USF Libraries
USF Digital Collections

Educational policy analysis archives


Material Information

Educational policy analysis archives
Physical Description:
Arizona State University
University of South Florida
Arizona State University
University of South Florida.
Place of Publication:
Tempe, Ariz
Tampa, Fla
Publication Date:


Subjects / Keywords:
Education -- Research -- Periodicals   ( lcsh )
non-fiction   ( marcgt )
serial   ( sobekcm )

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E11-00392
usfldc handle - e11.392
System ID:

This item is only available as the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
mods:mods xmlns:mods http:www.loc.govmodsv3 xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govmodsv3mods-3-1.xsd
mods:relatedItem type host
mods:identifier issn 1068-2341mods:part
mods:detail volume mods:number 12issue 43series Year mods:caption 20042004Month August8Day 1818mods:originInfo mods:dateIssued iso8601 2004-08-18

xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam a22 u 4500
controlfield tag 008 c20049999azu 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E11-00392
0 245
Educational policy analysis archives.
n Vol. 12, no. 43 (August 18, 2004).
Tempe, Ariz. :
b Arizona State University ;
Tampa, Fla. :
University of South Florida.
c August 18, 2004
Democratic education across school types : evidence for the U.S. from NHES99 / Clive R. Belfield.
x Research
v Periodicals.
2 710
Arizona State University.
University of South Florida.
1 773
t Education Policy Analysis Archives (EPAA)
4 856


EDUCATION POLICY ANALYSIS ARCHIVES A peer-reviewed scholarly journal Editor: Gene V Glass College of Education Arizona State University Copyright is retained by the first or sole author who grants right of first publication to the Education Policy Analysis Archives EPAA is a project of the Education Policy Studies Laboratory. Articles are indexed in the Dir ectory of Open Access Journals ( Volume 12 Number 43 August 18, 2004 ISSN 1068-2341 Democratic Education Across School Types: Evidence for the U.S. from NHES99 Clive R. Belfield Teachers College, Columbia University Citation: Belfield, C. R. (2004, August 18). Democratic education across school types: Evidence for the U.S. from NHES99. Education Policy Analysis Archives, 12 (43). Retrieved [Date] from http :// Abstract This article reports on the differences in democratic education across school types, using the US National Household Education Survey (NHES) of 1999. We replicate the estimation approach of Campbell (1998) and find a strongly positive effect from attendance at Catholic school or private independent schools on community service participati on, civic skills, civic confidence, political knowledge and political tolerance. The results are reasonably robust to alternative specifications. We consider the implications of these results for policy. Introduction Privatization and the Social Good A major public purpose of schooling in a democr atic society is the adequate provision of a common educational experience that will orient all students to grow to adulthood as full participants in the social, political, and econom ic institutions of our society. A democracy requires that its members master the skills an d knowledge necessary for civic and economic participation including one’s rights and res ponsibilities under the law, the principles of democratic government, and an understanding of the overall economy and preparation for productive roles (for a general discussion, see Ravitch and Viteritti, 2001). In general, this is


