5. How Did They Get There? Religious Trajectories in the Lives of Emerging Adults (with Jonathan Hill)
(with Jonathan Hill)
throughout this book , we have met several of the young people who participated in our study. Each has a unique story to tell; yet, among them we can find common threads of experience related to their religious and spiritual journeys. On the surface, Kristen—the highly religious woman we met in —and Josh, the nonreligious young man we met in , appear to have very little in common. Yet, both are examples of young people whose religious journeys are marked by consistency and stability.
Kristen was raised in a home where her evangelical faith was a daily presence—in family interactions, practices such as church attendance, prayer and Bible reading, and her schooling. Overall, across each interview, Kristen demonstrated a commitment to understanding and living her life by the teachings she first learned while a child. Mapped as a trajectory, her story reflected consistently high religiosity from adolescence to adulthood. Josh also had a consistent religious trajectory, although at the other end of the spectrum from Kristen. Josh was raised in a home where religion was not present at all—neither of his parents were religious—although he was encouraged to pursue any interests he might have about religion while growing up. Throughout each of four interviews with Josh, his responses to our questions about religion remained remarkably consistent: he was curious about religion, particularly whether the claims could be rationally “proved,” but in the end, nothing he ever investigated met his criteria for truth.
Many of the emerging adults in our study have similar stories—religious lives that have remained relatively stable over time. But for other young people, their religious journeys include shifting toward or away from religion and doing so in gradual or (less often) dramatic fashion. We met Heather in . Taking a snapshot of the religiousness of Heather and Josh at Wave 4 would reveal similar levels of religious practice and commitment, in that neither of them is religious nor professes any connection to religion at this point in their lives. However, taking the longer view of their stories and the paths that led them to where they are today illustrates some of the different religious trajectories that young people follow from adolescence into adulthood. Heather’s life started out looking more like Kristen’s, with a religiously devoted mother, regular attendance at religious services, and socialization into the faith, resulting in confirmation in the Catholic church. But when she got older, Heather jettisoned the religious practices and commitments of her Catholic upbringing and adopted a nonreligious identity. In doing so, Heather illustrates one of the most common religious pathways among emerging adults—that of religious decline over time.
Thinking about Kristen, Josh, and Heather, we can see that in order to fully understand the religious experiences of emerging adults, we have to account not just for where people started religiously and where they are today but also for the path they followed to get from one point to the other. In this chapter, we sketch a picture of the various paths—or, more precisely, the religious trajectories—that young people follow as they move from adolescence to their mid-20s. We explore the most common trajectories and examine the factors associated with each type of spiritual journey.
Often, trends are represented by showing changes in the average of a given measure over time. For example, if we want to document how religious service attendance changes between adolescence and emerging adulthood, we can examine the mean of our measure of service attendance at each wave of the National Study of Youth and Religion (NSYR) (. Alternatively, we could examine how a proportion changes over time. For example, we might want to know what percentage of all young people attend church at least weekly at each wave of the NSYR (. Both of these data points tell us something important about the overall shape of religious practice associated with young people as they age.
Figure 5.1 Mean change in religious service attendance by age, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Figure 5.2 Percentage attending religious services weekly or more by age, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
While this type of analysis is helpful, it is remarkably easy for the reader to be left with the impression that the “typical” young person will follow a similar trajectory. But this simply isn’t the case. A brief thought experiment will illustrate. The first figure could consist of three types of young persons: The first attends religious services every week and never deviates from this in all four waves of the data. The second type is like the first, except never attends religious services. The last type begins like the first, but at some point along the way, stops attending religious services altogether, and ends up like the second group. Such a pattern could, hypothetically, produce the first figure we examine (although we know, from examining the data, that this is not the case). could also be produced by everyone following exactly the same pattern and attending religious services less frequently at each wave. But which one is correct? Distinguishing between these scenarios is certainly important if we want to accurately understand the patterns of religiousness and spirituality from adolescence onward.
Parsing out the common and uncommon trajectories requires a different approach. In this chapter we use a method known as growth mixture modelling (GMM), which sorts individuals into various groups based on similarities in their trajectories. In other words, it’s a method designed to distinguish between the different scenarios that might produce the same overall mean-level or proportion-level changes over time. While not everyone will fit perfectly in one group or another, this approach attempts to find a balance between the complexity of individual cases and the (over)simplicity of single-trajectory summaries.
