SEARCH

SEARCH BY CITATION

Keywords:

  • education;
  • volunteering;
  • emerging adulthood;
  • faith-based schooling;
  • adolescence

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

Research on civic engagement shows that volunteering rates decline as young people move from adolescence into emerging adulthood. Using panel data spanning this period of the life course, we examine the impact of secondary schooling type—public, Catholic, Protestant, private nonreligious, and homeschool—on sustaining volunteering into emerging adulthood. We apply a framework that posits pathways between secondary schooling and sets of opportunities to volunteer embedded in institutions and social networks. We also posit a link between schooling type and durable motivational dispositions to volunteer. Results indicate substantial differences by schooling type, although our measures of opportunity structure and motivation do not adequately account for these differences. Those educated in Protestant secondary schools are considerably more likely than other young people to continue to volunteer, even accounting for potential spurious influences. Those schooled at home or in private nonreligious settings are significantly less likely to continue volunteering. We conclude by discussing two alternative accounts that should be addressed in further research: one focused on the role of habituated social practices and the other focused on differences in organizational efforts to link adolescent volunteering to emerging adult volunteering.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

By most measures, civic engagement among American adults has eroded over the past half-century (see, e.g., Macedo 2005; Putnam 2000; Skocpol 2004). As a result, scholars have given considerable attention to the correlates of volunteering, charitable giving, political activity, and other forms of participation in public life. In most rankings of these correlates, education and religion appear near the top as predictors of robust engagement. But the mechanisms that link education or religion to civic engagement are profoundly complex and often counterintuitive. Nowhere is this more evident than in efforts to explain their role in shaping civic-oriented volunteering during the transition from adolescence to adulthood. The general scholarly assumption is that the effects of education and religion on this type of volunteer activity are indelible; attitudes about civic obligation and engagement carry from the dynamic formative years into early adulthood and perhaps the remainder of the life cycle. For many government agencies and nonprofits, this assumption is an important reason for partnering with schools and faith-based groups as key sites for developing early habits of good citizenship among adolescents.

In this article, we are particularly interested in youth volunteering as a specific form of civic engagement that may persist into adulthood. While some volunteering lacks a civic purpose (and some civic engagement does not entail volunteering), our use of variations on the term “volunteer” refers to a citizen's intentional, organized efforts to address matters of public concern.1 Many studies have examined the role of religion or education in youth civic engagement, but we suggest that two aspects of the relationship of education and religion to youth volunteering need further development. First, scholars know more about how religion or education generates youth volunteering than how those factors sustain volunteering over time. While adolescent volunteer activity clearly influences civic-oriented volunteering later in life, the steep decline in rates of this volunteering among young adults in their early 20s indicates that volunteering in adolescence does not inevitably produce emerging adults who volunteer (see Musick and Wilson 2008:234–35). Second, we know more about the independent influences of religion and education than how those factors might interact to affect the likelihood of sustained volunteering over time. If both religiosity and education separately help to sustain civic volunteerism over time, then it is a plausible expectation that the two factors in combination will heighten the effect. Our goal in this research is to examine the effect of that interaction on sustaining volunteering from adolescence into early adulthood.

More specifically, we ask whether the secondary education of youth in varied contexts—public, faith-based, nonreligious private, or homeschooled—helps generate momentum for civic volunteering that carries into adulthood. Numerous studies have found that adolescents’ attitudes about and engagement in civic life vary with schooling type (Campbell 2001; Coleman and Hoffer 1987; Dee 2005; Greene, Giammo, and Mellow 1999; Greene, Mellow, and Giammo 1999). The general message of this research is that many forms of religion-based education have a neutral to modestly positive effect on a range of civic outcomes, including political knowledge, tolerance, and volunteering (for an overview, see Wolf 2007). But these studies tend to examine how school type shapes the civic experience of adolescents while still in school, rather than the role of school type, if any, in fostering volunteering and other forms of civic engagement in adulthood. We build on this earlier research by examining whether an adolescent's school type helps to sustain his or her likelihood of volunteering as young adult.

Our analysis includes robust controls for possible spuriousness as well as potential mechanisms of influence. We examine the role of parental and youth religion, parental modeling and valuing of volunteering, the nature and extent of adolescent volunteering, potential life course trajectories that include employment and postsecondary schooling, prosocial orientation and empowerment, as well as structured opportunities to volunteer based on organizational activity and friendship networks. We unpack how these components help us specify the impact of secondary schooling type on sustaining volunteering into emerging adulthood. The results suggest that the religious identity of an adolescent's school can activate or suppress engagement in adulthood, and often in ways that defy portrayals of some types of schools as socially isolating, inward-looking, and fragmented. Indeed, we conclude that schooling type predicts different levels of volunteering even among individuals who otherwise share key social characteristics and orientations. This suggests that schools can have a powerful and independent influence on adult engagement.

Civic Engagement, Social Networks, and Religious Salience

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

Unlike political participation, which is directed toward the influence of government, the concept of civic engagement has a much more capacious—and often amorphous—connotation. But scholars have increasingly turned to the concept to help explain shifting patterns in civic involvement across generational cohorts. For example, in their path-breaking work on citizenship and participation, Zukin et al. (2006:7) define civic engagement as “organized voluntary activity focused on problem solving and helping others.” Based on that definition, they suggest that younger people are increasingly civically engaged, even though they are declining in their political participation. For Zukin's team and others who make similar arguments (see, e.g., Dalton 2009), volunteering exemplifies these emerging patterns of youth civic engagement. Young adults are more likely than older cohorts to see volunteering rather than conventional politics (e.g., voting) as a key way to address public problems through concerted action.

For most of these young adults, exposure to volunteering of various kinds begins in adolescence. One estimate suggests, for example, that as many as 75 percent of adolescents in the United States will participate in formal volunteering opportunities before adulthood (Johnson et al. 1998). Moreover, contrary to the broad pattern of disengagement across various age cohorts, the trend line for adolescent volunteering appears to be moving up. By the mid-2000s, nearly a third of older teenagers (ages 16–19) reported some volunteering activity in a given year, a rebound of more than twice the volunteering rate for that age group in 1989 (Grimm et al. 2006). While most of these volunteers are “episodic” (contributing 99 or fewer hours a year, according to the definition in Grimm et al. 2006), one data analysis suggests that 11 percent of 12th graders volunteer on a weekly basis (Smith and Faris 1999).

The two most common venues for adolescent volunteering are religious and educational organizations, which each capture about a third of volunteer activity (Grimm et al. 2006). In the United States, this is not surprising. Religious institutions are the most prominent set of associations within civil society; some estimates suggest that they account for fully half of all volunteering, charitable giving, and other civic activities in the country (Putnam 2000; Smidt et al. 2008). We also know that adolescents volunteer more frequently in both religious and secular settings when they belong to a religious group, say that religion is important to them, hold spirituality as a high value, or attend religious services regularly (Crystal and DeBell 2002; Gibson 2008; Sundeen and Raskoff 2000).

While historically religious institutions have consistently mobilized volunteer labor, the role of educational institutions in fostering volunteerism has increased in the past few decades (Spring, Dietz, and Grimm 2006). Social theorists and empirical social scientists have begun to realize that civic education and engagement are not merely a matter of garnering adolescent support for the political system; participating in public life is also a crucial part of the adolescent's developmental process (Dudley and Gitelson 2002; Flanagan 2003). That realization has prompted a movement among civic educators to couple traditional civics courses with high impact volunteering experiences, including service learning, as incubators of good citizenship (Andolina et al. 2003; Kahne, Chi, and Middaugh 2006; Spring, Dietz, and Grimm 2006).

