Family Formation and Returning to Institutional Religion in Young Adulthood

Abstract

Institutional religious involvement wanes during young adulthood, but evidence suggests life-course factors such as family formation bring people back to religion. Using the National Longitudinal Study of Adolescent to Adult Health (Waves 1, 3, and 4), we examine how often young adults who were involved in institutional religion as adolescents return—measured by religious service attendance and religious affiliation—after leaving in emerging adulthood, and how this return is patterned by family formation. The majority of young adults who leave do not return to regular religious service attendance, regardless of their family formation. But single parents, married parents, and childless married individuals are more likely, and childless cohabiting couples less likely, to return to religious communities than those who are both single and childless. Only married parents are more likely than childless singles to reaffiliate, though there is marginal evidence that childless married adults may also be more likely. Thus, the institutions of religion and family are still linked, even though overall levels of religious return are not as high as expected.

Full text

Family Formation and Returning to Institutional Religion among Young Adults*
Jeremy E. Uecker
Department of Sociology
Baylor University
Damon Mayrl
Department of Social Sciences
Universidad Carlos III de Madrid
Samuel Stroope
Department of Sociology
Louisiana State University
August 10, 2015
WORD COUNT: 11,088 words
KEYWORDS: religious return, marriage, cohabitation, childbearing, religiosity, religious
affiliation
*ACKNOWLEDGMENTS: This research uses data from Add Health, a program project directed
by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen
Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-
HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human
Development, with cooperative funding from 23 other federal agencies and foundations. Special
acknowledgment is due to Ronald R. Rindfuss and Barbara Entwisle for assistance in the original
design. Information on how to obtain the Add Health data files is available on the Add Health
website: http://www.cpc.unc.edu/addhealth. No direct support was received from grant P01-
HD31921 for this analysis.

Family Formation and Returning to Institutional Religion among Young Adults
Abstract
Institutional religious involvement wanes during young adulthood, but evidence suggests life
course factors such as family formation bring people back to religion. Existing research is dated,
however, and often fails to examine returning to religion after having left. Using the National
Longitudinal Study of Adolescent to Adult Health (Waves I, III, and IV), we examine how often
young adults who were involved in institutional religion as adolescents return to religion—
measured by religious service attendance and religious affiliation—after leaving in emerging
adulthood, and how this return is patterned by family formation. The majority of young adults
who regularly attended religious services as adolescents do not return to regular religious service
attendance, regardless of their family formation. But single parents, married parents, and
childless married individuals are more likely, and childless cohabiting couples less likely, to
return to religious communities than those who are both single and childless. In terms of
reaffiliation, only married parents are more likely than childless singles to reaffiliate, though
there is marginal evidence that childless married adults may also be more likely. Thus, the
institutions of religion and family are still linked, even though overall levels of religious return
are not as high as expected.

1
Family Formation and Returning to Institutional Religion among Young Adults
Young adults today are less involved in religious institutions than in previous generations.
Recent evidence suggests they are less likely to attend religious services on a regular basis or to
identify with a religion at all (Wuthnow 2007; Putnam and Campbell 2010; Smith and Snell
2009; Sherkat 2014; Twenge et al. 2015).
1
For example, just 16% of those born between 1971
and 1994 attended church weekly or more as young adults, compared to 21% of those born
between 1944 and 1955 (Sherkat 2014). Similarly, 20-30% of those who came of age in the
2000s did not identify with a religion as young adults, compared to about 5% of those who came
of age in the 1950s or before (Putnam and Campbell 2010). These trends have significant
implications for our understanding of American religion and theories of religious behavior, as
well as for American religious communities perennially concerned with their ability to influence
the broader culture and transmit their faith to the next generation.
Nevertheless, strong conclusions about religious decline could be premature. As Sherkat
(2014:39) explains, “Caution is warranted when making inferences about long-term changes
based on generational differences, because age and life-course statuses intersect with generations
in repeated cross-sectional data.” In particular, life-course events such as marriage and
parenthood may bring back to religious communities young adults who have left. Wuthnow
(2007:62-63) finds that declines in religious service attendance among young adults are
concentrated entirely among the unmarried, a pattern he says is “consistent with the conventional
wisdom that younger adults return to church once they settle down and begin thinking about
raising a family.” Another study acknowledges generational shifts in religion but also notes that
1
Although the preponderance of recent studies support this view, Smith and Snell (2009) caution that declines in
service attendance may be concentrated among Catholic and mainline Protestant adults, and that other religious
indicators, such as frequency of personal prayer and belief in life after death, have remained more stable over time.

2
life course and age effects may end up making today’s young adults look more similar to earlier
generations than we might imagine (Bengtson, Putney, and Harris 2013).
This emphasis on family formation as a catalyst for religious return is not without
evidence. Marriage is one of the most important predictors of religious involvement and
affiliation (Hout and Fischer 2002; Merino 2012; Thornton, Axinn, and Hill 1992; Uecker,
Regnerus, and Vaaler 2007). Those who never marry display higher rates of apostasy (Hadaway
and Roof 1988; Sandomirsky and Wilson 1990) and are less likely to ever return to religion
(Roof 1990). Similarly, having children is associated with being religiously affiliated and
attending services more regularly (Baker and Smith 2009; Sandomirsky and Wilson 1990;
Uecker et al. 2007; Schwadel 2010). Longitudinal studies have indicated that there appears to be
something causal about family formation that brings disaffiliates back to the church: Not only are
religious people more likely to marry and have children, but the act of marrying and having
children tends to bring people back to the religious fold (Thornton et al. 1992). Young adults
who marry and have children in their twenties and thirties tend to increase their rates of religious
service attendance (Myers 1996; Petts 2009; Stolzenberg et al. 1995; Thornton et al. 1992) and to
reaffiliate with their religious tradition at higher rates (Wilson and Sherkat 1994). A stronger
childbearing effect exists for parents who begin having children in their mid-twenties
(Stolzenberg et al. 1995). The age of children also matters. As children approach school (or
Sunday school) age, parents’ religious participation tends to increase (Stolzenberg et al. 1995;
Argue, Johnson, and White 1999). More recent evidence suggests that having school-age
children—not marriage or childbearing itself—may be the causal factor driving religious service
attendance (Schleifer and Chaves 2014).

3
In this study, we focus not on the causal impact of marriage and parenting on
participation and identification with institutional religion generally, but rather on how these life-
course events predict a return to religion among young adults. Specifically, we focus on those
who were involved with institutional religion as adolescents, but subsequently significantly
reduced or cut ties with institutional religion in emerging adulthood.
2
As Schleifer and Chaves
(2014) point out, family formation may have very little effect on the religiosity of those who
were never religious to begin with. Thus, we are interested in assessing the specific claim that
family formation brings this contemporary cohort of young adults back to institutional religion
after a time away in emerging adulthood—despite family formation patterns, as we discuss
below, that are substantially different from those of previous generations.
We assess this claim using three waves of data from the National Longitudinal Study of
Adolescent to Adult Health (Add Health), a panel study of more than 15,000 respondents
surveyed from adolescence to young adulthood. We provide new evidence that the institutions of
religion and family remain linked among contemporary young adults, even though overall levels
of religious return are not high in absolute terms. Our analysis also suggests that diversification
of family forms in contemporary society may be producing diversity in patterns of religious
return, as different types of families may return out of distinct motivations. Before turning to our
analysis, we consider why family formation may be important for religious return, and discuss
why the relationship between family formation and religious return may be different than it was
for previous generations.
WHY FAMILY FORMATION LEADS TO RELIGIOUS RETURN
2
Although “emerging adult” (Arnett 2014) and “young adult” are often used synonymously, in this study we will
use “emerging adulthood” to refer to ages 18-24 and “young adulthood” to refer to ages 25-31.

4
Sociologists have theorized a variety of reasons why family formation might lead young adults to
return to religion. These can be broadly organized into four main groupings: religious, social,
practical, and cultural.
Religious Explanations. Religious explanations focus on the appeal of religious beliefs
and practices, and argue that family formation kindles desires to more regularly access or make
greater use of those beliefs and practices. Married couples may return to religious communities
in search of religious teachings that can strengthen their marriage by giving it a spiritual purpose
or that can provide guidance and emotional support for relationships, pregnancy, and parenting
(Jesse, Shoneboom and Blanchard 2007; Mahoney 2010). But religious explanations are most
often evoked in the context of parenting. Large majorities of American parents, across all
denominations, indicate that they want their children to receive religious training (Dillon and
Wink 2007). Many new parents thus return to religious institutions because they want to
encourage their children’s religious socialization (Albrecht et al. 1988; Hoge 1981; Ingersoll-
Dayton, Krause, and Morgan 2002; Stolzenberg et al. 1995; Sullivan 2008). New parents may
also return to religion so that their children may participate in rituals such as baptism or bar
mitzvah (Manning 2013; Petts 2007), or because they wish to become better religious role
models for their children (Sherkat 2006).
Social Explanations. Social explanations focus on how religious return may expand and
strengthen social bonds. Religious congregations are eminently social institutions, which provide
multiple opportunities for couples to socialize (Stolzenberg et al. 1995). The social dynamics of
congregations may be more welcoming to families than they are for single people (Wilson and
Sherkat 1994). A recent qualitative study of Mormon undergraduates found that many said
marriage made service attendance more appealing because participating as a couple was more