To begin, we calculated a simple summary measure of religious service attendance, frequency of prayer, and self-rated importance of faith at each wave of the NSYR. This measure, although not perfect, is consistently captured at all four waves of the NSYR and taps into different dimensions of religious faith (public practice, private practice, and personal salience). Each measure is weighted equally and then added together to create a scale of religiousness that ranges from zero to 12. The mean level scores at each wave are reported in .
Figure 5.3 Mean change in composite religiousness measure by age, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Figure 5.4 Results from growth mixture model of six class trajectories by age, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
The GMM analysis tells us that the mean-level trajectory in is not the typical pathway for most. Rather, our analysis suggests there are six different “typical” religious pathways from adolescence into emerging adulthood, and young people do not all take the same pathway (. While the overall means at each wave suggest declining religious practice and salience, the picture in this figure complicates a simple story of religious decline. It would be more accurate to say that decline occurs in different ways. Of the three trajectories that exhibit some sort of downward trend, only one indicates a complete disengagement from religion.
This steep decline from high to low accounts for about 13 percent of the young people in our study. This group averages a moderate to high religiousness on our 12-point scale at the beginning, but ends more than seven points lower, with a religiousness score comparable to the fully disengaged group. The other two decline trajectories are less severe. One group starts at a moderate level (a little below eight on our 12-point scale) and ends at a moderate-low level (a little above five on the same scale). This group makes up about one-fourth of the young people. The other group starts near the bottom, at less than three, and ends at less than one. This group is disengaged religiously throughout the timeframe of the study, but still shows some movement, from a smattering of religiousness for some of them to virtually nothing by their mid-20s. This group also makes up one-fourth of the sample.
The most stable trajectory reflects young people who exhibit moderate to high religiousness throughout their late teens and 20s. This group is also the largest, at 29 percent. The remaining two groups are the smallest of the six, and both show a general movement toward increased religiousness over time. The smallest group, at only two percent, shows a substantial incline, moving from relatively low levels of religiousness (slightly above four on the 12-point scale) to relatively high levels (slightly below 10). The other group begins with a similar increase, but levels out and ends up with a moderate level of religiousness (a little less than six on the scale). This group represents six percent of the young people in our study. describes each of the various trajectories in terms of where they begin and end.
Table 5.1 Median Values of Religion Measures for Each Class Trajectory at Waves 1 and 4
Where They Begin (Wave 1 Medians)
Where They End Up (Wave 4 Medians)
High Stable (29%)
Attends church once a week Faith is very important Prays about once a day
Attends church once a week Faith is extremely important Prays about once a day
Shallow Decline (25%)
Attends church two to three times a month Faith is very important Prays a few times a week
Never attends church Faith is somewhat important Prays a few times a week
Steep Decline (13%)
Attends church once a week Faith is very important Prays a few times a week
Never attends church Faith is not important at all Prays a few times a year
Steep Incline (2%)
*
Attends church once a week Faith is very important Prays multiple times a day
Moderate Shallow Incline (6%)
Attends church a few times a year Faith is somewhat important Prays one or two times a month
Attends church a few times year Faith is somewhat important Prays a few times a week
Low Declining (25%)
Attends church a few times a year Faith is not very important Prays less than once a month
Never attends church Faith is not at all important Never prays
“Many times” a year is the category that falls between a “few times a year” and “once a month.”
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Stepping back for a moment, it is worth noting how little drastic religious change occurs for most young people. Four of the groups, comprising 85 percent of all young people, show some sort of stability or incremental change that never exceeds more than three points on our scale. Although the more typical trend is still downward overall, they are not experiencing a radical shift in their religious life one way or the other (at least using these particular measures). For the 15 percent who do undergo substantial shifts in religiousness, the vast majority belong to the group that is becoming less religious (13 percent) rather than more religious (two percent).
Establishing typical and atypical religious trajectories is useful at a purely descriptive level; however, we also want to dig a little deeper into the nature of these trajectories. Just who are the people in these different groups? Are there factors in adolescence that seem to set the stage for belonging to one group versus the others? And what about the consequences of these trajectories? Does belonging to one or another of these groups make any difference in the life of young people in their mid-20s? It is to these questions that we now turn.
Demographics and Family Structure
We begin this exploration by looking at the demographic factors of gender, race, and social class. through 5.7 show the proportion of each class trajectory that is female, persons of color, and from higher socioeconomic strata, respectively. Because we use similar figures throughout the remainder of this chapter, it is worth pausing for a moment to explain the interpretation of the figures.
