The Dynamics and Correlates of Religious Service Attendance in Adolescence

Abstract

This study examines changes in religious service attendance over time for a contemporary cohort of adolescents moving from middle to late adolescence. We use two waves of a nationally representative panel survey of youth from the National Study of Youth and Religion (NSYR) to examine the dynamics of religious involvement during adolescence. We then follow with an analysis of how demographic characteristics, family background, and life course transitions relate to changes in religious service attendance during adolescence. Our findings suggest that, on average, adolescent religious service attendance declines over time, related to major life course transitions such as becoming employed, leaving home, and initiating sexual activity. Parents’ affiliation and attendance, on the other hand, are protective factors against decreasing attendance.

Keywords

church attendance emerging adulthood/adult transition longitudinal design

Interest in the religious lives of youth has been increasing in recent years, spurred by a large body of research extolling the benefits of religious involvement for young people’s well-being, attainment, and community involvement (e.g., Adamczyk, 2012; Glanville, Sikkink, & Hernández, 2008; Jang & Johnson, 2011; Pearce & Denton, 2011; Smith, 2003a, 2003b). Of particular interest in this article is young people’s pattern of religious service attendance in the transition to adulthood.

Religious service attendance is an important outcome of study because it has been linked to so many life outcomes for youth. Smith (2003b) suggests that religious communities offer the sort of closed network of adults that Coleman (1988) described as producing social capital. Congregation members may share information about community resources, schools, and local news, which parents may use to help guide their children’s academic and other decision making. In addition, youth may gain cultural capital and social skills through their religious involvement. Religious congregations often offer youth opportunities to learn important life skills through public service or community work, service to the congregation, public speaking, group decision making, or event planning (Smith, 2003a). These skills can then be transferred outside the congregation and benefit youth in other areas of their lives. Research consistently finds youth religious service attendance to be related to numerous positive psychosocial, educational, and behavioral outcomes, including academic achievement (Muller & Ellison, 2001; Regnerus, 2000), involvement in volunteering and other service activities (Youniss, McLellan, & Yates, 1999), and feelings of positive self-worth and efficacy (Bergin, 1983; Donahue & Benson, 1995; Markstrom, 1999). Furthermore, numerous studies have found evidence of negative relationships between religiosity and drug use, delinquency, (Cochran, 1992; Pearce & Haynie, 2004), and sexual activity (Billy, Brewster, & Grady, 1994; Rostosky, Regnerus, & Wright, 2003) among adolescents.

Although most research finds that religious service attendance is associated with positive outcomes among youth, there is evidence of some negative associations as well. For instance, recent work demonstrates that frequent religious service attendance is associated with lower self-rated intrapersonal well-being and less positive parental relationships among older adolescents, although this relationship was reversed among younger adolescents (Good, Willoughby, & Fritjers, 2009). There is also mixed evidence, with some showing negative associations, for the relationship between religiosity and mental health (see Cotton, Zebracki, Rosenthal, Tsevat, & Drotar, 2006 for a review). The influence of adolescent religiosity may also depend on context. Youth from fundamentalist Protestant families have lower educational outcomes than both other religious and nonreligious teens (Beyerlein, 2004; Lehrer, 1999), while youth who differ from their parents in their level of religious service attendance exhibit lower levels of educational attainment and higher levels of delinquent behavior than their peers (Pearce & Hardie, 2008; Pearce & Haynie, 2004).

Although there is much evidence regarding the benefits and drawbacks of religious institution attachment and an assumption that this attachment decreases in adolescence, there is still more to learn about how and why religious involvement changes during late adolescence. In this study, we explore changes in religious service attendance over time using two waves of a nationally representative sample of American youth fielded 3 years apart. We test the relationships between young adults’ attendance over time and demographic characteristics, family background, and major life course transitions to construct a comprehensive portrait of youth religious attachment and its correlates. We focus on religious service attendance rather than other indicators of youth religiosity, such as the frequency of prayer or importance of religion, because religious service attendance is a unique aspect of religiosity involving public practice and social interactions that are directly tied to the extension of adolescent social and cultural capital.

Dynamics of Religious Participation in Adolescence

Religious service attendance appears to vary across the life course. One study found that 37.5% of a nationally representative sample of adults had stopped attending religious services for at least 2 years at some point, and that 40% of these “dropouts” had stopped attending services in adolescence (Roozen, 1980). One of the most influential early studies of change over time was conducted by Ozorak (1989), who analyzed data from a regional study of three suburban high schools in the Northeast. High school- and college-aged youth reported their present and past religious affiliations and behaviors. Based on the evidence from this study, Ozorak concluded that “polarization” occurred in adolescence, such that highly religious youth became more religious, while others became less religious. This polarization trend was expected for most adolescents, but was enhanced by other factors. In particular, Ozorak argued that for many adolescents, developmental changes in cognition might lead to questioning their religious beliefs and either straying from or reaffirming their theological commitments. The presence of close networks of religious adherents were expected to increase religious observance, while positive or negative emotional experiences in adolescence could lead young people toward or away from religion. However, in addition to relying on retrospective data, this study used a sample of almost exclusively White, middle class youth, the majority of whom were Catholic. Using a more recent nationally representative sample, Pearce and Denton (2011) find that cognitive development is reflected in many adolescents’ desire to make “a faith of their own.” Sometimes this results in a decrease in religious adherence, but more often than not, youth report feeling more religious, because it means more to them.

In another study, Wilson and Sherkat (1994) used a nationally representative and longitudinal study of young men and women to investigate change in religious attachment over time. They found that youth religiosity, measured by “expression of a religious preference” (Wilson & Sherkat, 1994, p.152), declined in adolescence before stabilizing or increasing in response to important life course transitions such as marriage and childbearing. Although suggestive of the importance of life course events, particularly for older adolescents, the data for this study were drawn between the years of 1965 and 1982 and may be dated. Furthermore, measurement of an “expression of religious preference” is distinct from measuring the frequency of religious service attendance. Religious identity is often more stable than religious practice, with the latter having the potential to be more influenced by life events and contextual circumstances. Integration into a religious congregation through public religious practice is the mechanism through which individuals accrue many of the benefits outlined above.