5
enjoyable than participating alone (Chou 2010). Additionally, the programming of many
religious communities is geared toward families, especially those with small children (Wilcox,
Chaves, and Franz 2004). The finding that parents with school-aged children, and those who
have them during normative childbearing ages, have the highest rates of return likely reflects the
fact that “churches best serve conventional families” (Stolzenberg et al. 1995:100; see also
Argue et al. 1999). Finally, religious return may also strengthen bonds within couples by
allowing them to spend more time together; marrying a religious spouse often leads nonreligious
individuals to increase their participation (Hout and Fischer 2002; Merino 2012).
Practical Explanations. Practical explanations focus on the variety of concrete benefits
and supports that congregations provide for young families (Chaves 2004; Edgell 2006). These
practical supports may be particularly appealing to young people grappling with the demands of
parenting. A survey of pastors in upstate New York found that congregations provided a wide
range of services for parents, including babysitting, parenting classes, discussion groups, and
daycare (Becker and Hofmeister 2001). Parents may also receive a great deal of informal support
from co-religionists (including other parents) who can provide various forms of practical help,
including childcare, secondhand items, parenting advice, and other useful information. This
formal and informal practical support can be especially valuable for parents, and might
encourage those who have left to return to a religious community in order to benefit from it.
Cultural Explanations. Cultural explanations focus on how ideas, norms, scripts, and
roles circulating in American culture help encourage religious return. At the broadest level,
scholars have pointed out that religion and family have traditionally been mutually-supportive
institutions in American society, sustaining one another as bulwarks against market forces and
state control (Pankhurst and Houseknecht 2000). Religions have incorporated family imagery

6
and rituals, while families draw on the supportive rhetoric and practices of religious
organizations (Becker 1999; Sandomirsky and Wilson 1990). Accordingly, “the family is a
commitment mechanism for the church” (Wilson and Sherkat 1994:158), and “churches are
important institutional carriers of the norms, strategies of action, and social networks that keep
the practice of marriage alive” (Wilcox and Wolfinger 2007:583). Many congregations,
especially conservative Protestant ones, promote an ideology of “religious familism” wherein the
family is portrayed and promoted as the central unit of the social order (Edgell 2006; Wilcox
2004). The pronuptial and pronatalist messages espoused by these institutions idealize and
reinforce middle-class views of “the good family” (Edgell 2006; Mahoney 2010; Wilcox 2004).
Young adults who form families may thus be drawn to congregations because their message and
practices support and sustain the family commitments they have made. Married young adults
may also simply view religious participation as part of a “good” family life (Wilcox, Cherlin,
Uecker, and Messel 2012).
Briefly, two additional related cultural explanations also suggest pathways from family
formation to religious commitment. The first has to do with the symbolic equation of marriage,
family, and having “arrived” as an adult. Alongside a college degree and full-time employment,
both marriage and parenthood signal being “settled” and “established” (Becker and Hofmeister
2001; Wilcox et al. 2012). To the extent that young adults enter into marriage as a means of
“settling down,” they may also be more likely to join a religious community as part of the
package—that is, making one social commitment associated with maturity may make another
more likely (Greeley 1989). Similarly, family formation can encourage religious return by
activating new social roles that encourage religious participation. This may be especially so for
men, for whom family formation can encourage increased religiosity by activating expectations

7
about men as financial providers and fathers (Edgell 2006; Wilson and Sherkat 1994). A recent
study found that becoming a father leads men to reexamine their priorities, focus on other family
members, and turn to religious doctrine for guidance and assistance (Petts 2007).
In short, family formation may encourage young adults to return to the fold by providing
them with beliefs and rituals, social opportunities, practical supports, and cultural scripts and
roles.
REASONS TO ANTICIPATE CHANGING PATTERNS OF RELIGIOUS RETURN
Although existing studies have demonstrated that getting married and having children can lead
young adults to return to religion after an absence, and have generated some influential theories
about why it does so, this scholarship has grown somewhat stale. The classic longitudinal studies
demonstrating the reciprocal influence of family formation on religious behaviors draw on data
collected from individuals who came of age and formed families in the 1970s and early 1980s, if
not before (e.g., Argue et al. 1999; Sandomirsky and Wilson 1990; Stolzenberg et al. 1995;
Thornton et al. 1992; Wilson and Sherkat 1994).
3
Yet profound transformations in the family, the
economy, and religion have all occurred over the past thirty years, raising the possibility that the
effect of family formation on religious affiliation and attendance may have changed.
A handful of more recent studies draw on local samples or focus on limited subsections
of the population to address these questions (e.g., O’Connor et al. 2002; Petts 2007), but only
two nationally representative longitudinal studies examine the relationship between family
formation and religious participation in more recent years. The first, Petts’ (2009) analysis of the
National Longitudinal Survey of Youth through 2004, focuses on the period from early
3
Of these, Wilson and Sherkat (1994) provide the closest analogue to our study. However, they only assessed
reaffiliation (and not changes in religious service attendance), did not distinguish between single and cohabiting
unmarried young adults, and analyzed data from young adults who came of age in the 1960s and 1970s.

8
adolescence through early adulthood, and does not analyze the trajectories of study participants
beyond age 25. Yet studies show that the mean age of return to religion among apostates is
around 28 (O’Connor et al. 2002; Hoge, Johnson, and Luidens 1993), and marriage and first
birth now typically occur after age 25 (Arroyo et al. 2012; US Census Bureau 2014). It is thus
possible that Petts’ analysis does not capture the important effects of family on religious return
later in young adulthood. The second, Schleifer and Chaves’ (2014) analysis of panel data from
the General Social Survey, examines the causal effect of marriage, having a child, and having a
school-age child, but does not address the question of religious return. Because existing
scholarship is either outdated, does not focus on young adults in the prime of their family-
formation years, or does not answer the question of returning to religious institutions after a
period of detachment, there is an important need to assess whether scholarly knowledge about
family formation and religious return still holds for young adults in the twenty-first century.
Over the past thirty years, changes in the family and culture—spurred in no small part by
economic shifts—have transformed the context in which young adults make decisions about
religious participation. Below, we review three—delayed marriage and childbearing, rising rates
of cohabitation and nonmarital childbearing, and changing gender roles—which are grounds to
anticipate changes in the relationship between family formation and religious return.
Delayed marriage and childbearing. As mentioned above, the age at first marriage and
first birth have been steadily increasing. Although most young adults eventually marry
(Goldstein and Kenney 2001), young adults used to move rather seamlessly from their family of
origin into marriage and parenthood. Today, however, there is a significant gap between these
events (Furstenberg et al. 2004). The lengthening time between high school graduation (typically
at age 18) and family formation could alter the effects of family formation on religious return. If

9
young adults become accustomed to life outside of religious institutions during this interim
period, they may come to see less benefit to returning after starting a family. Additionally,
married young adults used to view marriage as a means of establishing themselves financially,
but now marriage typically occurs after a couple has established themselves financially (Kefalas
et al. 2011). For these couples, the practical benefits of congregational life may be less necessary
than in previous generations.
Cohabitation and nonmarital childbearing. The expectation that coupling and
childbearing should be done in the context of marriage has been weakening for at least half a
century, and a variety of arrangements have emerged that compete with the “traditional” married
family as models for family formation (Cherlin 2009). The practice of cohabitation has grown
dramatically, either as a precursor or as an alternative to marriage. Surveys reveal that nearly
three in five American women cohabit by the age of 24 (Schoen, Landale, and Daniels 2007).
Cohabitation violates the teachings and sexual norms of many religious traditions, and adults
who cohabit may avoid returning to religious communities out of fear that they may face
sanctions or ostracism (Petts 2009; Thornton et al. 1992). Thus, we expect that cohabiting,
especially as an alternative to marriage, will discourage apostate young adults from returning to
religion. This expectation is in line with existing studies that show that young adults who cohabit
are less likely to attend services (Stolzenberg et al. 1995; Thornton et al. 1992) and more likely
to drop their religious affiliation (Uecker et al. 2007).
Nonmarital childbearing has also risen dramatically over the last 50 years, and now about
two out of every five births are to unmarried women—with more than half of these nonmarital
births occurring in cohabiting relationships (Curtin, Ventura, and Martinez 2014). Like
cohabitors, single parents may fear that returning to religion will subject them to stigma from

10
religious communities oriented to the traditional family (Petts 2012; Sullivan 2008). Other single
parents, especially poor ones, may simply face financial and logistical challenges that limit their
ability to participate regularly (Sullivan 2008, 2012). The effects of parenting on religious return
appear to vary by marital status; indeed, Wuthnow (2007) reports that fewer single parents than
childless single adults attend religious services almost weekly among 20-29-year-olds, perhaps
owing to these marriage norms or resource constraints. At the same time, if young adults return
to religious communities for practical benefits, some single parents may return to institutional
religion in order to receive practical support.
Changing gender roles. Scholars have shown that gender shapes both family formation
and the propensity to return to religion. Men’s attendance patterns are more strongly linked to
life-course transitions like marriage and parenthood than are women’s (Sandomirsky and Wilson
1990; Stolzenberg et al. 1995); married men are more likely than single men or married women
to return to the fold after marriage, and the effect of parenthood on religious participation is
stronger for fathers than mothers (Wilson and Sherkat 1994). Family formation brings men’s
religious participation into parity with women’s; as Becker and Hofmeister (2001:718) put it,
family formation “makes men ‘more like women’ in their religious involvement.” These
differences may occur because unchurched men are more likely to marry religious women than
vice versa, or it may come about because family formation activates new cultural roles or scripts
(e.g., breadwinner) that find affirmation in religion for men (Becker and Hofmeister 2001;
Wilson and Sherkat 1994).
Recent shifts may have made these two pathways less robust, however. The predictive
power of gender on service attendance has been declining across cohorts over the twentieth
century (Schwadel 2010), so the proportion of more-religious women to less-religious men may