Democratic education across school types 2 usually interpreted as necessitating common elements of schooling with regard to curriculum, values, goals, language, and political orientation (although precisely what should be taught is a matter of debate, see Soder, 1996; Hirsch, 1987; Goodlad, 1997; Cuban and Shipps, 2000; Guttman, 1986). By virtue of being publicly funded, schools are expected to fulfill these objectives (Carnegie Corporation and CIRCLE, 2003). Yet, recent reforms to privatize the U.S. education system may influence the capacity of schools – and the education system as a whole – to foster such social cohesion. Privatization reforms essentially give families the option to distance themselves from the traditional neighborhood public school system. Charter schooling laws allow communities and corporations to set up schools that are distinct from local public schools: Cobb and Glass (1999) find that charter schools in Ariz ona enroll significantly higher proportions of white students than neighboring public school s. The legalization of home-schooling allows families to quit the public school system entirely as do vouchers for private schools. Finally, education tax credits offer subsidies to fam ilies to spend on private educational resources (see Levin, 2002). Together, these reforms co uld affect the school choices of many families and lead to an educational system which is orga nized very differently and with very different outcomes in terms of social goods produced. However, whether privatization will reduce or enhance civic cohesion is an open question within the US context. Traditional public schools have an advantage in producing a common curriculum and instruction. But, privatization – by allowing more private choices – need not automatically reduce civic cohesion. Private choices (either charter or private schools) may indeed lead to de facto segregation, and new choosers tend to be more affluent than those who do not have choice. However, if students from environments that are highly segregated socio-economically and racially are given a voucher to exercise their choice of schooling, privatization may raise civic cohesion. Others have argued that public schools are not “public”– in the sense of “open to all” – but are often organized to keep social groups from integrating (Ryan and Heise, 2002). Pu blic schools may be residentially segregated, such that there is wide variation in their prod uction of civic cohesion. Also, civic cohesion may increase if private schools either do a b etter job of teaching civic values, as some authors have found (Greene, 1998; Godwin et al ., 2001), or raise attainment, which is positively correlated with civic participation (Dee 2003). Finally, privatization programs will differently impact on civic cohesion, depending on the programs’ design (e.g. who is eligible for additional funds or services). Privatizati on includes many different reforms, and these may be targeted at specific student groups to raise (or reduce) civic cohesion. Empirical Tests for the Relationship Betw een Privatization and the Social Good There are two approaches to investigating soci al cohesion and education privatization. The first investigates what types of schools students enroll at, what motivates their decisions, and what are the characteristics of the other stud ents at these chosen schools (Martinez et al., 1996; Schneider et al., 2000; Fairlie and Resch, 20 02). This approach is useful, although it often leaves unresolved the normative issue of what choices are permissible or benign, versus those that are pernicious and divisive. In general, this evidence indicates that those who choose are among the more highly educat ed (of the target population) and that these choices are often motivated by racial or social rather than strictly academic, factors. The second approach, which is the one adopted here, looks directly at the outcomes of students in different circumstances, cond itional on the type of school they attend or the


Education Policy Analysis Archives Vol. 12 No. 43 3 choices available.1 This approach has the advantage of comparing actual amounts of social cohesion across different groups, and specifically in comparing public and private schools. (We note its limitations below). Empirical research comparing social cohe sion across school types is growing. Using similar approaches, however, Mocan et al. (200 2) and Figlio and Ludwig (2001) find discrepant results on the behavior of Catholic school students relative to public school students: the former researchers identify no sub stantive behavioral differences, whereas the latter find lower arrest rates and hard drugs usage (but no effect on gang involvement or marijuana usage). From data from the 1996 Youth Civic Involvement survey, Smith (2003, 114) reports higher levels of tolerance, civi c capital efficacy, and participation in private schools, although when these correlations are adjusted for student and community characteristics only private independent schools show an advantage. This paper adds to this research base. It begins by re-estimating – with more recent but harmonized data – the analysis reported by Campbell (1998). Our results are very similar. We then follow this research with a detailed inquiry into both the internal and external validity of these initial results. In ternal validity is checked through sensitivity analysis. External validity is considered in terms of what the results mean for educational policy. We conclude with a discussion of th e implications for policy on the promotion of social goods. Empirical Estimation Data The data for analysis are taken from the National Household Education Survey (NHES) of 1999. The survey asks questions about the educ ational experiences of families, both children and adults. The survey is large-scale, nationally representative, and harmonized with previous NHES from 1992 and 1996 (although there are some slight differences in the questionnaire and the sampling scheme). Youth are asked directly to identify the typ e of school they attend: assigned public school (76.1%); magnet or choice school (13.7% ); Catholic religious school (4.5%); nonCatholic religious school (2.7%); or private inde pendent school (3.0%). At issue is whether there are differences in civic cohesion across students of different types. The specific measures of civic education follow exactly those used by Campbell (1998). Community Service is a binary variable indicating whether the youth participated in “any community service activity or volunteer work at school or the local community.” Civic Skills is an index based on how many of the three following actions the student had performed during the school year: written a le tter to someone they did not know; given a speech or an oral report; taken part in a deba te or discussion to persuade others about one’s point of view. Civic Confidence is an index based on whether the student feels he or she could effectively write a letter to someone in govern ment about something of concern and make a comment or statement at a community meeting. Political Knowledge is an index based on responses to factual questions about American politics. Ten questions were included on the 1 A related approach is to compare the educati onal processes (e.g. pedagogies, cultures, classroom interactions, and textbooks) of private ve rsus public schools (see Peshkin, 1986; Brint et al., 2001). As well, qualitative inference about adults’ behaviors may be considered.


