Home | Research & Policy Library | Bringing Thou Back In

Bringing Thou Back In

Religious Schools in the United States and Pro-Social Orientations

March 5, 2018

David Sikkink


Research Report

This report examines the correlation between religious school attendance, pro-social attitudes, and civic involvement. Are religious school graduates more likely to vote, attend a community meeting, or engage in some other action that contributes to the common good? Author David Sikkink at the University of Notre Dame examines these issues through an analysis of a major American longitudinal data set.

Executive Summary

Moral and civic formation of students is one of the most important aspects of a religious school’s mission. Parents’ decisions to send their children to religious schools reflect a concern not only for academic success but also for moral socialization and religious education (Bulman 2004). The question is whether religious schools can fulfill this mission in a way that influences students into adulthood.

Limited evidence shows some longer-term positive effects of religious schools on civic commitments (Dee 2005; Dill 2009), but most studies are limited to analysis of the years during high school or are burdened by the high data demands for investigating private-school sectors in detail. Thus far, the Cardus Education Study has revealed important associations between attending a religious high school and various forms of generosity and civic engagement in young adulthood (Pennings et al. 2011; Pennings et al. 2014; Van Pelt et al. 2012). Religious-school attendance, for example, is associated with greater involvement in organized volunteering, and especially in charitable giving. What is less well-explored is the issue of informal helping and what many would call “pro-social” orientations, which include concern for the good of persons in need. Does religious schooling have an impact on these outcomes? We would expect that the religious, moral, and character emphases within religious schools (Vryhof 2004) would touch on issues of obligations to the disadvantaged. Religious language includes concern for the weak, the stranger, and the outcast—in biblical language, the orphan, widow, and alien. Practices within the school, including obligations to classmates who are considered “outsiders,” providing for the needs of families in the school who experience tragedy and hardship, and facilitating community service to the poor or elderly in the local community, may further instill student concern for those in need. Religious schooling may provide a language and instill habits that reflect the importance of making personal sacrifices in everyday life for the disadvantaged as well as being oriented to the good of others in personal interactions. This may extend to a sense of obligation to the collective good of communities beyond individual self-interest. Being willing to ignore the free-rider dilemma—in which self-interest calls us to let others do the work of building collective goods—and take self-sacrificial action for the civic good may result from a religious-school education. Religious schools may not only explicitly teach concern for others and integrate moral and civic formation into the activities of the school, but also generate a community of trust and social cohesion in which helping others is modelled and appreciated. Socialized within a caring community, the religious-school student may be more likely to carry an orientation of self-sacrifice for the good of others into their adult lives.

This report takes up the question of whether religious-school attendance is associated with pro-social attitudes and informal helping behaviours as well as the types of civic involvement, such as voting or attending a community meeting, that require a commitment to the collective good over self-interest. We do this through an analysis of survey data from a major longitudinal data set in which the influence of school sector and pro-social attitudes and behaviours is relatively unexplored.

A Community Focus on Others?

Religious schools have a strong incentive to spend time and resources on learning to care, since combining academic goals with moral and religious formation of students defines the religious-school-market niche in the United States. In religious schools, we expect several strengths in moral formation, including those that forge attitudes and practices that include sacrificial commitments to others. For many religious schools, commitment to others less fortunate is considered an important opportunity for religious and spiritual formation, which has a central place in the school’s mission. Daily expressions of selflessness within the school are often expected and encouraged, even if in small ways, such as reaching out to a student marginal to the school community. This would extend to expressions of care and concern, often formally organized by the school, for families that experience some kind of hardship or tragedy. Religious schools tend to facilitate other opportunities for outreach to the disadvantaged, including service learning courses and school community-service days, in which student identity is directed toward the good of others rather than the self alone. In addition, religious schools tend to develop partnership in moral formation with parents, which likely improves the extent that religious admonishment to care for others is reinforced for students (Damon 1996). The importance and value of pro-social orientations is embodied in religious understandings of the teacher role, in which the financial and career sacrifices of teachers in working at an underfunded religious school become a model for students of a meaningful life. Last, religious schools may create the kind of moral communities in which pro-social orientations make sense and are understood because they are rooted in a religious language and narrative (Hunter 2000).

