66 EDUCATION NEXT / S U M M E R 2 0 0 7 www.educationnext.org

(Figure 1a)

58

political tolerance

voluntarism

political knowledge

political participation

social capital

105

05

10

Number of findings (rigorous studies)

Traditional Public

School AdvantageNeutral

Choice SchoolAdvantage

1 1

1

21

31

2

Studies That Use Basic Statistical Methods

Studies That Are Experimental/Statistically Rigorous

6political tolerance

55

voluntarism

political knowledge

political participation

social capital

civic skills

105 0 5 10

Number of findings (basic studies)

2 2

41

24

patriotism1

1 1

Traditional Public

School AdvantageNeutral

Choice SchoolAdvantage

(Figure 1b)

Private Means,Public Ends (Figure 1)

Even the most rigorous

studies with few exceptions,

studies consistently find

that schools of choice instill

civic values to a higher or

the same degree as

traditional public schools.

Note: The 21 studies included in this review had a total of 59 findings. A study had multiple findings if it reported effect sizes for multiple types of choice school(e.g., secular private, Catholic, charter).

SOURCE: Author’s tabulations

FL

AG

GR

AP

HIC

: B

RU

CE

SA

ND

ER

S D

ES

IGN

/ P

HO

TO

S:

GE

TT

Y I

MA

GE

S

www.educationnext.org S U M M E R 2 0 0 7 / EDUCATION NEXT 67

Do assigned public schools have a comparative advantage over public schools ofchoice and private schools in steeping their charges in the civic values necessaryfor democratic citizenship? The theoretical argument in favor of such an advantage is bothintuitive and popular. As free government schools, open to all on equal terms, public schoolsmake an important statement about equality, a fundamental democratic value. Former edu-cation secretary Richard Riley aptly captured this perspective, noting that civic values are “con-veyed not only through what is taught in the classroom, but by the very experience of attend-ing [a public] school with a diverse mix of students.”

Many supporters of school choice argue that neighborhood assignment to public schools results notin what public school advocates celebrate but in just the opposite: schools that are less likely to contain adiverse mix of students and that are more internally segregated along racial lines than are schools of choice.In recent years, a number of empirical studies of the effects of school choice on civic values have been pub-lished. As the extent of school choice in American education continues to grow—the latest data from theDepartment of Education show that 26 percent of American students attended a school other than theirclosest neighborhood public school—it is time to take stock of the evidentiary record on whether assignedpublic schooling better prepares students for their responsibilities as citizens in a democracy.

Studying the Effects of Choice on Civic ValuesFor this review, I examine the results of 21 quantitative studies regarding the effects of school choice onseven civic values that relate to the capacity of individuals to perform as effective citizens in our represen-tative democracy. The values, in order from the most studied to the least studied, are political tolerance,

Schools of choice boost civic values

research

BY PATRICK J. WOLF

Civics Exam

PHOTOS: GETTY IMAGES

68 EDUCATION NEXT / S U M M E R 2 0 0 7 www.educationnext.org

voluntarism, political knowledge, political participation,social capital, civic skills, and patriotism.

The studies are divided into two categories, based on thestatistical rigor with which the investigation was conducted.To qualify for inclusion in this review, a study had to be a quan-titative analysis that controlled for observed differences in thebackgrounds of the students attending different schools. Tobe classified as rigorous, the study also had to attempt to cor-rect for the tendency of students and families to sort them-selves into different schools and school sectors based onunobserved factors, a research challenge commonly referredto as selection bias. Those studies classified as rigorous usedexperimental data or employed sophisticated statistical tech-niques that credibly adjust for the possibility of selectionbias when analyzing nonexperimental data. These more-rig-orous studies should be weighted more heavily in any assess-ment of school sector impacts.

Most of the studies included in the analysis focus on stu-dents in private schools. Only three studies present results forstudents in charter or magnet schools. Therefore, the resultsdescribed below primarily map out the effects of privateschooling on civic values.

Findings are divided into three categories. A finding iscategorized as signaling a traditional public school advantageif the evidence suggests that such a schooling arrangement pro-duced a statistically significant (at the 90 percent confidencelevel or better) increase in the realization of the particular civicvalue. A finding is classified as supporting a choice schooladvantage if attendance at a public or private school of choicegenerated a statistically significant positive effect on a civicvalue. Findings of no significant difference between tradi-tional public and choice schools are classified as neutral.

