[ If recent trends in religious switching continue, Christians
could make up less than half of the U.S. population within a few
decades]
[[link removed]]

MODELING THE FUTURE OF RELIGION IN AMERICA  
[[link removed]]


 

Stephanie Kramer, Conrad Hackett, Marcin Stonawski - Primary
Researchers
September 13, 2022
Pew Research Center
[[link removed]]


*
[[link removed]]
*
[[link removed]]
*
*
[[link removed]]

_ If recent trends in religious switching continue, Christians could
make up less than half of the U.S. population within a few decades _

, Nicholas Smith/Getty Images

 

Since the 1990s, large numbers of Americans have left Christianity to
join the growing ranks of U.S. adults who describe their religious
identity as atheist, agnostic or “nothing in particular.” This
accelerating trend is reshaping the U.S. religious landscape, leading
many people to wonder what the future of religion in America might
look like.

What if Christians keep leaving religion at the same rate observed in
recent years? What if the pace of religious switching continues to
accelerate? What if switching were to stop, but other demographic
trends – such as migration, births and deaths – were to continue
at current rates? To help answer such questions, Pew Research Center
has modeled several _hypothetical_ scenarios describing how the U.S.
religious landscape might change over the next half century.

The Center estimates that in 2020, about 64% of Americans, including
children, were Christian. People who are religiously unaffiliated,
sometimes called religious “nones,” accounted for 30% of the U.S.
population. Adherents of all other religions – including Jews,
Muslims, Hindus and Buddhists – totaled about 6%.1
[[link removed]]

Depending on whether religious switching continues at recent rates,
speeds up or stops entirely, the projections show Christians of all
ages shrinking from 64% to between a little more than half (54%) and
just above one-third (35%) of all Americans by 2070. Over that same
period, “nones” would rise from the current 30% to somewhere
between 34% and 52% of the U.S. population.

What is religious switching?

Switching, which in some cases could be described as religious
conversion, is defined in this report as a change between the religion
in which a person was raised (in childhood) and their present
religious identity (in adulthood).

Current rates of switching are based on responses from more than
15,000 adults to two questions posed in a 2019 Pew Research Center
survey: “In what religion, if any, were you raised?” and “What
is your present religion, if any?”

In many cases, switching does not happen in a single moment. Religious
“nones” often describe their disaffiliation as a gradual process,
and some may never have felt a strong connection to a religious
identity, even though they describe themselves as having been raised
in a faith tradition.

However, these are not the only possibilities, and they are not meant
as predictions of what _will _happen. Rather, this study presents
formal demographic projections of what _could_ happen under a few
illustrative scenarios based on trends revealed by decades of survey
data from Pew Research Center and the long-running General Social
Survey.

All the projections start from the current religious composition of
the U.S. population, taking account of religious differences by age
and sex. Then, they factor in birth rates and migration patterns. Most
importantly, they incorporate varying rates of religious switching –
movement into and out of broad categories of religious identity – to
model what the U.S. religious landscape would look
like _if_ switching stayed at its recent pace, continued to speed up
(as it has been doing since the 1990s), or suddenly halted.

Switching rates are based on patterns observed in recent decades,
through 2019. For example, we estimate that 31% of people raised
Christian become unaffiliated between ages 15 to 29, the tumultuous
period in which religious switching is concentrated.2
[[link removed]] An
additional 7% of people raised Christian become unaffiliated later in
life, after the age of 30.

While the scenarios in this report vary in the extent of religious
disaffiliation they project, they all show Christians continuing to
shrink as a share of the U.S. population, even under the
counterfactual assumption that all switching came to a complete stop
in 2020. At the same time, the unaffiliated are projected to grow
under all four scenarios.

Under each of the four scenarios, people of non-Christian religions
would grow to represent 12%-13% of the population – double their
present share. This consistency does not imply more certainty or
precision compared with projections for Christians and “nones.”
Rather, the growth of other religions is likely to hinge on the future
of migration (rather than religious switching), and migration patterns
are held constant across all four scenarios. (See Chapter 2
[[link removed]] for
an alternative scenario involving migration.)

Of course, it is possible that events outside the study’s model –
such as war, economic depression, climate crisis, changing immigration
patterns or religious innovations – could reverse current religious
switching trends, leading to a revival of Christianity in the United
States. But there are no current switching patterns in the U.S. that
can be factored into the mathematical models to project such a result.

None of these hypothetical scenarios is certain to unfold exactly as
modeled, but collectively they demonstrate how much impact switching
could have on the overall population’s religious composition within
a few decades. The four main scenarios, combined with four
alternatives outlined in Chapter 2
[[link removed]],
show that rates of religious switching in adulthood appear to have a
far greater impact on the overall religious composition of the United
States than other factors that can drive changes in affiliation over
time, such as fertility rates and intergenerational transmission
(i.e., how many parents pass their religion to their children).

The decline of Christianity and the rise of the “nones” may have
complex causes and far-reaching consequences for politics, family life
and civil society. However, theories about the root causes of
religious change and speculation about its societal impact are not the
focus of this report. The main contribution of this study is to
analyze recent trends and show how the U.S. religious landscape would
shift if they continued.

Why non-Christian religions are not projected individually

This report focuses on Christians and the religiously unaffiliated,
the two most common, very broad religious identities in the United
States today. People with all other religious affiliations are
combined into an umbrella category that includes Jews, Muslims,
Hindus, Buddhists and a diverse array of smaller groups that together
make up about 6% of the U.S. population. In 2015, Pew Research Center
projected
[[link removed]] the
growth of several of these groups separately, both in the U.S. and
around the world, and the Center may do so again in the future. But,
because data on religious switching and intergenerational transmission
is less reliable for groups with small sample sizes in surveys,
non-Christian groups are not shown separately in this report.

The report also does not project change for subgroups of Christians,
such as Protestants and Catholics, or for subgroups of “nones,”
such as atheists, agnostics and people who describe their religion as
“nothing in particular.” For the latest figures on the religious
composition of the U.S., including some subgroups, see our 2021
report, “About Three-in-Ten U.S. Adults Are Now Religiously
Unaffiliated
[[link removed]].”

Scenario assumptions and projection results

The four main scenarios presented here vary primarily in their
assumptions about the future of religious switching among Americans
between the ages of 15 and 29 – which are the years when most
religious change happens. Only a modest amount of switching is modeled
among older adults.

Fertility and mortality rates are held steady, as are rates of
intergenerational transmission. In each scenario, the groups begin
with their current profiles in terms of age and gender. Christians,
for example, are older than the religiously unaffiliated, on average,
and include a higher share of women.

Finally, the models assume that migration remains constant, which
helps explain why non-Christian groups follow the same trajectory in
each of the four scenarios. Immigration has an outsized effect on the
composition of non-Christian groups in the U.S. because adherents of
religions like Islam and Hinduism make up a larger share of new
arrivals than they do of the existing U.S. population.

Chapter 2
[[link removed]] presents
four additional scenarios that explore the impact of the factors held
constant here. These additional projections show how the U.S.
religious landscape might change if current switching patterns held
steady, but intergenerational religious transmission occurred in 100%
of cases; there were no fertility differences by religion; there was
no switching after age 30; or there was no migration after 2020.

The alternative scenarios are intended to help isolate – and thereby
illuminate – the impact of various factors. One might think of the
projections as an experiment in which some key drivers of religious
composition change are turned on or off, sped up or slowed down, to
see how much difference they make. For more information about modeling
assumptions and results, see Chapter 2
[[link removed]] and
the Methodology
[[link removed]].

Scenario 1: Steady switching – Christians would lose their majority
but would still be the largest U.S. religious group in 2070

SWITCHING ASSUMPTION: Switching into and out of Christianity, other
religions and the religiously unaffiliated category (“nones”)
continues among young Americans (ages 15 to 29) at the same rates as
in recent years. Most significantly, each new generation sees 31% of
people who were raised Christian become religiously unaffiliated by
the time they reach 30, while 21% of those who grew up with no
religion become Christian.

OUTCOME: If switching among young Americans continued at recent
rates, Christians would decline as a share of the population by a few
percentage points per decade, dipping below 50% by 2060. In 2070, 46%
of Americans would identify as Christian, making Christianity a
plurality – the most common religious identity – but no longer a
majority. In this scenario, the share of “nones” would not climb
above 41% by 2070.

Scenario 2: Rising disaffiliation with limits – ‘nones’ would be
the largest group in 2070 but not a majority

SWITCHING ASSUMPTION: Continuing a recent pattern, switching out of
Christianity becomes more common among young Americans as each
generation sees a progressively larger share of Christians leave
religion by the age of 30. However, brakes are applied to keep
Christian retention (the share of people raised as Christians who
remain Christian) from falling below about 50%.3
[[link removed]] At
the same time, switching into Christianity becomes less and less
common, also continuing recent trends.

OUTCOME: If the pace of switching before the age of 30 were to speed
up initially but then hold steady, Christians would lose their
majority status by 2050, when they would be 47% of the U.S. population
(versus 42% for the unaffiliated). In 2070, “nones” would
constitute a plurality of 48%, and Christians would account for 39% of
Americans.

Scenario 3: Rising disaffiliation without limits – ‘nones’ would
form a slim majority in 2070

SWITCHING ASSUMPTION: The share of Christians who disaffiliate by the
time they reach 30 continues to rise with each successive generation,
and rates of disaffiliation are allowed to continue rising even after
Christian retention drops below 50% (i.e., no limit is imposed). As in
Scenario 2, switching into Christianity among young Americans becomes
less and less common.

OUTCOME: If the pace of switching before the age of 30 were to speed
up throughout the projection period without any brakes, Christians
would no longer be a majority by 2045. By 2055, the unaffiliated would
make up the largest group (46%), ahead of Christians (43%). In 2070,
52% of Americans would be unaffiliated, while a little more than a
third (35%) would be Christian.

Scenario 4: No switching – Christians would retain their majority
through 2070

SWITCHING ASSUMPTION: This scenario imagines no person in America has
changed or will change their religion after 2020. But even in that
hypothetical situation, the religious makeup of the U.S. population
would continue to shift gradually, primarily as a result of Christians
being older than other groups, on average, and the unaffiliated being
younger, with a larger share of their population of childbearing age.

OUTCOME: If switching had stopped altogether in 2020, the share of
Christians would still decline by 10 percentage points over 50 years,
reaching 54% in 2070. The unaffiliated would remain a substantial
minority, at 34%.

Which scenario is most plausible?

The scenarios in this report present a wide range of assumptions and
outcomes. Readers may wonder which scenario is most plausible. While
there are endless possibilities that would lead to religious
composition change that is different from the plotted trajectories, it
may be helpful to consider how closely the hypothetical switching
scenarios adhere to real, observed trends. 

The “no switching” scenario (No. 4) is not realistic – switching
has not ended and there is no reason to think it will come to an
abrupt stop. The purpose of this scenario is to show the influence of
demographic factors (such as age and fertility) on religious
affiliation rates. Still, if fewer future young adults switch from
Christianity to no affiliation, or if movement in the opposite
direction increases, the future religious landscape might resemble the
results of this projection.

The “steady switching” scenario (No. 1) is conservative. It
depicts moderate, steady “net” switching (taking into
consideration some partially offsetting movement in both directions)
away from Christianity among young adults for the foreseeable future,
rather than the extension of a decades-long trend of increasing
disaffiliation across younger cohorts. Even long-standing trends can
be unsustainable or otherwise temporary, and this scenario best
represents what would happen if the recent period of rising attrition
from Christianity is winding down or already has ended.