Not much evidence exists of higher pro-social orientations among religious-school graduates. While in religious schools, students show evidence of other-directedness (Hunter 2000). The communal organization of many religious schools, in which the flourishing of persons and commitment to the common good of the school community is emphasized (Bryk, Lee, and Holland 1993), at least provides a context in which pro-social orientations could grow. There is some evidence that civic orientations are quite strong among religious-school students (Campbell 2001; Godwin, Ausbrooks, and Martinez 2001; Greene 1998; Wolf et al. 2001). Student behaviour problems tend to be limited within religious schools.

(Jeynes 2012), and this may contribute to an effective social setting in which to develop pro-social orientations. The parent-school relationship within religious schools may contribute to moral formation as well. Parental involvement in religious schools tends to be very high (Noel, Stark, and Redford 2015), and schools and families are more likely to expect to work together for the moral formation of students (Bryk, Lee, and Holland 1993). The structure of relationships between family and school, and among school parents (Bryk, Lee, and Holland 1993; Coleman and Hoffer 1987), should contribute to a positive environment for developing pro-social orientations.

Data and Measures

The survey data we consider in this report is from a large-scale, government-funded panel survey of American families with school-aged children, the National Longitudinal Study of Youth 1997 (NLSY97). This survey is funded through the United States Bureau of Labor Statistics, and is widely regarded by education and family researchers as one of the top two or three US studies in terms of the breadth and quality of panel data. This study includes a random sample of about 8,400 children in 1997, most of whom were in the sixth to eleventh grade. These students have been re-interviewed at least every other year since that time.

The analysis presented here uses data from survey questions related to pro-social orientations and civic engagement. The NLSY includes several measures on attitudes toward helping those in need. The following NLSY questions provide the data for the analysis of pro-social attitudes:

These days people need to look after themselves and not overly worry about others.

Personally assisting people in trouble is very important to me.

Those in need have to learn to take care of themselves and not depend on others.

People should be willing to help others who are less fortunate.

The measures capture an individualistic orientation to helping, an inward-looking focus that sees no collective responsibility for helping, as well as a generous orientation that calls for reaching beyond self-interest to consider the interests of those in need.

An additional set of questions administered in 2007 probes commitment to collective goods within the local community. The questions are phrased in terms of the importance of participating in society.

Here are some ways that we can participate in our society. Please tell me if you think it is very important to do these things, somewhat important to do these things, or not at all important to do these things.

Serve on a jury if called?

Report a crime you may have witnessed?

Keep fully informed about news and public issues?

Vote in elections?

Of course, these four questions are not tapping actual behaviours but the respondent’s sense of the importance of serving the common good. The context of the question likely focuses the respondent on the importance of going out of their way for the benefit of the community.

NLSY97 provides measures of informal helping behaviours as well. These included questions on blood donation, giving money to a homeless person, and allowing someone to cut in line. The following questions about the respondent’s activities in the last twelve months provided data on informal helping for this report:

Let someone you didn’t know well borrow an item of value like dishes or tools.

Allowed a stranger to go ahead of you in line.

Returned money to a cashier after getting too much change.

Given food or money to a homeless person.

Donated blood.

Though frequency on each of these does depend on opportunities, a lack of opportunities could indicate a lifestyle that looks inward rather than easily crosses the boundary into others’ worlds.

NSYR97 includes numerous variables designed to capture important aspects of personality. Of course, personality traits may be an important factor in parents’ decisions to place a child in a particular school sector. Selection factors on these variables would need to be kept in mind. But some of these variables may provide a window on moral formation in a religious-school environment. For example, the question of whether the respondent tends to break school rules may reflect in part socialization within a more disciplined religious-school environment. And the experience of a more disciplined schooling environment may have a longer-term impact on contributions to civic life. Thus the analysis presented below includes as dependent variables all of the personality variables available in NSYR97. The wording of some of these variables is shown below.

I support long-established rules and traditions.

I do not intend to follow every little rule that others make up.

I do what is required, but rarely anything more.

I do not work as hard as the majority of people around me.

Even if I knew how to get around the rules without breaking them, I would not do it.