Taking as our example, the top bar of the graph shows the overall proportion of adolescents at the first wave who are female. In fact, there are two bars grouped together. The top one is the “raw” percentage estimate—this is just the straightforward statistic based on the NYSR sample (adjusted for the usual survey weights). The second bar is the adjusted percentage estimate. This is the estimated percentage after other factors like family income, educational attainment of parents, family structure, region, race/ethnicity, and age are statistically controlled (for most graphs, we also statistically controlled for gender, but this would not be appropriate for . We will typically refer to the raw estimate, but occasionally we point out the adjusted results, when there are differences worth noting between the two.
The top bar serves as a reference for the other bars. The bars below are the estimates for survey respondents in each of the six growth curve groups. Each bar also includes error bars. These show the 95 percent confidence interval for each estimate. In other words, they show the range we might expect the true value to reasonably fall within. They also help gauge whether we can be confident that one group is different from another. For example, we can clearly see from that the Stable High group is disproportionately female (57 percent) and the Low Declining group is disproportionately male (34 percent female). We can also be pretty confident that these two groups are different from one another statistically and that they are different from the overall percentage estimate for females (49 percent).
Figure 5.5 Raw and adjusted percentage who are female by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, and age.
A similar pattern emerges when examining the racial/ethnic composition of the religious trajectory groups. People of color are more likely to be found in the groups characterized by higher levels of religiousness. In we see that people of color make up more of the religious categories (both those who are stable and those who increase), with about four out of 10 adolescents in these groups being people of color. The Steep Decline group, and especially the Low Declining group, are more likely to be white (with only about two out of 10 being nonwhite in the latter). Here, we also see that, unlike the estimates for gender, when we adjust for other factors (especially parents’ income and education), some of the differences between the groups are less dramatic—particularly in the Steep Decline and Steep Incline groups.
Figure 5.6 Raw and adjusted percentage of people of color by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, gender, and age.
Directly examining parents’ education and income corroborates this (. High socioeconomic status is similar across most of the trajectory groups, with the exceptions of those in the Steep Incline and Steep Decline categories (with a rather dramatic difference in the former). While 23 percent of teens can be classified as being raised in households with relatively privileged economic and educational capital, fully 30 percent of those in the Steep Decline group, and only five percent of those in the Steep Incline group, meet these criteria.
Figure 5.7 Raw and adjusted percentage living in a high socioeconomic status (SES) household in adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: High SES households have one or more parents with a college degree and had a household income of at least $80,000 in 2002–3.
N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for family structure, region of residence, race/ethnicity, gender, and age.
Finally, we include a basic measure of family structure (. Here we are examining the percentage of young people who, as teens, were raised in homes with two biological or adoptive parents. The differences here are fairly minor, with those in the Steep Incline category less likely to be in such households (although the adjusted percentage shows that this is at least partially due to the differences in socioeconomic status shown in the previous figure—single parents have, on average, fewer economic resources).
Figure 5.8 Percentage living with biological or adoptive parents by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N=3, 290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, region of residence, teen race/ethnicity, gender, and age.
Let’s pause for a moment to take stock of what these first few figures tell us. The emerging picture relates to social location and resources. Men, whites, and those with high socioeconomic status are overrepresented in groups that end up ranking low religiously in emerging adulthood. This includes those who measured high on our religion scale as teens (the Steep Decline group) and those that were low to begin with (the Low Declining group). Those without these markers of social advantage are more likely to stay religious and are much more likely to be overrepresented in the group that becomes substantially more religious in emerging adulthood, even though this group is relatively small. Family structure, in and of itself, is relatively unimportant compared with gender, race, and social class.
These findings align with a variety of views that suggest religion may serve as a social or psychological resource for individuals with less material, social, and cultural capital. Religious congregations have often been cited as places that can function as social safety nets, provide venues to teach civic skills, extend social networks across status barriers, and provide a site for the arts to flourish. At a psychological level, religion has been cited as a source of comfort and stability when “existential security” is low or as a source of help from a superhuman power. While we don’t want to read too much into these few variables—after all, we are not using any direct measure of social or psychological deprivation—we do find that the general picture that emerges is consistent with approaches that emphasize these factors.
