Home | Research & Policy Library | Making the Transition: The Effect of School Sector on Extended Adolescence

Making the Transition: The Effect of School Sector on Extended Adolescence

April 17, 2018

David Sikkink

Sara Skiles


Research Report

Looking at several large social surveys in the U.S., this paper finds that religious school isn’t a bulwark against extended adolescence. However, American religious school grads do follow some different patterns going into adulthood when compared to their public school counterparts.


The millennial generation, young adults currently in their twenties or early thirties, have received a great deal of attention in both academic and popular presses thanks to their unique path to adulthood. Many studies have argued that young adults—for rea- sons that have complicated structural and cultural roots—have established a new stage in the life trajectory that is marked by instability, searching, and a general incoherence related to a lack of meaning, purpose, and direction in life. As a group, millennials have higher rates of college-degree attainment than do previous generations, and are waiting longer to establish a home away from their parents, to get married, and to have children, giving them time to focus on higher education, identity formation, and career exploration (Arnett 2001; 2011; Smith, Davidson, Christoffersen, and Herzog 2011; Smith and Snell 2009). This unprecedented period of transition gives these so- called emerging adults the opportunity to learn about themselves before committing to long-term relationships, and time to investigate and prepare for emerging career fields. This freedom of choice is accompanied, for many millennials, with uncertainty, doubt, and a lack of direction. When it comes to milestones such as completing one’s education, establishing a career and financial independence, living independently, and being married and having children, the latest cohorts of young adults are far behind earlier cohorts. At the same time, this generation is characterized by a higher rate than in previous generations of identification as non-religious and by lower rates of religious-service attendance (Dean 2010; Smith et al. 2011; Smith and Snell 2009).

Are religious-school graduates no different? This project will investigate various aspects of the lost-in-transition phenomena to understand how religious schools may create different expectations and trajectories in young adulthood. We examine school-sector effects on young-adult outcomes such as family formation, employment, and how individuals are finding meaning and purpose in their lives. Does graduating from a Catholic high school predict a different path to adulthood than does graduat- ing from a Protestant high school? A non-religious private school? Do religious high school graduates have a different sense of purpose and meaning compared to other emerging adults? We would expect that at least in terms of marriage and family formation, religious-school graduates are less likely to be lost in transition (Engelberg 2016; Uecker and Hill 2014; Uecker 2009; Uecker and Stokes 2008). In what follows, we investigate public- and private-school-sector differences in the United States that are related to concerns regarding the fluid life trajectories of young adults.

We will draw on the National Survey of Youth and Religion, the National Longitudinal Survey of Youth 1997, and Cardus Education Survey data related to job stability, work and educational history, orientations to purpose and meaning, marriage, and other related measures to assess variation in life trajectories during young adulthood. Each of these surveys provides a unique window on recent young adults in the United States. We consider the type of job held by religious schoolers as well as the timing and progression of an educational career. We will also consider the religious correlates of whether young adults have established their own households and families, which tends to reduce problems associated with the “lost-in-transition” stage (Martin, Blozis, Boeninger, Masarik, and Conger 2014).


The data for this paper came from three sources: the Cardus Education Survey (CES), the National Longitudinal Survey of Youth 1997 (NLSY97), and the National Study of Youth and Religion (NSYR). The CES has been fielded twice in the United States in 2011 and 2014, each time utilizing an online survey of a nationally representative sample of high school graduates with roughly 1,500 respondents per wave. The goal of the CES was to investigate differences in various outcomes by sector of high school graduation. To that end, graduates of non-public high schools were oversampled to ensure large-enough sample sizes in the smaller private-school sectors. Respondents in the CES were aged twenty-four to thirty-nine at the time of data collection, although only respondents younger than twenty-nine were included in the results presented here to capture the young adult period. Results of analyses of educational attainment, marriage and family formation, and work status by school sector using CES data are presented in this paper.

Findings from the CES sample are corroborated and supplemented with findings from similar analyses using NSYR data. The NSYR was fielded four times between 2002 and 2013 using the same sample over time. For this paper, we primarily rely on data collected from respondents in 2013 when they were in their mid-twenties (twenty-three to twenty-eight years old), as well as data collected from their par- ents in 2002 regarding respondents’ school sector during adolescence as well as parents’ characteristics. The NSYR data set used for this paper includes about 2,100 respondents. In addition to examining sector differences in educational attainment, family formation, and work status, the NSYR data allowed us to examine differences by sector in how respondents are finding meaning and purpose in their lives.