I have high standards and work toward them.

[Would you say you are]

. . . Extraverted, enthusiastic

. . . Critical, quarrelsome

. . . Open, complex

. . . Calm, emotionally stable

. . . Reserved, quiet

Although these measures are primarily intended to capture personality orientations, some of these traits and outlooks may be reinforced in religious schools. In turn, orientations to rules and openness to others may facilitate pro-social actions in relation to neighbours and the local community.

The school-sector information for each respondent in NLSY97 is quite extensive. School sectors available include public technical school, traditional public school (TPS), magnet school, charter school, and alternative schools. Each of these is included in the models below, with TPS serving as the comparison group. The private sectors include private non-religious, Catholic, other religious, and homeschool. Unfortunately, the publicly available NLSY97 does not provide a further breakdown of the non-Catholic religious schoolers. A large majority of respondents in the non-Catholic religious sectors would be in the evangelical Protestant, or “Christian school,” sector, but there would also be Jewish schools and a few mainline Protestant schools represented. As a check on this, a separate variable, “other non-Catholic religious school,” was created for non-Catholic religious-school respondents whose responding parent reported in 1997 low levels of religiosity or religious service attendance. This assumes that most evangelical Protestant schools at the turn of the century were attracting predominantly evangelical Protestant parents, who tend to report high levels of religiosity. In addition, non-Catholic religious-school respondents who had a responding parent who reported being Jewish, Muslim, or any of the Eastern religions, such as Buddhist, were moved to the separate “other” non-Catholic religious-school category. If the responding parent reported a denomination that has a religious-school tradition, including Episcopalian and Disciples of Christ, the non-Catholic religious-school respondent was moved into the “other” non-Catholic religious-school category as well. Altogether, about 50 of the approximately 250 cases reporting a non-Catholic religious school were placed in this category. Similarly, homeschool respondents were split into two categories, one with respondents whose responding parent reported in 1997 low religiosity or religious service attendance.

A further complication in assigning sectors is that school-sector information is available for multiple years of the NLSY97. For this analysis, separate variables were created for each sector, and a respondent was coded as “1” for that sector if the respondent was schooled in that sector for at least one year. Some respondents were in different sectors during their school career, and that would be reflected in the variables included in the models. Each school-sector estimate is net of the other variables in the model, including experience in other school sectors.

Catholic School Attendees

The findings for Catholic schoolers provide some evidence of generous attitudes and actions, but are not consistent across all the dependent variables.

Pro-social orientations are one example. On the positive side, Catholic schoolers are significantly more likely than public schoolers to say that it is important to take care of people who are in trouble. This may reflect experiences in a Catholic-school community that attempts to “rally around” families in the school who fall on hard times or experience tragedy. On the other hand, we find no difference in the extent that Catholic schoolers think we should help the less fortunate. Nor are there public–Catholic school differences on questions of whether people should take care of their own selves and take care of their own problems. Individualistic views of responsibility for self and others appear to coexist with some sense of obligation to others in need.

On orientations to community responsibilities, Catholic-school differences are muted. The analysis finds no difference in the extent that Catholic schoolers sense an obligation to vote, to serve on a jury, to report a crime, or to be informed on public issues. But there are differences in actual levels of community involvement. Frequency of attending a community meeting is higher and statistically significant for Catholic schoolers. Catholic schoolers are consistently strong on voting across all survey years. Overall, it appears there is a positive effect of Catholic schooling on community responsibilities, though there is no positive or negative effect of Catholic schoolers on a sense of obligations to the community.

Interestingly, Catholic schoolers are less likely to report that they gave blood in the last twelve months. Yet they are more likely to vote, and this effect is very strong across survey years, with the exception of 2006. Overall, evidence is mixed on whether Catholic schooling works to overcome the tendency to let others create collective goods.

Finally, Catholic schoolers are somewhat distinctive on personality orientations that may affect pro-social behaviour, or at least provide a foundation for outreach to others. For example, Catholic schoolers report in comparison to public schoolers that they are both extroverted and enthusiastic and not quiet and reserved. In addition, in 2008 and 2010, they are less likely to report that they broke school rules. In those survey years, Catholic schoolers are also less likely to report that they only do what is required but rarely anything more. On a similar question, whether the respondent does not work hard compared to others, Catholic-school respondents are less likely to agree compared with public-school respondents. Thus it appears that Catholic schoolers are developing moral orientations that would prepare them to make positive contributions to the common good.