Religious Environment in Adolescence
Although there are clear fault lines along race, class, and gender, demographics are obviously not destiny. The religious environment in which the teen is raised, perhaps unsurprisingly, remains the most powerful predictor of the various religious trajectories we document. In this section, we use various measures of this environment to investigate how they map onto the trajectories.
The next four figures () are measures of the most common religious traditions (measured by the affiliation of the congregation the teen attended or identified with). breaks down the proportion of Conservative Protestants in each of the trajectory groupings. The pattern is striking. Moving down the figure, each group has progressively fewer Conservative Protestants. Although Conservative Protestants account for three in 10 young people in the NSYR, they make up more than four in 10 (42 percent) of those who remain consistently religious as they move into emerging adulthood. However, Conservative Protestants are also well represented in the two categories that decline religiously. This tells us that belonging to a Conservative Protestant tradition does not somehow ensure that teens will stay strong in their faith commitments as they get older. It does mean that Conservative Protestant teens are overrepresented in the trajectories that begin with moderate to high rates of religious service attendance, prayer, and importance of faith in daily living.
On the other hand, Conservative Protestants are relatively unlikely to be found in the groups that begin with moderate to low levels of religiousness in the teenage years. At the very bottom of the figure, only 14 percent of those in the Low Declining category were Conservative Protestants as teens. In groups that show some level of increase in religiousness over time, about two in 10 could be classified as Conservative Protestants at the first wave of the NSYR.
Figure 5.9 Percentage Conservative Protestant during adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, race/ethnicity, gender, and age.
Black Protestants, although often similar to Conservative Protestants on certain theological positions, reveal a pattern that diverges somewhat from that of their theological cousins (. While they are nearly twice as likely to be found in the Stable High group as in the general population (19 vs. 11 percent), they are much less likely to be found in the decline groups—particularly the Steep Decline category (five percent). Like Conservative Protestants, they are quite unlikely to be found in the Low Declining category (three percent). Unlike Conservative Protestants, however, they are also well represented in the categories that show an upward religious trajectory over time. There are nearly as many Black Protestants in the small Steep Incline category as Conservative Protestants (16 and 18 percent, respectively), despite the fact that Conservative Protestants outnumber Black Protestants almost three to one in the NSYR sample.
Figure 5.10 Percentage Black Protestant during adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, gender, and age.
Turning now to , Mainline Protestants make up the same proportion of young people in the first wave of NSYR as Black Protestants. Despite this, their faith trajectories are quite different. For most categories, Mainline Protestants are not particularly overrepresented or underrepresented. They are equally likely to be in the Stable High category as the Low Declining category (both nine percent). Even a category where they appear underrepresented, the Steep Incline group, reveals that the differences are not so dramatic after our statistical controls are introduced (four percent without controls and seven percent adjusted for controls). There is one clear exception: there are nearly twice as many Mainline Protestants in the Steep Decline category (21 percent) as in the overall sample. So, while Black Protestants are half as likely to be found in this category, Mainline Protestants are nearly twice as likely.
Figure 5.11 Percentage Mainline Protestant during adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, race/ethnicity, gender, and age.
The final religious tradition we examine is Catholic (. A little more than one in four teens were Catholic in Wave 1 of the NSYR. And, in most of the trajectories we’ve mapped out, Catholics make up about the same proportion of each group (ranging from 27 to 32 percent). Like Mainline Protestants, there is an exception. This time it is the High Stable group that stands out: only 18 percent of this group is Catholic. Again, this doesn’t mean that there are no Catholic teens that continue to go to Mass, pray, and say that their faith is important as they transition into their 20s. But we do see that Catholics are underrepresented among those young people who are maintaining their religious faith into adulthood.
Figure 5.12 Percentage Catholic during adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, race/ethnicity, gender, and age.
In summary, religious traditions are not equally represented in the different faith pathways. While Conservative Protestants and Black Protestants make up about four in 10 teens, they make up more than six in 10 teens who stay religious as they get older. And they account for only 17 percent of the young people who appear to be consistently disconnected from or uninterested in religious faith (the Low Declining group). Mainline Protestants and Catholics also make up nearly four in 10 teens, yet they make up more than half of the young people we would classify as exhibiting a steep drop in measures of religious faith and practice. In addition to this, they only account for around one-fourth of the group that remains at least moderately religious throughout the four waves of the NSYR (the High Stable group).