Democratic education across school types 4 questionnaire, although each individual respondent was asked only five of them (so that parents and children in the same households did not receive the same questions); so the index is therefore a score out of 5. Political Tolerance is an index based on answers to two questions: (1) If a person wanted to make a speech in your community against churches and religion, should he or she be allowed to speak? (2) 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? Responses to these questions are coded so that a “tolerant” response equals one [that is, yes to (1) and no to (2)] and an “intolerant” response equal zero. The responses were then added to together to produce a two-point scale, although the responses to ea ch individual question are also investigated directly. Initially, we re-estimate the relationships re ported by Campbell (1998), as nearly as the dataset will allow. We then supplement this analysis by examining issues of internal and external validity. Results Tables 1-6 report the differences across school types. Raw differences are reported, along with predicted values for each school type ba sed on probit or ordered probit estimation (see Notes to Table 1). Table 1 shows that the incidence of Community Service is considerably higher in the private sector, religious or secular, even when a large set of statistical controls are included. Table 2 shows that Civic Skills vary somewhat across school typ e, although only the Catholic school students report a statistically signific ant advantage over the other types. However, the ordered probit estimation has extremely low po wer and predicts the level of civic skills very poorly (see Row 2 of Table 2). These re sults and frequencies are almost exactly the same as those of Campbell (1998, Table 12-1).2 Table 1 School Type and Percentage of Student s Participating in Community Service Type of School Assigned Magnet Religious, Private Community Service public public Catholic nonCatholic secularWithout statistical controls 50 51 73** 72** 70** With statistical controls 50 50 75** 70** 70** N 3509 634 207 126 138 Notes: ** statistically significant difference from assigned public mean. Weighted data. Statistical controls are: age; gender; race (4 dummy variables); academic performance; college expectations; interest in the 2 Campbell controls for whether the school mand ates community servi ce, and the extent of the civic engagement of the parents. Unfortunately, there is limited information on the characteristics of the schools these students attend.


Education Policy Analysis Archives Vol. 12 No. 43 5 news; hours worked at part-time job; parental education level (4); household income (10); two-parent household; >20% black in local community; school size (4); students’ opinions matter (1); school courses require attention to politics; student government at school. For similar approach, see Campbell (1998, Appendix). Table 2 School Type and Civic Skills Index Type of School AssignedMagnet Religious, Private Civic Skills Index (0-3) public public Catholic nonCatholic secularWithout statistical controls 1.74 1.83 1.90** 1.63 1.81 With statistical controls 0.46 0.54 0.75** 0.36 0.58 N 3509 634 207 126 138 Note: See Table 1. Table 3 shows few differences between the unadjusted Civic Confidence levels across school types. When statistical controls are a dded, the religious, non-Catholic and private secular school types appear to promote more ci vic confidence than assigned public schools. In this case, Campbell (1998, Table 12-3) finds that all private schools report higher levels of civic confidence. Table 3 School Type and Civic Confidence Index Type of School AssignedMagnet Religious, Private Civic Confidence (02) public public Catholic nonCatholic secularWithout statistical controls 1.73 1.75 1.77 1.76 1.86 With statistical controls 1.75 1.85 1.81 1.95** 2.29** N 1909 384 108 68 72 Note: See Table 1. Tables 4 and 5 report on Political Knowledge and Political Tolerance Unadjusted means show that political knowledge is higher in the private sector; this advantage is reduced when statistical controls are added, such that onl y the secular schools convey an advantage. Tolerance is greater in Catholic schools, but a ppears to be lower in other religious schools. This last result is also found by Campbell (1998, Table 12-5).