By contrast, the scenario of rising disaffiliation without limits (No.
3) assumes there is a kind of ever-increasing momentum behind
religious switching. The visible rise of the unaffiliated might induce
more and more young people to leave Christianity and further increase
the “stickiness” of an unaffiliated upbringing, so that fewer and
fewer people raised without a religion would take on a religious
identity at a later point in their lives.

How we measured intergenerational transmission

Intergenerational transmission is the passing of religious identity
from parents to children. It occurs (or fails to occur) in childhood.
In this study, transmission rates are calculated based on the share of
children who inherit their _mother’s_ religion (or their
mother’s unaffiliated identity) because mothers tend to successfully
transmit their religious identities more often than fathers do.
Also, roughly a quarter of children under 18
[[link removed]] live
in single-parent households, which are overwhelmingly headed by
mothers.

The Center’s data shows the vast majority of teens (about 85%) have
the same religious identity as their mother, while 16% report a
different identity. Religious transmission, as measured in this study,
can fail to occur for many reasons and in either direction. For
example, if a mother doesn’t identify with any religion but her
14-year-old child identifies as Christian, it’s counted as
a _non-_transmission of religious identity – just as it would for a
Christian mother with a religiously unaffiliated teen.

Intergenerational transmission differs from switching because it
describes what happens before the age of 15 and is measured by
comparing the religious affiliation of mothers with the affiliations
reported by their teenage children. Switching, by contrast, describes
a change that happens after the age of 15; it is measured by comparing
the religions in which respondents say they were raised with the
affiliations they report today.

On the other hand, highly religious parents tend to raise highly
religious children who are less likely than children of less religious
parents to disaffiliate in young adulthood. As a result, there may
continue to be a self-perpetuating core of committed Christians who
retain their religion and raise new generations of Christians. It may
be useful to consider the experience of other countries in which data
on religious switching is available. In 79 other countries analyzed
(with a variety of religious compositions), most of the 30- to
49-year-olds who report that they were raised as Christians still
identify as Christian today; in other words, the Christian retention
rate in those countries has not been known to fall below about 50%.4
[[link removed]] The
“rising disaffiliation with limits” scenario (No. 2) best
illustrates what would happen if recent generational trends in the
U.S. continue, but only until they reach the boundary of what has been
observed around the world, including in Western Europe. Overall, this
scenario seems to most closely fit the patterns observed in recent
years.

None of the scenarios in this report demonstrate what would happen if
switching into Christianity increased. This is not because a religious
revival in the U.S. is impossible. New patterns of religious change
could emerge at any time. Armed conflicts, social movements, rising
authoritarianism, natural disasters or worsening economic conditions
are just a few of the circumstances that sometimes trigger sudden
social – and religious – upheavals. However, our projections are
not designed to model the consequences of dramatic events, which might
affect various facets of life as we know it, including religious
identity and practice. Instead, these projections describe the
potential consequences of dynamics currently shaping the religious
landscape.

Is switching only for the young?

Most people don’t change their religious identity. But among those
who do, the switch typically happens between the ages of 15 and 29.
That is why this report focuses on switching among young Americans.

However, since the rise of the “nones” began in the 1990s, a
pattern has emerged in which a measurable share of adults ages 30 to
65 also disaffiliate from Christianity. The Center’s analysis of
U.S. and international data indicates that modest levels of
disaffiliation among older adults could be a stage that
Christian-majority countries go through when Christian identity stops
being widely taken for granted – until about 30% of those raised
Christian already have shed Christian identity by the time they reach
30.

Today, among Americans who recently turned 30 and grew up Christian,
disaffiliation rates are already above 30%, so the projection models
assume that, on average, they will not switch religions again.
However, among groups of older adults born after World War II, we
model ongoing switching in which 7% of Americans who were raised
Christian will switch out between the ages of 30 and 65. This rate of
switching among older adults is held constant in each projection model
except the no-switching scenario, which does not include any switching
among older or younger adults. Switching by religiously unaffiliated,
older Americans into Christianity is not modeled in the projections
because there is no clear trend in this direction.

For more details on later adult switching, see the Methodology
[[link removed]] and Appendix
B
[[link removed]].

Religious change in context

These projections indicate the U.S. might be following the path taken
over the last 50 years by many countries in Western Europe that had
overwhelming Christian majorities in the middle of the 20th century
and no longer do. In Great Britain, for example, “nones” surpassed
Christians to become the largest group in 2009, according to
the British Social Attitudes Survey
[[link removed]].5
[[link removed]] In
the Netherlands, disaffiliation accelerated in the 1970s
[[link removed]],
and 47% of adults now say they are Christian. 

While the change in affiliation rates in the United States is largely
due to people voluntarily leaving religion behind, switching is not
the only driver of religious composition change worldwide. For
example, differences in fertility rates explain most of the
recent religious change in India
[[link removed]],
while migration has altered the religious composition of
many European countries
[[link removed]] in
the last century. Forced conversions, mass expulsions, wars and
genocides also have caused changes in religious composition throughout
history.

Moreover, the scenarios in this report are limited to
religious _identity_ and do not project how religious _beliefs and
practices_ might change in the coming decades.

Along with the decline in the percentage of U.S. adults who identify
as Christian in recent years, Pew Research Center surveys have found
declining shares of the population who say they pray daily or
consider religion very important in their lives
[[link removed]].
Still, it is an open question whether the Christian population in the
future will be more or less highly committed than U.S. Christians are
today.

On the one hand, within each generation, Christians with lower levels
of religious commitment may be most likely to shed their identity and
become religiously unaffiliated, while new converts may bring greater
zeal. These dynamics could lead to rising levels of commitment in the
remaining Christian population. On the other hand, religious
commitment could steadily weaken from generation to generation if
people continue to identify as Christian but are less devout than
their parents and grandparents. This dynamic could lead to steady or
declining levels of belief and practice.

Meanwhile, religiously unaffiliated Americans today are not uniformly
nonbelieving or nonpracticing. Many religious “nones” partake in
traditional religious practices despite their lack of religious
identity, including a solid majority who believe in some kind of
higher power or spiritual force. It is also unclear how this may
change in the future, and whether connections to these beliefs will
weaken if disaffiliation becomes even more common in the broader
society. At the same time, many observers have wondered what kinds of
spiritual practices, if any, may fill the void left by institutional
religion. We plan to continue exploring this question in future
research.

This report marks the first time Pew Research Center has projected
religious composition in the United States under multiple switching
scenarios, and the first time that differing rates of religious
transmission from parents to children have been taken into account.

These population projections were produced as part of the
Pew-Templeton Global Religious Futures project, which analyzes
religious change and its impact on societies around the world. Funding
for the Global Religious Futures project comes from The Pew Charitable
Trusts and the John Templeton Foundation.

1. How U.S. religious composition has changed in recent decades

Only a few decades ago, a Christian identity was so common among
Americans that it could almost be taken for granted. As recently as
the early 1990s, about 90% of U.S. adults identified as Christians.
But today, about two-thirds of adults are Christians.6
[[link removed]] The
change in America’s religious composition is largely the result of
large numbers of adults switching out of the religion in which they
were raised to become religiously unaffiliated.

In other words, a steadily shrinking share of young adults who were
raised Christian (in childhood) have retained their religious identity
in adulthood over the past 30 years. At the same time, having no
religious affiliation has become “stickier”: A declining
percentage of people raised without a religion have converted or taken
on a religion later in life.

While religious switching is the focus of this report, other
demographic forces that can cause religious change – transmission,
migration, fertility and mortality – will be briefly discussed in
the second half of this chapter.

Switching gained significant momentum in the 1990s, according to
the General Social Survey [[link removed]] (GSS) – a large,
nationally representative survey that has consistent data on religious
affiliation going back several decades. In 1972, when the GSS first
began asking Americans, “What is your religious preference?” 90%
identified as Christian and 5% were religiously unaffiliated. In the
next two decades, the share of “nones” crept up slowly, reaching
9% in 1993. But then disaffiliation started speeding up – in 1996,
the share of unaffiliated Americans jumped to 12%, and two years later
it was 14%. This growth has continued, and 29% of Americans now tell
the GSS they have “no religion.”7
[[link removed]]

Pew Research Center has been measuring religious identity since 2007
using a slightly different question wording – “What is your
present religion, if any?” – as well as a different set of
response options. Since 2007, the percentage of adults who say they
are atheist, agnostic or “nothing in particular” in the Center’s
surveys has grown from 16% to 29%. During this time, the share of U.S.
adults who identify as Christian has fallen from 78% to 63%.

There are many theories on why disaffiliation sped up so much in the
1990s and how long this trend might continue. For example, some
scholars contend that secularization is the result of increasing
“existential security” – as societal conditions improve and
scientific advances allow people to live longer lives with fewer
worries about meeting basic needs, they have less need for religion to
cope with insecurity (or so the theory goes).8
[[link removed]] Others
say that in the U.S., an association of Christianity with conservative
politics has driven many liberals
[[link removed]] away
from the faith. Still other theories involve declining trust in
religious institutions
[[link removed]], clergy
scandals
[[link removed]], rising
rates of religious intermarriage
[[link removed]],
smaller families, and so on. When asked, Americans give a wide range
of reasons for leaving religion behind
[[link removed]],
Pew Research Center has found.

Generational ‘snowball’

Whatever the deeper causes, religious disaffiliation in the U.S. is
being fueled by switching patterns that started “snowballing” from
generation to generation in the 1990s. The core population of
“nones” has an increasingly “sticky” identity as it rolls
forward, and it is gaining a lot more people than it is shedding, in a
dynamic that has a kind of demographic momentum.

Christians have experienced the opposite pattern. With each
generation, progressively fewer adults retain the Christian identity
they were raised with, which in turn means fewer parents are raising
their children in Christian households.

One way of gauging the momentum behind the U.S. switching trend is to
look at the other side of the coin – the rate at which Americans
retain the religion in which they were raised, as opposed to switching
out. By studying retention patterns, researchers can determine whether
a religious identity is becoming more or less sticky.

Until recently, Christian identity was stickier than unaffiliated
identity, which means that the share of people who remained Christian
after being raised as Christians was greater than the share of people
who remained unaffiliated after being raised with no religion.

Today, Christianity still is the stickier affiliation for older
Americans. But among younger adults, the unaffiliated identity has
become the stickier one. Among people who are 40 and older, 80% of
those raised as Christians are still Christian today, compared with
just 56% of those who were raised unaffiliated (in childhood) and
still do not identify with a religion today (in adulthood). However,
among people in their 30s, only 66% of those raised Christian are
still Christian today, compared with 73% of those raised unaffiliated
who still are today.

An analysis of GSS data by birth decade shows a similar pattern:
Roughly 90% of people who were born in the 1960s and raised Christian
were still Christian when they turned 30. Among those born in the
1970s, fewer than 85% remained Christian at 30. Among those born in
the 1980s, it is about 80%. Too few of those born in the 1990s have
turned 30 to estimate their switching patterns, but Christians in this
youngest cohort appear to be disaffiliating even more than older
cohorts.

Disaffiliation among older adults

The “snowballing” dynamic is being driven by an acceleration in
switching among young Christians – those ages 15 to 29. People under
30 tend to grapple with identities of all kinds, and young adulthood
is often a time of major change, when many people leave their
parents’ household, start careers and form lasting romantic
partnerships.

But there is a second dynamic that began in the 1990s that added a new
layer of change:
Starting in the mid-1990s, it became more common for adults in middle
age and beyond to discard Christian identity. Before that, changing
religions after 30 was rare.