The NLSY97 supplements some of the analysis on marriage and family expectations during young adulthood. The NLSY97 includes data from a large US sample of teenagers who have been followed over the course of their lives. In 2001, when respondents would range from about seventeen through twenty-one, the NLSY97 included questions on whether the respondent expected to be married and have children within a year or within five years. It also asked whether they expected to complete a college or university degree by age thirty. The survey included detailed questions on marital status throughout the young adult years. We use these data to provide some information on whether young adults have a sense of where they see marriage, family, and education fitting into their lives in their twenties. We include extensive controls in this analysis, including the age of the respondent. Additional controls include variables for the religion, net worth, and educational levels of parents, as well as demographic characteristics, including rural/urban residence and region. Specifically, controls are included for race and ethnicity, gender, age, number of children in the household under six years old, number of children in the household under eighteen years old, citizenship, education of each parent, whether the respondent lived with both biological parents, net worth, the religious tradition of both parents, region of residence in 1997, region of current residence, whether the respondent currently lives in a rural or urban area, and whether the respondent lives in a standard metropolitan statistical area (as defined by the census). The parent religious tradition variables included as controls are Catholic, high-attending Catholic, evangelical Protestant, and high-attending evangelical Protestant. These are included for each residential parent.


Educational Attainment

One important concern in the transition to adulthood is whether young adults are on track educationally. We can think of “drifting” young adults as less committed to attending college or university immediately after high school and finishing in the expected four-year window.

First, we consider what young adults project in the years immediately after high school. Based on the NLSY97 data, we find that Catholic school graduates are more likely than public school graduates to expect that they will have completed a bachelor’s degree by the time they are thirty years old, and that Catholic school graduates are less likely to think that they will be married in five years. Other findings for Catholic school graduates are not statistically significant in this sample, but are in the expected direction. For example, Catholic school graduates tend to think they will have a large family, but they do not expect that they will be married in the next year or in the next five years. Thus the emphasis on staying “on track” with education and career seems to hold for Catholic school graduates. Evangelical Protestant school graduates are more likely than public school graduates to expect that they will have a large family with three or more children. None of the other findings are statistically significant in this sample for Protestant school graduates. However, the direction of the effects are consistent with the view that evangelical Protestant school graduates are focused on marriage and family during the young adult years. Evangelical Protestant school graduates are more likely to say they will be married in five years than are public school graduates, and female evangelical Protestant school graduates are more likely to expect that they will be pregnant in five years. We do not find a positive effect of evangelical Protestant schooling on the expectation of completing college or university by age thirty. Females from religious homeschools are more likely to expect to be pregnant in five years compared to public school graduates. None of the other findings for religious homeschool graduates are statistically significant, nor do they seem consistently to follow a traditional marriage and large-family track.

Some of these expectations seem to carry through in the actual educational careers of the respondents in particular school sectors. Overall, we do not find consistent effects for private school graduates in terms of educational timing and completion. Based on both CES and NSYR data, how- ever, we find that Catholic school graduates are significantly more likely than public school graduates to complete a bachelor’s degree by the age of twenty-four. This would seem to be consistent with the sense that Catholic schools emphasize diligence and planning regarding their education- al and occupational trajectories. The historic role of Catholic schools in the social mobility of immigrants (Bryk, Lee, and Holland 1993; Greeley 1967; 1982; Greeley, Rossi, and Pinto 1964) may continue to push graduates through the difficult transition from high school to university. The well-worn pathways between Catholic schools and some private colleges and universities may also pave the way for a timely transition to and completion of post-secondary work.

In contrast, homeschool graduates are significantly less likely to complete a bachelor’s degree by age twenty-four than are public school graduates. As we have seen in other studies (Sikkink and Skiles 2015), homeschool graduates on average tend to be less enamored of simply gaining an educational credential, and less willing to adhere to life-course norms for post-secondary education. It is not clear that this finding should be viewed in terms of the “lost-in-transition” framework, since many homeschool graduates simply take a more practical approach to post-secondary education—they will get as much as they need for their job or career, and only when they need it. That view of educational credentials doesn’t require living according to educational trajectories normative in the broader society.