Studies using repeated cross-sectional data found that the religious service attendance of high schoolers declined between the years 1976 and 1986, before stabilizing in the late 1980s and 1990s (Smith, Denton, Faris, & Regnerus, 2002; U.S. Department of Education, Office of Education Research and Improvement, 1993; Wallace, Forman, Caldwell, & Willis, 2003). Samples drawn from particular religious subgroups have also shed some light on changes in religious attachments over time. In one study of Seventh Day Adventists, Dudley (1999) found that 25% of an original sample of 15- and 16-year-old active church members had dropped out 10 years later. Another study of Latter Day Saints found that young people were particularly likely to decrease their participation in religious services in their late adolescents through early twenties (Albrecht, Cornwall, & Cunningham, 1988).

More recent research has drawn on nationally representative longitudinal samples to examine religious change over time among adolescents and young adults. Smith and Snell (2009) examined religious change from youth to young adulthood and identified several different trajectories of change. Their work highlights the role of parents and congregations in shaping the religious futures of youth. The measure used in this analysis was a combined measure of religious salience, belief, and practice. This provided an overview of the religiosity of youth, but not the mechanisms of change specifically related to attendance.

Regnerus and Uecker (2006) and Uecker, Regnerus, and Vaaler (2007) use Add Health data to examine the factors related to declines in religious service attendance among youth and young adults. They found evidence for declines in religious practice as youth move into late adolescence and early adulthood. Yet Regnerus and Uecker (2006) measure only dramatic declines in adolescent attendance over a 1-year interval. Uecker et al. (2007) use a dichotomous measure to compare any decline in attendance to no decline in attendance as respondents entered early adulthood. Both analyses suggest that family structure and parental religion, as well as life experiences such as education and risk behaviors contribute to a decline in the religious participation of youth and young adults.

We extend the research described above by examining a contemporary, nationally representative sample at two points in adolescence, 3 years apart, to measure the magnitude of change in religious service attendance during adolescence. We examine detailed measures of religious service attendance to capture both large and small fluctuations in participation. In addition to examining demographic characteristics and family background, we take a closer look at the common life transitions experienced in this period of life where youth are gaining autonomy but have not yet entered fully into early adulthood.

Correlates of Youth Religious Participation

Demographic Correlates of Youth Religious Participation

Several demographic characteristics have been tied to youth religiosity, although there is less evidence regarding their affect on change in religiosity over time. In many studies, young women are more likely to be religious than young men and participate more actively in religious institutions (Gunnoe & Moore, 2002; Smith et al., 2002; Smith, Sikkink, & Bailey, 1998; Wallace et al., 2003; Willits & Crider, 1989). Among adult women, this difference has been explained by women’s reduced participation in the labor force (De Vaus & McAllister, 1987). However, the demonstrated gender gap in religious service attendance between adolescent boys and girls has not been statistically accounted for in previous research.

Race and ethnicity are also related to religious service attendance. Black adolescents and adults report, on average, higher levels of religiosity than other racial groups, both overall and controlling for other factors (Gunnoe & Moore, 2002; Wallace et al., 2003). They also exhibit higher levels of religious service attendance and religious youth group participation than Whites (Smith et al., 2002; Wallace et al., 2003). Some of these differences may be explained by the variance in the cultural importance that religious institutions and involvement play in the lives of African Americans.

Region of residence is also significantly related to religious service attendance and religiosity. Young people and adults in the South are more likely than people living in other parts of the United States to report high levels of religious attachment (Gunnoe & Moore, 2002; Smith et al., 2002; Wallace et al., 2003). Smith et al. (1998) found that this applies primarily to Southerners who lived in the region in both adolescence and adulthood, and somewhat for those who moved to the South after age 16. Southerners who move out of the region demonstrate significantly less religious attachment compared to nonmigrant Southerners.

Family Background Correlates of Youth Religious Participation

Parental characteristics are important predictors of adolescent religious service attendance. Parents’ education is positively related to stated religious importance and religious attendance among 8th, 10th, and 12th graders (Wallace et al., 2003). The relationship between income and religious service attendance is somewhat more complex, with middle income households appearing more likely to participate in religious services than either poor or wealthy households (Pearce, Foster, and Hardie, forthcoming). Given that many religious traditions place an emphasis on the importance of having two parents in the home, this family structure may also be correlated with more regular religious service attendance (Edgell, 2006). Adolescents are also more likely to attend religious services when their parents do (Dudley, 1999; Ozorak, 1989), particularly when parents and children are close (Gunnoe & Moore, 2002). Parents’ affiliation and religiosity provide a context in which adolescents’ own religious beliefs are augmented. Religious service attendance may be particularly related to parental attitudes and behavior, because attendance is a visible and concrete act which parents may insist upon (Hoge, Petrillo, & Smith, 1982). Yet it is not only the presence of parents that promote youth attendance, as this effect remains after children have left their parents’ home (Myers, 1996).

Life Course Transition Correlates of Youth Religious Participation

The transition to adulthood is a key period of the life course in which young people develop attitudes and behaviors they will carry into adulthood. As described above, the cognitive growth that occurs during this period may lead many young people to question their religious beliefs and those of their religious communities (Kagan, 1971; Ozorak, 1989). Yet it is also an unstable time, marked by high levels of migration (Garasky, 2002), economic instability (Hill & Yeung, 1999), and separation from one’s parents (Garasky, Haurin, & Haurin, 2001). As a result, adolescents may experience uneven disruptions in their attachment to religious institutions depending on their individual life course trajectories, at a time when such attachment may be particularly beneficial.

Young people make important decisions about postsecondary education, work, and family formation in adolescence. These transition experiences may inhibit religious service attendance as they compete for the time and attention of adolescents (Uecker et al., 2007). Work and schooling create time demands on young people’s schedules, which may reduce their attendance. In addition, adolescents’ religious service attendance is strongly related to their parents’ attendance (Cnaan, Gelles, & Sinha, 2004), and when they move away, they may lose one incentive to go to religious services. Lastly, adolescence is often a time of initiation into alcohol use and sexual activity. These behaviors may create cognitive dissonance for those who regularly attend religious services, possibly leading one to choose continuing one but not the other (Regnerus & Uecker, 2006).