Democratic education across school types 6 Table 4 School Type and Political Knowledge Index Type of School Religious, Political knowledge index (0-5) Assigned public Magnet public Catholic nonCatholic Private secularWithout statistical controls 1.81 1.77 2.23** 2.45** 2.88** With statistical controls 2.27 2.22 2.51 2.52 3.00** N 3230 578 223 80 102 Note: See Table 1. Table 5 School Type and Political Tolerance Index Type of School Religious, Political Tolerance Index (0-2) Assigned public Magnet public Catholic nonCatholic Private secular Without statistical controls 1.42 1.44 1.62** 1.26** 1.57** With statistical controls 1.37 1.48 1.59** 1.14** 1.51** N 1909 384 108 68 72 Note: See Table 1. Finally, Table 6 shows the results for allowing unpopular books and speaking out against religion. Here, Catholic schools a ppear the most tolerant, followed by secular schools; the outlier school type is the religio us non-Catholic schools, where unpopular books receive less than majority support.


Education Policy Analysis Archives Vol. 12 No. 43 7 Table 6 School type and Individual Political Tolerance Items Type of School Religious, Tolerance Item Assigned public Magnet public Catholic nonCatholic Private secular Allow unpopular book Without statistical controls 55 59 67** 44** 71** With statistical controls 54 60 59 44** 66** Speak against religion Without statistical controls 88 86 95** 82 86 With statistical controls 87 87 94** 86 82 N 1909 384 108 68 72 Note: See Table 1. Table 7 reviews these 1999 findings, and gives an overall comparison to Campbell’s 1996 results which are in square brackets where they differ. Given the substantial increase in educational privatization even over the short period of the 1990s, the robustness of these results is of interest. As with Campbell, we find little difference between the assigned and the magnet public schools. We find moderately positive results for Catholic school students, weaker than Campbell’s uniformly positive eviden ce. For schools of other religions, there is more community service, but otherwise little difference to public schools. Finally, the independent schools report somewhat more civic education, for three of the five items. Overall, the results are consistent with those of Campbell, lending some credence to the idea that privately-run schools do foster civic educ ation to a greater extent than publicly-run schools. Table 7 School Type and Five Facets of Civic Education Relative to Assigned Public School Facet of civil education Magnet public Catholic Religious, nonCatholic Private secular Community Service .. +ve +ve [..] +ve [..] Civic Skills .. +ve .. .. Civic confidence .. .. [+ve] .. [+ve] +ve Political knowledge .. .. [+ve] .. +ve [..] Political tolerance +ve [..] .. [+ve] .. [-ve] .. [+ve]


Democratic education across school types 8 Notes: Results are from probit and ordered probit estimations (details available from author). +ve or –ve indicates that there was a statistically significant difference (p<0.10) from the assigned public school category; “..” indicates no statistically significant difference. Where they differ, Campbell’s results (1998, Table 12-7) are given in square brackets. Sensitivity Analysis Internal Validity The results from NHES1999 affirm extant eviden ce, albeit where a similar method has been applied. Here, we test for the robustness of these results. We test for the correlation between the measures of civic education, an d consider alternative measures which are available from the NHES. Next, we test for omitted variable bias and for whether the model is sensitive to outliers and sample weights. Construct Validity for Civic Education It is possible that all these measures do not re present discrete behaviors, but a single measure of “civic education.” One might expect ci vic skills to promote civic confidence, for example. All the six measures are highly co rrelated (p<0.02), and if they are measuring the same behavior then a statistically significant co efficient may arise because of sampling error. However, applying principal components factor an alysis yields only two factors with Eigen values above zero, but both are below 1 (h igh weights are on the community service and civic variables for one factor and political k nowledge and tolerance for another factor). Thus, it is plausible to assume these are discrete behaviors. Of course, this correlational analysis may simply capture the simultaneous determination of school type and civic cohesion: parents who want their children to be active citizens may choose particular schools. The models here do control for possible sources of endogeneity, but we are also able to look at whether the schools compel students to engage in civic behaviors. Such compulsion mitigates in part the charge of endogeneity: families choose schools for many reasons, and may be unlikely to switch school just because the child is compelled to undertake community se rvice, for example. (Another approach to address endogeneity is taken below). Many private and religious schools require community service, and this strongly influences the results in Table 1. When community service is narrowed to only the youth who report regular service (23% of the sample), differences across school types are stronger: students from all types of private school report more regular community service than students in public schools. But, when the sample is restricted to only those who attend schools where community service is non-mandatory at the school-level, students in assigned public schools perform more service, relative to private independent, Catholic, and magnet schools. Further inspection reveals that this result is non-robust depending on the terms under which the community service is conducted. When it is non-mandatory for the individual student, no differences in school type are evident; although when the community service is not for credit, then all types of private school report higher levels.