One of the chief goals of both faith-based and school-based experiences is to foster volunteering in adulthood. But do these experiences succeed? Longitudinal studies confirm some basic facts. First, participation in a religious community persists from adolescence to adulthood as a predictor of volunteering and other forms of engagement (see Toppe, Kirsch, and Michel 2002:91). Second, exposure to volunteering opportunities in secondary education also increases the likelihood of volunteering and other forms of participation (e.g., voting) later in life (Hart et al. 2007).

While these facts are straightforward, the mechanisms underlying them are not. We posit two basic frameworks—one focused on opportunity structures, the other on motivation—that link the religious and educational experiences of young people to adult volunteering. We explore these mechanisms in some detail below.

Opportunity Structures

Religious organizations and schools are the primary institutions that directly recruit young people to volunteer. In a national sample of adolescents, Sundeen and Raskoff (2000) find that roughly a third of teens report volunteering because of their participation in a formal organization or group. Of these teens, a sizable majority (65 percent) report recruitment through a place of worship. Likewise, among all teens, nearly 40 percent report that they have volunteered for one reason or another because of their school. Why are these institutions so frequently the sites of recruitment? On the one hand, these institutional spaces are particularly well suited for groups to “prospect” for potential volunteers (Brady, Schlozman, and Verba 1999; Sundeen and Raskoff 2000). In the United States, churches and other houses of worship are attractive sites for prospecting, particularly when they combine strong emotional connections among parishioners with an expectation of social outreach (Becker and Dhingra 2001). Schools, too, are propitious venues for recruitment, partly because they present an opportunity for broad and relatively easy access to potential volunteers. Very few institutions, outside of schools and places of worship, contain a comparable pool of potential volunteers. Just as important, both houses of worship and schools encourage volunteering because of broadly conceived theological or humanistic visions of human relations and a just society. Without a broader narrative about why volunteering is an inherent good, perhaps even a duty, pools of potential recruits do little good.

How does this relate to the question of sustaining volunteering that is central to the present research? There are three ways that we expect religious and educational organizations to alter the opportunities to volunteer as young people transition to emerging adulthood. First, organizations can structure opportunities by encouraging (even requiring) volunteering directly during adolescence. The repeated exposure to opportunities to volunteer, and, for many, actual volunteering experiences will help to “normalize” volunteering for young people. Second, these types of organizations provide a context for developing certain types of social networks. These networks matter because (1) they often persist beyond adolescence, and (2) they have the potential to reinforce a positive orientation toward civic voluntarism. Finally, schools and religious organizations in adolescence can be expected to influence young adult volunteering by filtering young people into other institutions, such as colleges, churches, and political organizations, which open up additional opportunities to volunteer. While both schools and places of worship undoubtedly use each of these channels to influence volunteering, it is not as clear what sort of outcomes we should expect when religion and schooling overlap, as is the case with faith-based education. We will turn to some hypotheses shortly, but first we deal with the other primary hypothesized mechanism of influence: motivation.

Motivation

The role of education and religion in shaping volunteering has largely been examined through this lens of opportunities and social networks. But this is only half the story. To use a criminal justice metaphor: committing a crime requires motive as well as opportunity. Here we review how religious and educational institutions might shape the motivations that sustain youth volunteering into emerging adulthood.

We can think about motivation as both intrinsic and extrinsic. Intrinsic motivation to volunteer comes through the internalization of certain norms and values. One approach to explaining intrinsic motivation is to examine whether and how adolescents exhibit a prosocial orientation. Penner and Finkelstein (1998:526) define this orientation as “the tendency to think about the welfare and rights of other people, to feel concern and empathy for them, and to act in a way that benefits them.” From a psychological perspective, adolescents with this kind of orientation tend to be more self-reflective about their own goals and personal values; their civic activity is less habitual and more self-aware and intentional (Hart and Fegley 1995). From a developmental perspective, Janoski, Musick, and Wilson (1998) argue that adults who were socialized as youth to have prosocial attitudes are even more likely to volunteer and give than those adults who were given practical volunteering experiences without much reflection about the reasons for their work.

We expect that being socialized into a prosocial orientation will be dependent on the larger philosophical or religious vision of human good promoted in an institutional setting. We note that some scholars have been skeptical about religious beliefs as a source of a prosocial orientation, suggesting instead that the social dimensions of religious networks are the best explanation of civic activity. In their recent analysis, for example, Putnam and Campbell (2010:472) argue that “religiously based social networks … alone account for most of the apparent effects of church attendance” on the entire range of generosity, good neighborliness, and civic engagement. While they find that these network effects are strongest within religious institutions, they insist that those effects result from the peculiar “morally freighted” social interactions within those institutions rather than specific theological convictions or values (Putnam and Campbell 2010:477). Still others contend that religious motivations are necessary but insufficient conditions for engagement. For example, both Smidt et al. (2008) and Loveland et al. (2005) argue that private devotional activity interacts with public religious participation to produce higher levels of civic participation than either category alone. The upshot of these studies is that the intrinsic commitment to religious faith adds value to civic engagement beyond taking part in the social environment of religion.

It is important to note that feeling concern for the welfare of others is not a sufficient explanation of intrinsic motivation. One might have this concern yet feel entirely ill equipped to act on it. Hence intrinsic motivation to volunteer combines a prosocial orientation with a sense of empowerment. It is useful here to draw from what political scientists typically call internal political efficacy, that is, an individual's belief that he or she has the competency to engage effectively in political activity. It is almost a social scientific truism that those with higher self-efficacy are more likely to participate (Pollock 1983; Rudolph, Gangl, and Stevens 2000), and some recent scholarship associates higher levels of self-efficacy with belonging to schools and other communities that take participation seriously (Anderson 2010; Kahne and Westheimer 2006). The underlying assumption of this scholarship is that an individual's sense of empowerment can be located along a spectrum defined by the individual's sense of control. At one extreme an individual believes his or her own efforts can address public problems; at the other extreme an individual believes those efforts are fruitless when compared to what other people and institutions can do.

However, both religious and educational institutions can provide various incentives that induce extrinsic motivations to volunteer as well. In schools, the rewards for volunteering are often tied to course requirements, school-sponsored club activities, and college application expectations (Sundeen and Raskoff 1994). Similarly, places of worship often expect and reward various forms of volunteer activity (Becker and Dhingra 2001). In the real world, separating intrinsic and extrinsic motivation to volunteer is rather difficult. Young people may not even be fully aware of the variety of social forces that pull and push them into volunteering. Nevertheless, intrinsic and extrinsic motivations are conceptually distinct, and they very well may have different consequences as many adolescents leave their home, high school, and church as they enter emerging adulthood.

Secondary Schooling and Adolescent-to-Adult Engagement

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

We now turn to more precisely hypothesizing the impact of secondary schooling type on sustaining volunteering into emerging adulthood. As already noted, research confirms a strong association between adolescents’ volunteering and (1) the intensity of religious beliefs, environment, and social network and (2) the nature and extent of exposure to public life in schools. Some forms of secondary schooling combine these two factors through private education grounded in specific faith traditions. We expect the combination of the two influences will leave a civic imprint that lasts into adulthood.