One long-standing notion has been that a college education has a secularizing effect and that young adults who attend college are more likely to experience religious decline. However, recent studies have found that whatever secularizing effect college has had in the past may now be diminishing and college attendance may even facilitate continued religious engagement (Smith & Snell, 2009; Uecker et al., 2007). It may be that the higher level of structure offered by 4-year college campuses and access to campus religious groups that serve to reinforce the religious practices of students maintain young people’s ties to religion. Youth who transition to other types of postsecondary schools or to the labor force may have fewer resources and thus less access to institutional support. As college enrollment continues to rise in the United States, it is important to understand the implications of this trend for the religious lives of adolescence. We suggest that there may be an important relationship between major life transitions, including the transition to postsecondary education, and patterns of religious service attendance.

Current Study

For this study, we are particularly interested in how religious service attendance changes during adolescence in response to key life course experiences and transitions. Based on prior research, we test the following hypotheses:

Hypothesis 1a (H1a): Consistent with the polarization hypothesis, we expect to find that highly religious youth begin attending religious services more frequently during adolescence, whereas less religious youth decrease their rate of attendance. This should emerge in bivariate comparisons, and may be explained by differences in parental religiosity and transition experiences.

Hypothesis 1b (H1b): Consistent with the overall decline hypothesis, we expect to find that religious service attendance declines over time for all adolescents.

Hypothesis 2 (H2a): Consistent with the secularization hypothesis, we expect to find that young people who begin attending postsecondary school will experience a higher drop in religious service attendance than those who do not enter postsecondary schooling.

Hypothesis 2 (H2b): Conversely, we expect to find that changes that reduce the structural supports conducive to remaining active in religious communities—including leaving home, entering the labor force, and enrolling in community or technical/vocational postsecondary institutions—will result in a sharper decrease in religious service participation than those who remain at home or enter a 4-year college or university.

Hypothesis 3 (H3): Given the cognitive dissonance that may arise from violating religious teachings, we expect to see marked declines in religious service attendance among young people who begin engaging in sexual activity.

Hypothesis 4 (H4): Lastly, we expect that experiencing a traumatic life event may increase young people’s religious service attendance.

Method

We use two waves of survey data from the National Study of Youth and Religion (NSYR). The first wave of the survey was conducted in 2002 through 2003, and included a telephone survey of one adolescent and one parent in 3,290 English and/or Spanish speaking households nationwide. The sample, obtained through a random-digit dial method, was designed to represent all U.S. households with at least one adolescent between the ages of 13 and 17. An additional oversample of 80 Jewish households was included, but we delete these cases for the purposes of our analyses to present nationally representative results.

The second wave of the NSYR longitudinal telephone survey is a resurvey of the Wave 1 youth respondents, conducted from June 2005 through November 2005 when respondents were ages 16 to 21. Every effort was made to contact and survey all original NSYR respondents, including those out of the country and in the military. Of the original respondents, 2,604 participated in the second wave of the survey resulting in an overall retention rate of 78.6%. The predominant source of attrition in the second wave was nonlocated respondents. The refusal rate for Wave 2, calculated as the number of eligible respondents that refused to take part in the survey, was 4.0%. The overall combined response rate for waves 1 and 2 of the NSYR telephone survey is 44.8%. This response rate is consistent with national and international trends toward lower survey response rates (de Leeuw & de Heer, 2002). Diagnostic analyses comparing NSYR data with U.S. Census data on comparable households and with other nationally representative adolescent surveys—such as Monitoring the Future, the National Household Education Survey, and the National Longitudinal Study of Adolescent Health—confirm that the NSYR provides a nationally representative sample without identifiable sampling and nonresponse biases of U.S. adolescents ages 13 to 17 and their parents living in households (see Smith & Denton, 2005 for more detailed information). We use a survey weight in all descriptive analyses to adjust for the probability of inclusion in the study at Wave 1 and the extent to which census region and income is related to attrition between Waves 1 and 2.

Dependent Variable

Our key dependent variable measures the frequency with which adolescents attend religious services. In both waves, attendance was measured using the following response categories (a) never, (b) a few times a year, (c) many times a year, (d)once a month, (e) two to three times a month, (f) once a week, and (g) more than once a week. See the online supplement for a table containing descriptive statistics of all dependent and independent variables.

Independent Variables

We account for demographic and family background characteristics of NSYR respondents, all measured at the first survey wave. Sex is measured using an indicator of whether the respondent was female or not. Age of respondent was included as a continuous variable, after testing for nonlinear effects in previous analyses. Race/ethnicity was measured using indicator variables for respondents who were of White (reference), Black, Latino, or Other racial background. Finally, we include dummy variables indicating the respondent lived in the Midwest, South, West, or Northeast (reference) regions of the United States.

Parents’ education is included as a series of dummy variables: high school degree or less (reference), vocational degree or some college, associate’s degree, bachelor’s degree, and graduate or professional degree. Parents were asked to indicate their family income by selecting an income category within a 10-point scale. We transformed this into five dummy variables representing less than US$30,000 per year (reference), US$30,000 to US$50,000, US$50,000 to US$80,000, over US$80,000, and missing income. We control for whether the respondent was living with both biological or adoptive parents at the first survey and for the number of other children below age 18 in the household.

We also used measures of the surveyed parent’s religious attendance and affiliation. Primarily following the RELTRAD strategy (Steensland et al., 2000), we include dummy variables of affiliation for Evangelical Protestants (reference), Mainline Protestants, Catholics, other religious affiliations, and adolescents with no affiliation. We reassigned members of Black Protestant denominations to either Evangelical or Mainline congregations to prevent multicollinearity problems resulting from including controls for Black Protestant and race/ethnicity in our analyses. Parent religious service attendance was measured identically to adolescents’ attendance, with seven categories of attendance. All parent variables were measured through the Parent Survey conducted at Wave 1.