The predicted probabilities for each school sector provide some sense of the strength of the effects on educational attainment. Here we are interested in the probability that a person from a given school sector will complete a bachelor’s degree by the age of twenty-four. We find in the NSYR data that Catholic school graduates have a 52 percent likelihood of graduating with a bachelor’s degree by age twenty-four compared with a 38 percent chance for public sector graduates. The respective figure for Protestant school graduates is 47 percent; for private non-religious school graduates, 41 per- cent; and for homeschool graduates, 18 percent. These are predicted probabilities after account- ing for all the control variables in the model. The higher socioeconomic status of the non-religious private school graduates perhaps explains why, net of family-background factors, their probability of obtaining a bachelor’s degree by age twenty-four is relatively low. Another way of thinking about this is that the private non-religious school effect on this outcome is relatively small, especially relative to the effect of family, or that the school context does not add to the already strong likelihood that non-religious private school graduates will graduate from college in relatively quick fashion.

Another approach to education in young adult- hood is to consider the role of school sector in the likelihood of achieving a bachelor’s degree by the age of twenty-nine (since the CES sample is limit- ed to those under thirty). On this measure, we see the same positive and significant Catholic school effect in the regression results, and a negative and significant result for homeschool graduates. Thus the school-sector effects on staying on track have a longer-term effect on levels of educational attainment.

We also ran analyses that predicted the age at which respondents graduated from university, which provides a further assessment of whether school experiences influence the actual age that people complete their degrees. But these analyses of the both the CES and NSYR data did not reveal any school-sector differences. That is consistent with the finding that school-sector effects on completing the degree before age twenty-four and completing the bachelor’s degree by age twenty-nine are nearly identical.

In terms of education, then, Catholic school graduates are significantly less likely to be lost in transition, and are strongly oriented to obtaining university credentials in the usual timeframe during young adulthood.

Family Formation

Another concern of the theorists of young adult- hood is that the millennial generation delays entering marriage and having children—two mile- stones that were so important for transitions to adulthood in earlier generations.

One possibility is that young adults move back in with parents, which is often related to delay- ing marriage and family formation. The boomerang effect, in which young adults return to live with their parents during their twenties, might be less likely among religious school graduates, who may respond to norms for early marriage and childbearing. We ran a boomerang analysis using NSYR data, but did not find significant differences by school sector. Perhaps the religious factor works both ways on this score, since the importance of family and family commitments for religious school graduates, especially among evangelical Protestants, may facilitate returning home as much as leaving home.

Another related concern is that sex outside of marriage is the norm for millennials, which seems to be part of the package in the new phase of young adulthood. In extended adolescence, marriage  is delayed but sexual experiences are not (Freit- as 2008; 2013; Smith et al. 2011; Wade 2017). Sexual experiences during the teenage years tend to determine union formation later in life (Raley, Crissey, and Muller 2007). Interestingly, the NLYS97 provides some insight on the relation- ship between school sector and sex. Both Catholic and evangelical Protestant school graduates are much less likely to report having sex before the age of twenty. The negative evangelical Protestant school effect is nearly three times as strong as the Catholic school effect. The size of the estimates show that evangelical Protestant school graduates are nearly half as likely as public school graduates to have had sex during their teenage years. These findings are net of the significant negative effect of being raised in an evangelical Protestant or Catholic family and being raised in a traditional intact family. The school context may influence students through peers as well. The NLSY97 asked students in 1997 to estimate the percentage of their peers who were sexually active. Catholic, evangelical Protestant, and non-religious private school graduates all reported much lower levels of sex among their peers than did public school graduates. Again, the evangelical Protestant effect was nearly three times the size of the other private school sectors. Compared to the public school graduates, evangelical Protestant school graduates are nearly a point lower on this scale after taking into account the control variables. On this five- point scale, that is the equivalent of a difference between 50 versus 75 percent of one’s peers who are reported by the respondent as having sex. Of course, a student’s sense of what his or her peers are doing is affected to some extent by what is normative in the school. Still, a negative relationship between religious school attendance and sex at a young age may in part reflect student expectations of peer behaviour as well as the likely response of peers to sex in dating or “hookup” relationships during high school or early college.

Regarding marriage and family formation, we used both CES and NSYR data, looking for sector effects on both marrying and becoming a parent in young adulthood. We find no significant differences by sector of high school graduation on being married in early adulthood. 1 1 Note that this contrasts with other analyses, including one study that predicts the age of first marriage using the first US wave of the CES (Uecker and Hill 2014). The Uecker and Hill study is fairly consistent, however, with the NLSY97 findings we report here. It is also possible that the relationship between religious schools and timing of first marriage is changing over time. We will revisit these relationships with 2018 data in the near future.  That is, we find that about 40 percent of young adults in our samples were (or had once been) married at the time of data collection, but one group of graduates was not more or less likely than another to be married.