About 95% or more of people who were born prior to the 1940s and were
raised Christian were still Christian from ages 30 to 65. But among
those born in the ’40s, ’50s and ’60s, there has been more
substantial movement away from Christianity after age 30. For example,
91% of Americans born in the 1960s were still Christian at age 30, but
83% identify as Christian today. Recent switching among older U.S.
adults may be the result of a period effect (when something about the
environment affects people of all ages for a period of time, such as
the COVID-19 pandemic’s
[[link removed]] consequences
for the mental and physical health of people of all ages). It might
also be the result of a tipping point: Once Christians began to lose
their overwhelming majority, people of all ages who had ties to
Christianity – but did not attend church, pray often or see religion
as an important part of their lives – may have begun to identify as
unaffiliated in larger numbers. As “nones” grew in size and
visibility, being unaffiliated may have become more socially
acceptable
[[link removed]] in
some circles, opening the floodgates to further disaffiliation. 

While this pattern is new – and it is unclear how long it might last
– it indicates that disaffiliation is extending into segments of the
population that may have been unaffected in the past. (For more
information about late-adult switching, see Appendix B
[[link removed]].)

Education, politics and geography tied to differences in religious
switching

A closer look at the characteristics of adults who have left
Christianity and are now religiously unaffiliated indicates that other
traits – such as age, gender, education, political identity and
region of residence – also are tied to disaffiliation.

Age

U.S. adults who have moved away from Christianity are younger, on
average, than those who have remained Christian after a Christian
upbringing. More than a quarter of former Christians (27%) are under
30, compared with 14% of all adults who were raised Christian and
remain Christian. This age pattern aligns with a decades-long trend in
which each cohort of young adults is less religiously affiliated than
the preceding one.

Gender

Americans who have moved away from Christianity are more likely to be
men, while women are more likely to retain their Christian identity. A
slight majority of U.S. adults who were raised Christian and are now
unaffiliated (54%) are male. Among people who have remained Christian,
57% are women.

Education

People who have become unaffiliated after a Christian upbringing are a
little more likely to have graduated college than those who remain
Christian, with 35% and 31%, respectively, holding college degrees.
This reflects a broader pattern: In the U.S., people with higher
levels of educational attainment tend to be less religious by some
traditional measures, such as how often they pray or attend religious
services
[[link removed]].

Politics

Seven-in-ten adults who were raised Christian but are now unaffiliated
are Democrats or Democratic-leaning independents, compared with 43% of
those who remained Christian and 51% of U.S. adults overall. Some
scholars argue that disaffiliation from Christianity is driven by an
association between Christianity and political conservatism that has
intensified in recent decades.9
[[link removed]]

Geography

People who have left Christianity are underrepresented in the South,
where 33% of former Christians live, compared with 42% of people who
have remained Christian and 38% of U.S. adults overall.10
[[link removed]]

Those who have disaffiliated after being raised Christian are more
likely than others to live in the West (28% live there, compared with
20% of those who remain Christian and 23% of all U.S. adults). Surveys
often find that U.S. adults tend to be more religious, on a number of
measures, in the South
[[link removed]],
and less so in the West and Northeast. This may indicate that people
adapt to the religious contexts in which they live and/or sort
themselves into like-minded communities.

Other drivers of change

Switching is the primary, but by no means the only, process causing
religious change in the U.S. Populations can grow or shrink through a
few other mechanisms. Patterns of religious transmission, migration
and fertility explain some of the shift in the religious landscape in
recent decades.

Transmission

The share of Christians is in decline partly because religion is not
always transmitted by Christian parents to their children.

For the purposes of the projections in this report, religious
identities are considered to be “transmitted” when children are
raised in their parents’ religion and identify with it as early
adolescents. There are a variety of reasons why children of
religiously affiliated parents may be raised without a religion and,
therefore, that religion is not transmitted. For example, a child may
have parents without strong religious commitment, or parents with
different religions, or parents who have decided to let children
explore and make decisions about religion on their own.

Consider the hypothetical case of an adult survey respondent who says
her mother was Christian, her father was Jewish, she was not raised in
any religion, and she currently does not identify with any religion. A
person like this has not _switched _religions, since switching is
defined as leaving the religion in which one was raised. However, in
this example, neither parent _transmitted_ their religion.

By the same token, not all unaffiliated parents transmit their
identity. For example, a 14-year-old child of unaffiliated parents
could acquire a Christian identity outside the parental home in
various ways, such as from other family members, a teacher or a
friend.

In this study, transmission rates are calculated based on the
mother’s religion because mothers tend to successfully transmit
their religion more often than fathers do, and roughly a quarter of
teens live in single-parent households, which are almost exclusively
headed by mothers.

Today, transmission of the mother’s religious identity happens in
the vast majority of families. In a 2019 Pew Research Center survey
of teens and their parents
[[link removed]],
an overwhelming majority of both Christian and unaffiliated mothers
had transmitted their religious identities to their teenagers. More
than eight-in-ten Christian mothers had Christian teens, while 17% of
their teens identified as atheist, agnostic or “nothing in
particular,” and less than 1% said they were members of another
religious group.

The teens of unaffiliated mothers show a similar pattern: 88% are
unaffiliated themselves, 11% are Christians, and 1% identify with a
non-Christian religion. The survey sample did not contain enough
mothers who belong to non-Christian religions to report on their
precise transmission rates, but their patterns seem broadly similar to
those with Christian and unaffiliated mothers – the vast majority of
teens raised by mothers of “other religions” also identify with a
religion in this category.

Even though the shares of Christian and religiously unaffiliated
mothers who transmit their affiliation (or lack thereof) are fairly
similar, the impact of failed transmission in Christian families is
far greater, numerically, because there are more than twice as many
Christian mothers as unaffiliated ones. At these rates, and as long as
Christians are the substantially larger group, many more people will
adopt a religiously unaffiliated identity rather than a Christian one
during childhood, which in turn increases the population share of the
unaffiliated.

Migration

Migration contributes to U.S. religious change because the composition
of immigrants and emigrants is not identical to that of the overall
U.S. population.

About a million immigrants come to the U.S. each year,
and one-in-seven people in the U.S. were born elsewhere
[[link removed]].
In the 1990s and early 2000s
[[link removed]],
the largest number of recent arrivals to the U.S. were from Mexico and
other Christian-majority countries in Central and South America.

Today, new arrivals are more likely to come from Asia. In 2018,
the top country of origin
[[link removed]] for
new immigrants was China (which is majority unaffiliated), followed by
India (which is majority Hindu). Most of the world’s people who
identify as religiously unaffiliated, Hindu, Buddhist, Sikh and Jain
live in either China or India, and this is reflected in the changing
profile of immigrants.

Christians still make up a majority of immigrants to the U.S.,
including a majority of immigrants from Mexico, the third-largest
source of new immigrants in recent years. But the estimated share of
new immigrants who are Christian (55%) is lower than the Christian
share of the existing U.S population (64%), meaning that immigration
is not boosting the Christian population share. The same is true of
religiously unaffiliated people: 12% of new immigrants are estimated
to be religiously unaffiliated, compared with 30% of the existing U.S.
population.

But immigration is leading to growth in the share of other religions
like Hindus and Muslims – 32% of new immigrants are estimated to be
adherents of other religions (versus 6% of the U.S. population),
according to recent data on the origin and size of migrant flows to
the U.S. and an earlier Pew Research Center analysis
[[link removed]] of
the typical religious composition of migrants from each country.11
[[link removed]]

Fertility

In countries with wide differences in fertility rates between
religious groups, those differences can cause significant changes in
religious composition over time.

Recently, religiously unaffiliated women in the U.S. have tended to
have fewer children than Christians and women of other religions. In
this report’s models, the average unaffiliated woman is expected to
have 1.6 children in her lifetime, while the average Christian woman
will have 1.9 children, and the average woman of other religions (an
umbrella category that includes Jews, Muslims, Hindus, Buddhists and
many smaller groups) will have 2.0 children (see Methodology
[[link removed]] for
more details).

Since the U.S. has a very large population and mothers tend to
transmit their religions to children, these small differences can add
up to noticeable changes over time. However, higher fertility among
Christians compared with the religiously unaffiliated has not been
nearly enough to maintain the Christian share of the population,
although it has slightly offset some of the impact of disaffiliation.

Age structures and mortality

The youthfulness of religious groups has an impact on the future that
is intertwined with fertility because young populations have higher
shares of people who are in, or soon will enter, their reproductive
years. In other words, they have more growth potential than older
populations. If two groups have identical total fertility rates, the
group with the younger age structure can grow more rapidly because of
the population momentum produced by having a larger share of women of
reproductive age. Young populations also tend to have a smaller share
of people who die each year.

Christians are older, on average, than the unaffiliated or people of
non-Christian religions. The average U.S. Christian is 43, compared
with an average age of 33 among the unaffiliated and 38 among people
of other religions. More than 80% of Americans older than 75 are
Christian, compared with roughly half of people in their prime
childbearing years (ages 20 to 34), many of whom will transmit their
religion to the next generation, if past patterns hold. More than 40%
of Americans between 20 and 34 are religiously unaffiliated, compared
with under 15% of the oldest Americans. These are among the reasons
why religious “nones” are projected to grow as a share of the U.S.
population even in the scenario with no further religious switching.

Due to a lack of sufficient data on mortality differences between
people in the three religious identity categories studied in this
report, each group is assumed to have the same mortality patterns. In
other words, for purposes of these projections, life expectancy is
assumed to be similar among members of each group at a given age. It
is also assumed to be rising over time, despite a dip caused by the
coronavirus pandemic.12
[[link removed]]

2. Projecting U.S. religious groups’ population shares by 2070

The first half of this chapter provides details on the assumptions and
results of each of the four main scenarios. These are not predictions
for the future. Rather, projections show what would happen under a
number of hypothetical scenarios.

Some scenarios are intentionally implausible and meant only to
illustrate the impact of different demographic forces. Scenario 1 is
conservative, because it assumes that disaffiliation will not speed up
beyond the current rate, even though in recent decades each generation
of people raised Christian has disaffiliated more than the generation
before it. Scenarios 2 and 3 may be more realistic because they assume
that switching will continue to accelerate. Scenario 4 is implausible
because it imagines that all switching ended in 2020; it is
nevertheless revealing because it demonstrates how demographic
dynamics (such as the higher average age of Christians) would cause
the Christian share of the population to continue to decline even
without any further switching.

Though some scenarios are more plausible than others, the future is
uncertain, and it is possible for the religious composition of the
United States in 2070 to fall outside the ranges projected.

The second half of this chapter presents four additional projections
(Scenarios 5-8), to demonstrate the effects of factors other than
switching. These projections show how the U.S. religious landscape
might change if current switching patterns among young adults held
steady, but with religious transmission set to 100%, no fertility
differences by religion, no switching after age 30, or no migration.

None of the scenarios project growth in the Christian share of the
U.S. population because we do not have empirical measures of any
recent switching patterns that favor Christianity in the United
States. In other words, there is no data on which to model a sudden or
gradual revival of Christianity (or of religion in general) in the
U.S. That does not mean a religious revival is impossible. It means
there is no demographic basis on which to project one.13
[[link removed]]

Baseline assumptions

All the scenarios in this report start with a 2020 data baseline.
Estimates of U.S. migration, fertility, age and sex structure, and
mortality patterns are based on the United Nations’ demographic
estimates. Religious differentials in fertility are from the National
Survey of Family Growth. Baseline religious composition data comes
from Pew Research Center surveys. (For a full list of sources and
explanations on how baseline data was prepared for analysis,
see Methodology
[[link removed]].)