11
have also been declining. More broadly, feminism and the movement for greater gender equality
have challenged the traditional gendered division of labor in many households. More women are
working outside the home, although they continue to devote more time to household labor and
child care than do men (Bianchi et al. 2000; Hochschild and Machung 1989). These changes
have the potential to unsettle the relationship between family formation and religious return by
altering cultural norms around working and childcare, thereby attenuating the activational effects
of marriage and parenting on religious participation. To the extent that the traditional roles which
reconnected individuals to religion through family formation are no longer culturally salient,
changing gender roles could reduce the likelihood of religious return.
DATA
The data for this study come from Waves 1, 3, and 4 of the National Longitudinal Study of
Adolescent to Adult Health (Add Health). Wave 1 was conducted in 1994 and 1995 and
consisted of in-depth interviews with 20,745 American youth in grades 7 – 12. Schools included
in the study were chosen from a sampling frame of U.S. high schools and were nationally
representative with respect to size, urbanicity, ethnicity, type (e.g., public, private, religious), and
region. A total of 132 schools participated in the study, ranging in size from 100 to over 3,000
students. Wave 3 was conducted in 2001 and 2002, when respondents were 18–28 years old, and
consisted of interviews with 15,197 of the Wave 1 respondents. Wave 4 was collected in 2007
and 2008 and included interviews with 15,701 of the original respondents. Respondents were
24–34 years old in this wave.
In order to avoid overlapping ages in Waves, we drop respondents who were older than
age 24 in Wave 3 and younger than age 25 in Wave 4. We also drop the few respondents who
were older than age 31 in Wave 4. Additionally, many of the Wave 4 respondents were not

12
interviewed in Wave 3, and others did not have a valid sample weight. Thus, the N for
participants in all three waves who have a valid sample weight and meet our age restrictions is
11,479. After these sample restrictions were applied, we employed multiple imputation to
account for missing values for all variables. Missing data were rare in all the samples we
employed; once sample restrictions were employed no observation was missing more than two
values. Additionally, the percentage of cases with missing values was small. In the sample for
attendance trajectories (N = 11,479), only .8% of cases had a missing value. Just 1.8% in the
affiliation trajectory sample (N = 11,479) had a missing value. We imputed 10 data sets.
For our analysis of family formation, we drop respondents who were cohabiting, married,
or a parent at Wave 3 in order to establish time order for the effect of forming a family on
religious characteristics. After limiting the regression sample to those who were religiously
involved as adolescents but not in emerging adulthood, the regression analysis for religious
service attendance has Ns of 1,366 (weekly or more adolescent attenders) and 2,289 (monthly or
more adolescent attenders), and for affiliation the N is 970. We again imputed 10 data sets to
account for missing values, and missing values were again rare. In these samples, only 1.0%
(return to weekly attendance sample; N = 1,366), .7% (return to monthly attendance sample; N =
2,289), and .2% (reaffiliation sample; N = 970) had missing values. We do not use data from
Wave 2 because those who were high school seniors at Wave 1 were excluded from the Wave 2
data.
MEASURES
Key Dependent Variables
We examine five outcomes in this study. The first is a measure of religious service attendance
trajectories across the three waves. This variable is a composite of one religious service

13
attendance measure at each wave. At Wave 1, respondents were asked, “In the past 12 months,
how often did you attend religious services?” At Waves 3 and 4, respondents were asked, “How
often have you attended church, synagogue, temple, mosque, or religious services in the past 12
months?” At Wave 1 the response categories were “never,” “less than once a month,” “once a
month or more, but less than once a week,” and “once a week or more.” At Waves 3 and 4, the
response categories were “never,” “a few times,” “several times,” “once a month,” “2 or 3 times
a month,” “once a week,” and “more than once a week.” We created three categories of religious
service attendance based on these responses—Hi, Mid, and Lo. “Hi” reflects respondents who
attended religious services weekly or more, “Mid” reflects those who attended 1-3 times per
month (or 1 time per month through less than weekly), and “Lo” reflects those who attended less
than once per month. From this, we create a variable for every combination of Hi, Mid, and Lo
across the waves (e.g., Hi-Lo-Hi, Mid-Lo-Lo).
We follow a similar procedure with religious affiliation. At each wave Add Health
respondents were asked, “What is your present religion?”
4
Those who responded “None” were
considered to be unaffiliated; those who identified a religion were considered to be affiliated.
Wave 3 included a follow-up question for those who responded that they had an “other” religion,
and those who identified as “none/atheist/agnostic” to this question were also coded as
unaffiliated. We create a variable for every combination of affiliation across each of the waves
(e.g., Yes-Yes-Yes, No-Yes-No).
In the multivariate analysis, we examine two different binary variables for religious
service attendance at Wave 4: weekly or more, and monthly or more. We examine affiliation at
Wave 4 as a dichotomous variable.
Key Independent Variables
4
The word “present” was not included in the Wave 1 question.

14
We constructed a series of dichotomous variables measuring respondents’ relationship and
parental status at Wave 4: cohabiting and has a child, cohabiting and does not have a child,
single and has a child, single and does not have a child (reference), married and has a child, and
married and does not have a child. Using the detailed information provided by respondents in the
household roster, we assigned respondents to these categories based on whether they listed a co-
resident spouse or romantic partner and a co-resident biological, step, or adopted child.
Control Variables
Due to their well-established importance in analyses of both family structure and
religious characteristics, we included controls for gender, age, region, urbanicity, and
race/ethnicity. Gender, race-ethnicity, urbanicity, and region are taken from Wave 1. We also
include educational attainment indicators: less than high school degree, high school degree
(reference), some post-secondary education, and college degree or more. As measures of
respondents’ family background and marital history, we included indicators for whether the
respondent lived in an intact family at Wave 1 and whether the respondent was ever divorced or
separated. Finally, to control for religious subcultures, we relied on the Steensland et al. (2000)
RELTRAD scheme to construct religious tradition indicators: conservative Protestant
(reference), Mainline Protestant, Black Protestant, Catholic, and “other” religion. Descriptive
statistics for study variables can be found in Appendix A.
ANALYTIC APPROACH
We begin with descriptive statistics detailing the religious service attendance patterns of young
adults from adolescence to emerging adulthood to young adulthood (Table 1). We then limit our
attention to those who attended (1) weekly or more in adolescence and (2) monthly or more in
adolescence, and examine by age in order to assess potential differences at different stages of

15
young adulthood (Table 2). Then, we present the percentage of young adults who return to (1)
weekly and (2) monthly attendance by family structure after attending less than once per month
in emerging adulthood (Table 3). We restrict our sample to those who were single and childless
in emerging adulthood in order to better establish the time order of family formation and
religious return. We also show results stratified by gender. Table 4 displays our multivariate
models. We report odds ratios from logistic regression models predicting returns to weekly
(Models 1 and 2) and monthly (Models 3 and 4) religious service attendance in young adulthood
using the same sample from Table 3. We include bivariate models between family structure and
religious service attendance as well as models with controls.
Tables 5-8 parallel the first four tables but examine religious affiliation instead of
religious service attendance. Table 5 shows patterns of affiliation from adolescence to emerging
adulthood to young adulthood, Table 6 presents patterns of affiliation just among adolescent
affiliates and by age, Table 7 reports the percent affiliated among those who were adolescent
affiliates but unaffiliated in emerging adulthood, by family structure and by gender, and Table 8
presents results from logistic regression models predicting religious affiliation in young
adulthood among those who had disaffiliated between adolescence and emerging adulthood.
RESULTS
Table 1 about here
Religious Service Attendance
In order to situate our discussion of religious return in a broader context, Table 1 displays the
trajectories of religious service attendance from adolescence to emerging adulthood to young

16
adulthood.
5
To make sense of 27 different trajectories, we sort these trajectories into five larger
groups: Return Trajectories, Decline Trajectories, Stable Trajectories, Increase Trajectories, and
Fluctuating Trajectories.
6
Return Trajectories, defined by increasing one’s religious service
attendance in young adulthood after attending less than once per month in emerging adulthood,
occurs among only about 9% of young adults.
7
Much more common are the Decline Trajectories
(which describe about 36% of young adults). Also common is stable religious service attendance
(especially stably low attendance), which describes about 39% of young adults. Another 9%
have an Increase Trajectory, and 7% have a Fluctuating Trajectory. Thus, it appears decline and
stability are far more common than religious return among young adults.
Table 2 about here
Table 2 reports the trajectories for those who attended religious services weekly or more
as adolescents (Panel A) and those who attended 1-3 times per month as adolescents (Panel B).
This approach limits the analytic sample only to those who were “at risk” for decline in emerging
adulthood. We also report results by age groups to evaluate whether certain trajectories become
more common at older stages of young adulthood. In Panel A, we see echoes of the results from
5
We conducted parallel analyses for religious service attendance for those who stopped attending religious services
altogether at Wave III. It could be argued that these are the people who really “left” religion. Returns to religion are
even rarer among these respondents, with just 9% returning to weekly attendance and 19% to monthly or more
attendance. In regression models, single and married parents are more likely than childless singles to return to
weekly attendance, while cohabiting parents, married parents, and childless married adults are more likely to return
to monthly attendance. Full results from our parallel analyses are available from the first author upon request. We
ultimately opted against this approach, however, for both methodological and conceptual reasons. Methodologically,
this reduced our analytic sample for returning to weekly religious service attendance to just 365 and for returning to
monthly attendance to just 628. In other words, only 365 respondents went from weekly religious service in
adolescence to never attending in emerging adulthood, and only 628 went from monthly or more attendance in
adolescence to never attending in emerging adulthood. Conceptually, we are interested in those who removed
themselves from consistent involvement in religious communities. Limiting the sample to those who never attended
in emerging adulthood would eliminate respondents who, for example, attended with their parents when visiting
home or only on major holidays. This level of involvement still represents a dramatic decline in participation that
substantially curtails the influence of and role played by religious institutions in their lives.
6
Return Trajectories are a type of Fluctuating Trajectory but are treated as separate here because of our focus on
religious return.
7
Even if we include the Hi-Mid-Hi trajectory in this group—which could be considered a religious return under a
broader definition—the Return Trajectories only include a little more than 10% of young adults.