However, there is an important objection that has been raised to the above expectation. Perhaps as a response to the political ascent of the Religious Right in the 1980s and early 1990s, there has been ample theoretical speculation about the effects of private religious education on a citizen's dispositions and participation. Much of that speculation has come in the form of normative assessment of various forms of nonpublic education, which some critics fear will diminish or fragment shared public values in the United States (e.g., Dwyer 1998; Gutmann 1999; Rose 1988). These critics have raised a range of concerns about the role of education in shaping good citizens—and, in the abstract, rightly so. After all, the polity has a stake in a set of institutions that have a demonstrable effect on civic outcomes (Andolina et al. 2003; Campbell 2006; Zukin et al. 2006). Moreover, private religious schools serve a significant proportion—nearly 10 percent—of students in the United States (National Center for Education Statistics 2011:22).

Critics often base their normative assessments in what are properly empirical assumptions that private religious education is frequently—perhaps fundamentally—illiberal and isolating, at best, and downright threatening to the body politic, at worst. Some of those assumptions have been challenged by systematic analysis. Several observational and quasi-experimental studies have found that adolescents who attend religious schools generally match or exceed their public school counterparts on measures of tolerance, political knowledge, and rates of volunteering (see Wolf 2007). While most of these studies focus on adolescents themselves, recent analysis does suggest that type of school is correlated with positive civic and educational outcomes in adulthood (Pennings et al. 2011). Citizens who attend Catholic schools or private nonreligious schools tend to be indistinguishable from public school graduates in their levels of volunteering, financial giving, and other forms of engagement in adulthood, and indeed these three types of graduates tend to report similar attitudes about life purpose and faith commitment. In contrast, graduates of Protestant schools are significantly more engaged in volunteering and giving in both religious and secular settings, they are more self-reflective about the spiritual dimensions to their civic work, and they also tend to have a stronger sense of civic obligation and purpose than others. It is worth noting that despite these conclusions, Pennings et al. (2011) found that Protestant-school-educated adults were less likely to engage in political activity than any other education type, which creates an intriguing contrast with Dee's (2005) finding that Catholic education increases the likelihood of voting in adulthood.

These findings clearly suggest a role for school type in shaping civic engagement among adolescents, and they more tentatively point to an effect of school type in adulthood. But they do not focus specifically on the empirical question that motivates our research: Does the combination of religion and education in the form of schooling type help to sustain civic engagement from adolescence to early adulthood? We are particularly interested in whether different types of educational settings explain adult volunteering even among those adults who were particularly active as adolescents. This focus provides insight into whether adolescent volunteering is by itself a sufficient impetus to adult activity, or whether certain forms of socialization are necessary ingredients to sustained involvement later in life.

We frame our analysis around four key hypotheses. The first essentially posits a spurious relationship between school type and adult volunteering:

H1

Key social characteristics that coincide with school type—and not the schools themselves—are associated with the likelihood of volunteering in adulthood.

It is quite possible, given the evidence that we have reviewed, that secondary schooling itself has little impact on sustaining volunteering. In this scenario the type of students who self-select into private religious schooling, private nonreligious schooling, and home schooling are all different on a number of religious and familial characteristics from adolescents who attend public schools. If it is these characteristics, and not schooling itself, that produces different rates of volunteering in emerging adulthood, then we have simple case of spuriousness.

Alternatively, perhaps secondary schooling does influence volunteering as young people transition to adulthood, even after accounting for religious and familial factors. As we have outlined above, this could happen through two different classes of social mechanisms:

H2

An adolescent's school type will be more or less likely to provide opportunities that lead to volunteering in early adulthood.

And/or

H3

An adolescent's school environment will socialize some young people into certain motivational dispositions that alter the likelihood of volunteering in emerging adulthood.

We deal more precisely with how we measure religion, family, opportunities, and motivations later. Before moving to the analysis, though, we note that these hypotheses are not mutually exclusive. It is plausible that some of the influence of schooling type on sustaining volunteering is spurious, some occurs through opportunities, and some through altering motivations.

Finally, in light of the specific social mechanisms that we hypothesize link school type with volunteering, we develop general expectations on secondary school sector differences in impact. More precisely:

H4

Adolescents educated in Catholic and Protestant secondary schools will be more likely to sustain volunteer activity into emerging adulthood compared to adolescents educated in other secondary schools.

The logic is quite straightforward here. We expect, on average, that structured opportunities, the types of friendship networks, the distinct life course trajectories, and the types of motivations nurtured by faith-based institutions will lead to higher rates of volunteering during the transition to adulthood because of the unique religious and educational mission of these schools. The type of moral education and community generated in faith-based institutions cannot be easily replicated in the public sector. At the same time, we expect home schoolers to be at a disadvantage when it comes to the number of opportunities they have to volunteer. Because of this, we expect that Catholic and Protestant schooled emerging adults will continue volunteering at higher rates when compared to other emerging adults.

Data and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

Our data source is the National Study of Youth and Religion (NSYR), a longitudinal, multiwave survey that includes extensive batteries of questions on youth religion, education, and engagement. The NSYR began as a nationally representative telephone survey of 3,290 teenagers between the ages of 13 and 17 (English- and Spanish-speaking) and a nonrepresentative oversample of 80 Jewish households, bringing the total survey size to 3,370. The teen respondents and one of their parents were initially interviewed between July 2002 and April 2004. The second and third waves of the study were resurveys of English-speaking adolescent respondents from Wave 1. The second wave was conducted in 2005, when respondents were between the ages of 16 and 21; the third was conducted in 2007 and 2008, when respondents were between 18 and 23 years old. Because our interest is the sustaining of civic volunteering from adolescence to adulthood, all the data used in the present study are taken from Wave 1 and Wave 3 of the study. The retention rate from Wave 1 to Wave 3 was 77.1 percent. (For more information about the survey, see Smith and Denton 2005.)

In the multivariate analysis, our key dependent variable is a dichotomous measure of volunteering. We computed the volunteering variable by identifying those 18–23 year olds (Wave 3) who had reported volunteering at age 13–17 (Wave 1). The relevant question at Wave 1 asked the respondent: “In the last year, how much, if at all, have you done organized volunteer work or community service?” Adolescents who responded “never” were coded as “0”; those who responded any volunteer activity, from rare to regular, were coded as a “1.”

We acknowledge here that the volunteering question in the NSYR is not precisely focused on distinctively “civic” activity in contrast to other types of volunteering. Our response is threefold. First, by evoking “community,” the question does prompt respondents to think in civic terms. Second, we know from other studies that the setting for most volunteering reflects a civic purpose behind a great deal of volunteer activity (see, e.g, Grimm et al. 2006). Finally, as discussed later, we apply some controls for volunteer setting to strengthen the assumption that we have measured largely civic-oriented volunteering. Taken together, these responses make a strong case that our dependent variable measures volunteering as a form of civic engagement.

Our primary predictor variable is educational setting at the secondary level. The survey at Wave 1 asked a parent of each adolescent a series of questions to identify the adolescent's predominant setting for secondary schooling. The result is a sizable array of responses, including traditional public, private nonreligious, magnet/charter, Catholic, Baptist, Lutheran, Jewish, homeschooled, or other. Because several options received less than 1 percent of the total responses, we identified plausible affinities among options and collapsed them into new categories. The result is a nominal-level variable with the following categories: public, Catholic, Protestant, private nonreligious, homeschooled, and other.