Pew Research Center estimates that in 2020, Christians made up 64% of
the U.S. population (including children) while “nones” accounted
for 30% and other religious groups 6%. Based on observed data, the
baseline scenario assumes that 34% of people who grew up Christian
will discard their religion by the time they turn 30 (including 3% who
switch to a non-Christian religion and 31% who identify with no
religion) and among some older cohorts of Christians, 7% will leave at
a later age.14
[[link removed]] Meanwhile,
21% of people who were raised religiously unaffiliated will become
Christian by the time they reach 30.

Each scenario assigns different future switching rates to young people
(ages 15 to 29) but holds switching rates steady for the small share
of older adults who switch after turning 30. (This later-adulthood
switching applies only to cohorts born before 1990; disaffiliation has
become so common in subsequent decades that the models assume that
people born after 1990 who will switch already have done so by age
30.15
[[link removed]])

All four main scenarios assume that transmission of religion from
mothers to children continues at recent rates, migration remains
constant, religious differences in fertility stay stable, and there
are no mortality differences among religious groups. Switching among
people raised in non-Christian religions is assumed to hold steady
under all but the “no switching” scenario, because there is not
enough data on people in the “other religions” category to model
shifting retention rates across age cohorts.

The population structure for 2020 differs from earlier projections

Readers familiar with Pew Research Center’s 2015 global projections
[[link removed]] might
note that the 2020 U.S. religious landscape described in this report
differs markedly from what was projected seven years ago. In fact, the
present religious landscape is similar to the 2015 projection for
2050
[[link removed]],
with Christians representing about two-thirds of the population and
the religiously unaffiliated making up more than a quarter. The 2015
global projections assumed stability in switching patterns in each
country and did not anticipate an acceleration in the U.S. switching
rate. The pace of disaffiliation in the U.S. increased continuously
between 2010 and 2020. The “steady switching” projection scenario
in this report is most similar to the assumptions modeled in the
earlier projections. Other scenarios in this report attempt to capture
what would happen if U.S. religious switching continues to speed up.

In our 2015 global projections report, it would have been impractical
to present customized scenarios that might be appropriate for
individual countries (or regions), such as the additional U.S.
switching scenarios included in this report. For the sake of
feasibility and comparability, a steady switching assumption was
applied to every country for which switching data was available, even
though steady switching was not the likeliest path for all countries.
At the time, we recognized that this approach was likely conservative
for projecting the growth of “nones” in the United States. In this
new report, we delve into the complex switching dynamics in the U.S.
with greater specificity.

Updates to projections are appropriate when facts on the ground
change. For example, soon after our 2015 global projections report was
released, a large wave of asylum seekers came to Europe, resulting in
a rapid increase in the region’s Muslim population that had not been
anticipated in our models. In 2017, we updated our European
projections in a report exploring how Europe’s Muslim population
could continue growing
[[link removed]] under
a variety of new migration scenarios.

Scenario 1: Steady switching

The first scenario differs from the others in that it assumes
religious switching will continue at recent rates across all age
groups. That is, in each new generation, 31% of people who were raised
Christian become religiously unaffiliated between the ages of 15 and
29, while 21% of those who grew up with no religion become Christian.
Moreover, 7% of people who were raised Christian disaffiliate between
the ages of 30 and 65.16
[[link removed]]

Other demographic forces that can affect religious composition –
migration patterns, differences in religious groups’ birth rates,
and the rate at which parents transmit religious identity to their
children – are held steady. (These factors remain constant in each
of the four main scenarios, but not in Scenarios 5-8, which begin
later in this chapter.)

In these circumstances, the largest amount of change among Christians
and the religiously unaffiliated would occur by 2050. Christians would
decline as a share of the population by a few percentage points per
decade, dipping below 50% by 2060. In 2070, 46% of Americans would
identify as Christian, making Christians a plurality – the most
common religious identity – but no longer a majority. In this
scenario, the share of “nones” would reach 41%, and other groups
would make up the remaining 13%.

(Scenario outcomes for non-Christian religions are closer than they
may appear due to rounding. For example, people of other religions are
projected to make up 12.53% of the U.S. population in 2070 under
Scenario 1 and 12.48% under Scenario 2.)

Scenario 2: Rising disaffiliation with limits

In this scenario, leaving Christianity is assumed to become more
common across each successive cohort of young adults, continuing
recent trends. However, this scenario also assumes that Christian
retention – that is, the share of people who were raised Christian
and still identify as Christian after young adulthood – can only go
as low as 50% for men and 55% for women.

This artificial “floor” is roughly equal to the lowest retention
rate observed in an analysis of 79 other countries
[[link removed]].
Great Britain has a Christian retention rate of 49%, the lowest in
this analysis, followed by France at 52%. The “limits” in this
scenario only assume there may be a cap on how rapidly change can take
place within a generation for people raised Christian or raised with
no religion. No predetermined limit is imposed on the eventual size of
populations of Christians, the religiously unaffiliated or of all
other religions.

The scenario also assumes that at least 5% of people raised without a
religion will convert into one, either to Christianity or another
faith. Having no religious affiliation is very quickly becoming
“stickier,” but this assumption recognizes that retention among
the unaffiliated is unlikely ever to reach 100%. 

With these assumptions, the trend in which each cohort of young adults
disaffiliates from Christianity at higher rates than the preceding
cohort continues and reaches an imposed “floor” retention rate of
53% overall (50% for men and 55% for women) in 2050, among the cohort
born between 2016 and 2020.

Under these conditions, people who do not identify with any religion
would become the largest group around the year 2060, though they would
not represent a majority of Americans. By 2070, “nones” would be a
plurality of 48%, Christians would account for 39% of the U.S.
population, and 12% of Americans would belong to other religions.

Scenario 3: Rising disaffiliation without limits

Since the 1990s, each cohort of young adults has disaffiliated from
Christianity at higher rates than the one before it. At the same time,
steadily fewer people have become affiliated with any religion after
growing up with no religion. If this trend of change between cohorts
continues, growing shares will continue to become unaffiliated during
young adulthood, and declining shares will offset this attrition
through conversion into a religion.

In this scenario, the share of Christians who disaffiliate by the time
they reach 30 continues to rise with each successive generation and is
allowed to grow without any imposed limit.

If the rate of disaffiliation among young adults continues to increase
unabated, there would be a steep, rapid drop in the share of
Christians who retain the identity they were raised with. The
unaffiliated would surpass Christians as the largest group by 2055,
when 43% of Americans would be Christian, compared with 46% who would
be religiously unaffiliated. By 2070, a slim majority of Americans
(52%) would be unaffiliated, while a little over a third (35%) would
be Christian.

By 2070, this scenario projects that the pace of disaffiliation would
have sped up so that 65% of Americans who were raised Christian would
switch out before the age of 30 (up from 31% in 2020), mostly in the
direction of the unaffiliated. Christian retention – the share of
people raised Christian who are still Christian – would fall to 35%.
Under these conditions, the share of young adults switching from an
unaffiliated upbringing to a Christian adulthood would fall to 2% by
2070, compared with the recent rate of 21%.

These assumptions result in the largest shift toward disaffiliation of
any projection scenario. This scenario projects an unaffiliated
plurality in 30 years and is the only one to result in an unaffiliated
majority in 2070.

Scenario 4: No switching

For the sake of demonstrating the impact of switching on religious
change, this scenario assumes that all religious switching stopped in
2020. If all other factors of change – including intergenerational
transmission, migration and fertility – continued steadily but there
was no further movement into and out of any religion, Christians would
still make up a 54% majority of Americans in 2070. This is the only
scenario modeled under which Christians maintain a majority over the
next 50 years.

But the share of Christians would still decline, while the religiously
unaffiliated would continue to edge up (from 30% to 34%). This is
largely because Christians are older than the other groups, on
average. About half of Americans in their early 20s are Christian,
compared with more than three-quarters of those in their early 60s and
even greater shares of older adults. As older people die, and as
younger, unaffiliated people become parents to unaffiliated children,
the Christian share of the population would naturally fall due to
Christian deaths outnumbering Christian births.

Other religions would remain on their trajectory to reach 12% by 2070,
mainly because immigration is still projected to continue in this
scenario.

Why the flow of people out of Christianity may eventually lose
momentum

There are many different ways of measuring the momentum of switching
into and out of Christianity. This analysis focuses on switching among
people under 30, because that is when most religious switching
happens, and also because it helps researchers quantify the trend
among young people to project forward.

Another helpful way of measuring switching momentum is by looking at
all U.S. adults who have left or joined Christianity, regardless of
when they did so. This wider lens amplifies an interesting fact about
the way demographic change works: When a majority group starts losing
members to a minority group, even small _percentages_ of people
leaving the majority group can have a large impact on the _total
numbers_ in the minority group. However, as the groups become more
similar in size, the net impact of roughly equal percentages of people
switching from each group to the other diminishes. Accordingly, the
change in the projected sizes of these populations becomes more
gradual over time in most projection scenarios, as the groups converge
in size.

Let’s take switching rates for all U.S. adults as an example. In
2019, among all U.S. adults, 23% of all people who had been raised
Christian had become unaffiliated, while 27% of those who had been
raised without a religion had become Christian.

Though these percentages are similar (and may even seem to favor
movement into Christianity) they represent vastly different numbers of
people and a much larger gain for the unaffiliated. Only 12% of adults
– approximately 30 million – were raised unaffiliated, and 83% –
or roughly 215 million – were raised Christian. This means that by
2019, about 50 million adults (23% of 215 million) had discarded a
Christian identity, and fewer than 8 million (27% of 30 million) had
become Christian after an unaffiliated upbringing. Even if 100% of
adults who were raised unaffiliated (all 30 million) had become
Christians, the unaffiliated category still would have gained more
members than it lost.

The same dynamic applies to the transmission of religion from mothers
to children. In 2019, 17% of Christian mothers were raising children
who – by the time they reached the teenage years – were not
affiliated with any religious group. A similar share of unaffiliated
mothers (11%) had teens who identified as Christians. The consequences
of this 6-point gap are exaggerated because there are so many more
Christian mothers than unaffiliated mothers. The difference in
the _number_ of children moving into each group would be much
smaller if the two groups were of comparable size.

As the unaffiliated grow to represent a share of the population that
is similar to the Christian share, relatively modest shifts in
transmission and retention rates could reverse the groups’
trajectories. For example, among all U.S. adults, Christianity has
attracted a greater share of people raised unaffiliated than vice
versa (27% compared with 23%). The pattern is opposite among young
adults, for whom being raised unaffiliated is “stickier” than
being raised Christian. However, a shift back to the pattern currently
observed among all adults (with higher retention rates among
Christians than among the unaffiliated) could be enough – depending
on other patterns – to begin growing the Christian population from a
new starting point at which Christians are smaller in number than, or
similar in size to, the unaffiliated. While this bottoming out and
regrowth of Christianity is theoretically possible, it would require a
reversal of the current trends in switching.

PROJECTING TO 2100
If trends are projected for an even longer period, change slows under
most scenarios. Even in the most extreme switching scenario, in which
each cohort of young adults disaffiliates more than the one before it,
with no floor imposed for Christian retention, Christians would still
represent about a quarter of the population in 2100. Under other
scenarios, the rate of growth of the religiously unaffiliated (and
decline of Christians) is curbed by 2080. This is due to switching
dynamics. If the Christian and unaffiliated populations become similar
in size – an eventuality under most scenarios – and if the gap
between their retention rates remains small, then the growth of the
unaffiliated eventually would slow, and the religious groups could
reach equilibrium rather than one group ascending completely and the
other disappearing.

Additional scenarios: What if migration stops or people stop switching
after the age of 30?