17
Table 1. Among weekly adolescent attenders, decline and stability are much more common than
return. About one third (32%) of weekly attenders diminish their attendance to less than once per
month in both emerging adulthood and young adulthood. Another 17% remain weekly attenders
throughout the transition to adulthood. Only about 14% of weekly attenders follow one of the
return trajectories (i.e., Hi-Lo-Mid or Hi-Lo-Hi). Many more young adults who went from
weekly attendance in adolescence to less than monthly attendance in emerging adulthood, then,
remain disengaged from religious communities than reengage (32% vs. 14%). This basic pattern
persists even among the oldest young adults (ages 30-31). Here 26% diminish their attendance to
less than once a month in emerging adulthood and remain at that frequency of attendance in
young adulthood, while about 13% diminish follow one of the return trajectories.
In Panel B, nearly half of all semi-regular attenders (1-3 times per month) decrease their
religious service attendance to less than once a month in emerging adulthood and then remain at
this level in young adulthood. Another 12% reduce their attendance in emerging adulthood but
return to semi-regular attendance in young adulthood, and an additional 5% return to weekly or
more frequent attendance in young adulthood after curbing their attendance in emerging
adulthood, meaning about 17% follow one of the trajectories of return. The age-stratified
analysis suggests that religious return may be slightly more common at older ages, but the
differences seen here are not drastic. Again, persistent decline is far more common than religious
return.
Table 3 about here

18
We now turn our focus solely to those individuals who were involved in religious
communities during adolescence but not in emerging adulthood.
8
Table 3 reports the percentage
of young adults returning to (1) weekly religious service attendance and (2) monthly religious
service attendance in young adulthood by their current family structure. Somewhat surprisingly,
the young adults who are most likely to return to weekly church attendance are single parents.
About one third of single parents who attended religious services weekly or more as adolescents
return to weekly (or more) religious service attendance in young adulthood after a break from
religious communities in emerging adulthood. Married parents are the second most likely to
return at about 23%. Just over 16% of married young adults without children return to weekly
attendance, followed by 9% of single, childless young adults and cohabitors with children, and
just 3% of cohabitors without children.
The second column reports the percentage of monthly or weekly adolescent attenders
who return to monthly or more attendance in young adulthood after attending less than once a
month in emerging adulthood.
9
Here, married parents are the most likely to return at 48%,
followed by 34% of single parents, 33% of both childless married adults and cohabiting parents,
25% of single childless adults, and 12% of childless cohabiters. Even by this broader definition
of religious return, most young adults do not return to religious communities after leaving in
emerging adulthood. Indeed, regardless of how we define a return to religious service
attendance—as either a return to weekly attendance or a return to monthly attendance—Table 3
shows that the majority of young adults whose religious involvement declines after adolescence
8
The samples in Table 3 only include respondents who were single and childless in emerging adulthood so as to
more accurately identify those whose family formation may have been the catalyst for their return to frequent church
attendance.
9
This includes weekly adolescent attenders who returned to monthly or more attendance.

19
do not return to their adolescent attendance patterns in their late 20s and early 30s—irrespective
of family formation.
Table 3 also shows differences across family structures by gender. Because of small cell
sizes that would violate the Add Health data user agreement, not all comparisons are possible.
Nevertheless, both men and women who are single and childless are less likely than childless
married young adults, married parents, and single parents to return to weekly or monthly
religious service attendance.
Table 4 about here
Table 4 displays odds ratios from logistic regression models predicting returns to weekly
and monthly religious service attendance. As in Table 3, the respondents in these samples were
single and childless in emerging adulthood. Model 1 shows the bivariate relationships among
family structure and returning to weekly or more attendance. As was seen in Table 3, single
parents who attended weekly or more as adolescents have nearly five times the odds of returning
to weekly attendance as childless single young adults. Married parents are nearly three times as
likely, and childless married young adults almost twice as likely. Childless cohabitors, by
contrast, have about 70% lower odds. There is no significant difference in returning between
childless singles and cohabitors with children.
Model 2 adds the control variables, which alters the size of the family formation effects,
though they are mostly still significant and in the same direction. In Model 2, married parents
and single parents who were adolescent weekly attenders have more than three times the odds as
childless singles of returning to weekly attendance, and childless married young adults are more
than twice as likely to attend weekly. Notably, the effect for single parents remains even after
controlling for gender (which itself is not statistically significant). This suggests that the findings

20
regarding single parents are not simply an artifact of the fact that single parents are more likely
to be women and women are more likely to be involved in religious communities than men.
Some other dynamic appears to be in play. Childless cohabitors have 61% lower odds of
returning to weekly attendance than childless singles, but the difference is no longer statistically
significant (p = .10) net of controls. An additional model (not shown) interacting gender with
family structure revealed only one significant interaction—childless cohabitation and gender,
such that cohabitation without a child had a negative effect for men, but not for women. Thus,
gender does not seem to be a consistent moderator of family formation effects in these models, in
contrast to what might have been expected from previous research on gender, family formation,
and religiosity (e.g., Sandomirsky and Wilson 1990).
Model 3 shows the bivariate relationships among family structure returning to monthly or
more frequent religious service attendance among those attending monthly or more in
adolescence. With this more inclusive sample, married parents clearly distinguish themselves
from childless singles (OR = 2.87, p < .001). Married childless adults have marginally (p < .10)
higher odds of returning to monthly attendance, and cohabiting childless adults have
significantly lower odds of returning (OR = .44, p <.001). Once controls are added in Model 4,
the effect of being a married parent is slightly stronger, as is the effect of being married but
childless (and it is now significant at p < .05). Single parents are not statistically different from
single childless adults in either model, and cohabiting childless adults remain significantly less
likely to return to monthly religious service attendance net of controls in Model 2. An additional
model (not shown) interacting gender and family structure found no significant interaction
effects.
Table 5 about here

21
Religious Affiliation
Table 5 reports the trajectories of religious affiliation from adolescence to emerging adulthood to
young adulthood.
10
Unlike religious service attendance, where consistently low religious
participation was the most common pattern, about two-thirds of young adults (66%) report
having a religious affiliation at all three waves. Just 7% were affiliated in adolescence,
unaffiliated in emerging adulthood, and reaffiliated in young adulthood. Another 7% drop their
affiliation in emerging adulthood and remain unaffiliated in young adulthood, while 7% more
maintain their affiliation in adolescence and emerging adulthood but drop it in young adulthood.
Five percent were consistently unaffiliated, and only 7% were unaffiliated as adolescents and
then affiliated later in the life course, meaning more unaffiliated adolescents identify with a
religion in emerging or young adulthood than remain unaffiliated throughout the transition to
adulthood.
Table 6 about here
Table 6 reports the trajectories of religious affiliation among only those who were
affiliated during adolescence. More than three-fourths of adolescent affiliates remain affiliated at
both emerging and young adulthood. Just 9% drop their affiliation in emerging adulthood and
reaffiliate in young adulthood. About 8% drop their affiliation in emerging adulthood and remain
10
We also analyzed a return to the same affiliation as one’s adolescence. The findings are similar to those reported
here for our simple measure of affiliation, with the exception that single parents have significantly higher odds than
single childless adults of reaffiliating with the same religious tradition in the bivariate model, and marginally higher
odds of reaffiliating in the multivariate model. Additionally, married childless young adults are not marginally more
likely to reaffiliate with the same tradition. Unfortunately, however, for this analysis we had to restrict our sample to
those respondents we could confidently classify, which reduced our sample size by about a third to 617. This results
from considerable measurement error at both Waves I and IV on account of respondents identifying as “Christian,”
“Just Christian,” “Just Protestant,” or “other religion.” Additionally, 2,170 of the full Wave I sample identified as
“Christian Church (Disciples of Christ),” a small mainline denomination with which only 23 respondents identified
at Wave 4. Clearly many respondents thought this was simply a generic Christian identity. Because of this, and
because we are interested in general returns to institutional religion, and because most of the scholarly research and
concern from religious communities is about young adults leaving religion (not merely switching traditions), we
only report findings for general reaffiliation. Our ancillary analyses, however, are available upon request from the
first author.