Finally, we include indicators of schooling transitions, moving away from one’s parents, work transitions, sexual activity, and the presence of traumatic life events. We focus on four schooling transitions: no transition (remaining in secondary school), transition to noncollege adult education (vocational, community or junior college, part-high school/part-college, or other), transition to a 4-year college or university (reference), and transition to the labor force. We next include an indicator for whether the respondent moved out of the home of the parent-figure he or she was living with in Wave 1. To capture change over time in work schedules during the school year, we include two dummy variables: same or reduced hours of work between Wave 1 and Wave 2 and increased work from Wave 1 to Wave 2. The reference category is not working in either wave. Next, we included two dummy variables indicating whether the respondent had become sexually active between Waves 1 and 2, or had been sexually active in both waves, compared to not engaging in sexual activity in either wave. Finally, we included an indicator of whether the respondent said they had experienced a traumatic life event (such as the death of a family member or serious illness) between the first and second wave.

Analytic Strategy

We first describe changes in religious service attendance between Wave 1 and Wave 2 of the NSYR survey overall and within categories of our independent variables. We then use lagged dependent variable models to estimate the change in religious service attendance from Wave 1 to Wave 2. Lagged dependent variable models estimate the effect of an event occurring between two waves of data on a change in the outcome variable. In our analyses, we ask how life course transitions affect a change in religious service attendance between Wave 1 and Wave 2. In addition, we are able to estimate the relationship between time invariant variables on a change in attendance, such as demographic factors, family background, and parent religiosity. Our models control for demographic factors, family background, and parent affiliation, all of which minimize the risk of unmeasured exogenous factors.

Results

Our results point to an overall decline in religious service attendance over time. Figure 1 reports the raw percentages of adolescents reporting each category of religious service attendance in Wave 1 and Wave 2. In 2002 to 2003, 17.7% of adolescents reported never attending religious services, and in 2005, 27.8% fell within this category, indicating a 10% rise in nonattenders. At the other end of the spectrum, about 42% of adolescents in Wave 1 reported attending religious services at least once a week compared to 29% two and one-half years later. Only small changes in the middle categories were observed.

Figure 1.

Religious service attendance of youth in the United States, 2002 and 2005.

Although changes in religious service attendance appear most clearly at the very low and very high categories of attendance, it may be that adolescents who report moderate levels of religious service attendance also exhibited change. This change may be masked by even numbers of adolescents entering and exiting these categories. Figure 2 displays the percentage of adolescents who attended religious services less, the same, and more in Wave 2 (2005) than in Wave 1 (2002-2003), by category of attendance in Wave 1.

Figure 2.

Religious service change over time, by Wave 1 attendance.

Among adolescents who reported never attending religious services in Wave 1, 73% reported the same religious service attendance in Wave 2, but nearly 27% reported some attendance 3 years later. For adolescents who reported attending religious services a few times a year in 2002 to 2003, over 40% had stopped attending religious services by Wave 2. Among those who attended about monthly in Wave 1 (those in categories “many times a year” and “about once a month”), only about 10% continued to attend religious services at the same rate. Among those who attended “many times a year,” slightly over 40% attended less in the next wave of data, while 50% attended more. Among those who attended religious services “about once a month,” 70% reported less attendance in Wave 2 of the survey. Thus, it is not that little change takes place for adolescents in the middle categories of attendance—in fact, these are the adolescents who see the most change. Instead, it appears that the aggregate effect of change in attendance over time creates the least total change in percentage of youth attendance at these levels.

Altogether, the findings depicted in Figures 1 and 2 do not support the polarization hypothesis, whose thesis suggests that adolescents who exhibit high levels of religious commitment should increase their religious participation over time, while those at moderate and low levels should decrease their participation over time. Instead, the findings depicted in Figures 1 and 2 suggest a pattern of declining religious service attendance for youth at all prior levels of attendance.

Table 1 reports the means of religious service attendance in Wave 1 and Wave 2 for the overall sample and all of the independent variables. For the entire sample, the difference between Wave 1 mean attendance (3.27) and Wave 2 mean attendance (2.60) was .67, a decrease of about two thirds of an attendance category. There were some interesting differences between groups, however.

Table 1.

Paired t-test Comparisons of Religious Service Attendance in 2002 and 2005.

	Wave 1	Wave 2	Difference
Overall	3.27	2.60	.67
Sex
Young women	3.40	2.70	.70
Young men	3.03	2.49	.54
Age
13 Years Old	3.37	2.81	.56
14 Years Old	3.32	2.78	.54
15 Years Old	3.27	2.70	.57
16 Years Old	3.18	2.38	.80
17 Years Old	2.93	2.28	.65
Race
Black	3.26	2.81	.45
Latin	3.08	2.53	.55
Other	3.13	2.28	.85
White	3.23	2.57	.66
Region
Northeast	2.75	2.06	.68
Midwest	3.17	2.58	.59
South	3.57	2.90	.67
West	2.90	2.38	.52
Parents’ highest education level
High school degree or less	2.96	2.34	.63
Vocational college/some college	3.02	2.42	.60
Associates degree	3.26	2.62	.64
Bachelor’s degree	3.46	2.90	.57
Professional or graduate school	3.47	2.79	.68
Family income
Less than US$30,000	2.87	2.34	.53
US$30,000 to US$50,000	3.17	2.51	.66
US$50,000 to US$80,000	3.44	2.82	.62
Over US$80,000	3.29	2.65	.64
Missing	3.33	2.59	.75
Family structure
Two-parent family	3.45	2.87	.59
Nontwo-parent family	2.92	2.25	.67
Children in household
Only child	2.95	2.38	.57
1–2 siblings	3.31	2.63	.68
3 or more siblings	3.68	3.18	.50
Parent religious affiliation
Evangelical Protestant	3.71	3.14	.56
Mainline Protestant	2.96	2.20	.76
Catholic	3.11	2.32	.79
Other religion	2.89	2.52	.37
No affiliation	1.17	0.92	.26
Parent religious service attendance
Never	1.14	0.96	.18
Occasional	2.63	1.92	.71
Once a week or more	4.42	3.73	.70
School transitions
Remained in secondary schooling	3.35	2.77	.58
Entered postsecondary, not college	3.06	2.31	.75
Entered college/university	3.22	2.63	.59
School to labor force	2.68	1.89	.80
Moved away from parents	2.98	2.08	.89
Work transitions
No paid work, both waves	3.12	2.63	.49
Same or reduced work, W1 to W2	3.28	2.45	.83
Increased work, W1 to W2	3.29	2.62	.67
Transition to sexual activity
No sex, either wave	3.59	3.21	.37
No sex W1, sexual activity W2	3.06	2.21	.87
Sexual activity, both waves	2.64	1.85	.79
Experienced life trauma between waves	3.23	2.62	.61

All t tests show a statistically significant difference of .001, except for the difference in attendance for youth whose parents do not attend religious services or are not affiliated and those in the “Other religion” category, which was significant at the .01 level.