This report focuses on four main scenarios that explore how the U.S.
religious landscape might change if switching out of Christianity
among young adults were to speed up, keep a steady pace or stop
entirely. These scenarios seek to explore the effect of religious
switching in late adolescence and early adulthood, and they hold
steady other demographic forces that can cause a country’s religious
composition to change – namely intergenerational “transmission”
(the passing of religious identity from parents to children),
switching later on in adulthood, migration, fertility and mortality.

But how impactful are these other factors, and how could variations in
them affect the future of religion in America? To measure the relative
impact of some of the assumptions built into the model, researchers
created four more scenarios that turn off one mechanism of change at a
time. Otherwise, these four scenarios are identical to Scenario 1, in
that they assume a steady rate of religious switching among young
adults, without any acceleration. (This does not necessarily mean that
Scenario 1 is the most plausible.) Turning off one mechanism of change
at a time is the best way to assess whether it has any meaningful sway
on the overall outcome.

The results of this statistical exercise produce small deviations from
Scenario 1, on which they all are based. In other words, even if we
vary our assumptions about migration – or fertility rates, or
transmission rates, or the future rate of switching among older adults
– the projections would be similar to those from Scenario 1.

Turning off any one of these mechanisms of demographic change results
in a projected Christian share ranging from 46% to 48% in 2070. The
religiously unaffiliated share would rise to between 39% and 45%,
depending on which component stopped. And people of all other
religious groups would be projected to make up between 8% and 13% of
the U.S. population in 2070.

The scenarios in this report demonstrate that switching is the driving
force behind religious change in the U.S. today. Scenario 4
illustrates that age structure is also consequential – Christians
are expected to shrink and the unaffiliated to grow in part because
Christians are older and the unaffiliated are younger, on average.
However, Scenarios 5-8 demonstrate that other demographic factors are
expected to have less impact on the overall direction and pace of
change.

Transmission is the process of children taking on the same religious
identity as their parents. Today, about 85% of teens with a religious
identity classified as Christian, unaffiliated or other religion have
a mother whose identity is classified in the same group, according to
analysis of surveys that measure the religious identities of both
teens and parents in the same household. In projections, transmission
is based on the share of _mothers_ who successfully transmit their
religion. Fathers matter, too, but mothers tend to be more successful
at transmission, and about a quarter of U.S. children live with
single parents
[[link removed]],
who are overwhelmingly mothers. 

If rates of religious transmission between mothers and children
increased to 100% – in other words, if every mother transmitted her
religious identity to every child she has – the unaffiliated would
grow slightly less, to 39% of the U.S. population by the end of the
projection period (compared with 41% under Scenario 1). This indicates
that over the long term, it makes little difference whether all teens
or just 85% inherit their mother’s faith (or lack thereof).

Scenario 6: What if religious groups had children at identical rates?

Over a 50-year period, the impact of fertility is also small. Scenario
6 is identical to the “steady switching” scenario except that
differences in fertility rates across religious groups are turned off.
If there were no differences in fertility among Christians, the
religiously unaffiliated and people of other religions, Christians
would be projected to shrink to 46% of the U.S. population in 2070
(the same as in Scenario 1), while the unaffiliated would grow by 1
percentage point more. The percentage estimate for people of other
religions would be 1 point lower without fertility differences, as
long as all other conditions mirrored Scenario 1.

Scenario 7: What if all immigration and emigration ceased?

People of non-Christian faiths make up a larger proportion of recent
immigrants than they do of the overall population, and their numbers
would be most affected in a “no migration” scenario. If no
migrants entered or left the U.S. after 2020, people of other (i.e.,
non-Christian) religions would grow to represent only 8% of the
population by 2070, rather than the 12% or 13% projected under
scenarios accounting for steady migration. The unaffiliated would make
up a slightly larger share of the population (45%) compared with
Scenario 1 (41%). The 2070 Christian share of the U.S. population is
similar in Scenarios 1 and 7, suggesting that continuation of the
current migration patterns would have a relatively small long-term
impact on the size of the Christian population, though immigrants are
adding substantially to the population that identifies with other
religions.

Scenario 8: What if older Christians stopped switching?

Most of the scenarios assume that about 7% of adults in some older
cohorts who were raised Christian leave Christianity between the ages
of 30 and 65, as they have since the 1990s. As noted above, however,
there are reasons to believe that this pattern of religious switching
in older adulthood is the result of a period effect and may not
continue forever.17
[[link removed]] Under
a “no switching after 30” scenario, future religious identity
change occurs only in young adulthood, steadily at recently observed
rates. The end result of Scenario 8 is similar to Scenario 1: 47% of
the U.S. population would be Christian in 2070 (versus 46% in Scenario
1), while the share of nones (41%) and people with other religions
(13%) is the same in both scenarios. Without ongoing defection from
Christianity among adults over 30, the Christian population would only
be slightly larger.

Acknowledgments

This report was produced by Pew Research Center as part of the
Pew-Templeton Global Religious Futures project, which analyzes
religious change and its impact on societies around the world. Funding
for the Global Religious Futures project comes from The Pew Charitable
Trusts and the John Templeton Foundation (grant 61640).

This report is a collaborative effort based on the input and analysis
of the following individuals. Find related reports online
at pewresearch.org/religion
[[link removed]].

PRIMARY RESEARCHERS

Stephanie Kramer, _Senior Researcher        _
Conrad Hackett, _Associate Director of Research and Senior
Demographer_
Marcin Stonawski, _Director, Center for Advanced Studies of
Population and Religion (CASPAR)_

RESEARCH TEAM_                    _

Anne Fengyan Shi, _Senior Researcher_
Yunping Tong, _Research Associate   _
Gregory A. Smith, _Associate Director of Research   _
Justin Nortey, _Research Analyst      _
Joshua Alvarado, _Research Assistant          _
Alan Cooperman, _Director of Religion Research     _

EDITORIAL AND GRAPHIC DESIGN

Dalia Fahmy, _Senior Writer and
Editor                   _
Michael Lipka, _Editorial
Manager                           _
Rebecca Leppert, _Editorial Assistant          _
Bill Webster, _Senior Information Graphics Designer_

COMMUNICATIONS AND WEB PUBLISHING

Stacy Rosenberg, _Associate Director, Digital_
Reem Nadeem, _Associate Digital Producer            _
Anna Schiller, _Senior Communications Manager_
Kelsey Beveridge, _Communications Associate        _

Others at Pew Research Center who provided guidance include Andrew
Mercer, senior research methodologist; Phillip Connor, former senior
researcher; and Jacob Ausubel, former research assistant.

Methodology

This study projects the future population sizes of Christians,
religious “nones” and people of other religions in the United
States. Since recent religious change in the U.S. has been driven
primarily by voluntary changes in religious identity – religious
switching – we modeled for the first time how the religious
landscape could change in scenarios with different patterns of future
switching, as well as additional scenarios quantifying the relative
impact of other dynamics, including migration, fertility and
transmission of religion from parent to child. Methodologically, the
study builds on the multistate cohort component methods used in
previous Pew Research Center studies projecting the future of world
religions
[[link removed]],
as well as how migration may shape Europe’s growing Muslim
population
[[link removed]].

Previous religious population projections by Pew Research Center
assumed that all children inherit their mother’s religion (or lack
of religion). We were aware that this is an oversimplification of the
process of intergenerational religious transmission, but in most
countries, we did not have input data on which to model more nuanced
transmission dynamics. We do not know of any prior population
projections that have attempted to model anything aside from perfect
transmission of religious identity from parent to child. This study
uses Pew Research Center data about levels of religious transmission
between parents and their children to model the potential breakdown of
intergenerational religious transmission.

While religious switching is typically concentrated among young
adults, Center researchers have found that in recent decades, there
also has been a modest amount of switching among adults over age 30.
In a theoretical innovation, this study posits that later adult
switching may be a part of the process whereby societies in which 90%
or more of adults once identified as Christian transition to a society
with significantly lower Christian retention rates. In a
methodological innovation, we model this phenomenon of later adult
switching continuing for cohorts that still had levels of Christian
identification over 70% at age 30.

This Methodology section provides details on the input data and
methods used in the projections. The first section explains how the
baseline (2020) religious composition estimates were derived. The
second section describes how key input data (age and sex composition,
fertility, mortality, migration, transmission and religious switching)
were gathered. The third part of this Methodology details the
projection methods. Appendix A
[[link removed]] describes
the survey sources used in this report. And Appendix B
[[link removed]] presents
U.S. religion trends from a range of surveys and discusses analysis of
international retention rates that influenced projection assumptions
in this report.

Input data

This is an overview of input data used for the baseline population and
projections in this report. In the projections, results for men and
women are modeled separately because men and women vary in religious
composition, switching rates, rates of inheriting their mother’s
religion (intergenerational transmission) and life expectancy. In the
report, these results are aggregated to report on men and women
together. We do not have enough data on people who say their gender is
different from their sex assigned at birth to separately project
religion trajectories for people who are transgender or nonbinary.

Baseline structure

The baseline religious composition of the U.S. by age and sex is based
on data from the American Trends Panel
[[link removed]] (ATP)
and Pew Research Center’s survey of teens (for the age group
including 15- to 19-year-olds). The religious composition of children
ages 0 to 14 is estimated based on the age structure of women in each
broad religious category, fertility patterns by religion, and the
assumption that children generally inherit their mother’s religion.
(In this step, we assume rates of religious inheritance similar to
those used in our projection modeling, as described later in this
Methodology.)

Religious composition is estimated for each five-year age and sex
group, such as women ages 25 to 29 and men ages 60 to 64. For age
groups older than 80, we have a modest number of respondents. The
religious composition of these older age groups was estimated
indirectly using a decomposition procedure, which involved
extrapolating trends observed in younger cohorts to older age groups.

The religious compositions of each adult five-year age and sex group
were smoothed and raked to match the overall religious composition
from the ATP. This procedure minimizes jumps between age groups that
may be due to statistical noise.18
[[link removed]] To
estimate the count of people in each religious group, we multiply
estimates of the share of Christians, “nones” and those who
identify with other religions in each age and sex group by the
estimated count of total people in each cohort, according to the
UN’s 2019 World Population Prospects statistics. The smoothed 2020
age, sex and religious composition is shown in the above table. This
age and sex composition is the baseline for each projection scenario.
(Refer to Appendix B for more information about sources of religion
data.)

Fertility

Population projections use estimates of current and future fertility
rates among women of different ages. In this and other Pew Research
Center projections of religious change, country-level estimates of
current and future fertility come from the United Nations (UN total
fertility rate projections under their medium fertility scenario).
According to this UN data, updated in 2019, the overall total
fertility rate in the U.S. is expected to increase very slightly, but
throughout this century, the UN’s expected fertility rate rounds to
1.8 children per woman.19
[[link removed]]

Data on religious differentials in fertility are based on several
waves of the National Survey of Family Growth (NSFG). Estimates of
fertility used in this report for Christians, religious “nones”
and other religious groups are the product of differentials observed
in the NSFG applied to current and future overall U.S. fertility rates
estimated by the UN. Total fertility rates used for the 2020-2025
period are 1.9 for Christian women, 1.6 for the religiously
unaffiliated and 2.0 for women of other religions, and these
differentials remain stable throughout the projection period.

Vital Statistics data from the U.S. reveals a recent decline in
fertility that is not reflected in the input data from the NSFG and
the UN. According to the CDC
[[link removed]], the fertility rate
has already fallen to 1.7 children per woman. However, Vital
Statistics data provides no information on religious differences in
fertility and it reports only on past fertility patterns.