Education Policy Analysis Archives Vol. 12 No. 43 9 Overall, from the set of social cohesi on variables available in the NHES99, it is unlikely that the results can be overturned by the charge of construct invalidity, although they may be somewhat undermined. Model Misspecification The model determining social cohesion ma y be misspecified, as may the estimation procedure. Three approaches – in relation to family income/resources, to parental characteristics, and to race – are considered. Also, sensitivity to outliers are considered. Social cohesion is likely to be higher with family income: volunteering one’s time (rather than studying or taking leisure) is probably a luxury good. In these data, civic education and household income are positive correlated, but weakly. Moreover, when the estimation is split into high-income and low-in come families, the Catholic school effects are evident only for the low-income familes. (Thi s finding of greater benefits for low-income groups is common, see Howell and Peterson, 2002). For the student, the opportunity cost to volunteering time is either leisure or pa y. However, there is a positive relationship between the student working at a job for pay within the school year and whether the student does community service and has high civic skills If there is an opportunity cost to civic engagement, therefore, it appears to be in terms of leisure foregone. Parental characteristics influence the leve ls of social cohesion reported by youth. There is a positive correlation between community service and family activities (visiting the library, an art gallery, museum or historical si te), and the family-level variables are strong predictors of the behaviors reported by th e youth. However, controlling for parental influences on civic education with a binary variable where the student talks with family at least once per week about politics or national issues has no effect on the coefficients across school types; and splitting the sample by parent al education level does not materially affect the conclusions. The results are moderately sensitive to treatment of outlier responses. For two variables, political tolerance and civic confid ence, there are very small samples (less than 10%) in the bottom cells. When these two variables are collapsed into binary indicators of tolerance and confidence, however, a premium for Catholic school students emerges for students with the lowest levels of social cohesi on. Finally, when the weights are not applied, there are no material changes to the conclusions reached above. It is readily possible that the models applied above are incomplete and that the effect of private school is attributable to variables not observed in the model. One simple way to test for the size of the bias of the unobservables is to compare the coefficients of the full model with a basic model only including the school type variable. If the coefficients are not much altered when additional important and observable variables are included, it is unlikely that the bias from unobservable variables will be significant. In fact, in this basic model the coefficients are around 25% larger than for the full model possible with these data. The bias from the unobservable characteristics would have to be of equivalent magnitude for the results presented here to be materially changed. In conclusion, therefore, there is certa inly no evidence that private schools generate less of these socially cohesive attributes th an public schools, and reasonable evidence of more social cohesion. External Validity We now turn to the substantive interpretation of school type on social cohesion. To begin, we should caution that the overall fit of the model is not very strong: more than 50% of the


