Bowling Together

Private schools, public ends
Illustration by Travis Foster

As the civic participation of young people continues to plummet, it becomes ever more important that we learn how schools can teach students to be active citizens. Indeed, “producing better citizens” was the original justification for creating America’s public schools. In the 1800s, Horace Mann and others successfully argued that the public schools could assimilate immigrants into the norms of American civic life. Today essentially the same objective remains. In a 1996 Phi Delta Kappa/Gallup Poll, 86 percent of Americans reported that they feel “preparing students to be responsible citizens” is a “very important” purpose of the nation’s schools; just 76 percent considered it very important that schools “help people become economically self-sufficient.”

Today a broad consensus exists on three objectives characteristic of an education that develops good citizens. The first objective is to equip the nation’s future voters with the capacity to be engaged in the political process. This is especially salient against a backdrop of declining rates of political activity, most notably among young people. The second objective is to have citizens not only participating in democratic institutions, but also doing so knowledgeably. Students should understand the nation’s history and political system. The third objective stems from political philosopher Amy Gutmann’s idea that the defining characteristic of a democratic education is that it imparts the “ability to deliberate” in a context of “mutual respect among persons.” The general public would seem to agree. According to a 1999 Phi Delta Kappa/Gallup Poll, 93 percent of Americans believe that the schools should teach “acceptance of people of different races and ethnic backgrounds.” Seventy-one percent stated that the schools should also teach “acceptance of people who hold unpopular or controversial political or social views.” In other words, an education that prepares students for democracy teaches them to respect the opinions of others and promotes social and political tolerance.

Different types of schools pursue various strategies for meeting these objectives, some with more success than others. Here I ask the simple question: How well do different types of schools promote civic education? For instance, do private schools foster social divisiveness, as their critics often claim? Does attending a religious private school rather than a secular one have different civic consequences?

Building Social Capital

There is only a small but nonetheless growing body of research on the civic effects of public versus private schools. Researchers have shown that Catholic schools are more racially integrated than public schools and that voucher programs do not have an adverse effect on integration. Another study found that Hispanic adults who were educated in private schools are more likely to participate in politics than those who attended public schools. Evidence from the National Education Longitudinal Study further demonstrates that students in private schools are more likely to participate in community service than are their peers in public schools. Similarly, private school administrators more often rate their schools as “outstanding in promoting citizenship” than do their public school colleagues. Research from the voucher programs in Dayton, Ohio, and Washington, D.C., found that parents of private school students were more civically engaged than parents of public school students.

Strong evidence has accumulated that nonpublic, particularly Catholic, schools are a private means to the very public end of facilitating civic engagement.

In 1966, before school vouchers were on the nation’s political agenda, Andrew Greeley and Peter Rossi used extensive survey data collected from American Catholics to argue that Catholic schools do not depress civic engagement or promote intolerance. The 1998 National Assessment of Educational Progress (NAEP) Civics Report Card for the Nation reports that students in private schools (both Catholic and non-Catholic) have higher average scores on the NAEP civics test than do their peers in public schools. However, this is without adjusting the data for background characteristics that may affect students’ level of civic knowledge, such as their parents’ educational level. James Coleman and Thomas Hoffer did control for family background and found that students in private schools, both Catholic and non-Catholic, scored higher on the High School and Beyond civics test than did public school students, although the results were not statistically significant. Taken together, these results give no reason to suspect that private schools do a worse job of providing a civic education than assigned public schools and some reason to think they do a better job. Yet reasonable doubt remains.

More broadly, Harvard University professor Robert Putnam’s research on civic participation provides reason to think that private schools should be better able to deliver civic education than are public schools. Since the publication of Putnam’s Making Democracy Work (1993) and the follow-up, Bowling Alone (2000), it has become increasingly common for political scientists to discuss political participation as driven by social capital. As Putnam defines it, “social capital refers to features of social organization such as networks, norms, and social trust that facilitate coordination and cooperation for mutual benefit.” Before Putnam employed the concept to explain differences in governmental performance between northern and southern Italy, Coleman and his colleagues developed it to theorize why students in Catholic schools excel academically relative to their public school peers. Perhaps the same characteristics that cause Catholic schools to excel academically enable them to produce better citizens.