In addition to secondary schooling, we identified a range of religion-oriented variables that the literature suggests are possible determinants of volunteering. Indicators of religious belief, behavior, and affiliation were constructed from both parental and adolescent responses at Wave 1 and young adult responses at Wave 3. The NSYR provides measures of religious tradition at Wave 1 (e.g., Catholic, mainline Protestant, evangelical Protestant, Jewish), which are included in the multivariate analysis to help distinguish the effect, if any, of religious education from other aspects of religious social context. Similarly, we include controls for parental and youth worship attendance (for youth, at Waves 1 and 3), youth's religious salience and private devotionalism (e.g., self-report of religion's importance and frequency of prayer and scripture reading) at Wave 1 and Wave 3, the frequency of talk about religious subjects at home, and whether or not the adolescent belongs to a religious youth group.

We also control for household support and modeling of volunteering through three measures from the parent survey. We include a measure of whether the parent encourages his or her teen to volunteer, whether it is important to the parent if the teen volunteers, and whether or not the parent does any volunteer work. Also from the parent survey, we include a measure of whether the parent thinks the teen's friends are a good influence or not.

Other measures that we incorporate into our models include the amount of volunteering that was done through a religious organization (measured at Wave 1), whether the respondent lives with his or her parents (measured at Wave 3), and whether the respondent is employed (measured at Wave 3). If the respondent is in college, we use the Integrated Postsecondary Education Data System (IPEDS) code to separate the respondents into attending religiously affiliated college and universities, elite universities (measured by the top 51 U.S. News and World Report rankings for national universities in 2007), other four-year colleges and universities, or associate's degree granting institutions. If the respondent was not currently in college at Wave 3, we created a dichotomous variable for those who had attended some college (we include those who have earned an associate's degree in this category) but not graduated, and a dichotomous variable for those who had earned a bachelor's degree or more. The reference category for all postsecondary schooling variables is those respondents who have never enrolled in any type of college. All categories for postsecondary schooling are mutually exclusive.

To unpack motivation to volunteer, we included the NSYR's rough gauge of prosocial orientation, which is measured in a series of questions that probes young adult attitudes about certain vulnerable or minority groups (the questions ask about the extent that a respondent “cares” about racial equality, the elderly, or the poor). In addition, we used several Likert items to construct a psychological measure of empowerment, labeled “locus of control.”2 The analysis also evaluates several variables that focus on the nature of volunteering itself, such as whether volunteering was ever required of the respondent (by school, parents, or the juvenile justice system) at Wave 1, how much volunteering was done through a religious organization (measured at Wave 1), and the percentage of a respondent's close friends who volunteer (at Wave 3). In order to get a direct measure of institutional opportunity to volunteer, we also include measures of the number of religious and nonreligious organized activities the respondent is involved in (Wave 3).

Standard demographic controls are also included beginning in Model 2. These include parents’ education, income, ideological self-identification (whether the parent respondent self-describes as a conservative or not), and the respondent's gender, age, and race/ethnicity.

Modeling Strategy

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

We begin with bivariate comparison of volunteering during adolescence (age 13–17) and volunteering during early emerging adulthood (age 18–23) sorted by secondary schooling type. This provides a baseline picture of the relationship between schooling type and volunteering over time.

We follow this initial analysis with a more complex multivariate procedure. Using only those who volunteered as adolescents, we test several models that seek to account for the possible influence of secondary schooling type on volunteering during early emerging adulthood. The first model includes only dummy variables for secondary schooling (omitting public schooling as the comparison category). The remaining models apply various potential spurious factors and social mechanisms that might help account for the basic correlations in the first model. The second model adds in several standard demographic characteristics, a series of youth and parent religious measures, and measures of parental encouraging and modeling of volunteering. All of the variables in the second model are designed to account for potential spurious factors (H1).

The third model adds in measures of required volunteering (likely required by the school) and a measure of the social influence of the adolescent's friendship network (as reported by the respondent's parent). Both of these measures are much more likely to be channels through which the school influences volunteering outcomes by potentially altering the opportunity structures to volunteer (H2). Model 4 continues modeling opportunity structures by including Wave 3 (age 18–23) life course outcomes that could plausibly be influenced by secondary schooling and that subsequently could influence volunteering opportunities. These include religious involvement, postsecondary schooling, living arrangements, and employment status.

The last two models are designed to most directly measure the two sets of social mechanisms outlined above. In Model 5, three variables designed to tap into a prosocial orientation, and one index designed to measure empowerment (locus of control), are included to see if they mediate the effects of secondary schooling. Finally, Model 6 includes two measures of institutional opportunities to volunteer and one measure of the percentage of close friends who do volunteer work. These are designed to tap into the degree of opportunities to volunteer in both organizational and social network form. We expect that these may decrease the influence of secondary schooling type (as well as some of the other measures of opportunity in Models 2 and 3).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

We began with a simple bivariate comparison of each age cohort's level of volunteering sorted by secondary schooling. For comparative purposes, these analyses only include respondents who participated in both Wave 1 and Wave 3 of the survey. Table 1 displays the results of the cross-tabulations. Looking at the top cross-tabulation, we see that adolescents begin with different rates of volunteering based on school type. Catholic school students are the most likely to volunteer during adolescence with only 13 percent reporting that they do no volunteer work. Catholic school students also report higher levels of required volunteering than other groups. These findings are consistent with the existing literature (Bryk, Lee, and Holland 1993:138–39; McLellan and Youniss 2003) that indicates overall high rates of volunteering (with much of it required) at Catholic schools. Home-schooled students are the least likely to volunteer with a full 37 percent having done no volunteering in the past year. Other groups fall somewhere in between.

Table 1. Bivariate association between schooling type and volunteering, weighted
 Secondary Schooling, Age 13–17
 PublicCatholicProtestantPrivate, NRHomeschool
Note
  1. Percentages may not add to 100 due to rounding and some respondents who were unsure or refused to answer.

  2. Source: National Study of Youth and Religion 2002–2003, 2007–2008.

Volunteering@W1     
None3413292337
All voluntary3321303035
Some voluntary2645263015
All involuntary721131713
Volunteering@W3     
None5334236177
All voluntary3132572316
Some voluntary92214123
All involuntary711643

The changes between the waves also reveal some notable patterns. In each category of schooling except Protestant, there is at least a 20 point increase from adolescence (Wave 1) to young adulthood (Wave 3) in the percentage of respondents claiming to have never volunteered in the past year, with corresponding decreases in each category in levels of voluntary service. The percentage of nonvolunteers in the Protestant category, by contrast, decreases by 6 percentage points from adolescence to young adulthood, with a corresponding increase in voluntary service.

The patterns hint at a role for type of secondary schooling as a predictor of volunteering, but they do not tell us whether schooling or other factors explain the variations across this early part of the life cycle. To test various possibilities, we turn to multivariate analysis. At the outset, it is important to note that we have limited the samples in these analyses to only those respondents at Wave 3 who had reported at least a minimal amount of volunteering at Wave 1 (even if it was all required of them). The logic behind this decision is straightforward: our goal is to examine whether type of education helps explain sustaining volunteering over time. Including the full sample would conflate sustaining volunteering with generating volunteering.3 Missing data (much of it on the measure of income) in the final models reduces the sample by about 8 percent. In order to recover the full sample of volunteers at Wave 1, we employ multiple imputation. The resulting sample includes all respondents who volunteered in Wave 1 and answered the volunteer survey question in Wave 3 (N = 1,717). Of these respondents, slightly less than half (47 percent) report at least some voluntary volunteering at Wave 3 of the survey. This means that more than half of those who volunteered as adolescents no longer did so during early emerging adulthood (or only did so to meet some sort of external requirement—6 percent of the sample making up this latter category).