In Pew Research Center’s 2015 projections of religion around the
world
[[link removed]],
the central projection scenario assumed that over the course of a
century, fertility differences among religious groups would diminish
and eventually disappear. However, past patterns suggest that the
extent to which fertility rates converge may vary significantly by
context. If we assume that the fertility rates of Christians,
“nones” and other religious groups in the United States will
converge with one another, this would lead to an uptick in the
fertility of the unaffiliated, who currently have the lowest fertility
rates. But we do not have clear precedent for an uptick in fertility
among the unaffiliated, and since we lack clear reasons to think that
fertility rates will converge among groups in the U.S., fertility
differentials are assumed to remain constant in all scenarios.

Mortality

Survival rates are calculated based on mortality data from the UN’s
2019 World Population Prospects report. Mortality rates vary over
time, by sex and five-year age group, but the same rates are applied
regardless of religious affiliation. The UN projects that life
expectancy at birth in the U.S. will increase from 79 in 2020 to 86 in
2070. The life expectancies of most other age groups are expected to
increase significantly during this period as well. For example,
today’s 40-year-olds can expect to live another 41 years, and people
who turn 40 in 2070 are expected to have another 47 years of life, on
average. 

Migration

The estimated size of future migrant flows into and out of the United
States are based on global migration flow estimates produced by
Shanghai University professor Guy Abel
[[link removed]], based on the UN’s 2015
estimates of migrant stocks by origin and destination. The religious
composition of migrants from other countries to the U.S. is based on
Pew Research Center analysis conducted for the 2012 report, “Faith
on the Move – The Religious Affiliation of International Migrants
[[link removed]],”
which drew on U.S. sources including the New Immigrants Survey and the
Center’s “Religious Landscape Study
[[link removed]].”
Data collected for the “Faith on the Move” report included the
religious composition of U.S.-born people living in other countries.
This aggregated information on the religious composition of emigrants
from the U.S. is used to estimate the religious composition of people
who will move away from the U.S. in future decades.20
[[link removed]]

Migration both into and out of the United States is assumed to stay
constant at 2010-2015 levels throughout the projection period. Based
on these estimates, migration will account for net population growth
of about 5 million people in each five-year period.

Transmission

New analyses across several datasets revealed that transmission of
religious identity from mothers to children falls short of the 100%
transmission rate assumed in previous projection models. Transmission
of mothers’ religious identity is less common when parents do not
share a religion or when neither parent is highly religiously
committed. The data on current transmission rates in the U.S. come
from Pew Research Center’s 2019 survey of teens. In this dataset,
there are many teens, often with loosely affiliated parents or with
one Christian and one unaffiliated parent, who do not share their
mother’s religious affiliation. This mirrors transmission patterns
observed in retrospective data from older cohorts in the General
Social Survey (GSS) and the National Study of Youth and Religion
(NSYR). Projection scenarios compare the impact of full transmission
(100% of children inherit their mother’s affiliation) to observed
transmission rates from the teens survey.

It would be ideal to have measures of religious socialization and
transmission in preteen years to clearly distinguish between the
phenomenon of parental transmission of religion and religious
switching that typically begins in adolescence and continues in young
adulthood. However, the teens survey was the best source available to
us about transmission that has (or has not) occurred by age 13. In the
teens survey, teens ranged from ages 13 to 17. We were concerned that
the religious composition of older teens might have been influenced by
the religious switching that often occurs in adolescence and young
adulthood. However, we found that older and younger teens in this
sample were about equally likely to share their mother’s religion.
Therefore, we used data from all teens in the survey to estimate
transmission rates. In reality, religious transmission is a complex
process that may involve a mix of intentional and unintentional
actions and messages communicated by parents and others to children
over the course of many years. In our model, patterns of whether teens
had adopted their mother’s religion in the teens survey are used as
input data to model rates of transmission as something that does or
does not occur at birth. Modeling transmission or non-transmission at
birth is, of course, a simplification for projection purposes.

Due to the large size of Christian and unaffiliated populations in the
U.S., we have data from many more Christian and religiously
unaffiliated mothers than mothers in the “other religion”
category. We model the imperfect transmission of religious identity
from mothers in the other religion category to their children based on
data from an effective sample size of fewer than 100 respondents.
Since we therefore have less confidence in other religion transmission
numbers than the transmission patterns for Christian and unaffiliated
mothers, transmission data for mothers affiliated with other religions
is not displayed in this report. Furthermore, transmission in the
other religion category may be difficult to interpret because it
combines all religious identities aside from Christian and religiously
unaffiliated identities.

The scenarios in this report assume one of five future switching
patterns: no switching at all, steady switching, rising switching
rates with limits, rising switching rates without limits, and no
switching after age 30. The pattern in which religious switching
continues to occur among young adults (ages 15 to 29) at constant,
recently observed rates is based on the amount of switching measured
among people ages 30 to 40 in the ATP. These panelists are the cohort
to have moved through young adulthood most recently, and also are a
large enough group to base estimates on. Within this group,
Christianity has retained the smallest shares of its members, with 61%
of men and 70% of women who say they were raised Christian still
identifying as Christian in their 30s. For the projections,
probabilities of switching are distributed equally across the young
adult switching period. For example, the 30% of women who would be
expected to leave Christianity during early adulthood are modeled as
equally likely to leave in the 15 to 19, 20 to 24, and 25 to 29 age
groups.

Two scenarios account for trends in switching across cohorts and
extrapolate them into the future, with or without caps on how high or
low switching and retention rates could go. This is because
disaffiliating from Christianity has become steadily more common
across successive cohorts, and this trend could continue. Switching
inputs for cohort models are based on observations in the GSS since
1973.21
[[link removed]] Rates
at ages 30 to 34 (right after the young adult switching period closes)
for five-year birth cohorts are estimated using rolling averages
across three observations. Depending on the scenario, this trend is
either allowed to continue in a linear model or to flatten out once
Christian retention reaches a low of 53% by 2050 and unaffiliated
retention reaches a high of 95% by the year 2100. These shares were
chosen as plausible boundaries based on the range of retention rates
observed in cross-national analyses of 79 other countries from Pew
Research Center, the International Social Survey Program (ISSP) and
the British Social Attitudes Survey (BSAS). (The most recent data was
analyzed when more than one dataset per country was available.) A
lower bound for Christian retention within each generation is
consistent with the supposition that the Christians remaining after a
period of widespread disaffiliation will be relatively committed,
rather than increasingly likely to defect.

In all of the scenarios, steady rates of switching are assumed for
members of other religions due to the limited number of respondents
available to provide data on patterns of retention and switching among
those who grew up in other religions.

Switching after age 30

Most religious switching happens in early adulthood, but older
American adults also have been disaffiliating from Christianity in
recent decades. We think that disaffiliation after age 30 may be
largely due to a period effect in the U.S., in which loosely
affiliated people of all ages are leaving the faith now that a
Christian identity is no longer so common that it is taken for granted
and perceived as socially necessary. An analysis by birth cohort in
the GSS revealed that, starting in the 1990s, Christian retention
rates declined by about 7 percentage points _after_ the young adult
switching period in several 10-year birth cohorts.

There is not much cross-national survey data that could show what
tends to happen when countries go through a period like this. The
highest-quality data comes from the British Social Attitudes Survey,
and Britain transitioned away from an 85-90% Christian majority much
earlier than the U.S. The BSAS has been conducted every year since
1983 with a similar sample size to the GSS. (The BSAS includes a
two-step question on religious affiliation, unlike the GSS.) Based on
the same method as the GSS analysis, there was little further
attrition from Christianity after age 30 in any Great Britain cohort
going back to those born in the 1930s. The highest Christian retention
rate for any cohort was about 80%.

In 20 ISSP countries where religious switching could be measured
across multiple years and (relatively large) age groups, there were
only two examples of further disaffiliation from Christianity among
adults older than 30 once retention was already at 70% or below:
France and Australia.

For scenarios that allow for more switching after age 30, an
additional 7-point drop in retention – based on the pattern in the
GSS over the past couple of decades – is applied to cohorts that
leave young adulthood with Christian retention rates higher than 70%.
This later attrition is applied evenly across ages 30 to 65. About
three-quarters of American adults who were born in the 1980s and were
raised Christian were still Christian in their 30s, so this will be
the last birth cohort to receive an adjustment allowing for further
adult attrition. People born in the 1990s and raised Christian are
entering their 30s with a lower retention rate than the prior cohort,
continuing the recent trends (refer to Appendix B
[[link removed]] for
details on trends that inform these modeling assumptions).

Since the analysis for this report was completed, new data from
Australia suggests that, at least in Australia, switching out of
Christianity may continue beyond age 30, even among cohorts already at
70% or lower retention by age 30.

A July 2022 Australian Bureau of Statistics report
[[link removed]] comparing
responses to Australia’s religion census question in 2016 and 2021
finds a considerable overall change in this short period. The report
says that in 2016, 52% of the population identified as Christian, 30%
identified with no religion, 8% identified as “other religion” and
9% didn’t answer the religion question.22
[[link removed]] In
2021, the Christian share of respondents dropped to 44%, the
religiously unaffiliated rose to 39%, 10% identified as “other
religion” and 7% didn’t answer the question.

The Australia census doesn’t measure childhood religion (measuring
childhood religion in any census is uncommon but Scotland included
such a measure in 2001), so it isn’t possible to calculate retention
rates directly from one year’s census data. But it is possible to
consider how counts for groups of Australians vary from one census to
the next. Because census data is gathered from the entire Australian
population and since there has been considerable change in this
period, it is possible to detect that among adults ages 30 and older,
at each 2016 age, fewer Christians were counted in 2021. For example,
170,425 50-year-olds identified as Christian in 2016. Five years
later, when this cohort was 55, they numbered only 157,097.
Conversely, counts of unaffiliated older adults rose. There were
79,468 50-year-olds who identified with no religion in 2016 and five
years later, their number swelled to 100,919.

In the Australia data, as in many previous studies, religious change
is once again concentrated in young adult years. However, this new
data reveals that in this country, where only about half of the
population identified as Christian in 2016 and Christian retention
among 30- to 49-year-olds had already dropped to 55% in 2018, modest
levels of switching from Christianity to no religion seem to have
continued among middle-aged and older adults.23
[[link removed]]

By contrast, as described previously, most scenarios in this Pew
Research Center report assume that older adult switching will cease
for Americans born in the 1990s and later. If these rising cohorts in
the U.S. instead follow this newly observed Australian pattern of
ongoing switching throughout the life course, then this report’s
assumption about the end of older adult switching will have been
premature.

Modeling of switching

Projections of switching used in the scenarios are based on data from
the 1973-2018 GSS and the 2019 ATP. Using these sources, we estimated
retention for subsequent cohorts averaging results at ages 30 to 34 at
each time _t_; 35 to 39 at time _t+5;_ and 40 to 44 at
time _t+10_,assuming that most switching events occur before age of
30. Thus, averaged results from ages 30 to 44 for cohorts give us good
approximation of retention. In the next step, we fit the mathematical
model to these estimates. We chose the four-parameter Weibull model,
which enables us to put the limits on retention levels used in some
chosen scenarios. Then, the models for each religious group are used
for projections of retention achieved in subsequent cohorts in the
future. The retention numbers are transformed into switching rates
equally distributed among five-year age groups of the main switching
period of life of individuals (ages 15 to 29).

The projection approach

The technical calculations for the projections in this report were
made in collaboration with the Center for Advanced Studies of
Population and Religion (CASPAR) and its director, Marcin Stonawski,
using an advanced variation of the standard demographic method of
making population projections. The standard approach is called the
cohort-component method, and it takes the age and sex structure of a
population into account when projecting the population forward in
time. This has the advantage of recognizing that an initial baseline
population can be relatively “young,” with a high proportion of
people in younger age groups (such as the unaffiliated) or relatively
“old,” with a high proportion of older people (such as
Christians).