Young women reported higher levels of religious service attendance than young men in Wave 1 and Wave 2. However, the decrease in religious service attendance between waves was more pronounced for women (.70) than men (.54). Older adolescents decreased their attendance much more than younger adolescents. Adolescents of “Other” races (primarily Asians) decreased their attendance the most over time. Whites experienced a sharper decrease in religious service attendance (.66) than Black (.45) or Latino (.55) adolescents, although they shared similar attendance patterns with Black adolescents in Wave 1. Regional differences were more prominent in starting values, with adolescents from the South attending religious services the most and adolescents in the Northeast attending the least. The South and Northeast regions saw similar decreases in attendance over time that were higher than the West or Midwest.

Youth religious service attendance was correlated positively with parental education in both waves. Youth whose parents have a professional or graduate education reported the highest levels of attendance in both waves, but exhibited the largest drop in attendance over time. Among youth whose parents ever attended religious services, occasional versus regular attendance did not appear to influence the degree to which adolescents decreased in their own religious service attendance.

Adolescents whose families reported an income of less than US$30,000 per year exhibited somewhat less change over time than those whose families reported higher incomes. These adolescents started at a lower level of religious service attendance overall, however. Among the other income categories, change over time was fairly consistent. Adolescents living with two married biological or adoptive parents at Wave 1 were less likely to decrease their religious service attendance than adolescents who were not, although the most obvious difference among these groups was in their initial rates of attendance at Wave 1. Adolescents from two-parent families attend religious services more often than those from other family structures. Finally, young people with three or more siblings attended religious services more often at Wave 1 than those with fewer siblings, and they also reported a smaller drop in attendance over time.

Adolescents whose parents were Mainline Protestant or Catholic saw the greatest decrease in religious service attendance, at .76 and .79, respectively, while those of “Other” religious affiliations saw the least, at .37 (with the exception of the “No affiliation” category). In both waves, children of Evangelical Protestant parents reported higher levels of religious service attendance than any other group.

Changes in adolescents’ religious service attendance also differed by the life course transitions they made between the first and second wave of the study. Adolescents who were attending secondary school in both waves began with the highest level of attendance and were the least likely to reduce their attendance over the two waves, with an average decrease of .58. Surprisingly, while young people who began attending a 4-year college between Waves 1 and 2 lost about the same level of attendance as those who remained in secondary school (.59). However, leaving secondary school to attend another type of postsecondary institution such as a 2-year college or vocational school is related to a greater drop in attendance (.75), as is moving from school to the labor force (.80). We also find that leaving one’s parent’s home is related to a much higher drop in religious service attendance, at .89, than continuing to live with one’s parents (.56).

As expected, increasing work hours over the two waves was accompanied by a larger drop in religious service attendance (.67) compared to youth who were not working at either time point (.49). Surprisingly, however, maintaining or decreasing work hours between waves was related to the greatest drop in religious service attendance, at .83.

Initiating sexual activity between the first and second wave was related to a larger than average decrease in attendance, at .87, while adolescents who reported not engaging in sexual activity in either wave decreased their attendance by only .37. Those who were sexually active in both waves decreased religious service participation by .79. Finally, experiencing a traumatic life event between waves 1 and 2 was related to an average drop of .61 in attendance.

Table 2 presents the results of our multivariate analyses of change in religious service attendance over time. Model 1 demonstrates that age, race/ethnicity, and region are related to religious service attendance change over time. For each year of age, adolescents experience about an additional one-tenth decrease in average religious service attendance, holding Wave 1 attendance constant. Black youth report about one fifth of a decline in religious services less than Whites, on average, while Latino and “Other” racial/ethnic groups exhibit similar change over time compared to Whites. This is consistent with past research showing higher levels of religious involvement among Blacks. Adolescents from the Northeast are more likely to report decreases in attendance than other youth. The greatest difference occurs between residents in the South, who attend religious services about one third of a category more often than those in the Northeast, holding Wave 1 attendance constant.

Table 2.

Wave 2 Religious Service Attendance on Family Background, Religious Affiliation, Demographic Characteristics, and Life Course Transitions.