We used logistic regression to predict volunteering among young adults who had volunteered as adolescents.4 To test for spuriousness as well as a variety of causal mechanisms, we specify six separate models. Table 2 displays odds ratios for each variable in these separate models.

Table 2. Logistic regression predicting voluntary volunteering at age 18–23 among respondents who volunteered at age 13–17 (odds ratios)
 (1)(2)(3)(4)(5)(6)
Notes
  1. +p < .1, *p < .05, **p < .01, ***p < .001.

  2. a

    Reference category is public schooling.

  3. b

    Other religious schooling and those who were unsure or refused to answer are controlled for but not shown.

  4. c

    Reference category is white.

  5. d

    Reference category is evangelical Protestant.

  6. e

    Those with an indeterminate religious tradition are controlled for but not shown.

  7. f

    Reference category is never enrolled in college.

  8. Source: National Study of Youth and Religion, 2002–2003, 2007–2008.

Wave 1 measures (age 13–17)
Secondary schoolinga,b
Catholic1.685*1.2781.3331.3671.3871.336
Protestant4.930***2.975*3.105**3.256**3.372**5.288***
Private, NR.420*.234**.245**.195**.204**.124***
Homeschool.463+.418*.410*.379*.400*.326*
Race/ethnicityc
Black .542+.578.520+.542+.464+
Hispanic .874.884.909.933.930
Asian .447+.494.427.438.527
Other 2.576**2.946**3.468***3.459***3.566***
Female 1.389*1.386*1.315*1.2481.559**
Age .970.971.936.933.952
Parents’ income 1.066*1.065*1.0411.0421.020
Parents’ education 1.173***1.167***1.142***1.135***1.095*
Parent politically conservative 1.1351.1301.0691.0771.082
Religious traditiond,e
Mainline Protestant 1.515+1.498+1.597*1.542+1.460
Catholic 2.077*2.013+1.948+1.915+1.996
Black Protestant 1.430+1.440+1.403+1.423+1.238
Jewish 5.894***5.592**5.556**5.263**5.097**
LDS 1.2871.277.959.958.885
Not religious 1.2631.2271.1331.0891.034
Other religion 2.381*2.279+2.606*2.563+1.942
Parent worship service attendance 1.0371.048.992.9911.001
Parent importance of faith .964.961.968.9851.035
Frequency of religious talk in home .982.988.976.970.969
Youth worship service attendance .986.977.955.949.933
Youth importance of faith .998.987.912.890.839+
Youth frequency of prayer 1.0371.0391.0081.0101.003
Youth frequency of scripture reading 1.174**1.175**1.147*1.151**1.127*
Attends a religious youth group 1.487*1.442*1.330+1.334+1.510*
Parent encourages teen to volunteer 1.0561.0401.0611.0711.038
Imp't to parent that teen volunteers 1.1051.0921.0351.021.999
Parent volunteers 1.2211.2171.2291.2311.241
Am't of volunteering through rel. org. 1.0511.0461.0391.0471.059
Volunteering required  .868*.856*.862*.887
Parent thinks teen's friends good infl.  1.165*1.144+1.151+1.142
Wave 3 measures (age 18–23)      
Religion      
Worship service attendance   1.155**1.167**1.138*
Attends rel. gp. other than worship   1.834**1.808**1.107
Importance of faith   1.0271.0051.044
Frequency of prayer   1.0561.0551.084
Frequency of scripture reading   1.000.985.907
Postsecondary schoolingf
Religious postsecondary schooling   3.522***3.329***2.289*
Elite postsecondary schooling   4.105***3.997***2.106
Other 4-year university or college   1.810**1.786**1.162
Associate's degree granting school   .956.937.844
Some college, not curr. enrolled   1.2931.2751.287
Earned bachelor's degree or higher   2.226*2.061*1.570
Living with parents   1.1461.1061.135
Employed   1.1081.1101.269
Cares about racial equality    1.145*1.219*
Cares about needs of elderly    1.1771.165
Cares about needs of poor    1.014.922
Locus of control    1.0491.119
# of org. activities run by NR org.     1.960***
# of org. activities run by rel. org.     1.627***
Percentage of close friends who volunteer     1.024***
N171717171717171717171717
McFadden's pseudo R2.02.14.14.19.20.31

As a baseline, Model 1 represents the predictor of primary interest: setting for secondary education. The reference category is public education, leaving Catholic, Protestant, private nonreligious, and home schooling as the key predictors (other religious schooling and those who refused to answer are controlled but not shown due to cell sparseness). The results echo the bivariate analysis. The odds that a respondent who attended Protestant secondary schools would continue to volunteer as an adult are nearly five times higher than the odds of a respondent educated in public schools. Catholic schooling also increases the odds (69 percent higher than public schools), but the effect does not approach the magnitude of Protestant schooling. By contrast, the odds of a private, nonreligious school attendee volunteering are less than half the odds of a public school student. All are statistically significant variables. Home-schooled emerging adults also have lower odds, and the variable nearly achieves significance at the .05 level.

The schooling effect in Model 1, however, may simply be spurious. To test H1, the results must be subjected to familial, religious, and other sociodemographic controls to distinguish the independent role of school type from relevant respondent characteristics that are associated with, yet independent from, school type. Model 2 includes these key controls. First, it includes parents’ socioeconomic status and ideology, and respondent's age, gender, and race/ethnicity; second, Model 2 accounts for religiosity at Wave 1, including religious tradition (evangelical Protestant is the reference category), the frequency of religious talk in the home, parental and youth worship attendance, importance of faith, measures of private devotionalism, and youth group attendance. In several instances, religious tradition has a statistically significant impact on volunteering. In fact, all groups—including those with no religion—have a higher predicted rate of volunteering during emerging adulthood when compared to evangelical Protestants (although not all reach statistical significance). This tells us that, net of measures of social class and personal religiosity, belonging to an evangelical Protestant denomination is associated with lower levels of sustaining volunteering into emerging adulthood. This is consistent with the overall lower levels of “bridging” civic engagement that have been reported in conservative Protestant congregations (Chaves 2004). Attendance and private devotionalism have little influence, with the exception of scripture reading. Somewhat surprisingly, the additional measures of religion have no discernible influence on volunteering with the exception of private scripture reading and youth group attendance.

Finally, we add several measures of parental support for and modeling of volunteering to Model 2 from the first wave of the NSYR. We also include a measure that reports how much volunteering during adolescence is through a religious organization. These are arguably the most important, and direct, measures that control for spuriousness. Although all of these variables positively influence emerging adult volunteering, none reach statistical significance.

While there is an appreciable influence on the predictors in Model 1 when adding the variables in Model 2, the two models follow a similar pattern (although Catholic schooling is no longer statistically distinct from public schooling). Protestant schooling, which had the strongest initial influence on subsequent volunteering, decreases by roughly 40 percent with the addition of the controls for spuriousness but remains positive and significant. The other noticeable change between the first two models is the increasingly negative impact of private nonreligious schooling on sustaining volunteering into emerging adulthood. We consider the coefficients from Model 2 to be the best estimate of the influence of secondary schooling type on long-term volunteering.