We note, however, that while we include numerous controls in the models to isolate the school effect, Catholic-school selection effects for personality characteristics cannot be ruled out entirely.

Protestant Schoolers

Evangelical Protestant schoolers have been shown in other research to be very active in their church and to be very generous in charitable giving (Pennings et al. 2014). While we would like to test for this sector these and other pro-social behaviours with the NLSY97 data, we are limited by a category that captures the non-Catholic religious schoolers who were raised in religious homes and who had parents who reported a religious tradition of Protestant or Catholic. We cannot know for sure how close our decision rule comes to the category of evangelical Protestant-school respondents, but we can compare the two sectors, non-Catholic religious schoolers whose parents are largely Protestant and involved in religious practices, and non-Catholic religious schoolers who are not. That may give some confidence that the Protestant-school effect—despite some measurement error—is a reasonable approximation of the “Christian school” category. First we discuss the findings for the “Protestant” schoolers.

The findings are quite consistent in pointing to pro-social orientations among Protestant schoolers. The impact of Protestant-schooling experience on the sense of obligation to help the less fortunate is positive and significant. Protestant schooling has a very strong and positive effect on the view that it is important to help people in trouble. On the other pro-social measures that capture individualistic orientations, such as whether people need to look out for themselves, there are no significant differences between Protestant and public schoolers. That is somewhat surprising given the emphasis on individualism in the evangelical tradition (Emerson and Smith 2000).

Interestingly, given the sense of responsibility to others shown in earlier research on evangelical Protestant schoolers (Pennings et al. 2011), the NLSY97 confirms that Protestant schoolers have a higher sense of obligation to participate in the community. For example, they are more willing than public schoolers to say that they have an obligation to vote and an obligation to accept jury duty. Protestant schoolers in this sample are not, however, statistically different on obligations to report a crime or be an informed citizen.

Regarding generous practices, the NSLY data presents a mixed picture for Protestant schoolers. They are not significantly more or less likely to give blood. Most of the effects on informal helping are strongly positive but not statistically significant. A small negative but not significant estimate is found on the measure of loaning something valuable to another. On the other hand, the Protestant-school effect on actual voting is very strong and positive, though that only applies to voting in 2004 and 2006. Voting propensity is positive in 2008, but not statistically significant. No Protestant-school effect is found for attending a community meeting, though we should keep in mind that the question mentions political and environmental meetings. Protestant schoolers have shown some reticence to be involved in these types of causes and organizations (Pennings et al. 2011).

A few of the questions on personality indicate that evangelical respondents may have dispositions useful for pro-social behaviour and civic involvement. In particular, Protestant schoolers are more likely to disagree that they are “critical, quarrelsome.” Like Catholic schoolers, Protestant schoolers are more likely to see themselves as willing to work above and beyond what is required.

"Other" Non-Catholic Religious Schoolers

This category of schoolers is admittedly a hodgepodge, including many mainline Protestant and Jewish schoolers. Thus we can’t make too much of the findings. The results for the “sector” do provide some contrasts with the Protestant schoolers that may be of interest.

Overall, there are few significant effects for this group, including for the findings on orientations toward helping others. The results are not statistically significant regarding attending a community meeting, nor are there any significant effects on the community-obligation questions. Keep in mind, however, that the small sample size makes it very difficult to find statistically significant effects.

Regarding informal helping, the results show that this group of respondents is less likely to allow a stranger to cut in line. Of course, this could reflect a lack of opportunities, but controls for geographic context should help alleviate this concern. This group of respondents are also more likely to loan something valuable to someone.

The group is on average higher on a measure of a reserved personality, is less likely to say they have broken school rules, and is less likely to say that they have bent the rules.

Non-Religious Private Schoolers

The results for non-religious private schoolers tend to show lower levels of pro-social behaviour and civic commitments and action.