As can be seen in Figures 1a and 1b, the 59 findings fromexisting studies suggest that the effect of private schooling orschool choice on civic values is most often neutral or positive.Among the group of more-rigorous studies, 12 findings indi-cate statistically significant positive effects of school choice orprivate schooling on civic values and 10 suggest neutral results(see Figure 1). Only one finding from the rigorous evaluationsindicates that traditional public schooling arrangementsenhance a civic value.

The studies that employ only basic adjustments for likelyself-selection paint an even rosier picture of the positive effectsof school choice on civic values (see Figure 1). Of the 36 find-ings, 21 indicated a school choice advantage in promotingpreparation for citizenship. Thirteen neutral results appear inthis collection of analyses, and two findings show benefits fromtraditional public schooling. The reader is cautioned not todraw strong conclusions from these studies, however, since theyemployed only rudimentary methods for addressing the prob-lem of selection bias. We now consider the specific civic val-ues that appear to be affected by school choice arrangements.

Studies of Political ToleranceDemocratic citizenship requires that we respect the rights ofothers, even if we profoundly disagree with their opinions. Themost commonly used method of measuring such political tol-erance first asks respondents to either think of their least-likedpolitical group or select one from a list that includes suchgroups as the Ku Klux Klan,American Nazis, the religious right,and gay activists. It then asks whether respondents wouldpermit members of the disliked group to exercise constitutionalrights such as making a public speech, running for politicaloffice, and teaching in the public schools. Other studies sim-ply ask respondents whether they would permit various activ-ities from a group with whom they disagree, without firstasking them to choose their least-liked group. In either case,responses are aggregated into a tolerance scale.

With one exception, the findings regarding the effect ofschool choice on political tolerance are confined to the neu-tral-to-positive range. Eleven findings—five of them from themore-rigorous studies—indicate that school choice increasespolitical tolerance. For example, one experimental voucherstudy in Washington, D.C., found that nearly one-half of thestudents who switched to a private school said they would per-mit a member of their disliked group to live in their neighbor-hood, compared with just over one-quarter of the students inthe public school control group. Three studies that used sophis-ticated nonexperimental techniques to control for selectionbias also found positive effects of choice arrangements onpolitical tolerance. These studies were of 8th-grade studentsin Dallas–Fort Worth attending private secular and nonevan-gelical religious private schools, 8th-grade students in private

With one exception, the findings

regarding the effect of school

choice on political tolerance are

confined to the neutral-to-positive

range. Eleven findings—five of

them from the more-rigorous

studies—indicate that school

choice increases political tolerance.

www.educationnext.org S U M M E R 2 0 0 7 / EDUCATION NEXT 69

research

CHOICE WOLF

secular schools in New York City, and Massachusetts stu-dents attending secular private schools.

The more-rigorous studies produced eight findings thatschool choice arrangements neither increase nor decreasepolitical tolerance. For example, three experimental studiesfound neutral results of school vouchers on the political tol-erance of middle-school students, in the Washington programafter three years, a Dayton program after two years, and a SanFrancisco program after two years.

Six of eight findings from the less-rigorous studies of theeffects of school choice on political tolerance indicate a schoolchoice advantage. Three studies concluded that secular pri-vate schools have a positive effect on political tolerance. Oneanalysis reported that Catholic schooling boosted tolerance.Another study found that religious schools in general increasethe political tolerance of their students. A third report con-cluded that private schooling of any type improves politicaltolerance among Latinos.

An observational study by Jay Greene and his colleaguesreported no effect of school type on the political tolerance ofTexas adults, and David Campbell’s analysis of the NationalHousehold Education Survey (NHES) found that students intraditional public schools had higher levels of political toler-ance than students in non-Catholic religious schools.

Studies of VoluntarismThe ideal citizen not only tolerates dissent but also activelyserves the community. With one exception, studies regardingthe extent to which private schooling or school choice affectsthe likelihood that students or parents will volunteer their timein community enterprises range from neutral to positive.