Unlike the variables in Model 2, Model 3 attempts to measure two characteristics that are not likely causally prior to the adolescent's educational experience. The required volunteering variable reflects a plausible concern that “mandatory volunteerism” in secondary schools and universities breeds resentment and deters civic engagement (Stukas, Snyder, and Clary 1999)—a concern the data here appear to vindicate. The second variable in Model 3 shifts focus away from school curriculum to school culture, and more specifically to the influence of a teen's peer group. Measuring such influence is notoriously difficult, but the NSYR uses parents as informed observers. In Model 3, adolescents whose peer groups received high parental marks were slightly more likely to report volunteering in adulthood. But the most important conclusion is that neither variable had a substantial effect on the odds ratios for the various schooling types. We can confidently conclude that the reasons for the differences between secondary schooling types in sustaining volunteering has little to do with requiring volunteering or the “good influence” from friends at certain types of schools.

Model 4 adds the young adult's religious activity, educational attainment, living arrangement, and employment status at Wave 3. This model is attempting to measure if the influence of secondary schooling type primarily works by filtering young people into different life course trajectories in emerging adulthood. Attending a house of worship or another religious group are both significant and increase the odds of volunteering, while salience (self-assessment of “importance of faith”) and devotionalism (prayer, scripture reading) fail the significance test. Being currently enrolled in a postsecondary institution (with the exception of an institution that primarily grants associate's degrees), or having graduated from college with a bachelor's degree or more, tends to have positive influences on reporting voluntary volunteer activity during emerging adulthood when compared to those who have never enrolled. This is particularly the case for students enrolled in elite or religious institutions. Neither living with one's parents nor employment status is statistically significant. Most importantly, these variables seem to have little overall influence on the coefficients for secondary schooling type. Once again, this suggests that the influence from schooling type has little to do with differences in subsequent life course trajectories.

Model 5 is the most direct specification of intrinsic motivation and sense of empowerment. While we must admit they are rough approximations, the three measures of “cares about…” provide some indication of a prosocial orientation. “Locus of control” is a psychological indicator of sense of empowerment, as discussed above. Only one of these variables is significant, and taken together they do not change the odds ratios for school type. This suggests that H2 is not borne out by the evidence.

Model 6 is the most direct specification of institutional opportunities for engagement, as well as the role that a young adult's social networks might play in that engagement. We distinguish religious from nonreligious organized activities (e.g., clubs, sports) in which a respondent is involved, and we also include a measure of volunteering among the respondent's closest peers. Each of these variables has a positive and significant effect on volunteering. But the results in Model 6 do not suggest that social network effects explain the odds ratios of schooling type in predicting volunteering, as H3 would predict. Indeed, the inclusion of these variables actually strengthens some of the existing patterns, with the odds ratios increasing to more than five for the Protestant schooling, and the odds ratios decreasing for both private nonreligious schooling and home schooling. We discuss the implications of these findings in the next section.

The influence of school type, despite robust controls and an attempt to account for the most plausible social mechanisms, is still highly predictive of whether or not an adolescent will continue to volunteer once he or she enters emerging adulthood. In order to get a better sense of the impact of schooling type on sustaining volunteering, we present a set of predicted probabilities in Figure 1. These are calculated from the coefficients of school type in Model 6, Table 2, holding all other variables at their mean. Also, so that we can display the possible range of true population values (and also to give a sense of how confident we can be in these estimates), we include error bars for the 95 percent confidence interval. The predicted probability for a public school student is about 48 percent, very close to the overall mean. For Catholic school students, it is slightly higher at 55 percent, although we cannot be confident that this is statistically different from those who attended public schools. Those who attended Protestant schools are substantially different. Our final model predicts that 83 percent of Protestant schooled emerging adults who volunteered in adolescence continue to volunteer—even after controlling for volunteering in the home, religious practice/beliefs, opportunities to volunteer, the number of close friends who volunteer, and other variables. We find this to be quite remarkable and will discuss the implications of these findings shortly. Nearly equally remarkable are the negative impacts from private nonreligious schooling and home schooling. Our models suggest that only 10 percent of emerging adults who attended private nonreligious schools continue to volunteer, and only 23 percent of home schoolers do the same.

image

Figure 1. Predicted probabilities of volunteering at age 18–23 by secondary school type (from Model 6, Table 2)

Source: National Study of Youth and Religion 2002–2003, 2007–2008.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

Our analysis examines a considerable range of factors that might explain what sustains adolescent-to-adult volunteering. But across models, a clear pattern persists: differences among educational settings in adolescence predict levels of adult volunteering. Despite a robust set of controls and the inclusion of variables that we anticipated would largely account for sector differences in school effects, the differences continue to exist. What do the data allow us to say and not say about this pattern?

First, these results cannot be explained away by arguing that existing family and religious differences in the propensity to volunteer account for apparent sector differences. It is true that parental education, income, race/ethnicity, and family religiosity partly explain some of the differences between schooling types, but substantial school differences remained. Not only this, but we introduced measures of parental modeling and support of volunteering. We would be hard pressed to come up with better measures of existing household differences in propensity to volunteer. Thus, H1—that schooling effects are actually spurious—is only partly true. We can conclude that schooling itself has a real impact on sustaining volunteering.

We posited two possible general categories of social mechanisms that might help explain school-based differences: opportunities and motivation. For opportunities, we expected some types of schools to provide more direct opportunities (or requirements) to volunteer than others, adolescents from some schools to form peer networks that supported volunteer activity, and certain schools to filter young people into different types of institutions and life course trajectories as they age that provided more opportunities to volunteer. We measured some aspects of each of these expectations, and while several had independent impacts on sustaining volunteering, they did not help account for secondary schooling influences. Moreover, our most direct measures of opportunity—the number of organized activities the respondent is involved in and the percentage of the respondent's five closest friends who volunteer—suppressed the influences of school type. This means that these school effects exist despite of differences in opportunities to volunteer and not because of them.

Our account of motivation fared no better. Measures of intrinsic motivation and a sense of empowerment had very little impact on sustaining volunteering and did not account for any of the secondary schooling effects. If the decline in volunteering among emerging adults who were home schooled or who attended nonreligious private secondary schooling is due to a narrowing of the circle of social concern, or a constrained sense of self-efficacy, then this is not captured at all by the indicators we have included in our models. Likewise, the lack of decline in volunteering for Protestant schooled emerging adults cannot be attributed to a widening of this circle of social concern. However, we cannot rule out the potential role of extrinsic motivations in secondary schooling influences (we simply had no adequate measure of this factor). It is possible that students who attend nonreligious private schools or are home schooled volunteer only when extrinsic social rewards are at stake. For example, the importance of volunteer work for college admission may motivate students who attend private nonreligious schooling because of the academic pressures in such an environment. Once college admission is no longer a motivation, volunteering can be expected to decline. We have some evidence of a similar phenomenon when volunteering is required for adolescents. Consistent with past research (Stukas, Snyder, and Clary 1999; Wuthnow 1995), required volunteering is negatively associated with subsequent voluntary volunteering. Therefore, we think it is possible that extrinsic rewards (e.g., grades or college admission) may play a role in motivating volunteering—although we doubt this would be able to explain the substantial schooling differences we have found.