The connection between religious traditions and the different religious pathways is not surprising, given our analysis of individual measures of religion in . As we noted in that chapter, the general pattern of declining religious engagement includes movement in both directions. And the solidification of religious commitment appears to be happening most commonly among those young people who hail from more conservative or culturally distinct religious traditions.
Moving beyond religious traditions, we now investigate several more direct measures of household faith and practice. The next two figures ( and examine reports of the religious service attendance of respondents’ parents. Forty-four percent of the parent respondents in the first wave of the NSYR reported that they attended church weekly or more. Two-thirds of the parent respondents of youth in the Stable High category reported weekly attendance, and about half of parent respondents reported this in the two categories that exhibit religious decline. As would be expected, regular worship attendance by a parent was much less common for those teens who started out less religious. About one-fourth of those in the categories that later saw increases in religion had a parent who attended weekly, and only 17 percent of young people who were consistently low on religious measures had a parent who attended church regularly.
The pattern is similar if we examine whether the teen and parent attend religious services together at Wave 1. This is not a measure of how frequently either parent or teen attends, but rather if they attend services at the same religious congregation. About two-thirds reported that they did. While this result is highest in the Stable High category (87 percent), it is not dramatically higher than the two declining categories (78 and 77 percent). It is less common in the other categories, the lowest being the Low Declining group (32 percent) and the Steep Incline group (31 percent).
Figure 5.13 Percentage with parents who reported attending services weekly by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Figure 5.14 Percentage with parents who reported attending services with teen by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
For the sake of space, we have not included a number of other analyses of household religious faith. The story is largely the same for all of them, however. We examined how often the parent prays for the teen, whether married parents belong to the same faith (religious homogamy), and how important faith is in the daily life of the parent. In each case, we saw a clear step-like pattern, with descending percentages in each category moving down the figure. In other words, teens who stay religious are the most likely to have parents who measure high on religious measures, and parental religious commitment becomes increasingly less likely as we move to the Low Declining category at the bottom of the figures.
Most of these measures of household religious environment do not dramatically distinguish which teens remain religious and which teens decline in religiosity. None of these measures seem to be the “key” to keeping teens in the faith. However, there was one exception. In Wave 1, the teens were asked how frequently their family talked about God, sacred texts, prayer, or other religious or spiritual things together. We examine these responses by religious trajectory in . About one-third (34 percent) of teens said that their family talks about these religious matters at least a few times per week. Those who are in the declining groups (Shallow Decline and Steep Decline) are roughly similar to this overall rate. However, teens who stay religious into emerging adulthood are much more likely to report their family talking about matters of faith (62 percent). We will return to this finding shortly.
Figure 5.15 Percentage who reported religious talk in home a few times per week or more as teen by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
First, though, we want to end this section with two figures that measure youth-focused church programming. The first ( measures youth group participation and the second ( religious mission trips. Although youth group participation is more common, the pattern is similar. In both figures the Stable High group has the largest proportion who participated in church programming. The most noticeable gap, however, is between the top three classes and the bottom three. Again, it does not seem to be the case that either of these—youth group or mission trips—clearly distinguish who stays religious and who does not over time.
Figure 5.16 Percentage attending a religious youth group in adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Figure 5.17 Percentage who participated in a religious mission trip in adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
In sum, the nature of the religious environment matters. Young people who have parents who say their faith is extremely important, attend church regularly (and with their teen), belong to the same faith tradition, pray regularly for their teen, and talk regularly about religious or spiritual matters in the household are all disproportionately represented in the Stable High category and underrepresented in the Low Declining category. We suspect this is capturing how deeply integrated religious faith is in the teen’s household. The extent that religion is a regular part of the day-to-day goings-on of family life has an influence on whether a young person continues with the faith as they enter emerging adulthood.
The measure of religious talk in the household also helps to explain some of the differences in religious traditions we discussed at the beginning of this section. Nearly half of Conservative Protestants (47 percent) and more than half of Black Protestants (56 percent) reported religious talk at least a few times per week. The equivalent statistic is only 23 percent for both Catholics and Mainline Protestants. Running a statistical model that controls for the frequency of religious talk in the household accounts for nearly half of the advantage Conservative Protestants have over Catholics and Mainline Protestants in keeping young people in the Stable High category, compared with one of the categories of religious decline. In other words, the fact that Conservative Protestants are more likely to remain religious may be explained by the integration of faith into household life that is more common within this religious tradition.
