of secularization assumes that human beings will become less religious over time as societies “modernize” and undergo economic development that results in important cultural changes. This evolutionary framework predicts that rising levels of education and autonomy in these societies lead people to eventually reject religion after which religious institutions become less significant and eventually wither away. If this theory is accurate we would expect to see religion to be strongest in the “developing” world and weakest in those countries that have “modernized” under capitalist and democratic systems. The problem for this theory continues to be reality. When applied to data it often does not work as neatly as it is assumed. One representation of this is shown below in a scatter plot of Pew’s country-level estimates of the religiously unaffiliated by each nation’s GDP per capita. There is an upward sloping straight line through these data—just not many observations around it that would represent a powerful relationship. However, if you look closely you can see some other potentially important patterns that may deserve further exploration.

Notice that most of the world’s Nones were not fostered in the bright light of “reason” blooming in “modernized” countries. Instead most have been raised in states that actively and in some cases brutally suppress religion.

About 19% of the world’s population lives in China. At the same time this single country is home to 62% of the global unaffiliated. When one adds in other similar states it is evident that two-thirds of the world’s unaffiliated live under a communist regime. An additional six percent live in formerly communist countries adjacent to the former Soviet Union or China.

One commenter to the Huffington Post’s story on the Pew data wrote ,“Yay! Go non-religion go! Reason is pulling ahead in the curve.” Maybe this comment was posted via an iPhone along with many other seemingly celebratory tweets about “Nones being the 3rd largest (non)affiliation in the world.” How many of these people knew that the person who assembled their phone lives in the country that is singularly responsible for that 3rd-place “title”? Remove China from the numbers and the unaffiliated fall to a distant 4th—behind Hinduism (...all of this completely ignoring that many unaffiliated consider themselves religious and believe in a creator).

Of course there are other causal factors to explore than communist state repression of religion. For example, many of the countries with the largest numbers of Nones have historically had very low or immeasurable levels of affiliation with the world’s Abrahamic religions (i.e., Judaism, Christianity, or Islam). Three-fourths of the world’s unaffiliated live in such a country and all of these are in Asia. Nations that have had an Abrahamic religious tradition represent a majority of the world’s population (55%) but contribute only 25% of the world's Nones.

Three-fourths of the world’s unaffiliated also reside in countries with a GDP per capita below $15,000. Yet the data may also reveal a weak pulse for secularization theory in the higher within-country unaffiliated percentages, on average, in nations with higher GDP per capita figures (i.e., last column in the table). Then again these percentages are in the teens and far below those seen in communist countries. “Modernization” is certainly not doing any of the heavy lifting in creating more of the religiously unaffiliated.

Stark once proposed that, “once and for all, let us declare an end to the social scientific faith in the theory of secularization, recognizing that it was the product of wishful thinking” (p. 269). Yet I don’t think the end is near. Predicting the demise of secularization theory is just as dangerous of predicting the demise of religion. Faith remains comparatively strong in both. I know some religion reporters who seem quite a bit fonder in the former than the latter which helps the theory of secularization thrive in popular culture even as belief in God in the U.S. has fallen only 5 percentage points since 1944.

The modern construction of the Mayan prediction (...no resemblance to what they believed or understood) was indeed wrong. It’s December 21st and you, me, God, and secularization theory are alive and well in America.  

The Future of Religious Affiliation in America: The other half of the news that is fit to print

I am just finishing a semester teaching a course on theory and evidence for secularization and preparing for next semester where I will teach a class on forecasting and prediction. The intersection of these two courses has me thinking about the future of religion in America and there is some new data out to review on the topic. Pew just released a study on global religious affiliation and non-affiliation. Approximately 16% of people around the world do not have a religious affiliation amounting to about 1.1 billion in all. Some 50.9 million of these people reside in the United States representing 4.5% of the world’s “Nones” (...75% of global Nones live either in China, Japan, Vietnam, Korea, or Taiwan... for more on international numbers see my next post).

This comes on the heels of Gallup releasing new data that concluded that “God is alive and well” in the United States. This research, based on a massive series of surveys in which one out of every thousand Americans was asked about their religion in 2012, has gone largely unnoticed by religion reporters. Not that reporters are required to cover anything but this is a bit odd as Gallup’s sample size is nearly ten times larger than many of the most highly regarded studies on this topic in the U.S. (e.g., Pew or ARIS). The sample is a hundred times larger sample than the one used by Robert Putnam and David Campbell used to write American Grace,which was widely covered by the media. The New York Times covered Pew’s research with “Study Finds One in 6 Follows No Religion” but made no mention of Gallup’s, which also provides some much needed insight on religion in America. Was it just not newsworthy? Here were some of the key Gallup findings:

• Sixty-nine percent of American adults are very or moderately religious.
• Religiousness increases with age, albeit not in a smooth path but rather in stages. Americans are least religious at age 23 and most religious at age 80.
• Trends in the age composition of the American public suggest that religion may become increasingly important in the years to come. This is mostly the result of the fact that the number of Americans who are 65 and older will essentially double over the next 20 years, dramatically increasing the number of older Americans. As long as these aging baby boomers become more religious as they age—following the path of their elders—the average religiousness in the population will go up.