In contrast, we do find a few significant differences when considering marital status using the NLSY97 data. Our first NLSY97 outcome is whether the respondent is married/widowed, never married, or divorced/separated in 2015. At this point, respondents would be between thirty and thirty-four years of age. The Catholic school graduates in our full models are not more or less likely than public school graduates to be never married, single, or divorced or separated. There is a negative effect of Catholic schooling on being divorced or separated, but it is not statistically sig- nificant. Evangelical Protestant school graduates, on the other hand, are more likely to be married than either of the other alternatives in this model. Evangelical Protestant school graduates are about 32 percent less likely than public school graduates to be divorced/separated rather than married, and are about 32 percent less likely to be single as well. Religious homeschool graduates are not significantly different from public school graduates on these outcomes. Non-religious homeschool graduates are more likely to be single than married in this analysis. The second outcome we investigated with NLSY97 data was the total number of years during young adulthood that the respondent re- ported being married. This slightly different out- come addresses the question of whether marriage tends to disrupt extended adolescence for graduates of religious schools. For this outcome, we find that evangelical Protestant school graduates are married for more years during young adult- hood than are public school graduates. The oth- er private school sectors are not significantly re- lated to the number of years of marriage during young adulthood. The Catholic school graduates are married fewer years during their twenties than public schoolers, but this is accounted for by family background characteristics. Similarly, we find evangelical Protestant school effects when considering the number of years during young adulthood that the respondent reported being separated or divorced. After controlling for family background and demographics as well as number of years be- ing single, we find that attending an evangelical Protestant school reduces the likelihood of separation or divorce during young adulthood.

Getting married at younger ages may disrupt the trajectory of extended adolescence. The question is whether religious school attendance matters for this outcome. When considering the age of first marriage, we do not find using the NLSY97 that school sector has a strong effect on age of first marriage after accounting for demographic and family-background factors. There is evidence, however, that Catholic school graduates marry later on average than public school graduates. The Catholic school effect is very strong, but drops by about two-thirds after accounting for parent education. That Catholic school graduates tend to marry at older ages is consistent with the normative trajectory of education followed by marriage that Catholic school graduates seem to uphold. Although the estimate for evangelical Protestant school graduates is negative, it is not particularly large and is not statistically significant. Similarly, the religious homeschool graduates are associated with a younger age of marriage, but this effect is not statistically significant. The non-religious homeschool graduates in this analysis, in contrast, are strongly and significantly related to a younger age of first marriage.

Interestingly, using NSYR data, we find that when asked about the factors that are most important to them in leading the type of life they want to lead, Catholic school graduates are the only sec- tor significantly more likely to say that having a fulfilling romantic relationship is important. That finding is not directly addressing marriage, but is generally consistent with the notion that Catholic school graduates are not averse to marriage relationships even while the life-course trajectory set by them is defined first of all by educational and occupational career.

Sector effects emerge in similar fashion when considering childbearing. Using both CES and NSYR data, we find that graduates of Catholic high schools are less likely to have children in early adulthood than are public school graduates. This is likely due in large part to the sector difference in educational attainment discussed previously, which in part reflects cultural expectations that working toward the educational credential should precede marriage and family. Public school graduates who have not earned a bachelor’s degree by age twenty-four have a 43 percent chance of having at least one child in early adulthood compared with only an 11 percent chance for Catholic school graduates. (It should also be noted that homeschool graduates without bachelor’s degrees by age twenty-four have a 46 percent chance of having at least one child in early adulthood, although the difference between Catholic school and homeschool graduates is not significant on this measure because of a low sample size of homeschool graduates.) Therefore, though the findings for age of marriage and Catholic schooling are mixed, we find that Catholic school graduates are delaying parenthood into later adulthood in order to earn bachelor’s degrees relatively quickly after completing high school.

Finding Purpose

Another concern in the young adult literature is whether millennials have any sense of direction in life. Do they have a sense of purpose, or a sense that there are some things worth fighting for, as Tolkien puts it in Lord of the Rings? Here we would expect that religious schools could make a difference, instilling a sense that God has a purpose for each of their lives and that they must work to find that sense of purpose or calling. How teachers talk about why they teach in a religious school and the development of a sense of obligation for others may stick with graduates, setting them apart from graduates of non-religious schools on the question of purpose and direction in life.