Comparing public and private school students is difficult, given that most Americans attend a public school in the elementary and secondary grades. In 1995, 91 percent of American secondary-school students were enrolled in public schools. It is even more difficult to make comparisons within the private sector, which comprises Catholic schools, religious schools sponsored by other faiths, and secular schools. Even within the public sector, there are schools to which students are assigned based on geography and schools they choose to attend (magnet and charter schools, for example). Here I use five types of schools: assigned public, magnet public, Catholic, secular private, and other religious. This last category combines schools from a broad range of faiths, including Christian fundamentalists, Quakers, and Muslims.

I used survey data from the 1996 National Household Education Survey, a large nationally representative survey of both parents and their children. This analysis draws on questions asked of students (and their parents) in grades 9 through 12, for a total sample size of 4,213. Even with a sample this large, the number of students attending other religious and secular private schools is still quite small (80 and 102 cases, respectively).

Throughout the analysis, except where noted, I adjusted the data to account for a variety of factors that might influence students’ civic education and knowledge other than the kind of school they attend. At the individual level, I controlled for the usual demographic factors: students’ age, gender, race, ethnicity, whether they spoke English, and whether they lived in the South. I also controlled for their academic performance, expectations of going to college, their expressed ability to take political action, their interest in the news, and the number of hours they spent at a job. Since adolescents’ family lives exert a strong influence on their values, I controlled for a variety of family-based factors. These included parents’ educational levels, family income, church attendance, whether students grew up in a two-parent household, parents’ volunteer work, parents’ political participation, parents’ civic skills, and parents’ political tolerance. I also considered factors such as the racial composition of a school, school size, whether the school arranges volunteer service, whether students’ opinions matter, whether courses with political content are offered (so-called civics classes), and whether the school has a student government. Owing to the richness of the control variables, there can be reasonable confidence that these results capture the effect of the school a student attends.

Civic Engagement

Voter turnout is the most commonly used measure of civic engagement, but it is inadequate here because very few secondary-school students are old enough to vote. Instead, three other measures can be used. One, voluntary community service, is a measure of what students are doing in the present. The second, acquiring “civic skills” in the classroom, is a measure of what students have learned to prepare them to be civically engaged in the future. The third is a measure of whether students feel confident that they could actually use their civic skills outside of the classroom.

Community Service. Past studies have found civic activity while young to be a “pathway to participation” in adulthood. Students were asked whether they engaged in “any community service activity or volunteer work at your school or in your community.” Without any adjustments to the data, the results show that, statistically, there is no difference between assigned public schools and magnet public schools or secular private schools. However, students in both Catholic and other religious schools are more likely to engage in community service than are students in assigned public schools. Forty-seven percent of assigned public school students perform community service, compared with 64 percent of students in other religious schools and 71 percent of students in Catholic schools. This is not surprising, given the religious character of these schools. Other research has shown that religious people are more inclined to volunteerism.

A reasonable objection to these results is that they are potentially misleading because many religious schools require their students to perform “voluntary” service. Seventy percent of students in Catholic schools report that their schools require community service in the 9th through 12th grades. This compares with 16 percent of students in assigned public schools, 22 percent in magnet public schools, 28 percent in other religious schools, and 38 percent in secular private schools. It is unclear what these differences mean for students’ long-term commitment to community service. On the one hand, students who are compelled to perform service may only grow resentful. On the other hand, requiring service in the community would presumably introduce to volunteer work students who would not otherwise have had that experience. They may find that they enjoy it and wish to continue even after they have fulfilled their school’s requirement. Regardless, after excluding from the analysis those students whose schools required them to perform community service, the results were very similar-with the notable exception of religious/non-Catholic schools. While more of their students participate in community service than do students in assigned public schools, the difference ceases to be statistically significant.