	Model 1	Model 2	Model 3
Wave 1 religious service attendance	0.57*** (33.64)	0.39*** (19.22)	0.38*** (18.65)
Female	–0.01 (–0.17)	0.10 (1.36)	0.07 (0.92)
Age	–0.09*** (–3.30)	–0.10*** (–3.96)	–0.00 (–0.02)
Race/ethnicity^a
Black	0.20* (1.96)	–0.03 (–0.30)	0.01 (0.12)
Latin	0.04 (0.29)	0.16 (1.19)	0.19 (1.35)
Other race	–0.24 (–1.41)	–0.19 (–1.16)	–0.28* (–1.73)
Region^b
Midwest	0.28** (2.33)	0.28** (2.36)	0.25** (2.17)
South	0.35*** (3.13)	0.24** (2.21)	0.26** (2.34)
West	0.25** (1.97)	0.23* (1.82)	0.24* (1.93)
Parents’ highest education^c
Vocational college/some college		0.02 (0.18)	–0.03 (–0.28)
Associates degree		0.01 (0.06)	–0.07 (–0.53)
Bachelor’s degree		0.15 (1.26)	0.04 (0.31)
Professional or graduate school		0.05 (0.41)	–0.12 (–0.91)
Family income^d
US$30,000 to US$50,000		–0.04 (–0.36)	–0.03 (–0.25)
US$50,000 to US$80,000		0.09 (0.75)	0.10 (0.84)
Over US$80,000		0.07 (0.55)	0.09 (0.64)
Missing income		–0.16 (–0.90)	–0.23 (–1.33)
Two-parent family		0.23*** (2.91)	0.13 (1.59)
Number of children in household		0.04 (1.11)	0.05 (1.41)
Religious affiliation^e
Mainline Protestant		–0.41*** (–3.80)	–0.40*** (–3.78)
Catholic		–0.42*** (–4.16)	–0.45*** (–4.50)
Other religion		–0.10 (–0.70)	–0.14 (–0.98)
Not religious		–0.15 (–0.91)	–0.16 (–0.97)
Parent religious service attendance		0.26*** (12.04)	0.25*** (11.47)
Schooling transitions, Wave 1 to Wave 2^f
Remained in secondary school			–0.19
Entered postsecondary, not college			–0.27* (0.15)
School to labor force			–0.37** (0.14)
Moved away from parents, Wave 1 to Wave 2			–0.20* (–1.95)
Working transitions, Wave 1 to Wave 2^g
Same or reduced work, W1 to W2			–0.13 (–1.09)
Increased work, W1 to W2			0.04 (0.45)
Transition to sexual activity^h
No sex W1, sexual activity W2			–0.59*** (–7.04)
Sexual activity, both waves			–0.70*** (–6.19)
Experienced life trauma between waves			0.14** (1.98)
Constant	1.79*** (4.34)	1.67*** (3.93)	0.88 (0.67)
N	2516	2445	2410
Adjusted R²	0.329	0.383	0.402

t statistics in parentheses.

One-tailed t-test: *p < .05. **p < .01. ***p < .001.

White.

Northeast.

High school degree or less.

Less than US$30,000.

Evangelical Protestant.

Entered 4-year college/university.

No paid work, both waves.

No sexual activity, both waves.

Surprisingly, Model 2 reveals no effect of parent education or income after demographic and other family background factors are controlled. Therefore, while parental education and income are both related to absolute levels of youths’ religious service attendance; they do not appear to predict changes in attendance over time. Adolescents from two-parent families attend religious services about one quarter of a category more, holding Wave 1 attendance constant, than those who do not live with their biological or adoptive parents. Finally, adolescents whose parents were Evangelical Protestants at Wave 1 attend religious services more often than Mainline Protestants and Catholics, holding initial attendance levels constant. This difference is equivalent to almost half of a category. In addition, parent religious service attendance is positively related to adolescent attendance at Wave 2, suggesting that lower levels of parental attendance may be associated with decreasing adolescent religious service attendance over time.

In Model 3, we include indicators of life course transitions. Holding other factors constant, leaving secondary school for either postsecondary school that is not a 4-year college or university or the labor force is related to a reduction in religious service attendance, compared to those who attend a 4-year college or university. This is equivalent to more than a quarter of a category of attendance less for those attending a non4-year postsecondary institution (–.27), and more than a third of a category for those who enter the labor force after exiting secondary school (–.37). As expected, remaining in high school offers stability in terms of religious service attachment, with no significant difference in change over time compared to those in college. Moving away from one’s parents was also related to a .20 greater decrease in religious service attendance over time in comparison to those adolescents who did not move.

Initiation into sexual activity between Waves 1 and 2 is related to a marked drop in religious service attendance over time, compared to those who do not engage in sexual activity in either wave. Those who initiated sexual activity reduced their attendance by .6, on average, compared to those who did not engage in sex in either wave. Those who had already engaged in sexual activity at Wave 1 experienced a drop in religious service attendance of .7. Finally, experiencing a traumatic life event was related to a .14 point higher level of religious service attendance in Wave 2, holding Wave 1 attendance constant. This means having experienced a traumatic experience is related to a smaller decrease over time than for those who have no traumatic experience in the past 3 years.

Conclusion

Religious institutions are a source of community, support, and friendship for many adolescents and their families. The importance of these religious communities and the resources they offer may be even more important in the transition to adulthood, as young people leave their homes and secondary schools. Recent research has found that nearly one fifth of young people between the ages of 16 and 24 are not engaged in either school or work (Danziger & Ratner, 2010). Separated from institutional structures that provide daily contact with adult mentors and educational programs designed to build social, intellectual, and practical skills, young adults who continue to attend religious services will have access to a network of adults and peers who can convey important information and resources (Smith, 2003a).

The overwhelming story of our bivariate results shows decline across all demographic groups and at all levels of religious attachment, as predicted by Hypothesis 1b, rather than the polarization thesis articulated by Ozorak (1989) and Hypothesis 1a. Although some young people may report stronger feelings of religiosity, this does not translate into higher levels of religious service attendance. Although prior religious participation, family context, and life course transitions lessened the decline in religious service attendance for some, it did not completely eliminate it. This is troubling given the importance of religious institutions as sources of support in adolescence and young adulthood.

We find support for Hypothesis 2b which posited that young people who have greater access to structural supports that facilitate religious service participation (living at home or entering a 4-year college or university) experience a smaller drop in attendance. We did not find support for Hypothesis 2a, which suggests that colleges and universities may have a secularizing effect on young people, decreasing their religious service attendance. Surprisingly, increased work hours do not appear to accelerate a decline in religious service attendance, despite evidence of a bivariate relationship. This may be due to differences in Wave 1 attendance among those who begin or increase paid work. Changes in paid work may also correlate with other important transitions, such as completing high school and leaving one’s household, which may better explain changes in religious service attachment.

The emotional and the psychosocial development of youth are also key factors in behavior change among adolescents. We find support for Hypothesis 3, which predict that young people who initiated sexual activity between waves decrease their religious service participation. Although the causal ordering is unclear, it is interesting that those who had already engaged in sexual activity at Wave 1 also experienced a drop in attendance. Whether a decline in participation in religious activities is followed by sexual activity or vice versa, it appears that religious service participation continues to drop after sexual activity has occurred. It may be that adolescents who have had sex feel uncomfortable attending religious services when premarital sexual relations are expressly proscribed. As predicted in Hypothesis 4, experiencing a traumatic life event appears related to an increase in religious service participation, controlling for other factors. Though we do not know the exact nature of this relationship or its mechanisms, it is not surprising that young people would seek guidance and reassurance in religious institutions in difficult moments.