Adolescent Relationships
We now turn from explicit measures of religion to other relational factors that might matter during the teenage years. The next few figures examine how the relationship with parents and peers might influence the religious trajectory of teens as they move into emerging adulthood. Let’s begin by considering how close the teen was to his or her parents. The first wave of the NSYR contains numerous measures of the parent-teen relationship. We created a composite index from the various measures. shows what percentage of each group scored in the top 30 percent of the index. In other words, these teens tend to score higher than the average in terms of feeling close to their parents, talking to them about personal subjects, and doing various activities together. The Stable High group is disproportionately made up of young people who were close to their parents as teenagers (37 percent), while the inclining classes and the Low Declining class all have lower than average numbers of respondents who were close to their parents. The Shallow and Steep Decline classes are close to the mean at 31 percent in each.
Figure 5.18 Percentage who felt close to parents as teens (top one-third of composite measure) by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Next, we used a similar technique to measure the degree to which parents monitor what their teen does and who she or he spends time with. shows the percentage in each class who score in the top 30 percent (high scores mean that parents more frequently monitor the teen). The results are remarkably similar to the previous figure, with the Stable High class having about double the proportion of high monitoring parents (39 percent) as the Low Declining class (19 percent).
Figure 5.19 Percentage whose parents monitored their activities and friends closely as teens (top one-third of composite measure) by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
The next two figures examine relationships with peers at the time of the Wave 1 survey ( and . The first is a parent-reported measure of the influence of the teen’s friends. Those in the Steep Incline category are more likely than the other groups to have friends whom the parent respondent considers to have a very or somewhat negative influence (22 percent in this group compared with 13 percent overall). The differences between the various groups is still relatively small. includes the teen’s assessment of how important it is to fit in with what other teens think is cool. This is designed to be another measure of peer influence, this time from the teen’s perspective. In this case, the Low Declining group includes disproportionate numbers of those who said it is “not important at all” to them. However, like the previous measure, differences in peer influence are not that dramatic between the various groups.
Figure 5.20 Percentage whose parents reported friends are generally “very negative” or “somewhat negative” influence in adolescence by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Figure 5.21 Percentage who reported it is “not important at all” to fit in with what teens think is cool by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 3,290. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Although this is not an exhaustive analysis of the influence of parents and peers, we do find that the religious trajectories young people travel are more strongly influenced by their relationships with parents than with peers. Having close, warm relationships with parents who regularly spend time with their teen and monitor the teen’s activities and relationships is most associated with the Stable High category and least associated with the Low Declining category. While the previous section helped to identify what religious activity in the home might be associated with certain trajectories, these graphs show that what matters is not simply the content of the religious environment but also the quality of relationships. How parents invest or do not invest in their teens’ lives has consequences for their religious faith as they transition into adulthood.
Postadolescent Life Events
Finally, we want to move away from the teenage years to consider important life events that occur after adolescence. In some ways, this means we are thinking less about the factors that might shape future trajectories and more about the way that some religious trajectories might be more (or less) likely to be accompanied by important educational, occupational, and familial changes in life.
Let’s begin with education. looks at the percentage of the NSYR sample respondents who have graduated from college with at least a bachelor’s degree by the fourth wave. Overall, this is 41 percent of the emerging adult sample. The measure of education is not noticeably different across religious categories, with one exception. Only 20 percent of the Steep Incline group have a bachelor’s degree. Controlling for our standard set of variables, however, changes this substantially. After additional factors are accounted for (especially parents’ income and education measured in adolescence), we could expect this group to have a degree acquisition rate of closer to 35 percent. To be clear, this is not the rate of college completion this group has; rather, it is simply what we might expect if socioeconomic status were similar across these religious trajectories. In other words, the lower educational attainment among the Steep Incline group can be explained in part by the socioeconomic background of the young people in this group.
Figure 5.22 Percentage who reported receiving a bachelor’s degree or higher by Wave 4 by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 2,064. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
The next three figures focus on marriage and romantic partnerships. shows the percentage of emerging adults in each group who have never married by Wave 4. This tends to be relatively low for the Stable High group (45 percent) and the Steep Incline group (49 percent) and relatively high for the Steep Decline (67 percent) and the Low Declining (69) groups. Another way to put this: those groups that end up being less religious by Wave 4 are also less likely to have ever been married. It appears to matter less where one starts religiously during the teenage years and more where one ends up in their 20s. As we will see in the next chapter, marriage and religious faith are mutually influencing factors in the lives of young people.