That final point sure seems a bit provocative and worthy of note or further exploration. The news “problem” with the Gallup’s study may be in that it looks at religiosity through the lens of the life-cycle rather than then the “linear” secularization model that is the dominant “conventional wisdom” (…even as this has largely been discarded in the academic world in both theory and evidence. Comte, Durkheim, Weber, Marx, and Freud are all dead, yet religion lives on). Both Gallup and Pew reveal similar things about religion in the U.S. There is a large and growing number of Americans who are without any attachment to organized religion—although many are still personally religious or spiritual and believe in God. But there also remains a much larger group that is religiously affiliated.

What about the future? Gallup has placed a bet on growth in religiosity given the demographic changes expected ahead. At the same time The New York Times and most other secular news organizations are fixated on perspectives about the demise of religion in America (…just follow the trail of headlines in the archives). Who will win that bet?

Gallup is correct that life-cycle effects are important (even if they are largely unnoticed) and are being “stretched” by a growing “adultolesence” that has taken the rather typical dip in religiosity many American have always experienced in their teens and 20s into the 30s. One simple piece of evidence can be found in the age structure of Nones. Look in any recent decade and you’ll find that Nones are always disproportionally young—as if they are the “Lost Boys” never aging. Where are all the senior citizen Nones who had no affiliation in their 20s in the late-1960s?

Adultolescence is only part of the puzzle. Religious affiliation is being affected in the same way that all kinds of membership organizations in America have been since the 1950s from PTOs to the Shriners. For Robert Putnam the primary cause for this membership decline was television (…Bowling Alone is one of the most important books I have ever read). People are not as likely to bowl in leagues anymore—even though bowling may be something they still enjoy. Similarly, there are many Americans who do not belong to a church anymore but who believe in God and consider themselves religious. Too many jump to the conclusion that a lack of affiliation means a lack of religiosity. The Gallup study is important in clarifying this common mistake. Just as Putnam highlighted the effects of TV (…still America’s #1 leisure time “activity” and growing by the year) I think we can now add in the hours we spend on iPhones, tablets, video games, Facebook, etc. to the demise of face to face participation in membership organizations (...on average, teenagers spend more than an hour and a half per day just texting... and a day is still only 24 hours!). We focus so much on the simple number of people who are not in churches on Sunday. Why don’t we ever study what they do instead? In CARA’s research we know many Catholics cite family obligations, work, illness, etc. It might be interesting to know how many are playing Assassin’s Creed instead of saying the Nicene Creed at Mass. We just don’t know. The Gallup study is betting that the 20-year-old without a religious affiliation playing video games on Sundays now will be in a church 35 years from now.

As always I’m really only interested in where the data lead. This requires one to try to disentangle period, life-cycle, and generational effects. In the figure below we see the trends in lack of religious affiliation by generation. This was very uncommon among Americans born before 1942. It was never common among the Lost Generation (born 1883 to 1900) and the G.I. Generation (born 1901 to 1924). Yet this begins to tick up among the Silent Generation (born 1942 to 1960) in the 1990s. This of course is a pattern that the Gallup life-cycle model would not predict. With each successive generation from the Boomers (born 1943 to 1960) to the Millennials (born 1982 or later) there is significant growth in non-affiliation as if parents are increasingly unable to pass on their affiliation. This coincides with what Putnam’s “Bowling Alone” model would predict. The generations raised with a television in their living rooms are less likely to be connected to organized religion (…or other secular institutions). Those born in the digital age are even more disconnected from real-world membership institutions. I don’t think the figure below leads one to be very confident that those playing video games on Sundays now will be praying in churches at mid-century when they are in their late 50s. 

As shown below, Protestant churches have felt the brunt of these changes. A majority of Baby Boomers self-identify as Protestant or Christian (excluding Orthodox Christians), yet only about four in ten Millennials identify as such. This is a reality that would be unrecognizable to the Lost Generation—the youth of a century prior who were 76% Protestant. This is visible linear change.

However, something quite different is occurring among U.S. Catholics. Here, there is remarkable stability across generations (even growth compared to the oldest generations). Some believe this pattern can be completely explained by immigration. Yet as I’ve noted elsewhere the numbers for that argument just don’t add up as nicely as it is assumed and the reality we see in survey data shows a significant number of Catholics who may lose affiliation for a time returning later in life as the Gallup life-cycle model predicts. Immigration is important (as it has always been for Catholics in the U.S.), but the religious life-cycle may be as well.

So I think Gallup may be partially correct in its predictions. Catholicism seems to “stick” more than Protestant affiliations and this may be important in understanding the future of religious affiliation in the U.S. The Church has a higher retention rate of its youth than individual Protestant denominations and some who left revert back later in life (...the None retention rate is even lower. Not the “Lost Boys” after all...). Immigration is expected to continue to be important as well even as some of this has been on the decline since 2007 (...and the Catholic population percentage has remained steady).

I am less hesitant about making one prediction. I am 100% sure religion and God will still be alive and well in America, as Gallup argues, when The New York Times prints its final edition (1, 2, 3). Some institutions survive cultural changes better than others. Those who don’t see the changes coming probably weren’t paying attention to all of the relevant data.