These results, however, still do not account for differences in the backgrounds and characteristics of students who attend these types of schools that might in turn affect whether they engage in community service. After again excluding students whose schools require community service, I took into account the various factors listed above. The results still show that students in Catholic schools are more likely to perform community service than those in assigned public schools (see Figure 1). Forty-eight percent of public school students participate in community service, compared with 59 percent of Catholic-school students. That is, Catholic schools contribute about as much to the likelihood of students’ providing community service as does having a parent or guardian in the home who participates in community service (which also increases the share of students participating in volunteer activity by about 11 percentage points). Catholic schools contribute almost twice as much to a student volunteering as does raising a parent’s educational level from a high-school diploma to a college degree. There is no statistically significant difference between students in assigned public schools and those in secular private, magnet public, or religious/non-Catholic schools.

This suggests an answer to one of the essential questions in the debate over voucher programs-that is, will sending students to private schools harm their civic education? The answer appears to be no; in fact students in Catholic schools are more likely to engage in voluntary service than are students in assigned public schools, and there were no significant differences between students in assigned public schools and those in the three other types of schools. These findings are consistent with the research of Anthony Bryk, Valerie Lee, and Peter Holland, who explored the relationship between Catholic schools and their communities. They report that in all of the Catholic schools they included in their study, community service work was available as an elective course. Furthermore, an ethic of service was frequently found among both the staff and students they interviewed. They write:

These service programs signify Catholic schools’ commitment to a just social community. One board member of a field-site school remarked, “A school should not call itself Catholic if it doesn’t have a volunteer service program.” The director of the program at St. Edward’s [one of their case-study schools] commented: “I’m a believer in service. It’s important for students to realize that the things they do make a difference. We can heal people and make their lives better. We can raise the awareness of others. Physical contact is vital for Christianity. Some of our students are sheltered from poverty and from people of different races. This program is important because it makes them more aware.”

Civic Skills. People differ in their capacity to perform the mundane tasks that constitute virtually all political activity-skills like giving speeches, holding meetings, and writing letters. Those who lack these skills are extremely unlikely to participate in politics. While the authors focused on how adults learn civic skills on the job or through participation in voluntary organizations, it is in school that people are most likely to learn them when young.

An index of civic skills was created to test for systematic differences across the five types of schools. The household survey asked students, During this school year, have you done any of the following things in any class at (your current) school:

• Written a letter to someone you did not know?

• Given a speech or an oral report?

• Taken part in a debate or discussion in which you had to persuade others about your point of view?

Once again, the result for students who attend a Catholic school is statistically significant at the .05 level. Given the litany of control variables included in this analysis, this variable has quite a statistical hurdle to clear to reach statistical significance. Compared with a student who attends an assigned public school, a Catholic-school student learns an average of .13 more civic skills. Not a dramatic difference, but completely consistent with the other findings reported here, it bolsters the evidence that Catholic schools deliver a high-quality civic education.

Civic Confidence. Learning civic skills is one thing; being able to use them is another. The household survey also asked respondents whether they feel that they could use two of the civic skills learned inside the classroom elsewhere. The questions ask:

• Suppose you wanted to write a letter to someone in the government about something that concerned you. Do you feel that you could write a letter that clearly gives your opinion?

• Imagine you went to a community meeting and people were making comments and statements. Do you think that you could make a comment or a statement at a public meeting?

Students in secular private, Catholic, and other religious schools are more likely than students in assigned public schools to have confidence in their ability to exercise civic skills if called upon to do so. Of these three, the religious/non-Catholic-school students display the greatest degree of civic confidence. Civic confidence is the only component of civic education included in this analysis for which each type of private school displays a positive and statistically significant effect. The bottom line is that public schools can take a lesson from private schools about how to prepare students for civic life, not only in providing skills, but also in providing the confidence to use those skills.

Political Knowledge

The second objective of a civic education is to teach future voters specific, factual information about American politics. Indeed, of the three objectives I have listed, this one is most clearly the province of the schools. While there may be disagreement over whether schools should require community service, presumably everyone agrees that schools should require the acquisition of knowledge. Without understanding the particulars of American politics, people are unable to engage fully in the political process. In fact, political scientist John Zaller argues persuasively that factual knowledge about politics is the best measure of political engagement.

Perhaps the same characteristics that cause Catholic schools to excel academically enable them to produce better citizens.

The National Household Education Survey includes a series of factual questions about American politics. Each respondent was asked five of the following ten questions. To avoid contamination effects, whichever five questions a student answered, her parent answered the other five:

• What job or political office is now [in 1996] held by Al Gore?