This leads us to conclude that some of the most plausible explanations for the apparent influence of school type turn out to be false. The relationship of school type to volunteering is not spurious, and the mechanisms that shape the relationship are not indirect, that is, the mechanisms are not the social networks, institutional opportunities, psychological motivations, or life course trajectories generated by school type. We take this to be our most important—and surprising—set of findings.

But if schooling type is not spurious and does not relate indirectly to adult volunteering through opportunities and motivation, what accounts for the relationship of schooling type to adult volunteering? We suggest what we believe are two plausible alternative accounts (which are unfortunately difficult to measure with standard survey indicators). It is important to underscore that what we are about to present is necessarily tentative, and further research is necessary to assess whether these accounts help make sense of the impact of schooling type on volunteering.

One mechanism could be that schools provide social scripts that respondents adopt at the crucial developmental stage of adolescence and follow unreflexively in adulthood. This interpretation would help explain why motivation and empowerment, both of which assume that respondents are reflective and intentional, do not show up as significant variables in the final models. It would also be consistent with work on civic engagement that suggests that “value internalization” leads adolescents to carry a sense of civic duty into adulthood (Bekkers 2007). The theoretical signpost points to practice theorists, who understand much of social action as “embodied” within certain social contexts and habitual routine (Bourdieu 1990; Gross 2009; Swidler 2001).

A different (though not mutually exclusive) mechanism focuses less on habituated action and more on the ways organizations bridge adolescence and young adulthood. The idea is that certain school types connect adolescents to voluntary organizations that are more effective at involving respondents from their teen years into adulthood. One can imagine, for example, that some evangelical parachurch groups that have a presence in both secondary school and higher education might maintain respondent engagement from adolescence into young adulthood. If those kinds of groups are more prominent in Protestant schools compared to the others, then they help explain the odds ratios of Protestant schools in our models. This explanation would assume that the setting (in contrast to frequency) of volunteering already at Wave 1 varies systematically with school type. Unfortunately, the NSYR does not provide questions that allow researchers to explore this possible mechanism fully.

The NSYR does, however, contain a survey item at Wave 3 that asks only those respondents who volunteered to report how much, if any, of the volunteering was organized by a religious organization or congregation (we use a version of this variable from Wave 1 in our models). If we look at this variable among emerging adults who volunteered at Wave 1 and are still volunteering at Wave 3, we see some differences by schooling type. For respondents with public school backgrounds who sustain their volunteering, about half report that at least some of their volunteering is religiously organized. A little more than half of those who attended Catholic schools report the same. However, more than three-fourths of respondents with Protestant school backgrounds report that some of their volunteering is religiously organized. We should note that very few in any group report that all of their volunteering is religiously organized (around one-fifth in the Protestant and Catholic school group, and around one-seventh of the public school group). This also brings up the question of whether the type of volunteering by those with Protestant school backgrounds is benefiting the broader community or more narrowly serving specific religious groups. We simply note here that we cannot draw any firm conclusions about the type of volunteering using the NSYR data.

Conclusion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References

In empirical terms, this study tells us more about what we can reject than what we can affirm. But in rejecting several hypotheses, we can say affirmatively that schooling type is an independent predictor of whether adolescent volunteers will become adult volunteers. By itself, that conclusion is intriguing and important for those who are invested in the civic engagement of young people.

The persistent result that religious identity of schools helps to explain volunteering is an invitation to further inquiry. We see the most fruitful approaches going in two different directions. First, we believe social scientists who study volunteering should attempt to strengthen our scholarly understanding about the mix of conscious motivations and habituated social practices that are associated with different types of schooling. Perhaps some schools are producing young people who nonreflexively adopt social scripts that do or do not involve volunteering. This would not be easily picked up by standard survey measures, yet it has the potential to be an important mechanism of influence. Second, future research should distinguish among types of volunteering and types of organizations that support volunteering among adolescents and young adults. It is quite plausible that the sort of volunteering, and the type of organizations that partner with schools to promote volunteering, are quite different at Protestant schools than at public schools (or Catholic schools for that matter). Understanding these distinctions may be the key to making sense of the sizable school effects found in this study.

  1. 1

    As we note later in the article, distinguishing volunteering as civic engagement from other types of volunteering presents both a measurement challenge for this study and an opportunity for future analysis.

  2. 2

    The four Likert items were the following: (1) you have little control over the things that happen to you; (2) you often feel helpless in dealing with the problems of life; (3) there is really no way you can solve some of the problems you have; and (4) there is little you can do to change many of the important things in your life.

  3. 3

    An alternative approach would be to include the entire sample of respondents at Wave 3 and include volunteering and giving at Wave 1 as a predictor of engagement at Wave 3. We rejected the design because the resulting analysis would conflate two separate research questions, namely, the effects of schooling on sustaining engagement and the effects of schooling on generating engagement. The underlying reason is that schooling interacts with volunteering at Wave 1; accordingly, coefficients for a model including all Wave 3 respondents would reflect both the weighted effect of schooling type on those who volunteered at Wave 1 and the weighted effect for nonvolunteers at Wave 1. This latter effect is not pertinent to our research question, which is the effect of schooling type on adult engagement net of the impact of volunteering in adolescence.