Figure 5.23 Percentage who reported having never married at Wave 4 by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 2,067. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
The next figure is restricted to those who have been married or are currently married (. Here, we examine whether any of the religious trajectories are associated with higher or lower rates of divorce or marital separation. Among those who have ever been married, about 17 percent reported a divorce or separation. Most religious trajectories are similar, and in fact the lower sample size means our estimates are less precise (notice the range of the error bars). Still, we can be fairly confident (statistically speaking) that the Stable High group has lower rates of divorce and separation than other trajectories.
Figure 5.24 Percentage who reported having had a divorce or separation at Wave 4 (includes only those who reported having ever married) by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 647. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Our last measure of marriage and relationship status includes only those who have never married, who are divorced, or who are currently widowed (. Here, we are investigating how many are currently living with their partner in a marriage-like relationship (cohabiting). Overall, the data suggest that about 13 percent are currently cohabiting at Wave 4. Here, again, we find a relationship with the various religious trajectories. The cohabitation rate is lowest among those who are currently the most religious (seven percent in the Stable High and six percent in the Steep Incline groups) and highest among those who are the least religious (16 percent in the Steep Decline and 18 percent in the Low Declining groups). In other words, we find precisely the opposite relationship that we found when examining marriage.
Figure 5.25 Percentage who reported currently cohabiting with partner at Wave 4 (includes only never married, divorced, or widowed) by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 1,481. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
Next, we want to explore political identification. The last decades of the 20th century saw a realignment of conservative politics with religious faith, especially among Protestants. Do we see a similar association among emerging adults? displays the proportion of emerging adults in each group who identify as politically conservative at Wave 4. The differences are striking. While 28 percent of all emerging adults identify as politically conservative, fully 46 percent of the Stable High emerging adults identify as such. The Steep Decline and Low Declining groups stand out in sharp contrast to this. Here, only 14 percent (Steep Decline) and 18 percent (Low Declining) identify as politically conservative.
Figure 5.26 Percentage who reported they consider themselves politically conservative at Wave 4 by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 2,065. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
To what extent might this be due to how they identified as teens? Unfortunately, we don’t have such a measure from Wave 1, but we do have the political identification of the parent respondents. We ran the models again, this time including parents’ political identification as a control factor in the adjusted results (not displayed). The findings remain virtually unchanged. Emerging adults in the Stable High group are much more likely to be politically conservative, while Steep Decline emerging adults are much less likely to be so. This suggests that young people are at least partially defining their political identities in ways that fit the religious trajectories they are traveling.
In the previous section, we saw that young people who were close to their parents as teens were more likely to fall into the Stable High religious group. What does this look like later in life? uses the Wave 4 report of closeness to parents. There are some striking differences between this and the previous measure. The Steep Decline and the Low Declining groups feel less close to their parents than the other groups (43 percent and 45 percent, respectively, compared with 54 percent overall). On the other hand, the Steep Incline and the Moderate Shallow Incline respondents are nearly as close to their parents, on average, as the Stable High group (58 and 59 percent for the two incline categories, compared with 64 percent for Stable High). In the previous figure, the Steep Incline cohort had the lowest levels of closeness to parents as teens. Again, this tells us that higher religiousness, at the time of reporting, tends to come packaged with closer relationships to parents. The small group of respondents who started off relatively low religiously, and relatively less likely to report being close to their parents, became both more religious and closer to their parents over time.
Figure 5.27 Percentage who reported being very or extremely close to parent(s) in Wave 4 by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 2,032. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
The last measure we include is an overall assessment of life at Wave 4 (. Respondents were asked to agree or disagree with the following: “In most ways your life is close to ideal.” More than six out of 10 emerging adults agree with this statement. Those in the Stable High group are the most likely to agree with this statement (69 percent), while those in the Steep Decline group are the least likely (53 percent). The differences between these groups are not large, but they do fit a clear pattern and are significantly different from a statistical perspective.
Figure 5.28 Percentage who agreed with the statement, “In most ways, [my] life is close to ideal,” at Wave 4 by class trajectory, weighted.
Source: National Study of Youth and Religion surveys 2002, 2005, 2008, 2013.
Notes: N = 2,070. Error bars display 95 percent confidence interval of estimate. Adjusted percent represents the average marginal effect from a logistic regression after controlling for parents’ income, education, family structure, region of residence, teen race/ethnicity, gender, and age.