• Whose responsibility is it to determine if a law is constitutional…the President, the Congress, or the Supreme Court?

• Which party now has the most members in the House of Representatives in Washington?

• How much of a majority is needed for the U.S. Senate and House to override a presidential veto?

• Which of the two major parties is more conservative at the national level?

• What job or political office is now held by Newt Gingrich?

• Whose responsibility is it to nominate judges to the federal courts…the President, the Congress, or the Supreme Court?

• Which party now has the most members in the U.S. Senate?

• What are the first ten amendments of the U.S. Constitution called?

• Which of the two major parties is in favor of the larger defense budget?

Before making any adjustments, the average scores of students in assigned public schools are lower than those in Catholic, religious/non-Catholic, and secular private schools. Students in assigned public schools got an average of 2.4 questions out of five correct, while students in Catholic, religious/non-Catholic, and secular private schools scored an average of 3.2, 3.4, and 3.2 respectively.

Once the statistical adjustments are made for all the factors that can influence students’ political knowledge except the type of school they attend, only students in Catholic schools still perform better than do students in assigned public schools. Therefore, we can conclude that only students in Catholic schools display more political knowledge when accounting for a slew of potentially confounding demographic factors, many of which are themselves statistically and substantively significant. Older students score better on the index, as do males, whites, and non-Hispanics. Having higher grades, expecting to attend college, expressing greater political interest, and spending more time watching or reading the news are all positively related to political knowledge. At the family level, both parents’ educational levels and political knowledge are positive factors predicting students’ greater political knowledge.

Political Tolerance

While all three objectives are equally important components of a civic education, it is the third-respect for opinions different from your own, or political tolerance-that may be most relevant to the debate over the civic consequences of attending private schools. This is often expressed as a concern that private (particularly religious) schools exacerbate social tensions. In the words of the late union leader Al Shanker, widespread voucher programs that send students to private schools “would foster divisions in our society; they would be like setting a time bomb.” Amy Gutmann stresses the need for students in religious schools to be taught democratic norms under direction from the state, presumably fearing that these schools cannot be trusted to provide instruction in a “common democratic character” on their own. Certainly, the concern for teaching a respect for universal civil liberties is well placed, as democracy is defined as much by respect for minority rights as by simple majority rule.

Forty-eight percent of public school students participate in community service, compared with 59 percent of Catholic-school students.

The question of adolescents’ attitudes and how they are related to enrollment in different types of schools has been virtually unexplored. The research literature suggests two alternative hypotheses, drawn from different perspectives on what fosters political tolerance. The first hypothesis is derived from distinguishing between schools as public versus private institutions. By this reasoning, private schools may be thought to foster an exclusivity among their students that translates into a disregard for minority opinions. In particular, religious schools may foster civic divisiveness, a fear reinforced by survey data that show religiosity to be negatively related to political tolerance. The second hypothesis follows from studies showing the positive academic effects of attending a private school. One widely noted study reports that education increases tolerance by enhancing students’ general cognitive proficiency. It would follow that tolerance is greatest in those schools where students display the strongest cognitive performance, generally private schools.

The household survey contains two questions gauging political tolerance:

• If a person wanted to make a speech in your community against churches and religion, should he or she be allowed to speak?

• Suppose a book that most people disapproved of was written, for example, saying that it was all right to take illegal drugs. Should a book like that be kept out of a public library?

The question about churches and religion particularly challenges the political tolerance of students in religious schools, since it confronts them with an opinion they will almost certainly reject. Students in Catholic and secular private schools have higher tolerance scores than students in assigned public schools, averaging 1.6 and 1.8 tolerant responses respectively, compared with 1.4 tolerant responses among assigned public school students. Students in other religious schools have an average score (1.2 tolerant responses) lower than that of public school students. Students in magnet public schools have slightly higher scores than assigned public school students, although the difference does not approach statistical significance.