22
unaffiliated in young adulthood, and another 8% remain affiliated in emerging adulthood but not
in young adulthood. The oldest young adults (ages 30-31) are not reaffiliating at much higher
rates than younger young adults.
Table 7 about here
Table 7 displays the proportion of adolescent affiliates who reaffiliate with religion after
disaffiliating in emerging adulthood by their family structure. Again, the sample is limited to
those who were single and childless in emerging adulthood. Overall, just under half (49%) of
emerging adult disaffiliates reaffilliate in young adulthood. Childless cohabitors (39%) and
childless singles (47%) are the least likely to reaffiliate, while all three types of parents--
cohabiting parents (57%), single parents (69%), and married parents (63%)—are the most likely.
Childless married young adults occupy the middle ground at 55% reaffiliation.
There are apparent gender differences in religious affiliation for childless married young
adults. Women who are married but not mothers look similar to women who are single and
childless, while men who are married but not fathers are more likely (61% vs. 47%) to reaffilate
than single and childless men. Nevertheless, in multivariate models (not shown), no gender-
family structure interaction term was statistically significant.
Table 8 about here
Table 8 reports odds ratios from logistic regression models predicting reaffiliation in
young adulthood among those who disaffiliated in emerging adulthood. In Model 1, which
includes no controls, married parents are about twice as likely as childless singles to reaffiliate
with religion. Single parents are more than twice as likely, but this is not statistically significant
(presumably on account of a relatively small cell size for this group in this sample). The controls
added in Model 2 do not alter the relationship between married parents and reaffiliation (if

23
anything, it is slightly stronger in Model 2), but the controls do explain most of the
(nonsignificant) effect of single parenthood on reaffiliation. There is also a slight suppression
effect seen in Model 2, such that childless married adults are marginally more likely to reaffiliate
than single childless adults once accounting for the control variables.
DISCUSSION
The institutions of religion and family have both undergone significant changes in the last
several decades, and we have theorized the ways in which these changes may have affected the
well-established relationship between family formation and religion in young adulthood.
Specifically, we suggested that delayed marriage and childbearing, widespread premarital
cohabitation, and changing gender roles have weakened the connection between marriage and
parenthood and religious return among more recent cohorts of young adults. Nevertheless, we
have argued, religious institutions continue to appeal —for religious, social, practical, and
cultural reasons—at least to some young adults as they settle into family life.
Family formation appears to have a stronger effect on active participation in a religious
community compared with affiliation with a religious tradition. Participation in a religious
community requires continued, consistent commitment and effort, whereas religious affiliation
requires mere assent, often with little cost. At the same time, there may be less benefit to
religious affiliation, given the reduced social pressure to adhere to a religious tradition in the
contemporary United States (Hout and Fischer 2002). As a result, it is in religious service
attendance that we see the most variation in religious trajectories, and where family formation
appears to play a more significant role in religious return.
Our data suggest that the diversification of the American family, which has accelerated
over the past thirty years, is complemented by a diversity of pathways of religious return.

24
Although, due to limitations of the dataset, we cannot conclusively evaluate the reasons why
religious return may be more common among young adults with different family forms, the
combination of the findings here suggest the following interpretation. Just as there are diverse
pathways to adulthood, there appear to be at least two pathways to religious return via family
formation: a single-parent path, perhaps especially spurred by practical and cultural reasons, and
the “traditional” married (with children) path, likely particularly encouraged by social and
cultural reasons. Both pathways may also include young adults who return for religious reasons.
Our findings for single parents are particularly interesting. Single parents who attended
religious services weekly as adolescents are much more likely than their childless single
counterparts to return to a religious institution, and they are similar to married parents in this
respect. These findings qualify Wuthnow’s (2007) conclusion that fewer single parents than
childless single adults regularly attended religious services. We find that more single parents
than childless single adults return to a religious community after a hiatus in emerging adulthood.
Additionally, as might be inferred from the clear effects of single parenting on returning to
weekly attendance but not monthly attendance, participation in a religious community is
something of an all-or-nothing endeavor for single parents.
11
Single parents who reenter religious
institutions either participate regularly or not at all in religious services.
These religiously active single parents likely have compelling reasons to return to
religious participation. They may be drawn by practical support, religious resources, and cultural
factors (Petts 2012; Sullivan 2008). Religious institutions can provide an assortment of practical
benefits for parents—if even just an hour of free childcare once a week. Single parents—when
they choose to engage in religious communities—may avail themselves of these resources as
11
This is confirmed by ancillary analysis (not presented) showing single parents who were weekly adolescent
attenders are far more likely to attend weekly than less than once a month, but not more likely to attend 1-3 times
per month than less than once a month.

25
frequently as possible because staying at home on Sunday morning provides less chance for rest
than it might for two parents who can take turns caring for children or dividing labor.
Additionally, the trials and high stakes of parenting may lead parents to reach out for religious
support and inspiration—adaptive coping methods often emphasized in religious communities
(Mahoney 2010; Sullivan 2008). Single parents may experience higher levels of stress due to
lack of a partner to share in parenting, and thus such religious resources may be an important part
of their re-immersion in religious life. Single parents also may have added urgency to return to
regular involvement in a religious community in order to find a partner with whom to share in
raising a child. In any case, our findings do not suggest that stigmatization or religious
disapproval are preventing religious return among single parents. This may indicate rising
acceptance of nontraditional family forms among some religious communities, and the influence
of religious doctrines of grace and redemption that might help single parents sanctify their
nontraditional situation.
Cultural explanations may also help explain the full-fledged return to weekly religious
participation seen among single (and married) parents. Returning to weekly attendance is a
significant change and—just like other significant behavioral changes (e.g., quitting smoking,
staying in at night)—having a child may be a powerful motivator for such a change. In other
words, religious participation may be seen as part of a package of reprioritizations and lifestyle
changes that accompany parenthood that are designed to improve parenting and childrearing
conditions (Petts 2007).
Although some of the above mechanisms may also be operative for married parents, the
pattern of findings for marrieds (both with and without children), hints at a somewhat different
pathway for family formation and religious return. The combination of marriage and parenthood

26
is strongly associated with religious return in all our models; childless marriage is only
marginally associated with a return to religious service attendance, and only marginally
significant for reaffiliation once control variables are introduced. It is likely that for married
couples with children, all four types of mechanisms (religious, social, practical, and cultural) are
highly operative, resulting in higher odds of returning. For childless married adults there are
likely religious, social, and cultural reasons for religious return as well, though fewer than for
married parents. For example, given that marriage in the absence of children is associated with
returns to monthly attendance and (marginally) associated with weekly returns, married adults
may view religious participation—even if it is not regular—as part of the package of adulthood
and a “good” family life even if religious devotion is not the mechanism driving their return.
Childless marrieds may also receive affirmation for their marriage, but not the repeated
appreciation expressed to parents regarding their child’s contribution to congregational life
(Gallagher 2007). Married couples without children also cannot take advantage of all family-
related offerings religious institutions feature, such as childcare, parenting classes, and more.
Importantly, childless marrieds do not have a child for whom to provide a consistent role model
of faithful weekly attendance (Sherkat 2006).
Taken together, these findings suggest Schleifer and Chaves’ (2014) recent finding that
only school-age children (and not marriage or childbearing) have a direct effect on religiosity
may reflect the fact that family formation has little effect on the religiosity of those who were
never religious to begin with. In other words, it may be that only (or at least primarily) those who
have been exposed to religious institutions (and their norms and benefits) are drawn back to them
after marriage or childbearing—the idea may not occur or appeal to those who never attended
religious services while growing up.

27
Cohabitation also seems to have its own unique, negative effects on religious return. This
is in keeping with previous research on cohabitation and religiosity (Thornton et al. 1992;
Stolzenberg et al. 1995; Uecker et al. 2007). Conservative religious groups are well known for
their opposition to premarital sex and cohabitation, and even mainline Protestant groups whose
official discourse is tolerant of diverse family forms are often pro-nuptial at the congregational
level (Edgell 2006). The continued and public violation of religious teachings implied by
cohabitation is a problem for religious involvement in ways even nonmarital childbearing and
single parenting are not. While single parenting may be viewed as the outcome of an isolated
“sin,” cohabitation is seen as an ongoing one. The negative effect of cohabiting on religious
service attendance, however, appears to be counteracted by being a parent. Cohabitation also has
no discernable effect on reaffiliating with religion, vis-à-vis being a single and childless young
adult. The fact that cohabiting parents are more involved in religious communities than
cohabiting childless couples suggests that the benefits parents receive from involvement in a
religious community may to a certain extent outweigh the costs of involvement in social
networks that are thought to be disapproving (if not openly, at least in principle) of one’s
relationship status, and their continued religious affiliation suggests cohabitors’ break from
religion is not total.
Although it is important to consider the relative frequency of religious return, the
absolute levels of return also matter. The numbers are most remarkable for religious service
attendance. When considering evidence of religious disaffiliation, scholars often caution that life
course factors may eventually act as catalysts for religious return—in other words, to say that
drawing firm conclusions about disengagement are premature because youth often return once
they form families (Bengtson et al. 2013; Sherkat 2014; Wright 2010; Wuthnow 2007). The

