September 1999 | EPI Book
Can Public Schools Learn From Private Schools?
Case Studies in the Public & Private Nonprofit Sectors
by Richard Rothstein, Martin Carnoy, and Luis Benveniste
Read the Introduction
SOCIAL, ECONOMIC MAKE-UP OF COMMUNITY, NOT CHARTER, IS KEY DIFFERENCE BETWEEN PRIVATE, PUBLIC ELEMENTARY SCHOOLS
Case study findings challenge conventional wisdom that private school practices are superior to public school practices
Washington, D.C. – Current policy debates about education are missing the most important reform issues by focusing on school quality based on its public or private charter, according to a report released today by the Economic Policy Institute (EPI). The most significant variations in schools are the social, cultural, and economic differences between communities, not between public and private schools in the same community, the report finds.
In Can Public Schools Learn From Private Schools? (co-published with The Aspen Institute’s Nonprofit Sector Research Fund), EPI Research Associate Richard Rothstein, Stanford University Professor Martin Carnoy, and Luis Benveniste of the World Bank, report on case studies of eight public and eight private elementary schools in California to determine whether there are any identifiable and transferable private school practices that public schools can adopt to improve student outcomes.
The private school practices examined in the report include accountability to parents, outcome expectations, clarity of emphasis on both academic and moral objectives, and teacher selection and retention policies. The report finds that the inner-city private schools shared more in common with public schools in low-income communities than with affluent suburban private schools. Suburban public schools shared more characteristics with suburban private schools than with urban public schools.
Despite commonly held generalizations about the superiority of nonprofit, private school practices, the authors made the following discoveries among public and private schools surveyed within the same community:
· Private elementary school personnel are not necessarily more accountable to parents than are public elementary school personnel.
· Private school outcome expectations for students are not more clearly defined than are public school outcome expectations.
· Private elementary schools do not necessarily aim to produce higher nonachievement outputs – behavior and values, for example – than do public elementary schools. Moreover, private schools do not always allocate a higher proportion of resources to these nonacademic objectives than do public schools.
· The report found no school, public or private, where formal evaluation, supervision, or mentoring of teachers was a meaningful indicator of variation.
· Private school innovations do not in every case stimulate improved practices at the public schools with which they compete.
These observations could have important implications for those who champion “choice” in public education as the basis for improving academic achievement. The authors find that a much greater complexity of factors must be considered when developing “fixes” for our nation’s schools.
“We found both public and private schools that had greater or lesser parent accountability, more or less well-defined expectations and student outcome goals, more or less emphasis on nonacademic goals, traditional or less traditional curricular materials, more or less rigorous teacher selection and retention policies,” according to the report. “Schools learn from each other, whether they are public or private.”
The Economic Policy Institute is a nonprofit, non-partisan think tank founded in 1986.
Richard Rothstein is a research associate of the Economic Policy Institute, a contributing editor of The American Prospect, and an adjunct professor of public policy at Occidental College. Recent publications include The Way We Were? The Myths and Realities of America’s Student Achievement, and Where’s the Money Going? Changes in the Level and Composition of Education Spending 1991-96.
Martin Carnoy is professor of education and economics at Stanford University. In the area of education, he is author of Schooling and Work in the Democratic State.
Luis Benveniste is an education specialist at the World Bank. He has conducted research on formal and informal school accountability mechanisms in public and private schools.
Can Public Schools Learn From Private Schools? is co-published by EPI and The Aspen Institute’s Nonprofit Sector Research Fund. The full-text of the report is available online at The Aspen Institute’s Nonprofit Sector Research Fund. Call 1-800-EPI-4844 to order copies at $13.95 each (plus shipping).
TABLE OF CONTENTS
Chapter 1: Accountability to parents
Resistance to parents at public Tatuna Point,
Olympic Charter, and Adams Charter Middle School
Resistance to parents at private Shalom Ieladim Jewish School
Structural limits to parent accountability:
the Liniers French School’s national curriculum and
proprietary ownership at Knuckleborough Private School
Mandating participation at parochial schools
Begging lower-income parents
to participate in public and private schools
Cases of formal accountability to parents
Accountability to parents: some observations
Chapter 2: Clarity of goals and expectations
The religious character of parochial schools
Broader educational goals vs. testable outcomes
Anchoring expectations in scripture:
a school falls apart
Clarity of goals: some observations
Chapter 3: Behavioral and value objectives
Chapter 4: Clear standards for teacher selection and retention
Hiring standards and teacher quality
Formal teacher evaluation
Informal teacher evaluation
Teacher retention and dismissal
Selection and retention: some observations
Chapter 5: Similarity of curriculum materials
Formal curricular similarities
Similarities in curricula, differences in pedagogical skill
Chapter 6: Competitive improvements
Chapter 7: Conclusions
Appendix A: Case study descriptions
About the authors
About EPI / The Nonprofit Sector Research Fund
Because private schools can select (and are selected by) their students, analysts have not been able to determine whether private schools’ apparently superior outcomes (like test scores) are attributable to superior private school practices or to more selective student bodies. If superior
private school practices are responsible, these practices are widely believed to include more accountability to parents, more clearly defined outcome expectations, a greater clarity of emphasis on both academic and behavioral/value objectives, and more efficient teacher selection and retention policies. Private schools mostly utilize the same curricular materials as public schools, and so private schools’ apparently superior outcomes should be able to inspire competing public schools to improve by imitating private school practices.