To begin with, the private non-religious schoolers are not significantly different from public schoolers on pro-social attitudes. The only statistically significant difference is on the measure of whether people should be responsible for taking care of themselves. On that score, private non-religious schoolers, taking up an individualistic stance, are in strong agreement. Community obligations are not strongly held either, including voting. Compared to public schoolers, private non-religious schoolers do not think that they are responsible for taking on jury duty or reporting a crime. They also are not different from public schoolers on the likelihood of voting. Private non-religious schoolers are negatively related to giving blood, though this effect is not significant.

Private non-religious schoolers are not significantly different from public schoolers on informal helping behaviour. The estimates show a small positive effect on giving to the homeless and loaning a valuable item, but these effects are not significant, and the other informal giving estimates are negative and non-significant. The frequency of attending community meetings is not related to non-religious private schooling.

On the personality measures, private non-religious schoolers tend to report being extroverted and open and complex, but disorganized and careless. They are also more likely than public schoolers to report that they tended to break the rules in school.

Overall, there is little evidence that non-religious private schooling has a positive effect on social or civic orientations that would contribute to a vital public life. The older notion of noblesse oblige in private non-religious schools seems to have disappeared.

Religious Homeschoolers

The religious-homeschooling sector is not strongly or consistently related to measures of pro-social and civic commitments and practices.

Religious homeschoolers are not different from public schoolers on measures of pro-social attitudes. They are not more likely to support community obligations. The only significant difference with public schoolers is a negative relation between religious homeschoolers and obligation to vote.

On informal helping behaviour, results do not show religious homeschooler differences.

Community-meeting participation is not related to religious homeschooling. Religious homeschoolers are strongly negatively related to voting, and this is consistent across nearly all elections. The relation of this sector to giving blood is negative but not significant.

Interestingly, religious homeschoolers in the 2008 survey are negatively related to breaking rules in school, and they are positive on a measure of whether they support long-established rules and traditions.

Overall, we notice the countercultural orientations of homeschooling, especially as expressed in much lower support for voting. The other outcomes are not significantly different from public schoolers. The lack of significance is no doubt partly due to the lack of strong distinctive orientations and practices among religious homeschoolers, but note that the relatively small sample size makes it more difficult to detect significant differences.

Non-Religious Homeschoolers

Similarly, for the few non-religious homeschoolers the results do not show many differences with public schoolers. Pro-social attitudes are not significantly different. Non-religious homeschoolers exhibit strong support for the claim that people need to take care of themselves, but this is not statistically significant. On informal helping behaviours, there appears to be a positive effect on giving to the homeless and allowing strangers to cut in line, but a negative relationship to loaning a valuable item to another person. But these are not statistically significant. The influence on blood donation is positive, but not significant either. The findings for voting tend to be negative, but not consistently significant.

The non-religious homeschoolers report that they are less calm and emotionally stable. Non-religious homeschoolers tend to agree that they will not be rule followers. They are also agree that they tended to flaunt school rules, but this is not statistically significant.

While some of the findings for non-religious homeschoolers seem consistent with what we would expect, there are few statistically significant findings. This reflects in part the very small sample size available in the NLSY97. Our conclusions about this sector, therefore, remain very tentative.


Overall, we find evidence of differences between religious-school attendees and public-school attendees on a number of measures of generous attitudes and practices toward others. Although Catholic schoolers are not higher on informal helping, they are more likely to attend community meetings and vote. The Catholic-school emphasis on service to the disadvantaged is evident in the positive relation of Catholic schoolers with personally helping those in need. Catholic schoolers are not immune to the view that people need to pull themselves up by their own bootstraps, yet the experience of Catholic schooling, whether in teaching or practices, seems to generate a willingness to sacrifice personal interests for participation in creating collective goods.

Protestant schoolers are similar. They are supportive of pro-social orientations that include an obligation to help the disadvantaged, and the findings on this score are more crisp than for the Catholic schoolers. They have a strong sense of community obligations, whether that is voting or jury duty. And there is evidence, even if somewhat inconsistent, that in practice Protestant schoolers are involved in meeting community needs through voting.