Cohorts are groups of people that had an experience in a particular
time. A birth cohort, the type of cohort referenced in this context,
comprises people born during a certain period. Birth cohorts can also
be described as males or females who have reached a certain age in a
particular year. For example, the cohorts of females ages 15 to 19 in
the year 2000 and males ages 15 to 19 in the year 2000 shared the
experience of being born between 1981 and 1985.

Components are the three ways in which populations grow or shrink: new
entrants via births, exits via deaths and net changes from migration.
Each cohort of the population is projected into the future by adding
likely gains – births and people moving into the country
(immigrants) – and subtracting likely losses – deaths and people
moving out (emigrants) – year by year. The very youngest cohorts,
those ages 0 to 4, are created by applying age-specific fertility
rates to each female cohort in the childbearing years (ages 15 to
49).24
[[link removed]]

The cohort-component method has been in existence for more than a
century. First suggested by the English economist Edwin Cannan in
1895, then further improved by demographers in the 1930s and ’40s,
it has been widely adopted since World War II. It is used by the
United Nations Population Division, the U.S. Census Bureau, other
national statistical offices, and numerous academic and research
institutions.

The advanced variant of this approach, multistate cohort component
projection, became viable starting in the 1970s thanks to the
availability of mainframe computers and work by the American
geographer Andrei Rogers, among others. The multistate approach
permits simultaneous projection of the religious groups included in
this study, taking into account variation by religion in age, sex,
childbearing patterns, and propensity and direction of migration. This
approach also enables modeling of religious switching as a transition
between religious “states.”

Appendix A: Sources of religion data

This report relies on religion data from several sources. Data on
current patterns of religious switching come from Pew Research
Center’s American Trends Panel
[[link removed]] (ATP).
Adult religious composition data in the ATP is weighted to results
from Pew Research Center’s National Public Opinion Reference Survey
(NPORS). Data on patterns of religious transmission between parents
and children come from analysis of a 2019 Pew Research Center survey
of teens and their parents
[[link removed]].
Religious switching trend data come from NORC’s General Social
Survey. Each of these sources is described below.

The American Trends Panel

The American Trends Panel, created by Pew Research Center, is a
nationally representative panel of randomly selected U.S. adults.
Panelists participate via self-administered web surveys. Panelists who
do not have internet access at home are provided with a tablet and
wireless internet connection. Interviews are conducted in both English
and Spanish. The panel is being managed by Ipsos.

The analysis in this report is based on demographic profile data for
all 15,494 active panel members as of Nov. 30, 2019. For 10,744
panelists who joined the panel prior to 2019, the data was collected
on the ATP’s 2019 annual profile survey conducted Aug. 7 to Sept.
29, 2019. For 4,720 panelists who were recruited in 2019, the data was
collected on the 2019 recruitment survey conducted Aug. 16 to Nov. 30,
2019.

The ATP was created in 2014, with the first cohort of panelists
invited to join the panel at the end of a large, national, landline
and cellphone random-digit-dial survey that was conducted in both
English and Spanish. Two additional recruitments were conducted using
the same method in 2015 and 2017, respectively. Across these three
surveys, a total of 19,718 adults were invited to join the ATP, of
which 9,942 agreed to participate.

In August 2018, the ATP switched from telephone to address-based
recruitment. Invitations were sent to a random, address-based sample
(ABS) of households selected from the U.S. Postal Service’s Delivery
Sequence File. In each household, the adult with the next birthday was
asked to go online to complete a survey, at the end of which they were
invited to join the panel. For a random half-sample of invitations,
households without internet access were instructed to return a
postcard. These households were contacted by telephone and sent a
tablet if they agreed to participate. A total of 9,396 were invited to
join the panel, and 8,778 agreed to join the panel and completed an
initial profile survey. The same recruitment procedure was carried out
in August 2019, from which a total of 5,900 were invited to join the
panel and 4,720 agreed to join the panel and completed an initial
profile survey.

The U.S. Postal Service’s Delivery Sequence File has been estimated
to cover as much as 98% of the population, although some studies
suggest that the coverage could be in the low 90% range.25
[[link removed]]

Weighting

The ATP data was weighted in a multistep process that accounts for
multiple stages of sampling and nonresponse that occur at different
points in the survey process. First, each panelist begins with a base
weight that reflects their probability of selection for their initial
recruitment survey. The base weights for panelists recruited in
different years are scaled to be proportionate to the effective sample
size for all active panelists in their cohort. To correct for
nonresponse to the initial recruitment surveys and gradual panel
attrition, the base weights for all active panelists are calibrated to
align with the population benchmarks identified in the accompanying
table to create a full-panel weight.

In addition to sampling error, one should bear in mind that question
wording and practical difficulties in conducting surveys can introduce
error or bias into the findings of opinion polls.

National Public Opinion Reference Survey

The ATP may underrepresent a portion of the population that would
participate by paper and pencil (if given the option) but would be
reluctant to be surveyed online, and which is presumably more highly
religious than the rest of the U.S. public. As a result, Pew Research
Center has decided, for the foreseeable future, to adjust (or
“weight”) ATP samples to religious affiliation parameters derived
from the NPORS. For more explanation, please refer to the report
“Measuring Religion in Pew Research Center’s American Trends Panel
[[link removed]].”
For details on the design and administration of NPORS, see the survey
methodology
[[link removed]].

Survey of teens and parents

The analysis of how religious identity is transmitted between parents
and children in this report is based on a self-administered web survey
conducted by Pew Research Center in 2019, among a sample of 1,811
pairs comprised of one U.S. adolescent ages 13 to 17 and one parent
per adolescent. The survey was conducted by Ipsos Public Affairs in
English and Spanish using KnowledgePanel, its nationally
representative online research panel. Here is the survey methodology
[[link removed]].

The General Social Survey

The General Social Survey (GSS), conducted by NORC at the University
of Chicago, began in 1972 and primarily collected survey data in
face-to-face interviews for most of its history. When the pandemic
made face-to-face data collection infeasible, in 2021 the GSS switched
to use address-based sampling for recruiting participants, who were
invited to take the GSS online. This change in mode could explain some
change in religious affiliation between 2018 and 2021. Complete GSS
documentation and data is available at [link removed].

Appendix B: Supplemental analyses

Pew Research Center conducted analyses not detailed elsewhere in this
report to make decisions that shaped projections. This appendix
contains an overview of some of these analyses.

Trends in other sources of religious composition data

This report considers how the rapid and continuous rise of religious
“nones” in the United States might continue or change in the
future. However, in 2021, one report
[[link removed]] suggested
the number may have already begun to decline. To evaluate this
possibility, data from the American National Election Survey, Gallup
and Public Religion Research Institute (PRRI) was compared with the
Pew Research Center and General Social Survey (GSS) trends that form
the basis of the report. While each of these data sources have shown
occasional small declines in unaffiliated shares from year to year,
the common pattern is a steady increase in unaffiliated shares over
the decades. What appear to be occasional small declines may be
statistical noise, including effects of changing survey and sampling
methodologies.

While the unaffiliated share reported by PRRI was lower in 2021 than
the organization reported in previous years, PRRI again reported an
increasing share of “nones” in 2022. Considering the general trend
across surveys and the declining rates of religious affiliation among
young cohorts, the preponderance of evidence suggests the share of
“nones” is not declining and, as described in the “no
switching” scenario of this report, should be expected to grow for
demographic reasons alone (e.g., the younger age structure of
“nones” compared with the religiously affiliated).

While different surveys show the same general rise of Americans
identifying with no religion, scholars have noted that the specific
estimate varies across surveys. In 2002, sociologists Michael Hout and
Claude Fischer noted that Gallup estimates were lower than estimates
from the National Election Study, Pew Research Center and the GSS.
They observed, “One important distinction between Gallup and the
other surveys: Gallup interviewers accept ‘no religion’ as an
answer but do not suggest it to their respondents.” The other
surveys offered explicit “no religion” options.26
[[link removed]] Gallup
has since adjusted its response options
[[link removed]],
though it continues to have a relatively low estimate of religious
“nones.”

International retention

The “rising disaffiliation with limits” scenario imposes a floor
of 50% retention for men raised Christian and 55% for women raised
Christian. This limit was chosen after an analysis of Christian
retention rates in 79 countries outside the U.S. This analysis relied
on data from the International Social Survey Program
[[link removed]],
the British Social Attitudes Survey [[link removed]],
and regional Pew Research Center surveys covering sub-Saharan Africa
[[link removed]], Latin
America
[[link removed]], Central
and Eastern Europe
[[link removed]],
and Western Europe
[[link removed]].
(We analyzed the most recent survey when multiple data sources were
available for a country.) Every survey included a question about which
religion respondents were raised in as well as a question about their
present religious affiliation.

Among people ages 30 to 49, the average country had a Christian
retention rate (the percentage of people raised Christian who still
identify as Christian) of 87%. The lowest rates of Christian retention
were in Great Britain (49%), France (52%) and the Netherlands (53%).
Countries with large Christian majorities also tended to have high
retention rates, but high retention rates were also found in some
countries in which Christians are a minority.

The report’s “rising disaffiliation with limits” scenario also
imposes a ceiling for unaffiliated retention at 95%. This threshold
arose from a similar analysis of 30- to 49-year-olds in 22 countries
(fewer countries had sufficient samples of people who were raised with
no religion to analyze). Great Britain and the Czech Republic had the
largest shares of people who remained unaffiliated after being raised
without a religion, with a little over 95% retention each. Across
countries, the average unaffiliated retention rate was 73%. 

Disaffiliation after age 30

Most scenarios assume that cohorts with members between the ages of 30
and 65 in 2020 will lose 7 percentage points in Christian retention by
age 65. This figure is based on an analysis of later switching by
10-year birth cohort. People born in the 1950s, ’60s and ’70s had
lost an average of about 7 points in Christian retention between the
year their oldest members turned 30 and 2018.27
[[link removed]]

This pattern of later attrition began in the 1990s among people born
in the ’40s and later, at about the same time as the broader trend
of disaffiliation from Christianity. It may be based on a period
effect in which adults of all ages who were loosely affiliated with
Christianity felt free to disaffiliate as the unaffiliated grew in
numbers. After this period, switching may return to being almost
exclusively a phenomenon of young adulthood, but it is unclear when
this period will end.

To get a rough idea of when later attrition could end, we looked for
later switching in 21 countries outside the U.S. with sufficient data
to compare retention rates across birth cohorts for at least two time
points. The best cross-national comparison to the U.S., both in terms
of conditions/histories and data quality, come from the British Social
Attitudes Survey (BSAS). This social survey is similar to the GSS in
sample size, and it has been conducted every year since 1983, but its
flaws are outlined in the earlier discussion of data issues. Based on
exactly the same method as in the GSS cohort data analysis, the
British data shows falling retention by cohort at age 30 (i.e.,
increases in attrition from Christianity during young adulthood) until
retention got down to about 50%, but no attrition from Christianity
among people ages 30 and older in any cohort going back to those born
in the 1930s. For example, rolling averages of three waves for
retention were between 73% and 80% for Brits born in the 1930s and
raised Christian in every survey year, with the highest retention rate
in 2016. People born in the 1940s started and ended the measurement
period with retention rates in the high 60s, and this pattern of
consistency held for 60% of people born in the 1950s, 50-57% born in
the 1960s, and about 50% for those born in the 1970s and 1980s.