After again making the statistical adjustments listed above, students in secular private schools scored substantially higher on the political tolerance index than students in assigned public schools, while students in religious/non-Catholic schools scored substantially lower (see Figure 2). Catholic-school students still score higher than assigned public students, though the difference is only a third as large as that between secular private schools and assigned public schools. It is possible that there is some credence to the concern expressed by critics of private education that it has the potential to foster political intolerance. While students in Catholic schools (the most common form of private education) and secular private schools are more politically tolerant than students in assigned public schools, the 2 percent of America’s students in other religious schools-an amalgam of schools sponsored by many different faiths-score lower on the political tolerance index.

One should not necessarily conclude that other religious schools breed intolerance. It may be that students’ attitudes are shaped more by family background than by the schools they attend. For all its virtues, the National Household Education Survey has poor measures of parents’ religious involvement. Because family tolerance is poorly measured and the sample size, for this group of students, is small, we should be cautious in drawing firm conclusions.

Evaluation of responses to survey questions about civil liberties must also take into account the content of the question. Because one of the questions on the tolerance index deals specifically with the rights of a speaker who is opposed to religion, students in religious schools might be expected to be especially wary of granting full freedom of expression. This is not to diminish the importance of respect for religious differences as an important component of political tolerance, but only to suggest that other questions might provide more of a “hard case” for students in secular schools. And lest it be thought that there is something inherent in religious education that breeds intolerance, remember that Catholic-school students show higher levels of tolerance than students in assigned public schools.

Make Democracy Work

The results reported here are consistent with four similar studies-the 1973 High School Seniors Cohort Study, the National Educational Longitudinal Study, the Latino National Political Survey, and data collected from participants in school-choice programs in Washington, D.C., and Dayton, Ohio. Few findings in social science can be replicated in five independent sources of data (six, if you count the Washington and Dayton surveys separately). In short, it seems that strong evidence has accumulated that private-particularly Catholic-schools are a private means to the very public end of facilitating civic engagement.

How “public” is a school in an exclusive suburb with high housing costs, especially when compared with a Catholic school that costs around $2,500 a year?.

This conclusion is admittedly provocative, if only because of the connotations the words “public” and “private” carry in contemporary discourse. In the United States, the word “public” is supposed to refer to the source of a school’s funding and not to the population served by a school. Nevertheless, critics of private education often implicitly extend the limited definition of “public” to mean the population served by the school. Critics speak of high-priced preparatory schools as though they are the only, or at least the most common, type of private education in the United States. Often all private schools are grouped together and caricatured as exclusive and insular. While it is true that privately funded schools have the prerogative to apply virtually any criteria they want for admissions, in practice Catholic schools, at least, are very inclusive.

Catholic high schools are not highly selective in their admissions. The typical school reports accepting 88 percent of the students who apply, and only about a third of the schools maintain a waiting list. Anthony Bryk and his colleagues also report that “religious affiliation is not a routine consideration” in admissions to Catholic schools. Even though Catholic schools charge tuition, 87 percent offer financial aid. By contrast, public schools almost exclusively enroll students who live in the geographic area surrounding the school. How “public” is a school in an exclusive suburb with high housing costs, especially when compared with a Catholic school that offers financial aid to assist with its tuition (which, in turn, is usually only around $2,500 a year)?

Critics also tend to define public schools as the only institutions providing an education that promotes publicly spirited citizens. By this definition, the evidence presented here suggests that Catholic and private secular schools are really more “public” than schools funded by the state. The claim that private organizations can contribute to the quality of a community’s public life is hardly original. Echoing Alexis de Tocqueville’s Democracy in America, Putnam found that measures of civic associational life, everything from choral societies to soccer clubs, are the primary explanation for effective governance across Italy’s regions. Voluntary associations like these produce social capital, and social capital “makes democracy work.” In fact, our public schools seem to have much to learn from private, especially Catholic, schools about what makes democratic education work.

-David E. Campbell is a Ph.D. candidate in government and a research associate of the Program on Education Policy and Governance at Harvard University.

Last Updated


Notify Me When Education Next

Posts a Big Story

Business + Editorial Office

Program on Education Policy and Governance
Harvard Kennedy School
79 JFK Street, Cambridge, MA 02138
Phone (617) 496-5488
Fax (617) 496-4428

For subscription service to the printed journal
Phone (617) 496-5488

Copyright © 2024 President & Fellows of Harvard College