Democratic education across school types 10 variation in civic activities is unexplained by the model. Nevertheless, the coefficients on school type appear to be substantively signific ant, relative to other possible changes in household characteristics. For example, switching from a public school to a private independent school has an equivalent eff ect on community service as would moving up either two deciles in household income or one quintile in parental education levels. The effects from switching to a Catholic private sc hool are even greater. Given that it may be easier to liberalize the school market than to increase household income, there may be public policy implications from these findings. However, there are several cautions about drawing straightforward policy conclusions from such analysis. First, outcomes-based analysis cannot explain why there are differences between school types in the amounts of social cohesion produced. (It is not possible to identify whether the survey data used here suffers from a bias caused by the desire to give socially acceptable answers to questions about civic atti tudes). Brint et al. (2001) argue that, in producing socialized students, (primary) sc hools differ in terms of: their organizational priorities (“maintenance of order, the minimiza tion of trouble, the encouragement of work effort, and the promotion of a sense of iden tification with the school by all students”, p.174); as well as “value messages originatin g in the broader society that are expressed primarily through the subject matter curricu lum and through the routine practices of everyday classroom life” (p.174). The first of these – organizational priorities – may be cultivated by different public school manage ment. But the latter depends more on the characteristics of local communities, and thes e are unlikely to be ameliorated simply by enrollment in private schools. As well, these in dividual-level measures of social cohesion may fail to identify the social nature of education insofar as it relates to the quality of interactions between individuals within social groups. Second, these outcome measures are not di rectly related to the curriculum that may be taught in public versus private schools. By definition, religious schools will be offering faith-based education, and the influence of this instruction on social cohesion is not easily identified in the constructs available in th e NHES. Private religious schools may influence the very knowledge that students acquire, by, for example, integrating religious and nonreligious materials into the curriculum; or social cohesion may be very subtlely introduced, through presenting one side of an argument as fact, or by including descriptive terms that disparage other (religious) groups, cr eating a sense of social invidiousness. To some, the very fact that religious education is being sanctioned by the state is an indicator that social cohesion has been impaired. Third, it is unknown whether expansion of private schooling would produce proportionately equivalent social cohesion. So, the student induced into the marginal private school may receive education of different quality to that currently received by students in the private sector. Or, the marginal student may not be equivalent to the average private school attendee. Conclusion The privatization of American education is on-going, and yet the full social and moral implications of this change are only just bein g identified (Wolfe, 2003). Private schools may be more effective at raising attainment, as most reviews suggest, but this is not the only goal of a publicly funded education system, and ther e is some expectation that private choices will


Education Policy Analysis Archives Vol. 12 No. 43 11 lead to socially undesirable outcomes. However when the actual behaviors of private school students are compared with public school students we can identify reasonable confidence that the propensity to undertake civic activities is not diminished. Notwithstanding, there are two main concer ns with such conclusions. The first is that there is little evidence as to what determi nes social cohesion (or the breakdown of social order). Without a theory of determination, it is difficult to predict how schools can have influence. The second concern is that social cohesion itself has not been measured properly. To alleviate this concern, further testing of alternative constructs in other datasets is necessary. References Brint, S., Contreras, M. F., and M. T. Matth ews. (2001). Socialization messages in primary schools: An organizational analysis. Sociology of Education 74 157-180. Campbell, D. E. (1998). Making democratic ed ucation work. In P. E. Peterson and D. E. Campbell (Eds). Charters, Vouchers and Public Education. Washington: Brookings Institution Press. Carnegie Corporation and CIRCLE. (2003). The Civic Mission of Schools Cobb, C. D. and G. V Glass. (1999). Ethnic segregation in Arizona charter schools. Education Analysis Policy Archives 7 1. Cuban, L. and D. Shipps. (2000). Reconstructing the Common Good in Education Stanford, CA: Stanford University Press. Dee, T. (2003). Are there civic returns to education? NBER Working Paper, w9588 Fairlie, R. W. and A. M. Resch. (2002). Is there “white flight” into private schools? Evidence from the National Educational Longitudinal Survey. Review of Economics and Statistics 84 21-33. Figlio, D. and J. Ludwig. (200 1). Sex, drugs and Catholic schools: Private schooling and adolescent behaviors. NCSPE Working Paper #30. Godwin, K., Z. Deng, V. Martinez, P. Wolf, & S. Wood. (2001). Comparing Tolerance in Public, Private, and Evangelical Schools. Denton, TX: Department of Political Science, North Texas State University. Goodlad, J. I. (1997). In Praise of Education New York: Teachers College Press. Greene, J. P. (1998). Civic Values in Public and Private Schools. In PE Peterson & BC Hassel (Eds). Learning from School Choice Washington, D.C.: The Brookings Institution. Guttman, A. (1986). Democratic Education. Princeton, NJ: Princeton University Press. Hirsch, E. D. Jr. (1987). Cultural Literacy: What Every American Needs to Know New York: Houghton-Mifflin. Howell, P. G. and P. E. Peterson. (2002). The Education Gap. Washington: Brookings Institution. Levin, H. M. (2002). Privatizing Education. Can the Marketplace Deliver Freedom of Choice, Efficiency, Equity and Social Cohesion. Boulder, CO: Westview Press. Martinez, V., Godwin, K. and F. Kemerer. (1 996). Public school choice in San Antonio: Who chooses and with what effects? In B. Fuller and R. Elmore (Eds). (1996). Who Chooses? Who Loses? Culture, Institutions, an d the Unequal Effects of School Choice. New York: Teachers College Press.