Four voluntarism findings emerged from rigorous stud-ies, with three of them favoring school choice. The study of8th graders in New York City and Dallas–Fort Worth foundthat private school students are 21 percent more likely to vol-unteer—and dedicate more hours to community service—than comparable public school students.A study using just thesample of 8th graders in New York City found that privateschooling promotes volunteer activity if the students attendreligious private schools. A third rigorous study looked at theeffects of school choice on the likelihood of parents volunteer-ing in New York City and the New Jersey suburbs and foundthat parents in school choice districts are about 6 percentmore likely to volunteer than are comparable parents in non-choice districts. Finally, students in private secular schools inNew York City are nearly 17 percent less likely to volunteer thancomparable public school students, a finding that favors tra-ditional public schooling arrangements.

Studies of voluntarism employing less-rigorous statisticalmethods produced 10 findings; half favor school choice,whereasthe other half report no effects of school type. The findings in

support of school choice show higher voluntarism among stu-dents in religious schools, parents of students in religiousschools, parents who home school, students in any type of pri-vate school, and students in public charter schools. Other stud-ies reported that voluntarism rates were similar between the stu-dents in secular private, non-Catholic religious, and magnetschools and their peers in traditional public schools. Two stud-ies reported similar rates of volunteering between the parentsof students in secular private and traditional public schools.

Studies of Political KnowledgePresumably, democratic citizens will be more active andeffective in public forums to the extent they are knowledge-able about politics and current events. Researchers typicallymeasure political knowledge by administering brief civicsquizzes of 3 to 10 items pertaining to central features of theU.S. Constitution as well as contemporary public figures.Five studies produced nine findings regarding the effect ofschool choice on political knowledge. Only three of these find-ings are from rigorous studies. R. Kenneth Godwin and FrankKemerer, in their analysis of students in schools in New YorkCity and Dallas–Fort Worth, found that choice studentsscored higher than traditional public school students onpolitical knowledge regardless of whether they attended pri-vate schools in general or evangelical private schools in par-ticular. The only experimental study of school choice andpolitical knowledge found no significant difference in aver-age political knowledge levels between recipients of vouch-ers and comparable students in public schools.

Two volunteer students with potted flowers at Prospect Park ,Brooklyn, New York

70 EDUCATION NEXT / S U M M E R 2 0 0 7 www.educationnext.org

Three studies that employed basic statistical methodsreport six findings regarding the effect of school choice onpolitical knowledge, two showing a choice school advantage.Richard Niemi and his colleagues drew on the NHES data toconclude that private schooling increases political knowl-edge. David Campbell’s more fine-grained analysis of thesame data reported that only Catholic private schools demon-strated a clear political knowledge advantage. Campbell foundthat students in non-Catholic religious, secular private, andpublic choice schools all evidenced political knowledge lev-els that were comparable to students in traditional publicschools. In the earliest known empirical study of the effect ofschool choice on civic values, James Coleman and ThomasHoffer concluded that students in Catholic and public schoolswere similar in their average levels of political knowledge.

Studies of Political Participation, Social Capital,Civic Skills, and PatriotismBeyond being tolerant, community minded, and wellinformed, we also expect well-trained citizens to be polit-ically active possessors of social capital with civic skillswho are loyal to their country. Unfortunately, relativelyfew studies have queried the extent to which school choice

arrangements foster such attributes. One rigorous study, byThomas Dee, concluded that Catholic schooling increasesvoter turnout as adults. Jay Greene and his colleagues con-ducted less-sophisticated studies that found that Latinoswho received all of their K–12 education in private schoolswere 16 percent more likely to say they voted in the lastpresidential election than comparable Latinos who wereeducated exclusively in public schools. They also reported thatTexas adults who were educated at least partly in privateschools were 9 percent more likely to have voted recently, allelse being equal. An observational study by Christian Smithand David Sikkink found that parents who enroll their chil-dren in private religious schools or who home school themare more politically active than are otherwise comparable par-ents who enroll their children in public schools. Parents ofstudents in private secular schools do not differ significantlyfrom public school parents in political participation.

Two rigorous studies reported findings regarding theeffects of school choice on social capital, typically defined asa close connection with one’s community via social networks,group norms, and cooperation for mutual benefit. MarkSchneider and his colleagues concluded that the responsibil-ity to choose their child’s school increases the social capital ofparents. Paul Peterson and David Campbell reported no dif-ference in the levels of social capital between voucher users andcontrol group members in their experimental analysis of theChildren’s Scholarship Fund.