28
evidence presented here suggests, however, that while this social process is still at work, family
formation does not lead to reengagement with religious communities in the majority of cases for
25-31-year-olds. Though family formation still affects rates of return, the majority of married
parents do not even return to monthly attendance once they have curbed their attendance in
emerging adulthood. Institutional religious return may still occur among these individuals later in
life—indeed, there are clear within-cohort increases in religiosity throughout the life course
(Sherkat 2014). That question can only be answered definitively as the current cohort of young
adults ages.
By contrast, the findings for religious affiliation paint a picture of religious consistency
across the transition to adulthood. Here the modal trajectory—found among nearly two thirds of
young adults, and more than three fourths of those who were affiliated as adolescents—is
consistent affiliation with a religion. Even among those who disaffiliate in emerging adulthood,
more than half pick up a religious affiliation again in young adulthood. This suggests that young
adults are not necessarily opposed to religious institutions altogether, but rather may have
relatively few reasons to commit deeply to those institutions. It also suggests that being a
religious “none” is not a particularly stable identity. Religious leaders concerned about the next
generation might consider how to re-engage these affiliated-but-uninvolved young adults and
might think creatively about the role of their institutions in the faith of young adults given shifts
toward more privatized understandings of religion in younger cohorts (Bengtson et al. 2013).
Conclusion
Explaining these absolute trends in reengagement and reaffiliation requires a historical
perspective that lies beyond the capacity of our data. Nevertheless, although we lack comparable
data on previous cohorts of young adults, it is likely our observed levels of religious return after

29
family formation are lower than they once were, as religious involvement itself has waned across
cohorts (Sherkat 2014). We theorize that delayed marriage and childbearing have created a
prolonged period of disengagement from religious communities that makes reengagement less
likely. Practical benefits may also be less useful as young adults now wait until they are already
financially stable to get married. Religious communities whose doctrine, programming, and
community life are geared toward traditional families may also be poorly positioned to appeal to
young adults increasingly likely to engage in premarital cohabitation and nonmarital
childbearing. Changing gender roles and weaker cultural norms about the roles of fathers and
mothers may also be a factor. Young adults also feel less social pressure than they once did to be
a part of a religious community after they marry (Hout and Fischer 2002; Sherkat 2014). Because
these cultural scripts have weakened, the decision to return to religion may be a more conscious
choice for today’s young adults than it was for previous cohorts.
There may be other reasons for this lack of return, however. These findings could reflect
a wider trend toward secularization (i.e., a period effect) where all cohorts are now less religious
than they used to be (Schwadel 2010). Or, religious return may simply happen later in the life-
course. With the age at first marriage and birth increasing (US Census Bureau 2011; Arroyo et
al. 2012), and the importance of school-age children for religious return (Stolzenberg et al. 1995;
Schleifer and Chaves 2014), young adults may eventually return to institutional religion—
perhaps in their late 30s or early 40s (Bengtson et al. 2013). Still, it is notable that even at ages
30 and 31 we see very little prevalence of religious return. Further studies, both historical and
prospective, will do much to clarify which (if any) of these explanations best fits the trends we
see. Additionally, our focus on attendance as a measure of religious participation may overlook
other kinds of religious behavior that are more salient for young adults, such as private

30
spirituality, which by some measures is less likely to decline compared to service attendance
(Mayrl and Uecker 2011; Uecker et al. 2007). Examining the role of family formation in these
more inward aspects of religion is a promising avenue for future research.
Several other lines of future research in this area are needed. Scholars should continue to
monitor the religious commitments of young adults as they and their children age. Other
extensions of the current research—such as variations across religious traditions, race-ethnicity,
and social class—may prove enlightening. And, although the Add Health survey has a large N,
cell sizes for some groups in our analysis are rather small once the sample restrictions are
applied. Thus, future surveys may specifically target those whose religiosity declined from
adolescence to emerging adulthood in order to gain a larger sample. In-depth interviews with
these individuals may also yield a clearer picture of the specific mechanisms at work.
For now, our study has shed considerable light on young adults’ commitment to religious
institutions during the transition to adulthood. Overall, the associations between family
formation and religious return reveal continued linkages between institutional religion and
family despite significant changes in both. We have found that though the majority of young
adults who disaffiliate in emerging adulthood do not return to their adolescent levels of religious
service attendance even after forming families, young adults who are married and childless,
married parents, and single parents are more likely than childless singles to return to religious
communities. Parents likely find a great deal of practical, religious, and cultural support for
raising their children, whether they are married or not, and married couples receive religious
validation and social opportunities while sending cultural signals to the wider community about
their accomplishments (since marriage is the “capstone” to adulthood [Cherlin 2009]). Religious
disaffiliation is far less common to begin with, but only married parenting (and in some instances

31
being married and childless) is associated with reaffiliation for young adults. The reasons for
affiliation and disaffiliation are less clear than they are for involvement in a religious community
and are thus more difficult to predict. Nevertheless, in terms of religious service attendance,
family formation—while still a factor in religious return—does not typically result in regular
religious service attendance among young adults who attended regularly as adolescents.

32
REFERENCES
Albrecht, Stan L., Marie Cornwall, and Perry H. Cunningham. 1988. “Religious Leave-Taking:
Disengagement and Disaffiliation among Mormons.” Pp. 62–80 in David G. Bromley, ed.,
Falling from the Faith: Causes and Consequences of Religious Apostasy. Sage.
Argue, Amy, David R. Johnson, and Lynn K. White. 1999. “Age and Religiosity: Evidence from
a Three-Wave Panel Analysis.” Journal for the Scientific Study of Religion 38:423–435.
Arnett, Jeffrey Jensen. 2014. Emerging Adulthood: The Winding Road from the Late Teens
Through the Twenties. New York: Oxford University Press.
Arroyo, Julia, Krista K. Payne, and Wendy D. Manning. 2012. Crossover in Median Age at First
Marriage and First Birth: Thirty Years of Change (FP-12-03). National Center for Family &
Marriage Research. Retrieved from http://ncfmr.bgsu.edu/pdf/family_profiles/
file107893.pdf.
Baker, Joseph O., and Buster G. Smith. 2009. “The Nones: Social Characteristics of the
Religiously Unaffiliated.” Social Forces 87(3): 1251-63.
Becker, Penny Edgell. 1999. Congregations in Conflict: Cultural Models of Local Religious Life.
New York: Cambridge University Press.
Becker, Penny Edgell, and Heather Hofmeister. 2001. “Work, Family, and Religious
Involvement for Men and Women.” Journal for the Scientific Study of Religion 40(4): 707-
22.
Bengtson, Vern L., Norella M. Putney, and Susan Harris. 2013. Families and Faith: How
Religion Is Passed Down Across Generations. New York: Oxford University Press.
Bianchi, Suzanne M., Melissa A. Milkie, Liana C. Sayer, and John P. Robinson. 2000. “Is
Anyone Doing the Housework? Trends in the Gender Division of Household Labor.” Social
Forces 79(1): 191-228.
Chaves, Mark. 2004. Congregations in America. Cambridge: Harvard University Press.
Cherlin, Andrew J. 2009. The Marriage-Go-Round: The State of Marriage and the Family in
America Today. New York: Alfred A. Knopf.
Chou, Hui-Tzu Grace. 2010. “The Perceived Relationship between Life Events and Religiosity
among Individuals Raised in a Mormon Community.” Mental Health, Religion, and Culture
13(5): 437-51.
Curtin, Sally C., Stephanie J. Ventura, and Gladys M. Martinez. 2014. “Recent Declines in
Nonmarital Childbearing in the United States. NCHS Data Brief No. 162. Hyattsville, MD:
National Center for Health Statistics.

33
Dillon, Michele, and Paul Wink. 2007. In the Course of a Lifetime: Tracing Religious Belief,
Practice, and Change. Berkeley: University of California Press.
Edgell, Penny. 2006. Religion and Family in a Changing Society. Princeton, NJ: Princeton
University Press.
Furstenberg, Frank F., Jr., Sheela Kennedy, Vonnie C. McLoyd, Rubén G. Rumbaut, and
Richard A. Settersten, Jr. 2004. “Growing Up Is Harder To Do.” Contexts 3:33-41.
Gallagher, Sally K. "Children as Religious Resources: The Role of Children in the Social Re‐
Formation of Class, Culture, and Religious Identity." Journal for the Scientific Study of
Religion 46.2 (2007): 169-183.
Goldstein, Joshua R. and Catherine T. Kenney. 2001. "Marriage Delayed or Marriage Forgone?
New Cohort Forecasts of First Marriage for U.S. Women." American Sociological Review
66:506-519.
Greeley, Andrew. 1989. Religious Change in America. Cambridge University Press.
Hadaway, C. Kirk, and Wade Clark Roof. 1988. “Apostasy in American Churches: Evidence
from National Survey Data.” Pp. 29–46 in David G. Bromley, ed., Falling from the Faith:
Causes and Consequences of Religious Apostasy. Sage.
Hill, Jonathan P. 2011. “Faith and Understanding: Specifying the Impact of Higher Education on
Religious Belief.” Journal for the Scientific Study of Religion 50:533-551.
Hochschild, Arlie Russell, with Anne Machung. 1989. The Second Shift. New York: Penguin.
Hoge, Dean R. 1981. Converts, Dropouts and Returnees. Pilgrim Press.
Hoge, Dean R., Benton Johnson, and Donald A. Luidens. 1993. “Determinants of Church
Involvement of Young Adults Who Grew Up in Presbyterian Churches.” Journal for the
Scientific Study of Religion 32:242–255.
Hout, Michael, and Claude S. Fischer. 2002. “Why More Americans Have No Religious
Preference: Politics and Generations.” American Sociological Review 67(2): 165-90.
Ingersoll-Dayton, Berit, Neal Krause, and David Morgan. 2002. “Religious Trajectories and
Transitions over the Life Course.” The International Journal of Aging and Human
Development 55:51–70.
Jesse, D. Elizabeth, Chantel Shoneboom, and Amy Blanchard. 2007. “The Effect of Faith or
Spirituality in Pregnancy: A Content Analysis.” Journal of Holistic Nursing 25: 151-58.