Democratic education across school types 12 Mocan, H. N., Scafidi, B. and E. Tekin. (2 002). Catholic schools and bad behavior. IZA Discussion Paper, 599. Peshkin, A. (1986). God’s Choice: The Total World of a Fundamentalist Christian School. Chicago: University of Chicago Press. Ravitch, D. and J. P. Viteritti (Ed). (2001). Making Good Citizens: Education and Civil Society. New Haven, CT: Yale University Press. Ryan, J. and M. Heise. (2002). The political economy of school choice. Yale Law Journal 2043. Schneider, M., Teske, P. and M. Marschall. (2000). Choosing Schools. Cons umer Choice and the Quality of American Schools Princeton, NJ: Princeton University Press. Smith, K. B. (2003). The Ideology of Education. The Commo nwealth, the Market, and America’s Schools State University of New York Press: Buffalo, NY. Soder, R (Ed). (1996). Democracy, Education, and the Schools San Francisco: Jossey-Bass Publishers. Wolfe, A. (2003). School Choice. The Moral Dimension. Princeton: Princeton University Press. About the Author Clive R. Belfield National Center for the Study of Privatization in Education, Box 181, Teachers College, Columbia University, Thompson Hall, W. 120th St., New York 10027 Dr. Clive Belfield is the Associate Directo r of the National Center for the Study of Privatization in Education, at Teachers Colle ge, Columbia University. His research is on the economics of education, vouchers, and cost-effectiveness.


Education Policy Analysis Archives Vol. 12 No. 43 13 Education Policy Analysis Archives http:// Editor: Gene V Glass, Arizona State University Production Assistant: Chris Mu rrell, Arizona State University General questions about appropriateness of topics or particular articles may be addressed to the Editor, Gene V Glass, or reach him at College of Education, Arizona State University, Tempe, AZ 85287-2411. The Commentary Editor is Casey D. Cobb: EPAA Editorial Board Michael W. Apple University of Wisconsin David C. Berliner Arizona State University Greg Camilli Rutgers University Linda Darling-Hammond Stanford University Sherman Dorn University of South Florida Mark E. Fetler California Commission on Teacher Credentialing Gustavo E. Fischman Arizona State Univeristy Richard Garlikov Birmingham, Alabama Thomas F. Green Syracuse University Aimee Howley Ohio University Craig B. Howley Appalachia Educational Laboratory William Hunter University of Ontario Institute of Technology Patricia Fey Jarvis Seattle, Washington Daniel Kalls Ume University Benjamin Levin University of Manitoba Thomas Mauhs-Pugh Green Mountain College Les McLean University of Toronto Heinrich Mintrop University of California, Los Angeles Michele Moses Arizona State University Gary Orfield Harvard University Anthony G. Rud Jr. Purdue University Jay Paredes Scribner University of Missouri Michael Scriven University of Auckland Lorrie A. Shepard University of Colorado, Boulder Robert E. Stake University of Illinois—UC Kevin Welner University of Colorado, Boulder Terrence G. Wiley Arizona State University John Willinsky University of British Columbia