Non-religious private schoolers do not provide any evidence of generous attitudes and actions toward the disadvantaged. The individualistic competitive culture of these schools perhaps explains these outcomes. A sense of obligation to civic life among these often elite schools seems to have atrophied. Without a moral community rooted even weakly in religion, moral and civic formation in private non-religious schools is not strong—at least according to the evidence available in the NSYR97 sample.

Homeschoolers present a mixed picture. Overall, however, community obligations do not appear to be high on their list of priorities. The countercultural orientations of the homeschooling movement, while laudable in many ways, do not seem to have any distinctive effect on whether homeschoolers are active in civic life. We should keep in mind, however, that the homeschoolers are not significantly different from public schoolers on any measure but voting, and that voting in a majoritarian electoral system perhaps gives way to direct action in petitioning officials and protesting policies for an educational minority. Homeschoolers may be active in community and political life in ways not measured in this report.

Bryk, Anthony S., Valerie E. Lee, and Peter Blakeley Holland.  1993. Catholic Schools And The Common Good. Cambridge, MA: Harvard University Press.

Bulman, Robert C. 2004. “School-Choice Stories: The Role of Culture.” Sociological Inquiry 74, no.4:492–519.

Campbell, David E. 2001. “Bowling Together.” Education Next 1, no. 3:55–61.

Coleman, James S., and Thomas Hoffer. 1987. Public and Private High Schools: The Impact of Communities. New York: Basic Books.

Damon, William. 1996. Greater Expectations: Overcoming the Culture of Indulgence in Our Homes and Schools. New York: Free Press.

Dee, Thomas. 2005. “The Effects of Catholic Schooling on Civic Participation.” International Tax and Public Finance 12, no. 5:605–25.

Dill, Jeffrey S. 2009. “Preparing for Public Life: School Sector and the Educational Context of Lasting Citizen Formation.” Social Forces 87, no. 3:1265–90.

Emerson, Michael O., and Christian Smith. 2000. Divided by Faith: Evangelical Religion and the Problem of Race in America. Oxford: Oxford University Press.

Godwin, Kenneth, Carrie Ausbrooks, and Valerie Martinez. 2001. “Teaching Tolerance in Public and Private Schools.” Phi Delta Kappan 82, no. 7:542–46.

Greene, Jay. 1998. “Civic Values in Public and Private Schools.” In Learning from School Choice, edited by Paul E. Peterson and Bryan C. Hassel, 83–106. Washington, DC: Brookings Institution.

Hunter, James Davison. 2000. The Death of Character: Moral Education In An Age Without Good or Evil. New York: Basic Books.

Jeynes, William H. 2012. “A Meta-analysis on the Effects and Contributions of Public, Public Charter, and Religious Schools on Student Outcomes.” Peabody Journal of Education 87, no. 3:305–35.

Noel, Amber, Patrick Stark, and Jeremy Redford. 2015. “Parent and Family Involvement in Education, from the National Household Education Surveys Program of 2012. First Look. NCES 2013-028.Rev.” National Center for Education Statistics.

Pennings, Ray, John Seel, Deani A. Van Pelt, David Sikkink, and Kathryn Wiens. 2011. Cardus Education Survey: Do the Motivations for Private Religious Catholic and Protestant Schooling in North America Align with Graduate Outcomes? Hamilton, ON: Cardus.

Pennings, Ray, David Sikkink, Ashley Berner, Christian Smith, Mark Berends, Julie Dallavis, and Sara Skiles. 2014. Cardus Education Survey: Private Schools for the Public Good. Hamilton, ON: Cardus.

Van Pelt, Deani A., David Sikkink, Ray Pennings, and John Seel. 2012. “Private Religious Protestant and Catholic Schools in the United States and Canada: Introduction, Overview, and Policy Implications.” Journal of School Choice 6, no. 1:1–19.

Vryhof, Steven C. 2004. Between Memory And Vision : The Case for Faith-Based Schooling. Grand Rapids: Eerdmans.

Wolf, Patrick, Jay Greene, Brett Kleitz, and Kristina Thalhammer. 2001. “Private Schooling and Political Tolerance.” In Charters, Vouchers, and Public Education, edited by Paul E. Peterson and E. Campbell David, 268–91. Washington, DC: Brookings Institution Press.