34
Kefalas, Maria J., Frank F. Furstenberg, Jr., Patrick J. Carr, and Laura Napolitano. 2011.
“‘Marriage Is More Than Being Together’: The Meaning of Marriage for Young Adults.”
Journal of Family Issues 32:845-875.
Mahoney, Annette. 2010. “Religion in Families, 1999-2009: A Relational Spirituality
Framework.” Journal of Marriage and Family 72(4): 805-27.
Manning, Christel J. 2013. “Unaffiliated Parents and the Religious Training of Their Children.”
Sociology of Religion 74(2): 149-75.
Mayrl, Damon, and Jeremy E. Uecker 2011. “Higher Education and Religious Liberalization
among Young Adults.” Social Forces 90:181-208.
Merino, Stephen M. 2012. “Irreligious Socialization? The Adult Religious Preferences of
Individuals Raised with No Religion.” Secularism and Nonreligion 1: 1-16.
Myers, Scott M. 1996. “An Interactive Model of Religiosity Inheritance: The Importance of
Family Context.” American Sociological Review 61(5): 858-66.
O’Connor, Thomas P., Dean R. Hoge, and Estrelda Alexander. 2002. “The Relative Influence of
Youth and Adult Experiences on Personal Spirituality and Church Involvement.” Journal for
the Scientific Study of Religion 41(4): 723-32.
Pankhurst, Jerry G., and Sharon K. Houseknecht. 2000. “Introduction.” Pp. 1-40 in Sharon K.
Houseknecht and Jerry G. Pankhurst, eds., Family, Religion, and Social Change in Diverse
Societies. New York: Oxford University Press.
Petts, Richard J. 2007. “Religious Participation, Religious Affiliation, and Engagement with
Children among Fathers Experiencing the Birth of a New Child.” Journal of Family Issues
28(9): 1139-61.
______. 2009. “Trajectories of Religious Participation from Adolescence to Young Adulthood.”
Journal for the Scientific Study of Religion 48(3): 552-71.
_____. 2012. “Single Mothers’ Religious Participation and Early Childhood Behavior.” Journal
of Marriage and Family 74: 251-68.
Putnam, Robert D., and David E. Campbell. 2010. American Grace: How Religion Divides and
Unites Us. New York: Simon & Schuster.
Roof, W. Clark. 1990. “Return of the Baby Boomers to Organized Religion.” Pp. 284-291 in
Constant Jaquet, ed., Yearbook of American and Canadian Churches. Abingdon Press.
Sandomirsky, Sharon, and John Wilson. 1990. “Processes of Disaffiliation: Religious Mobility
among Men and Women.” Social Forces 68:1211–1229.

35
Schleifer, Cyrus, and Mark Chaves. 2014. “Family Formation and Religious Service Attendance:
Untangling Marital and Parental Effects.” Sociological Methods & Research. doi:
10.1177/0049124114526376.
Schoen, Robert, Nancy S. Landale, and Kimberly Daniels. 2007. “Family Transitions in Young
Adulthood.” Demography 44:807-820.
Schwadel, Philip. 2010. “Age, Period, and Cohort Effects on U.S. Religious Service Attendance:
The Declining Impact of Sex, Southern Residence, and Catholic Affiliation.” Sociology of
Religion 71(1): 2-24.
Sherkat, Darren E. 2006. “Preference Structures and Normative Constraints in Movements
Outside, Between, and within Religious Organizations.” Pp. 146–64 in Theorising Religion:
Classical and Contemporary Debates, edited by J. A. Beckford and J. Walliss. Burlington,
Vermont: Ashgate.
______. 2014. Changing Faith: The Dynamics and Consequences of Americans’ Shifting
Religious Identities. New York: New York University Press.
Smith, Christian, with Patricia Snell. 2009. Souls in Transition: The Religious & Spiritual Lives
of Emergina Adults. New York: Oxford University Press.
Steensland, Brian, Jerry Z. Park, Mark D. Regnerus, Lynn D. Robinson, W. Bradford Wilcox,
and Robert D. Woodberry. 2000. “The Measure of American Religion: Toward Improving
the State of the Art.” Social Forces 79:291–318.
Stolzenberg, Ross M., Mary Blair-Loy, and Linda J. Waite. 1995. “Religious Participation in
Early Adulthood: Age and Family Life Cycle Effects on Church Membership.” American
Sociological Review 60:84–103.
Sullivan, Susan Crawford. 2008. “Unaccompanied Children in Churches: Low-Income Urban
Single Mothers, Religion, and Parenting.” Review of Religious Research 50(2): 157-75.
______. 2012. Living Faith: Everyday Religion and Mothers in Poverty. Chicago: University of
Chicago Press.
Thornton, Arland, William G. Axinn, and Daniel H. Hill. 1992. “Reciprocal Effects of
Religiosity, Cohabitation, and Marriage.” American Journal of Sociology 98:628–651.
Twenge, Jean M., Julie J. Exline, Joshua B. Grubbs, Ramya Sastry, and W. Keith Campbell.
2015. "Generational and time period differences in American adolescents’ religious
orientation, 1966–2014." PLOS ONE doi: 10.1371/journal.pone.0121454.
Uecker, Jeremy E., Mark D. Regnerus, and Margaret L. Vaaler. 2007. “Losing My Religion: The
Social Sources of Religious Decline in Early Adulthood.” Social Forces 85:1667–1692.

36
United States Census Bureau. 2011. “Estimated Median Age at First Marriage, by Sex: 1890 to
the Present.” Accessed online May 27, 2015 from http://www.census.gov/population/
socdemo/hh-fam/ms2.xls.
Wilcox, W. Bradford. 2004. Soft Patriarchs, New Men: How Christianity Shapes Fathers and
Husbands. Chicago: University of Chicago Press.
Wilcox, W. Bradford, Andrew J. Cherlin, Jeremy E. Uecker, and Matthew Messel. 2012. “No
Money, No Honey, No Church: The Deinstitutionalization of Religious Life among the
White Working Class.” Research in the Sociology of Work 23:227-250.
Wilcox, W. Bradford, Mark Chaves, and David Franz. 2004. “Focused on the Family? Religious
Traditions, Family Discourse, and Pastoral Practice.” Journal for the Scientific Study of
Religion 43:491-504.
Wilcox, W. Bradford, and Nicholas H. Wolfinger. 2007. “Then Comes Marriage? Religion,
Race, and Marriage in Urban America.” Social Science Research 36: 569-89.
Wilson, John, and Darren E. Sherkat. 1994. “Returning to the Fold.” Journal for the Scientific
Study of Religion 33(2): 148-61.
Wright, Bradley R.E. 2010. Christians Are Hate-Filled Hypocrites…and Other Lies You’ve Been
Told: A Sociologist Shatters Myths from the Secular and Christian Media. Bloomington,
MN: Bethany House.
Wuthnow, Robert. 2007. After the Baby Boomers: How Twenty- and Thirty-Somethings Are
Shaping the Future of American Religion. Princeton, NJ: Princeton University Press.

37
Table 1. Religious Service Attendance Trajectories from Adolescence to Emerging Adulthood to Young Adulthood
%
Return Trajectories (Total)a
8.7
Hi-Lo-Mid
3.1
Hi-Lo-Hi
2.3
Mid-Lo-Mid
2.3
Mid-Lo-Hi
1.0
Decline Trajectories (Total)
36.2
Hi-Lo-Lo
12.3
Mid-Lo-Lo
9.3
Hi-Mid-Mid
3.6
Hi-Mid-Lo
3.4
Hi-Hi-Mid
3.1
Hi-Hi-Lo
2.6
Mid-Mid-Lo
1.9
Stable Trajectories (Total)
38.5
Lo-Lo-Lo
30.7
Hi-Hi-Hi
6.4
Mid-Mid-Mid
1.4
Increase Trajectories (Total)
9.3
Lo-Lo-Mid
3.3
Lo-Lo-Hi
1.8
Lo-Hi-Hi
.8
Lo-Mid-Mid
1.1
Lo-Mid-Hi
.6
Mid-Hi-Hi
1.0
Mid-Mid-Hi
.7
Fluctuating Trajectories (Total)a
7.2
Lo-Hi-Lo
1.3
Lo-Hi-Mid
.6
Lo-Mid-Lo
2.2
Mid-Hi-Lo
.7
Mid-Hi-Mid
.7
Hi-Mid-Hi
1.7
N = 11,479
a
a Return Trajectories are a type of Fluctuating Trajectory but are shown here as a separate
category because of our conceptual focus on returning to religion.