What can we say about how the religious trajectories that we have been exploring relate to where young people end up in life by their 20s? In general, we see that those who are in the Stable High religious trajectory are more likely to have a college degree, be married (and less likely to be divorced or cohabiting), feel close to their parents, and be generally satisfied with the direction their life has taken them. They are also much more likely to be politically conservative. Many of these factors, particularly the marital and relational ones, are also associated with those who become more religious by their mid-20s (particularly the Steep Incline group). Notice that we are not arguing that there is something causal going on here. We are simply noticing that several of these factors—education, marriage, relationships, politics, and life satisfaction—tend to be part of the package of some religious trajectories.
Conclusion
In this chapter, we analyzed the complex pathways of religious faith that young people travel as they move from their teens to their 20s. Instead of looking at overall mean-level change for everyone, we used a technique (GMM) that separated out the most dominant trajectories in the data. We found that, although the general trend is downward on measures of religious service attendance, prayer, and importance of faith, a not insubstantial number of young people (nearly three in 10) remain moderate to high on these measures as they transition into adulthood. Moreover, about one-fourth of young people remain low on these measures throughout the four waves of the NSYR (although they continue to trend slightly downward throughout). This means that more than half of young people are not exhibiting substantial religious shifts—at least as measured by these three factors—as they move into adulthood. Moreover, those that are experiencing changes are not necessarily experiencing dramatic ones. Another 30 percent are experiencing small to moderate shifts over time, beginning toward one end or the other of the scale as teens, but ending up somewhere in the middle. This means that only about 15 percent exhibit dramatic religious change over time. Most of this is downward (13 percent), becoming less religious, but a small group becomes substantially more religious (two percent).
The rest of the chapter explored the social and religious factors associated with these various trajectories. Household religious factors appear to be most important for creating a context in which young people maintain religious engagement during their teen and emerging adult years. Having highly religious parents who are invested and active in the life of their teen is associated with a stable religious trajectory moving into adulthood. In particular, we found that teens who report regular discussions in the household about God, prayer, scripture, or other religious things are much more likely to stay religious themselves over time. We think this measure is likely capturing how faith is integrated into daily living and routines in some households. We also found that this factor was substantially different across religious traditions and helped at least partially to explain why Conservative Protestants and Black Protestants were much more likely to see their teens continuing to practice their faith into the emerging adult years.
Attempting to understand who takes up religious faith and practice as they enter emerging adulthood is a little more challenging. To begin, only a very small number (about two percent) of young people fit the Steep Incline trajectory. The data seem to suggest that those who do experience increased religious engagement are more likely to come from positions of social or economic disadvantage. Women, people of color, those without college-educated parents or middle-class incomes, and those who do not have particularly warm relationships with their parents are all overrepresented in this Steep Incline group. It is possible that, for these young people, their life opportunities may make a religiously upward trajectory appealing. Instability in the household, lack of resources, culturally marginal status, less support from mainstream institutions—all of this leads to greater than typical chances of turning to faith communities during the transition to adulthood.
The chapter finished by examining how various life outcomes related to these trajectories. We found that those who ended up more religious by the fourth wave, and especially those who stayed religious throughout the years measured in the NSYR, were more likely to be married, college educated, close to parents, and satisfied in life. They were also much more likely to identify as politically conservative.
In addition to revealing the correlates and covariates of religious trajectories, the pathways examined in this chapter provide another lens through which to see the big picture of the religious lives of emerging adults. Consistent with the previous two chapters, we acknowledge that, taken together, there is aggregate religious decline among emerging adults in our study.
However, parsing out this larger trend reveals distinctly different lived experiences among these young people. While many are moving away from religious engagement, there are distinct pathways by which they are making these shifts. Combining the quantitative analysis of the typical trajectories of religious change with the stories that emerge from our qualitative interviews, we see that there is a group of young people who are gradually drifting away from religious communities and practices that no longer seem relevant for their lives, while a smaller group seems to have shifted away from religion in a more intentional and decisive way. This chapter also highlights, however, the story of the emerging adults who are not drifting away from religion, but rather are maintaining or increasing their religious commitment. Examining the distinct trajectories by which emerging adults live their religious lives, we see again the pattern of polarization toward the religious and not religious ends of the continuum and a hollowing out of the religious middle.