In the first study of education and social capital, Colemanand Hoffer employed basic statistical methods and foundthat Catholic schooling was associated with higher levels ofsocial capital. Greene and his colleagues replicated thoseresults on a national sample of Latino adults.

No experimental studies have been conducted on theeffects of school choice on civic skills or patriotism. Twostudies that applied basic statistical methods to the 1996NHES data generated diverse findings regarding the effect ofschool choice on civic skills. The survey asked students, Dur-ing this school year, have you done any of the followingthings in any class at your school: Written a letter to some-one you did not know? Given a speech or an oral report? Takenpart in a debate or discussion in which you had to persuadeothers about your point of view? Students in private highschools were more likely to have engaged in these three activ-ities than comparable students in public high schools, accord-ing to one study. The second study found that students inCatholic schools scored slightly higher than comparable stu-dents in assigned public schools. No significant differencesin civic skills were uncovered between students in assignedpublic schools and comparable students in non-Catholicreligious or secular private schools.

An observational study of patriotism employed an indexthat includes five questions about students’visceral attachment

Twenty-two study results

show a school choice

advantage, suggesting the

secular private schooling

enhances political tolerance,

that charter schooling

increases voluntarism, and

that education at an

evangelical private school

increases political knowledge.

to their country and its symbols (such as the flag and the Pledgeof Allegiance). New York City 8th graders in private schoolsscored somewhat lower on patriotism, on average, than com-parable students in public schools. One weakness of thisanalysis is the patriotism scale employed, which could beinterpreted as a measure of national chauvinism or jingoism.

The Catholic Schooling EffectSeveral prominent scholars have claimed that Catholic school-ing may be largely responsible for the generally positive schoolchoice effects on civic values. Would thelikely effects of choice on political toler-ance, voluntarism, and other democraticvalues disappear or turn negative withCatholic schools out of the picture? Fig-ure 2 excludes all results based on compar-isons between public and Catholic schoolpopulations or that focus exclusively onthe experiences of Latinos (who, if pri-vately schooled, predominantly attendCatholic schools). Study groups identifiedsimply as “private religious”are consideredCatholic and excluded for purposes ofthis analysis, since most religious privateschools in the U.S. are Catholic. Twenty-two results showing a school choice advan-tage remain, suggesting that secular pri-vate schooling enhances politicaltolerance, that charter schooling increasesvoluntarism, and that education at anevangelical private school increases polit-ical knowledge. Twenty findings indicatethat school choice has no clear effect, pos-itive or negative, when schools other thanCatholic schools are chosen.

Three findings showing a traditionalpublic school advantage remain, sug-gesting that evangelical Protestant schoolsreduce political tolerance, that secularprivate schools decrease voluntarism,and that private schooling of any sortmay diminish a particularly passionateform of patriotism. As all the negativeeffects shown in Figures 1 also appearhere, it seems non-Catholic schools ofchoice are responsible for the few nega-tive effects of choice arrangements oncivic values observed here. However, non-Catholic schools ofchoice also appear to generate many positive outcomes regard-ing democratic values. These results suggest that the expan-sion of school choice is more likely to enhance than dimin-

ish the civic values of our next generation of citizens, even ifnone of the new choosers end up in communitarian-infusedCatholic schools.

DiscussionAll of the studies reviewed draw on data either about the var-ious school sectors as they existed in the 1980s and 1990sor from modestly sized school choice experiments. Thedemographic composition of the various school sectorsand the independent effects of private schooling and schoolchoice on the civic values reviewed here would likely change

somewhat under a complete or even larger-scale schoolchoice regime. One should therefore be cautious in draw-ing strong conclusions from the empirical record to date onschool choice and civic values.

www.educationnext.org S U M M E R 2 0 0 7 / EDUCATION NEXT 71

research

CHOICE WOLF

10

2

All Studies, Excluding Those Focused on Catholic Schools and Latinos

political tolerance

45

voluntarism

political knowledge

political participation

social capital

civic skills

515 10 0 5 15

Number of findings

1 1

1

41

34

patriotism

1

1

91 7

Traditional PublicSchool Advantage

Neutral Choice SchoolAdvantage

Not Just for Catholics (Figure 2)

A clear choice school advantage emerges even after removing studies of Catholicschools and Latino students, refuting the notion that Catholic schools are uniqueamong private schools in their capacity to teach civic values.