Democratic education across school types 14 EPAA Spanish & Portuguese Language Editorial Board Associate Editors Gustavo E. Fischman Arizona State University & Pablo Gentili Laboratrio de Polticas Pblicas Universidade do Estado do Rio de Janeiro Founding Associate Editor for Spanish Language (1998—2003) Roberto Rodrguez Gmez Universidad Nacional Autnoma de Mxico Argentina Alejandra Birgin Ministerio de Educacin, Argentina Mnica Pini Universidad Nacional de San Martin, Argentina Mariano Narodowski Universidad Torcuato Di Tella, Argentina Daniel Suarez Laboratorio de Politicas Publicas-Universidad de Buenos Aires, Argentina Marcela Mollis (1998—2003) Universidad de Buenos Aires Ana Ins Heras Monner Sans Universidad Nacional de Jujuy Jos Luis Bernal Agudo Universidad de Zaragoza Carlos Mora-Ninci Universidad Nacional de Crdoba Brasil Gaudncio Frigotto Professor da Faculdade de Educao e do Programa de Ps-Graduao em Educao da Universidade Federal Fluminense, Brasil Vanilda Paiva Lilian do Valle Universidade Estadual do Rio de Janeiro, Brasil Romualdo Portella do Oliveira Universidade de So Paulo, Brasil Roberto Leher Universidade Estadual do Rio de Janeiro, Brasil Dalila Andrade de Oliveira Universidade Federal de Minas Ge rais, Belo Horizonte, Brasil Nilma Limo Gomes Universidade Federal de Mina s Gerais, Belo Horizonte


Education Policy Analysis Archives Vol. 12 No. 43 15 Iolanda de Oliveira Faculdade de Educao da Universi dade Federal Fluminense, Brasil Walter Kohan Universidade Estadual do Rio de Janeiro, Brasil Mara Beatriz Luce (1998—2003) Universidad Federal de Ri o Grande do Sul-UFRGS Simon Schwartzman (1998—2003) American Institutes for Resesarch–Brazil Canad Daniel Schugurensky Ontario Institute for Studies in Education, University of Toronto, Canada Chile Claudio Almonacid Avila Universidad Metropolitana de Cien cias de la Educacin, Chile Mara Loreto Egaa Programa Interdisciplinario de Invest igacin en Educacin (PIIE), Chile Colombia Jorge Ossa Universidad de Antioquia Espaa Jos Gimeno Sacristn Catedratico en el Departamento de Di dctica y Organizacin Escolar de la Universidad de Valencia, Espaa Mariano Fernndez Enguita Catedrtico de Sociologa en la Un iversidad de Salamanca. Espaa Miguel Pereira Catedratico Universidad de Granada, Espaa Jurjo Torres Santom Universidad de A Corua Angel Ignacio Prez Gmez Universidad de Mlaga J. Flix Angulo Rasco (1998—2003) Universidad de Cdiz Jos Contreras Domingo (1998—2003) Universitat de Barcelona Mxico Hugo Aboites Universidad Autnoma Metropo litana-Xochimilco, Mxico


Democratic education across school types 16 Susan Street Centro de Investigaciones y Estudios Superiores en Antropologia Social Occidente, Guadalajara, Mxico Adrin Acosta Universidad de Guadalajara Teresa Bracho Centro de Investigacin y Docencia Econmica-CIDE Alejandro Canales Universidad Nacional Autnoma de Mxico Rollin Kent Universidad Autnoma de Puebla. Puebla, Mxico Javier Mendoza Rojas (1998—2003) Universidad Nacional Autnoma de Mxico Humberto Muoz Garca (1998—2003) Universidad Nacional Autnoma de Mxico Per Sigfredo Chiroque Instituto de Pedagoga Popular, Per Grover Pango Coordinador General del Foro Latinoamericano de Polticas Educativas, Per Portugal Antonio Teodoro Director da Licenciatura de Cincias da Educao e do Mest rado Universidade Lusfona de Humanidades e Tecnologias, Lisboa, Portugal USA Pia Lindquist Wong California State University, Sacramento, California Nelly P. Stromquist University of Southern Califo rnia, Los Angeles, California Diana Rhoten Social Science Research Coun cil, New York, New York Daniel C. Levy University at Albany, SUNY, Albany, New York Ursula Casanova Arizona State University, Tempe, Arizona Erwin Epstein Loyola University, Chicago, Illinois Carlos A. Torres University of California, Los Angeles Josu Gonzlez (1998—2003) Arizona State University, Tempe, Arizona