Finally, many previously discovered demographic and family background correlates of youth religious service attendance have a different relationship to this outcome once a longitudinal approach and controls for life course transitions are included. Although religious service attendance appears to drop more precipitously as youth age we find that life course transitions better explain the rate of religious service attendance decline. Additionally, while Black youth appear to experience the smallest drop in attendance between surveys, this difference is not significant when controlling for family background and religiosity in contrast to prior research (e.g., Smith et al., 2002). Region remains a strong predictor of attendance. Prevailing community norms may have a stronger influence over adolescents’ behavior than many family attributes. The religious context within the home, however, is also an important buffer against declining rates of attendance. Parental religiosity predicts a smaller decrease in religious service attendance over time, and conservative Protestant youth experience a smaller decline in attendance than Mainline Protestant or Catholic youth.

In sum, our findings point to the significant role of institutional structures in maintaining ties to religious institutions. Young adults who remain at home or enter 4-year colleges and universities are most likely to maintain their connection to religious institutions, while leaving secondary school for the labor force or community college or vocational schools predict a steeper decline in religious service attendance. Unfortunately, this suggests an accumulation of risk factors for young people who may leave the supportive environment of family, school, and a religious community at approximately the same time. Young adults who transition to college maintain ties to religious communities more readily, augmenting their sources of social and emotional support in the transition to adulthood.

Limitations and Future Directions

Our speculations regarding the importance of life course transitions and the structures available within different young-adult pathways, the cognitive dissonance arising out of contradicting religious teachings, and the role of traumatic life events in religious activity all offer fruitful avenues for further research. In particular, qualitative work could explore how young adults make sense of changes in their lives and concurrent changes in their relationship with religion and religious institutions. It also remains to be seen whether these rates of attendance will increase once these young people establish a career, marry, and have children of their own. It may be that adolescence is a vulnerable but intermediate stage in the life course, after which old routines of religious service attendance resurface. Alternatively, adolescents who experience a drop in attendance may have trouble reentering a religious community later in life. Continuing to examine the patterns of religious service attendance during the transition to adulthood will be beneficial in understanding this process and in determining how later transitions such as marriage and childbirth affect religious service attendance in adulthood.

There are also important limitations to our study which should be addressed in future research. Additional unmeasured factors may explain drops in religious service attendance among adolescents. Religious service attendance may be tied to the availability of transportation and parental monitoring, which could be influenced by changes in family structure and economic resources. It is also important to examine how different characteristics of religious communities may influence young people’s attendance patterns. Although the use of longitudinal analysis and family background controls mostly addresses these concerns, additional survey data could further examine these factors.

Footnotes

Acknowledgements

We are grateful to the Carolina Population Center for training support (T32 HD007168) and for general support (R24 HD050924).

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was funded by support to the first author from the Eunice Kennedy Shriver National Institute of Child Health and Human Development to the Pennsylvania State University Population Research Institute Postdoctoral Training in Family Demography and Individual Development (5 T32 HD007514).

Author Biographies

Jessica Halliday Hardie is an assistant professor of sociology at the University of Missouri-Kansas City. Her research focuses on inequality in adolescence and young adulthood, with particular attention to how young people’s social class, race/ethnicity, and gender shape their school, work, and family pathways in the transition to adulthood.

Lisa D. Pearce is associate chair and associate professor of sociology at the University of North Carolina at Chapel Hill. Her research focuses on religiosity in youth and the transition to adulthood, especially how it relates to aspirations, achievement, and family life.

Melinda Lundquist Denton is an associate professor in the Department of Sociology at The University of Texas at San Antonio. Her research examines the intersection of religion and family life in the United States, with a current focus on the religious lives of adolescents and emerging adults.

References

Adamczyk

(2012). Extracurricular activities and teens’ alcohol use: The role of religious and secular sponsorship. Social Science Research, 41, 412-424.

Albrecht

S. L.

Cornwall

Cunningham

P. H.

(1988). Religious leave-taking: Disengagement and disaffiliation among Mormons. In Bromley

D. G.

(Ed.), Falling from faith: Causes and consequences of religious apostasy (pp. 62-80). Newbury Park, CA: SAGE.

Bergin

(1983). Religiosity and mental health: A critical re-evaluation and meta-analysis. Professional Psychology: Research and Practice, 14, 170-184.

Beyerlein

(2004). Specifying the impact of conservative Protestantism on educational attainment. Journal for the Scientific Study of Religion, 43, 505-518.

Billy

J. O. G.

Brewster

K. L.

Grady

W. R.

(1994). Contextual effects on the sexual behavior of adolescent women. Journal of Marriage and the Family, 56, 387-404.

Cnaan

R. A.

Gelles

R. J.

Sinha

J. W.

(2004). Youth and religion: The Gameboy generation goes to “church.” Social Indicators Research, 68, 175-200.

Cochran

J. K.

(1992). The effects of religiosity on adolescent self-reported frequency of drug and alcohol abuse. Journal of Drug Issues, 22, 91-104.

Coleman

J. S.

(1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95-120.

Cotton

Zebracki

Rosenthal

S. L.

Tsevat

Drotar

(2006). Religion/spirituality and adolescent health outcomes: A review. Journal of Adolescent Health, 38, 472-480.

10.

Danziger

Ratner

(2010). Labor market outcomes and the transition to adulthood. The Future of Children, 20, 133-158.

11.

De Leeuw

de Heer

(2002). Trends in household survey nonresponse: A longitudinal and international comparison. In Groves

Dillman

Eltinge

Little

(Eds.), Survey Nonresponse (Chapter 3) (pp. 41-54). New York, NY: John Wiley.

12.

De Vaus

McAllister

(1987). Gender differences in religion: A test of the structural location theory. American Sociological Review, 52, 472-481.

13.

Donahue

M. J.

Benson

P. L.

(1995). Religion and the well-being of adolescents. Journal of Social Issues, 51, 145-160.

14.

Dudley

R. L.

(1999). Youth religious commitment over time: A longitudinal study of retention. Review of Religious Research, 41, 109-120.