  4. 4

    Given our use of a nonrandom restricted sample (including only those who volunteered as adolescents) we considered whether the use of a two-stage Heckman correction model would be appropriate. Ultimately, we decided against such a model because our coefficients for school type provide an unbiased estimate (assuming adequate controls for spuriousness) of the influence of school type on continuing to volunteer or not over time for the population that volunteer as adolescents. A two-stage model would be appropriate if we were using a restricted model and wanted to provide unbiased estimates for the entire adolescent population. However, given our research question about sustaining volunteering, our results should only apply to a restricted population.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Civic Engagement, Social Networks, and Religious Salience
  5. Secondary Schooling and Adolescent-to-Adult Engagement
  6. Data and Methods
  7. Modeling Strategy
  8. Results
  9. Discussion
  10. Conclusion
  11. References
  • Anderson, Mary R. 2010. Community psychology, political efficacy, and trust. Political Psychology 31(1):5984.
  • Andolina, Molly W., Krista Jenkins, Cliff Zukin, and Scott Keeter. 2003. Habits from the home, lessons from school: Influences on youth civic engagement. PS: Political Science and Politics 36(2):27580.
  • Becker, Penny Edgell and Pawan H. Dhingra. 2001. Religious involvement and volunteering: Implications for civil society. Sociology of Religion 62(3):31535.
  • Bekkers, René. 2007. Intergenerational transmission of volunteering. Acta Sociologica 50(2):99114.
  • Bourdieu, Pierre. 1990. Practical reason: On the theory of action. Stanford, CA: Stanford University Press.
  • Brady, Henry E., Kay Lehman Schlozman, and Sidney Verba. 1999. Prospecting for participants: Rational expectations and the recruitment of political activists. American Political Science Review 93(1):15368.
  • Bryk, Anthony S., Valerie E. Lee, and Peter B. Holland. 1993. Catholic schools and the common good. Cambridge, MA: Harvard University Press.
  • Campbell, David E. 2001. Making democratic education work. In Charters, vouchers, and public education, edited by Paul E. Peterson and David E. Campbell, pp. 24167. Washington, DC: Brookings Institution.
  • Campbell, David E. 2006. Why we vote: How schools and communites shape our civic life. Princeton, NJ: Princeton University Press.
  • Chaves, Mark. 2004. Congregations in America. Cambridge, MA: Harvard University Press.
  • Coleman, James S. and Thomas Hoffer. 1987. Public and private high schools: The impact of communities. New York: Basic Books.
  • Crystal, David S. and Matthew DeBell. 2002. Sources of civic orientation among American youth: Trust, religious valuation, and attributions of responsibility. Political Psychology 23(1):11332.
  • Dalton, Russell J. 2009. The good citizen: How a younger generation is reshaping American politics, rev. ed. Washington, DC: CQ Press.
  • Dee, Thomas S. 2005. The effects of Catholic schooling on civic participation. International Tax and Public Finance 12(5):60525.
  • Dudley, Robert L. and Alan R. Gitelson. 2002. Political literacy, civic education, and civic engagement: A return to political socialization? Applied Developmental Science 6(4):17582.
  • Dwyer, James G. 1998. Religious schools v. children's rights. Ithaca, NY: Cornell University Press.
  • Flanagan, Constance. 2003. Developmental roots of political engagement. PS: Political Science and Politics 36(2):257261.
  • Gibson, Troy. 2008. Religion and civic engagement among America's youth. Social Science Journal 45(3):50414.
  • Greene, Jay P., Joseph Giammo, and Nicole Mellow. 1999. The effect of private education on political participation, social capital, and tolerance: An examination of the Latino national political survey. Georgetown Public Policy Review 5(1):5367.
  • Greene, Jay P., Nicole Mellow, and Joseph Giammo. 1999. Private schools and the public good: The effect of private education on political participation and tolerance in the Texas poll. Catholic Education 2(4):42943.
  • Grimm, Robert, Nathan Dietz, John Foster-Bey, David Reingold, and Rebecca Nesbit. 2006. Volunteer growth in America: A review of trends since 1974. Washington, DC: Corporation for National and Community Service.
  • Gross, Neil. 2009. A pragmatist theory of social mechanisms. American Sociological Review 74(3):35879.
  • Gutmann, Amy. 1999. Democratic education. Princeton, NJ: Princeton University Press.
  • Hart, Daniel, Thomas M. Donnelly, James Youniss, and Robert Atkins. 2007. High school community service as a predictor of adult voting and volunteering. American Educational Research Journal 44(1):197219.
  • Hart, Daniel and Suzanne Fegley. 1995. Prosocial behavior and caring in adolescence: Relations to self-understanding and social judgment. Child Development 66(5):134659.
  • Janoski, Thomas, Marc Musick, and John Wilson. 1998. Being volunteered? The impact of social participation and pro-social attitudes on volunteering. Sociological Forum 13(3):495519.
  • Johnson, Monica K., Timothy Beebe, Jeylan T. Mortimer, and Mark Snyder. 1998. Volunteerism in adolescence: A process perspective. Journal of Research on Adolescence 8(3):30932.
  • Kahne, Joseph, Bernadette Chi, and Ellen Middaugh. 2006. Building social capital for civic and political engagement: The potential of high-school civics courses. Canadian Journal of Education 29(2):387409.
  • Kahne, Joseph and Joel Westheimer. 2006. The limits of political efficacy: Educating citizens for a democratic society. PS: Political Science and Politics 39(2):28996.
  • Loveland, Matthew T., David Sikkink, Daniel J. Myers, and Benjamin Radcliff. 2005. Private prayer and civic involvement. Journal for the Scientific Student of Religion 44(1):114.
  • Macedo, Stephen. 2005. Democracy at risk: How political choices undermine citizen participation, and what we can do about it. Washington, DC: Brookings Institution.
  • McLellan, Jeffrey and James Youniss. 2003. Two systems of youth service: Determinants of voluntary and required youth community service. Journal of Youth and Adolescence 32(1):4758.
  • Musick, Marc A. and John Wilson. 2008. Volunteers: A social profile. Bloomington: Indiana University Press.
  • National Center for Education Statistics. 2011. The condition of education: 2011. Washington, DC: United States Department of Education.
  • Penner, Louis A., and Marcia A. Finkelstein. 1998. Dispositional and structural determinants of volunteerism. Journal of Personality and Social Psychology 74(2):52537.
  • Pennings, Ray, John Seel, Deani A. Neven Van Pelt, David Sikkink, and Kathryn L. Wiens. 2011. Do the motivations for private religious Catholic and Protestant schooling in North America align with graduate outcomes? Hamilton, ON: Cardus.
  • Pollock, Phillip H. 1983. The participatory consequences of internal and external efficacy. Western Political Quarterly 36(3):40009.
  • Putnam, Robert D. 2000. Bowling alone: The collapse and revival of American community. New York: Simon and Schuster.
  • Putnam, Robert D. and David E. Campbell. 2010. American grace: How religion divides and unites us. New York: Simon and Schuster.
  • Rose, Susan D. 1988. Keeping them out of the hands of Satan: Evangelical schooling in America. New York: Routledge.
  • Rudolph, Thomas J., Amy Gangl, and Dan Stevens. 2000. Effects of efficacy and emotions on campaign involvement. Journal of Politics 62(4):118997.
  • Skocpol, Theda. 2004. Diminished democracy: From membership to management in American civic life. Norman: University of Oklahoma Press.
  • Smidt, Corwin, Kevin R. den Dulk, James M. Penning, Stephen V. Monsma, and Douglas L. Koopman. 2008. Pews, prayers, and participation: Religion and civic responsibility in America. Washington, DC: Georgetown University Press.
  • Smith, Christian and Melinda Lundquist Denton. 2005. Soul searching: The religious and spiritual lives of American teenagers. New York: Oxford University Press.
  • Smith, Christian and Robert Faris. 1999. Correlation found between religiosity and community service. Available at http://www.youthandreligion.org/news/6-4-2002.html, accessed January 11, 2013.
  • Spring, Kimberly, Nathan Dietz, and Robert Grimm. 2006. Educating for active citizenship: Service-learning, school-based service and youth civic engagement. Washington, DC: Corporation for National and Community Service.
  • Stukas, Arthur A., Mark Snyder, and E. Gil Clary. 1999. The effects of “mandatory volunteerism” on intentions to volunteer. Psychological Science 10(1):5964.
    Direct Link:
  • Sundeen, Richard A. and Sally A. Raskoff. 1994. Volunteering among teenagers in the United States. Nonprofit and Voluntary Sector Quarterly 23(4):383403.
  • Sundeen, Richard A. and Sally A. Raskoff. 2000. Ports of entry and obstacles: Teenagers’ access to volunteer activities. Nonprofit Management and Leadership 11(2):17997.
    Direct Link:
  • Swidler, Ann. 2001. Talk of love: How culture matters. Chicago: University of Chicago Press.
  • Toppe, Christopher M., Arthur D. Kirsch, and Jocabel Michel. 2002. Giving and volunteering in the United States, 2001: Findings from a national survey. Washington, DC: Independent Sector.
  • Wolf, Patrick J. 2007. Civics exam: Schools of choice boost civic values. Education Next 7(3):6672.
  • Wuthnow, Robert. 1995. Learning to care: Elementary kindness in an age of indifference. New York: Oxford University Press.
  • Zukin, Cliff, Scott Keeter, Molly Andolina, and Krista Jenkins. 2006. A new engagement? Political participation, civic life, and the changing American citizen. New York: Oxford University Press.