NSYR includes numerous questions on meaning, purpose, and direction in life, but we find few sector differences in these data. The NSYR Wave 4 included these questions:

How often, if ever, do you think about or plan for the future?

How often, if ever, does life feel meaningless to you?

How often, if at all, do you think about the meaning of life?

You don’t have a good sense of what it is you’re trying to accomplish in life.

Although these would seem to capture quite well a lack of direction and purpose in life, we do not find stark differences between graduates of different schooling sectors on these measures. One question was the exception:

Your life often seems to lack any clear goals or sense of direction.

On this measure, the Catholic school graduates disagreed more strongly than the public school graduates. This provides some evidence that the experience of a religious school can provide direction during the turbulent young adult time of life. Note that this is a Catholic school effect, which may be driven primarily by the educational and career directions that Catholic school students seem to imbibe. Still, it is reasonable to argue that a sense of religious obligation or commitment, perhaps related to the view that working toward a life goal that is meaningful and valued, explains why Catholic school students appear to uphold a strong life-trajectory norm that includes completing their education.

The concern about young adults’ direction in life goes beyond anomie to self-centred and egoistical orientations that seem characteristic of many millennials (Smith et al., 2011). The NSYR includes questions about what young adults value in life. Again, we would expect that religious schools spend a good deal of time talking about what is valuable in life, including family and sacrifice for others. And religious schools may emphasize practices, such as service learning and service opportunities for the religious community as a whole, that may contribute to pro-social orientations or at least to a willingness to think beyond the needs of the self, narrowly conceived. The NSYR questions include the following:

Below is a list of statements that describe things that some people feel help them have the kind of lives they want to. For each statement, please rate how important you personally feel it is for having the kind of life you want to live.

  • Being able to make a lot of money.
  • Having a fulfilling romantic relationship.
  • Having a good family life.
  • Helping others in need.
  • Having a close relationship with God.
  • Having a close set of friends you can count on.
  • Having meaningful and challenging work.
  • Becoming a famous or important person.
  • Travelling and having different experiences.

Surprisingly, we did not find school-sector differences on the emphasis on family, becoming famous, or travelling. We do see some school effects for other outcomes. As we mentioned earlier, Catholic school graduates are more likely than public school graduates to emphasize a fulfilling romantic relationship. Protestant school graduates are less likely than public school graduates to think that making a lot of money is an important direction for their lives. Instead, both homeschooling and Protestant school graduates favour a life in which they have a close relationship to God. That may provide a form of guidance during the young adult years for the Protestant and homeschooler compared to the average millennial. Protestant school graduates are also more interested in meaningful and challenging work than are the public school graduates. While helping others in need is not significantly related to religious school sector, we do find a negative relationship between non-religious private schooling and the life goal of helping others in need.

Overall, then, we don’t find consistent effects of religious schooling on views of the kind of life that is worth living. The lack of findings may reflect to some extent the difficulty of accurate measurements of life goals or outlook. Yet we expected that religious school graduates would value family life and perhaps also helping others. We can say at least that on these scores they are not less interested than public school graduates in these life goals, which is perhaps more encouraging than we can say for private non-religious school graduates, who are not focused on helping others. The findings do provide some evidence among the Protestant school graduates of a concern for meaningful work—perhaps the kind that fulfills a religious calling—and a lack of interest in simply accumulating wealth. Instead, the Protestant and homeschool graduates place a high value on a close relationship with God.

If religion offers a protective effect against being lost in transition, there is further evidence that religious school graduates are not likely to take entirely the same path as graduates of other sectors. The NLSY97 includes a measure in 2011 that asks respondents if they think God has nothing to do with what happens to them. Even after controls, Catholic and evangelical Protestant school graduates are not likely to agree to this. The evangelical Protestant school effect is over twice the Catholic school effect. The estimate places the likelihood that evangelical Protestant school graduates at half the likelihood that public school graduates will agree that God has nothing to do with their life. That religious school graduates see the hand of God in their life perhaps also reflects the greater likelihood that they will have a sense of direction and purpose during the young adult years.


We do not have conclusive or entirely consistent evidence that religious schooling is a bulwark against getting “lost in transition” during the young adult years. In many ways, the processes of young adulthood appear to be similar for graduates of each school sector. The estimates of school-sector effects on extended adolescence point to religious school differences, but given sample-size issues and perhaps greater variability in the religious school sector than we might expect, many findings are not statistically significant.