This may indicate that Britain – which became a country where
Christian identity could no longer be taken for granted far earlier
than the U.S. – reached a floor for attrition among people ages 30
and older by the mid-1990s. Since the British data starts with
retention rates in the high 60s to low 70s and none of the cohorts
show any substantial drops in retention after age 30, this may be a
reasonable threshold at which to assume that switching ends for adults
over 30.28
[[link removed]]

In 20 other countries for which we analyzed smaller sample sizes for
larger cohorts over two or three survey waves, there were only two
countries where Christian retention was 70% or below and there was
still decline within cohort after age 30: People born in the 1950s and
early ’60s in France who were raised Christian showed a drop in
retention from 56% to 52% between 2008 and 2018 – a relatively small
amount of change. The same cohort in Australia, however, declined from
61% to 46% retention. Data for these 20 countries comes from the ISSP,
and there are large and difficult to explain fluctuations in some
countries, but the patterns in most countries converge with the
retention floor suggested by the British data – i.e., the 70%
Christian retention floor at which point switching after 30 ends.

After the projections in this report were completed and shortly before
the report was released, we became aware of 2021 census data from
Australia showing that the count of Christians had dropped among
adults of all ages since 2016, though declines were concentrated among
young adults. At the same time, the count of religiously unaffiliated
Australians had increased among adults of all ages, though this
pattern too was concentrated among young adults (see discussion
in Appendix A
[[link removed]]).
After reading about these census patterns, we were able to gain access
to 2018 ISSP religion module data for Australia that was excluded from
the international release of 2018 ISSP religion module data.
Australia’s ISSP data confirmed the ongoing decline in retention
rates suggested by 2021 census data (see earlier discussion of ISSP
results).

If the United States follows Australia’s example of people who were
raised Christian continuing to disaffiliate after young adulthood even
among cohorts with retention rates below 70% at age 30, then the adult
switching patterns included in this report could prove to be
conservative. This report introduces new innovations in projections of
the religious future of the United States. We hope future studies will
build on this work, taking advantage of new data that will become
available in the years ahead.

1. Precise estimates are rounded to integer values in the text for
ease of reading. In 2020 – the starting point for the scenarios in
this report – U.S. children were slightly less likely than adults to
be religiously affiliated. (One reason for this is that parents of
childbearing age are more likely to be religiously unaffiliated than
are older adults). For the purposes of the models in this study, we
present estimated and projected shares of Americans of all ages,
including children. Throughout the rest of this report, figures that
apply only to the adult population are specified accordingly. In Pew
Research Center’s survey report on religious affiliation in 2021,
63% of U.S. adults (
[[link removed]]ages
18 and older) identified as Christian, 29% identified as religiously
unaffiliated, 6% identified with other religious groups, and 2% were
missing religious identity information. After excluding respondents
who were missing religion information, the shares from the 2021 survey
match our 2020 baseline (64% Christian, 30% religiously unaffiliated
and 6% other religion)

2. Our base model includes estimated rates of switching in all
directions, not just from Christianity to unaffiliated. For example,
we estimate that 3% of people raised Christian switch to a different
(i.e., non-Christian) religion during their young adult years, and
that 21% of people who are raised with no religion (i.e., as atheist,
agnostic or “nothing in particular”) become Christian in young
adulthood.

3. The retention rate is the percentage of people raised in a
religious group (in childhood) who remain in that group in adulthood.
A Christian retention rate of 66% would mean that two-thirds of people
raised as Christians remain Christians in adulthood. A Christian
retention rate below 50% would mean that fewer than half of people
raised as Christians remain Christians in adulthood. This “floor”
of 50% is roughly equal to the lowest retention rate observed in an
analysis of people ages 30 to 49 raised Christian in 79 other
countries. It is reasonable to expect that some people will retain a
Christian identity even if disaffiliation becomes more common. For
more details on this analysis, see Chapter 2
[[link removed]] and
the Methodology
[[link removed]].

4. Recent survey data indicates that the Christian retention rate in
Britain is 49%, the lowest found in any of the 80 countries, including
the U.S., for which data is available

5. The percentage of adults in Britain who identify as Christian
depends, in part, on how the question is asked. Identification with
Christianity remains higher in censuses than in surveys that measure
religion by first asking respondents whether they have a religion and
then, as a second step, collecting affiliation data only for those who
answer that, yes, they _do_ have a religion. Like census measures
that directly ask, “What is your religion, if any?” Pew Research
Center’s 2017 survey of religion across Western Europe
[[link removed]] relied
on a one-step question and found higher levels of affiliation than
two-step surveys like the British Social Attitudes Survey, the
European Social Survey and the European Values Survey.

* This chapter focuses on results of public opinion surveys of U.S.
adults. Most other population shares presented in this report are
estimates for Americans of all ages. See Methodology
[[link removed]] for
details on estimating the religious affiliations of children. 
* Prior to 2020, the General Social Survey (GSS) was conducted
primarily through face-to-face interviews. Due to the COVID-19
pandemic, in 2020 and 2021, the GSS gathered panel and cross-sectional
survey data primarily online. Since most data for the cross-sectional
survey was collected in 2021, NORC now describes the cross-sectional
survey data as the 2021 GSS (see page 4 of the codebook
[[link removed]] for
the latest GSS data). The change in the “mode” of survey
administration was concomitant with the GSS finding a rise in
religious “nones” from 23% in 2018 to 29% in 2021 and a
corresponding drop in the share of U.S. adults who identified as
Christian from 72% to 64%. Some of the change in the GSS between 2018
and 2021 may be due to this “mode effect.” For a basic explanation
of mode effects, see Pew Research Center’s video “Methods 101:
Mode Effects.
[[link removed]]”
For more details on the GSS data used in this report, see
the Methodology
[[link removed]]. 
* Norris, Pippa, and Ronald Inglehart. 2011. “Sacred and Secular:
Religion and Politics Worldwide
[[link removed]].” 
* Some research indicates that Americans tend to develop firm,
enduring political identities earlier than religious ones, and their
political views may influence their religious beliefs more than the
other way around. See Margolis, Michele. 2018. “From Politics to
the Pews: How Partisanship and the Political Environment Shape
Religious Identity.”
[[link removed]] In
addition, some scholars assert that the rise of the “nones” since
the 1990s is due in part to a reaction or “backlash” following an
increase in the visibility of the “religious right” and its
conservative positions on polarizing issues. See Hout, Michael, and
Claude S. Fischer. 2014. “Explaining Why More Americans Have No
Religious Preference: Political Backlash and Generational Succession,
1987-2012
[[link removed]].”
Sociological Science. 
* Regions are based on the U.S. Census Bureau’s definitions
[[link removed]] of
South, Northeast, Midwest and West. 
* Recent and future migration flows are estimated based on the most
recent five-year period with complete global migration data at the
time of analysis (mid-2010 to mid-2015). 
* Due primarily to the COVID-19 pandemic, life expectancy decreased
by 1.5 years between 2019 and 2020, according to the CDC
[[link removed]]. This reduction
is expected to be temporary, so unadjusted UN estimates were used in
projections. 
* Some social scientists expect that a large revival of Christianity
or another religion will not take place. According to the secular
transition model developed by University College London professor of
social science David Voas, big declines in religious affiliation are
unlikely to be reversed. Voas contends that societies experience, at
different times from country to country, a secular transition, a
permanent large-scale change. “A transition,” Voas says, “is not
cyclical or recurring; once out, the toothpaste will not go back in
the tube.” See Voas, David. 2008. “The Continuing Secular
Transition
[[link removed]].”
In Pollack, Detlef and Daniel V.A. Olson, eds. “The Role of Religion
in Modern Societies.” 
* Projections do not assume later switching into Christianity
because there is no clear trend in this direction. 
* Across scenarios that model switching at ages 30 to 65, the 7%
further disaffiliation is only assumed for cohorts with retention
rates above 70% at age 30. Of course, among cohorts in which Christian
retention rates have already dropped to 70% or lower by the time
members reach age 30, some individuals will still change religious
identity after turning 30, including changes into and out of
Christianity. However, such switching may be largely offsetting and
rare. 
* This later-adulthood switching applies only to cohorts born before
1990; disaffiliation has become so common in the decades since then
that the models assume people born after 1990 who are going to switch
will have done so before turning 30. Across scenarios that model
switching at ages 30 to 65, the 7% further disaffiliation is only
assumed for cohorts with retention rates above 70% at age 30. 
* Across scenarios that model switching at ages 30 to 65, the 7%
further disaffiliation is only assumed for cohorts with retention
rates above 70% at age 30, i.e people born between 1956 and 1989. 
* We smooth data using lowess (locally weighted scatterplot
smoothing) across age groups. Lowess smoothing is a common technique
used by statisticians and demographers. After smoothing the data,
iterative proportional fitting (raking) is used to match the ATP
measure of religious composition for the adult population. 
* After the analysis for this report was complete, the United
Nations released the first update to its 2019 World Populations
Prospects. The 2022 revision was released on July 11, 2022. In this
new update, the UN’s total fertility rate estimate for the U.S. in
2020 is 1.64 births per woman. 
* For details on deriving migration estimates see Abel, Guy J. 2018.
“Estimates of global bilateral migration flows by gender between
1960 and 2015
[[link removed]].”
International Migration Review. See also Abel, Guy J., and Joel E.
Cohen. 2019. “Bilateral international migration flow estimates for
200 countries [[link removed]].”
Scientific Data; and Abel, Guy J. 2013. “Estimating global migration
flow tables using place of birth data
[[link removed]].”
Demographic Research. 
* The first wave of the GSS, in 1972, did not ask about the religion
in which respondents were raised. 
* In the Australian Bureau of Statistics report, people who did not
answer the religion question are included in the denominator of all
percentages. If people who did not answer the religion question were
excluded, all percentages would be slightly higher. 
* Retention rate information comes from the 2018 ISSP religion
module included on the Australian Survey of Social Attitudes. As of
July 2022, this data was not available directly on the main GESIS ISSP
data archive. However, we were granted access to the data by the
Australian Data Archive. 
* The number of children ages 0 to 4 projected to join a population
is also influenced by infant and child mortality rates and migration
rates, which are incorporated in the projection model. 
* AAPOR Task Force on Address-based Sampling. 2016. “AAPOR Report:
Address-based Sampling
[[link removed]].” 
* See Hout, Michael, and Claude Fischer. 2002. “Why more Americans
have no religious preference: Politics and generations
[[link removed]].”
American Sociological Review 
* The most recent data available to us for analyzing switching
patterns in the GSS was from 2018. The 2020 GSS wave was delayed until
2021 because of the COVID-19 pandemic and when the 2021 data was
released, the variable with information on the religion in which
respondents were raised (relig16) was not initially available. 
* For another perspective, see the discussion in Appendix A
[[link removed]] about
switching patterns among older adults in a July 2022 report on
religion data in the Australian census. 

* religion
[[link removed]]
* Demographics
[[link removed]]
* Christian
[[link removed]]
* atheists
[[link removed]]
* Jews
[[link removed]]
* Muslims
[[link removed]]
* Hindus
[[link removed]]
* buddhists
[[link removed]]

*
[[link removed]]
*
[[link removed]]
*
*
[[link removed]]

 

 

 

INTERPRET THE WORLD AND CHANGE IT

 

 

Submit via web
[[link removed]]

Submit via email
Frequently asked questions
[[link removed]]

Manage subscription
[[link removed]]

Visit xxxxxx.org
[[link removed]]

Twitter [[link removed]]

Facebook [[link removed]]

 




[link removed]

To unsubscribe, click the following link:
[link removed]