15.

Edgell

(2006). Religion and family in a changing society. Princeton, NJ: Princeton University Press.

16.

Garasky

(2002). Where are they going? A comparison of urban and rural youths’ locational choices after leaving the parental home. Social Science Research, 31, 409-431.

17.

Garasky

Haurin

R. J.

Haurin

D. R.

(2001). Group living decisions as youths transition to adulthood. Journal of Population Economics, 14, 329-349.

18.

Glanville

J. L.

Sikkink

Hernández

E. I.

(2008). Religious involvement and educational outcomes: The role of social capital and extracurricular participation. The Sociological Quarterly, 49, 105-137.

19.

Good

Willoughby

Fritjers

(2009). Just another club? The distinctiveness of the relation between religious service attendance and adolescent psychosocial adjustment. Journal of Youth and Adolescence, 38, 1153-1171.

20.

Gunnoe

M. L.

Moore

K. A.

(2002). Predictors of religiosity among youth aged 17-22: A longitudinal study of the national survey of children. Journal for the Scientific Study of Religion, 41, 613-622.

21.

Hill

M. S.

Yeung

W. J.

(1999). How has the changing structure of opportunities affected transitions to adulthood? In Booth

Crouter

A. C.

Shanahan

M. J.

(Eds.), Transitions to adulthood in a changing economy: No work, no family, no future? (pp. 3-39). Westport, CT: Praeger Press.

22.

Hoge

D. R.

Petrillo

G. H.

Smith

E. I.

(1982). Transmission of religious and social values from parents to teenage children. Journal of Marriage and the Family, 44, 569-580.

23.

Jang

S. J.

Johnson

B. R.

(2011). The effects of childhood exposure to drug users and religion on drug use in adolescence and young adulthood. Youth & Society, 43, 1220-1245.

24.

Kagan

(1971). A conception of early adolescence. Daedalus, 100, 997-1012.

25.

Lehrer

E. L.

(1999). Religion as a determinant of educational attainment: An economic perspective. Social Science Research, 28, 358-379.

26.

Markstrom

C. A.

(1999). Religious involvement and adolescent psychosocial development. Journal of Adolescence, 22, 205-221.

27.

Muller

Ellison

C. G.

(2001). Religious involvement, social capital, and adolescents’ academic progress: Evidence from the national longitudinal study of 1988. Sociological Focus, 34, 155-183.

28.

Myers

S. M.

(1996). An interactive model of religiosity inheritance: The importance of family context. American Sociological Review, 61, 858-866.

29.

Ozorak

E. W.

(1989). Social and cognitive influences on the development of religious beliefs and commitment in adolescence. Journal for the Scientific Study of Religion, 28, 448-463.

30.

Pearce

L. D.

Denton

M. L.

(2011). A faith of their own: Stability and change in the religiosity of America’s adolescents. New York, NY: Oxford University Press.

31.

Pearce

L. D.

Hardie

J. H.

(2008). Family religious context and educational outcomes in late adolescence. Paper presented at the annual meeting of the Population Association of America, New Orleans, LA.

32.

Pearce

L. D.

Haynie

D. L.

(2004). Intergenerational religious dynamics and adolescent delinquency. Social Forces, 82, 1553-1572.

33.

Regnerus

(2000). Shaping schooling success: Religious socialization and educational outcomes in metropolitan schools. Journal for the Scientific Study of Religion, 39, 363-370.

34.

Regnerus

M. D.

Uecker

J. E.

(2006). Finding faith, losing faith: The prevalence and context of religious transformations during adolescence. Review of Religious Research, 47, 217-237.

35.

Roozen

D. A.

(1980). Church dropouts: Changing patterns of disengagement and reentry. Review of Religious Research, 21, 427-450.

36.

Rostosky

S. S.

Regnerus

M. D.

Wright

M. L. C.

(2003). Coital debut: The role of religiosity and sex attitudes in the Add Health survey. Journal of Sex Research, 40, 358-367.

37.

Smith

(2003a). Religious participation and network closure among American adolescents. Journal for the Scientific Study of Religion, 42, 259-267.

38.

Smith

(2003b). Theorizing religious effects among American adolescents. Journal for the Scientific Study of Religion, 42, 17-30.

39.

Smith

Denton

M. L.

Faris

Regnerus

M. D.

(2002). Mapping American adolescent religious participation. Journal for the Scientific Study of Religion, 41, 597-612.

40.

Smith

Sikkink

Bailey

(1998). Devotion in Dixie and beyond: A test of the “Shibley Thesis” on the effects of regional origin and migration on individual religiosity. Journal for the Scientific Study of Religion, 37, 494-506.

41.

Smith

Denton

M. L.

(2005). Soul searching: The religious and spiritual lives of American teenagers. New York, NY: Oxford University Press.

42.

Smith

Snell

(2009). Souls in transition: The religious and spiritual lives of emerging adults. New York, NY: Oxford University Press.

43.

Steensland

Park

Regnerus

Robinson

Wilcox

Woodberry

(2000). The measure of American religion: Toward improving the state of the art. Social Forces, 79, 291-318.

44.

U.S. Department of Education, Office of Education Research and Improvement. (1993). Trends in the well-being of American youth. Washington, D.C.: U.S. Government Printing Office.

45.

Uecker

Regnerus

M. D.

Vaaler

(2007). Losing my religion: The social sources of religious decline in early adulthood. Social Forces, 85, 1667-1692.

46.

Wallace

Jr. J. M.

Forman

T. A.

Caldwell

C. H.

Willis

D. S.

(2003). Religion and U.S. secondary school students: Current patterns, recent trends, and sociodemographic correlates. Youth & Society, 35, 98-125.

47.

Willits

F. K.

Crider

D. M.

(1989). Church attendance and traditional religious beliefs in adolescence and young adulthood: A panel study. Review of Religious Research, 31, 68-81.

48.

Wilson

Sherkat

D. E.

(1994). Returning to the fold. Journal for the Scientific Study of Religion, 33, 148-161.

49.

Youniss

McLellan

J. A.

Yates

(1999). Religion, community service, and identity in American youth. Journal of Adolescence, 22, 243-253.