However, we do find some important evidence that religious school graduates take on their twenties in a slightly different fashion than non-religious school graduates. Religious school graduates are less likely to have sex in their teenage years, which may be a result of a commitment to sex within marriage and, especially among Catholic school graduates, may reflect a concern that pregnancy would disrupt college or career goals. Religious school graduates remain more committed to marriage in their twenties, though Catholic school graduates are perhaps more committed in theory than in practice, since all evidence points to a commitment to following the norm of education and then career, followed by marriage and, much later, children. Catholic school graduates are oriented to the family, but their direction in young adulthood is more strongly oriented to staying on track for a successful educational and occupational career. That trajectory often leads to marriage opportunities in the workplace, though at a later stage in the life course (McClendon, Kuo, and Raley, 2014). In contrast, marriage likely does keep many evangelical Protestant school graduates from following the extended adolescence path.

Homeschool graduates show some inclination toward marriage and family, but the most solid conclusion that we can arrive at is that homeschool graduates are not concerned with normal educational trajectories. It would be difficult to say that is the result of being lost in transition, since most homeschool graduates on principle take a practical approach to the place of higher education in their lives. And they do seem to orient their life through religion. Evangelical Protestant school graduates appear to be concerned with staying on track with marriage and family, and in some ways show stronger inclinations toward a meaningful purpose and direction in life, which may reduce some of the aimlessness of the twenties. A focus on educational career seems to orient Catholic school graduates away from some of the tendencies of extended adolescence, while the focus on marriage and the family appears to do the same for evangelical Protestant school graduates.


Arnett, Jeffrey Jenson. 2001. Adolescence and Emerging Adulthood: A Cultural Approach. Upper Saddle River, NJ: Prentice Hall.

———. 2011. Debating Emerging Adulthood: Stage or Process. New York: Oxford University Press. Bryk, Anthony S., Valerie E. Lee, and Peter B. Holland. 1993. Catholic Schools and the Common Good.

Cambridge, MA: Harvard University Press.

Dean, Kenda Creasy. 2010. Almost Christian: What the Faith of Our Teenagers Is Telling the American Church. New York: Oxford University Press.

Engelberg, Ari. 2016. “Religious Zionist Singles: Caught Between ‘Family Values’ and ‘Young Adult- hood.” Journal for the Scientific Study of Religion 55, no. 2:349–64.

Freitas, Donna. 2008. Sex and the Soul: Juggling Sexuality, Spirituality, Romance, and Religion on Amer- ica’s College Campuses. New York: Oxford University Press.

———. 2013. The End of Sex: How Hookup Culture Is Leaving a Generation Unhappy, Sexually Unful- filled, and Confused About Intimacy. New York: Basic Books.

Greeley, Andrew M. 1967. The Catholic Experience: An Interpretation of the History of American Cathol- icism. Garden City, NY: Doubleday.

———. 1982. Catholic High Schools and Minority Students. New Brunswick, NJ: Transaction. Greeley, Andrew M., Peter H. Rossi, and Leonard J. Pinto. 1964. The Social Effects of Catholic Educa-

tion. Chicago: National Opinion Research Center, University of Chicago.

Martin, Monica J., Shelley A. Blozis, Daria K. Boeninger, April S. Masarik, and Rand D. Conger. 2014. “The Timing of Entry into Adult Roles and Changes in Trajectories of Problem Behaviors During the Transition to Adulthood.” Developmental Psychology 50, no. 11:2473-84.

Chen-Lan Kuo, Janet; R. Kelly Raley, and David McClendon. 2014. “Opportunities to Meet: Occupa- tional Education and Marriage Formation in Young Adulthood.” Demography 51, no. 4:1319–44.

Raley, R. Kelly, Sarah Crissey, and Chandra Muller. 2007. “Of Sex and Romance: Late Adolescent Rela- tionships and Young Adult Union Formation.” Journal of Marriage and Family 69, no. 5:1210–26.

Sikkink, David, and Sara Skiles. 2015. Homeschooling and Young Adult Outcomes: Evidence from the 2011 and 2014 Cardus Education Survey. Hamilton, ON: Cardus.

Smith, Christian, and Patricia Snell. 2009. Souls in Transition: The Religious and Spiritual Lives of Emerg- ing Adults. New York: Oxford University Press.

Smith, Christian, Hilary Davidson, Kari Christoffersen, and Patricia Snell. 2011. Lost in Transition: The Dark Side of Emerging Adulthood. New York: Oxford University Press.