SOURCE: Author’s tabulations

The empirical picture regarding the effects of schoolchoice on civic values raises some concerns. The lone studyon fostering patriotism indicates that public schools mayhold an advantage over schools of choice. In one study ofvoluntarism, attending private secular schools apparentlyreduced the likelihood of volunteering. Attending an evan-gelical Protestant school was found to decrease politicaltolerance in one study and increase political knowledge inanother, causing scholars such as Stephen Macedo to worrythat such schools may produce young adults who are stronglyequipped to act politically on their intolerance. The Madrassaschools of radical Islam remind us that private schools ofchoice can serve to undermine democratic values. It wouldseem reasonable to require some minimal oversight and reg-ulatory constraints on private schools that accept publicmonies, such as prohibitions against teaching hate. Asimportant as these concerns are, the record to date suggeststhat civic values tend to be enhanced, or at least not harmed,by the exercise of school choice.

What aspects of choice schools generate these modestlypositive civic values outcomes? No direct evidence yet existsregarding the specific conditions or practices of choiceschools relative to traditional public schools that wouldexplain this pattern of results. One theory is that schools ofchoice foster strong education communities typified by reg-ular parental involvement and a concern for the welfare ofall members. Yet several other plausible explanations alsodeserve attention. Teachers in private schools may be freerto infuse instruction with moral values and discuss contro-versial issues than public school teachers. Students who reg-ularly encounter value-based claims and perspectives maybecome more tolerant of people with value-based positionsthat differ from their own. They also may feel more moti-vated to volunteer for activities that seek to bring aboutsocial and political change.

The most intriguing explanation, in my opinion, for theapparent school choice advantage in promoting civic valuesis a generally higher level of order and discipline in schools ofchoice. Public charter schools and private schools tend to bemore well-ordered education institutions than neighborhoodpublic schools, especially in urban environments. A well-ordered and nonthreatening education environment likely con-tributes to students’ feelings of security and confidence. Suchfeelings might be a necessary precondition for young peopleto develop a willingness to tolerate potentially disruptivepolitical ideas and political groups and to venture out into thecommunity to promote social causes, an idea suggested by AlanPeshkin in his case study of a Christian fundamentalist school.There is a clear theoretical justification for linking a well-ordered education environment with stronger civic values, andI hope that future studies will explore this possibility.

Other aspects of schooling might also promote higher lev-els of civic values among students, be they in assigned publicschools or schools of choice. Effective instruction itself likelypromotes civic values, as better-educated citizens tend to bemore knowledgeable about politics, more tolerant, and moreactive in their communities. Some preliminary studies suggestthat students are more likely to embrace civic values as adultsif they had the opportunity to participate in student governanceor voluntary activities as students, or at least witnessed adultswho modeled proper civic behaviors in their schools. There isless empirical support for curricular interventions aimed atboosting civic values. Civics classes appear to increase civic val-ues such as tolerance only modestly, and only if they are cus-tomized to focus explicitly on that particular value. There is noevidence that taking a required civics course in junior high orsenior high school, in and of itself, enhances civic values.

In summary, the empirical studies to date counter theclaims of school choice opponents that private schoolinginherently and inevitably undermines the fostering of civicvalues. The statistical record suggests that private schoolingand school choice often enhance the realization of the civicvalues that are central to a well-functioning democracy. Thisseems to be the case particularly among ethnic minorities(such as Latinos) in places with great ethnic diversity (suchas New York City and Texas), and when Catholic schools arethe schools of choice. Choice programs targeted to such con-stituencies seem to hold the greatest promise of enhancingthe civic values of the next generation of American citizens.

Patrick J. Wolf is professor of education reform and 21st centurychair in school choice at the University of Arkansas College ofEducation and Health Professions.

A complete list of the studies is provided in the appendix to the unabridged version of this essay at www.hoover.org/publications/ednext/.

72 EDUCATION NEXT / W I N T E R 2 0 0 7 www.educationnext.org

The most intriguing

explanation for the apparent

school choice advantage

in promoting civic values

is a generally higher level of

order and discipline in

schools of choice.