38
Table 2. Religious Service Attendance Trajectories from Adolescence to Emerging
Adulthood to Young Adulthood among Adolescent Weekly Attenders, Overall and by Age
Full sample
Age 25-27
Age 28-29
Age 30-31
Panel A. Attended weekly or more as adolescents.
Hi-Lo-Lo
31.9
35.0
31.2
26.1
Hi-Hi-Hi
16.7
13.9
17.3
22.1
Hi-Mid-Mid
9.3
9.5
7.8
11.3
Hi-Mid-Lo
8.9
10.1
8.7
6.6
Hi-Lo-Mid
8.2
8.8
8.0
7.0
Hi-Hi-Mid
8.0
7.9
8.1
8.0
Hi-Hi-Lo
6.7
7.2
5.7
7.2
Hi-Lo-Hi
6.0
4.4
7.9
6.2
Hi-Mid-Hi
4.5
3.2
5.6
5.4
N
4,585
1,736
1,817
1,032
Panel B. Attended 1-3 times per month as adolescents.
Mid-Lo-Lo
49.1
49.2
49.6
48.3
Mid-Lo-Mid
11.8
9.6
13.4
13.0
Mid-Mid-Lo
10.1
12.6
9.1
7.8
Mid-Mid-Mid
7.4
7.2
6.8
8.4
Mid-Lo-Hi
5.4
4.5
5.8
6.0
Mid-Hi-Hi
5.3
4.5
5.7
6.0
Mid-Hi-Mid
3.8
3.5
4.0
4.1
Mid-Hi-Lo
3.7
6.0
1.7
3.1
Mid-Mid-Hi
3.5
3.0
4.1
3.3
N
2,205
682
839
683
Notes: Rows in bold are religious return trajectories.

39
Table 3. Percent Returning to Weekly and Monthly Religious Service Attendance by
Family Formation, Overall and by Gender
Return to weekly
attendance a
Return to monthly
attendance b
All young adults
Overall
12.2
28.2
Cohabiting, child
8.7
32.8
Cohabiting, no child
3.0
12.4
Single, child
33.8
34.4
Single, no child
9.4
24.6
Married, child
22.5
48.3
Married, no child
16.3
33.0
Women
Overall
14.8
31.6
Cohabiting, child
13.0
29.5
Cohabiting, no child
7.0
16.8
Single, child
32.2
31.8
Single, no child
9.5
28.1
Married, child
28.9
51.5
Married, no child
15.2
34.9
Men
Overall
10.3
25.8
Cohabiting, child
------c
35.3
Cohabiting, no child
------c
9.4
Single, child
39.3
48.0
Single, no child
9.3
22.7
Married, child
17.6
45.6
Married, no child
17.6
31.0
N = 1,366 (weekly) and 2,289 (monthly)
a Sample is those who attended religious services weekly or more as adolescents, less than once per month in
emerging adulthood, and were single and childless in emerging adulthood
b Sample is those who attended religious services monthly or more as adolescents, less than once per month in
emerging adulthood, and were single and childless in emerging adulthood
c Cell sizes too small to report.

40
Table 4. Odds Ratios from Logistic Regression Models Predicting Return to Weekly and Monthly Religious Service
Attendance in Young Adulthood
Return to
weekly or more attendance a
Return to
monthly or more attendance b
M1
M2
M3
M4
Cohabiting, child
.92
.85
1.50
1.66
Cohabiting, no child
.30*
.39
.44***
.50**
Single, child
4.94***
3.11*
1.61
1.24
Married, child
2.80**
3.06**
2.87***
3.08***
Married, no child
1.89†
2.14†
1.51†
1.64*
Female
1.42
1.25
Age
1.16*
1.05
West
.37**
.52**
Midwest
.69
.65*
Northeast
.30**
.52**
Urban
1.38
1.49*
Rural
1.18
1.20
Hispanic
3.01**
1.80*
Black
3.66*
1.88†
Asian
4.36**
1.39
Other race
.34
.95
Less than high school degree
1.29
1.17
Some post-secondary education
1.05
1.19
College degree or more
1.44
1.70*
Intact family, Wave 1
.70
1.21
Ever divorced or separated
1.20
1.26
Mainline Protestant
1.10
1.11
Black Protestant
.47
1.16
Catholic
.47
.88
Other religion
1.01
.88
N
1,366
1,366
2,289
2,289
*** p < .001 ** p < .01 * p < .05 † p < .10
Reference groups are single, no child; South; suburban; White; high school degree; conservative Protestant.
a Sample is those who attended religious services weekly or more as adolescents, less than once per month in
emerging adulthood, and were single and childless in emerging adulthood
b Sample is those who attended religious services monthly or more as adolescents, less than once per month in
emerging adulthood, and were single and childless in emerging adulthood

41
Table 5. Religious Affiliation
Trajectories from Adolescence to
Emerging Adulthood to Young Adulthood,
in Descending Order of Frequency
%
Yes-Yes-Yes
66.1
Yes-No-Yes
7.4
Yes-No-No
7.0
Yes-Yes-No
6.9
No-No-No
5.4
No-Yes-Yes
3.7
No-No-Yes
2.2
No-Yes-No
1.4
Notes: Row in bold is religious return trajectory. N = 11,479

42
Table 6. Religious Affiliation Trajectories from Adolescence to Emerging Adulthood to
Young Adulthood among Adolescent Affiliates, Overall and by Age
Full sample
(N = 10,106)
Age 25-27
(N = 3,368)
Age 28-29
(N = 3,951)
Age 30-31
(N = 2,787)
Yes-Yes-Yes
75.6
73.2
75.6
79.5
Yes-No-Yes
8.5
9.1
8.2
7.8
Yes-No-No
8.0
9.0
8.3
6.1
Yes-Yes-No
7.9
8.7
7.9
6.7
Note: Row in bold is religious return trajectory.

43
Table 7. Percent Reaffiliating with
Religion in Young Adulthood (N = 970)
%
All young adults
Overall
49.4
Cohabiting, child
57.4
Cohabiting, no child
39.1
Single, child
68.6
Single, no child
46.6
Married, child
63.4
Married, no child
54.7
Women
Overall
48.3
Cohabiting, child
54.1
Cohabiting, no child
38.8
Single, child
62.6
Single, no child
46.3
Married, child
61.3
Married, no child
46.8
Men
Overall
50.0
Cohabiting, child
59.6
Cohabiting, no child
39.3
Single, child
78.8
Single, no child
46.7
Married, child
65.3
Married, no child
61.4

44
Table 8. Odds Ratios from Logistic Regression Models Predicting Reaffiliation in Young Adulthood (N=970)
M1
M2
Cohabiting, child
1.54
1.22
Cohabiting, no child
.74
.83
Single, child
2.50
1.36
Married, child
1.98**
2.08**
Married, no child
1.38
1.69†
Female
1.01
Age
1.02
West
.67†
Midwest
.61†
Northeast
.57†
Urban
1.19
Rural
.81
Hispanic
1.36
Black
1.05
Asian
1.81
Other race
.85
Less than high school degree
.73
Some post-secondary education
.61
College degree or more
.52*
Intact family, Wave 1
1.17
Ever divorced or separated
1.82†
Mainline Protestant
.89
Black Protestant
2.76
Catholic
.78
Other religion
.53†
*** p < .001 ** p < .01 * p < .05 † p < .10
Reference groups are single, no child; South; suburban; White; high school degree; conservative Protestant.

45
Appendix A. Descriptive Statistics for All Study Variables, among [Adolescent Weekly Attenders/Affiliates] Who
Were Single and Childless in Early Adulthood and [Attended Religious Services Less than Once Per Month/Were
Unaffiliated] in Early Adulthood
Return to
weekly or
more
attendance
(N = 1,366)
Return to
monthly or
more
attendance
(N = 2,289)
Reaffiliation
(N =970)
Min/Max
Dependent Variable
Attends weekly or more, Wave 4
.12
----
----
0, 1
Attends at least once a month, Wave 4
----
.28
----
0, 1
Affiliated with religion, Wave 4
----
----
.49
0, 1
Independent Variables
Single, child, Wave 4
.04
.04
.03
0, 1
Single, no child, Wave 4
.49
.46
.48
0, 1
Cohabiting, child, Wave 4
.05
.06
.07
0, 1
Cohabiting, no child, Wave 4
.14
.15
.20
0, 1
Married, child, Wave 4
.13
.14
.12
0, 1
Married, no child, Wave 4
.15
.15
.10
0, 1
Female, Wave 1
.42
.42
.36
0, 1
Age, Wave 4
27.56
27.69
27.63
25-31
South, Wave 1
.38
.37
.28
0, 1
West, Wave 1
.15
.16
.22
0, 1
Midwest, Wave 1
.32
.32
.35
0, 1
Northeast, Wave 1
.16
.16
.15
0, 1
Suburban, Wave 1
.60
.61
.61
0, 1
Urban, Wave 1
.26
.25
.27
0, 1
Rural, Wave 1
.13
.14
.13
0, 1
White, Wave 1
.65
.66
.75
0, 1
Hispanic, Wave 1
.11
.12
.08
0, 1
Black, Wave 1
.16
.15
.10
0, 1
Asian, Wave 1
.05
.04
.04
0, 1
Other race, Wave 1
.03
.03
.03
0, 1
Less than high school degree, Wave 4
.06
.05
.06
0, 1
High school degree, Wave 4
.14
.15
.18
0, 1
Some post-secondary education, Wave
4
.40
.40
.38
0, 1
College degree or more, Wave 4
.40
.40
.38
0, 1
Intact family, Wave 1
.66
.64
.60
0, 1
Ever divorced or separated, Wave 4
.08
.08
.07
0, 1
Conservative Protestant, Wave 1
.20
.18
.18
0, 1
Mainline Protestant, Wave 1
.26
.27
.32
0, 1
Black Protestant, Wave 1
.15
.13
.09
0, 1
Catholic, Wave 1
.31
.32
.26
0, 1
Other religion, Wave 1
.09
.09
.15
0, 1