To explore these widespread beliefs about differences between private and public school practices, we conducted case studies of eight public and eight private elementary schools in California. We conducted extensive interviews with principals, other administrators, teachers, parents, and (in the case of some eighth graders) students. The schools were selected to typify a range of socioeconomic characteristics and included both sectarian and nonsectarian private schools. However, we make no claim that the sample of schools was scientifically chosen, or that the impressions we record here are statistically valid. The depth with which we probed each of these schools precluded a large enough sample to generate statistical conclusions.
Nonetheless, insights from these case studies tend to challenge widely held assumptions about differences between public and private schools. Inner-city private schools shared more characteristics with public schools in low-income communities than with affluent suburban private schools. Likewise, suburban public schools had more in common with suburban private schools than with urban public schools. Thus, policy debates about education may be missing the most important issues if they focus on whether a school’s quality can be deduced from its public or private charter. The most important variations between schools may be between schools of all types in different communities, not between public and private schools in the same community.
Among the 16 schools in this sample, private schools were not noticeably more accountable to parents than public schools. In low-income schools, public and private, teachers and administrators complained of the lack of parental involvement. Both public and private school teachers in these low-income communities felt challenged to involve parents even in a minimal way in their children’s education (for example, by supervising homework or attending school meetings). And in no sense did we observe either public or private schools in low-income communities acting as though they were “accountable” to parents. But an opposite phenomenon characterized both public and private schools in very affluent communities: here, staff in both types of schools complained of too much parent involvement – including interference in the daily curriculum and inappropriate challenges to school goals.
Policy makers often posit that private schools are more successfully organized around academic achievement objectives and are more successful in emphasizing behavioral goals. These case studies, however, include private schools organized around principles other than academic outcomes, such as religious beliefs, safety, or discipline; in some cases, academic achievement was a relatively low priority. The studies also found some public schools that were as successful as private ones in aligning themselves with academic goals, and some public schools that also emphasized behavioral or value objectives.
Many believe that an important public-private difference is laxity of teacher standards stimulated by public employee protections and unionization. Yet these case studies found no school, public or private, where formal evaluation, supervision, or mentoring of teachers was meaningful. Indeed, Catholic school procedures for terminating poorly performing teachers were nearly as cumbersome as public school procedures. Moreover, private schools in this sample were no more selective in teacher personnel policies than were public schools serving students of similar socioeconomic backgrounds. We observed both high- and low-quality classroom management and academic instruction in both public and private schools.
The social, cultural, and economic backgrounds of the parents and the community in which the school was located seemed to be the main determinant of variation, much more so than a school’s public or private character or, within the latter group, whether it was religious or secular. Within particular communities, similarities between schools and the problems they confronted overwhelmed the differences.
These cases point away from arguments that public schools can improve by adopting the greater accountability to parents and flexibility in hiring/firing teachers presumably characteristic of private schools. These may be good policies for all schools to follow, but public schools are as likely (or unlikely) to be accountable to parents as private schools serving similar student and parent populations. These observations, if confirmed by observations of a much broader group of schools, could have important implications for those who champion “choice” in public education as the basis for improving academic achievement. A much greater complexity of factors must be considered when developing “fixes” for our nation’s schools.
Private school outcomes are generally superior to public school outcomes. Private school students score higher on standardized tests than do public school students, and private school students are more likely to graduate and attend college.
But why is this the case? Do private school practices differ from practices of public schools? Or do private schools get better outcomes because they enroll students from more advantaged backgrounds? Even if private schools generate superior outcomes when student background characteristics are controlled, are these outcomes caused by practices that are transferable to public schools, or are they idiosyncratic to private schooling? For example, improved outcomes for disadvantaged students in private schools that spring from their association with more advantaged classmates would not be a transferable private school effect. Similarly, a private school advantage that derives from a sense of religious community or shared communal values may not be transferable to a secular public school.
Scholars differ on whether private schools should behave differently than public schools. Sociologists Richard Scott and John Meyer (1988) argue that the institution of public schooling has a different function than that of a private school, and so the two should have different organizational structures. A public school takes in all comers. Teachers and the school principal are accountable to multiple constituencies, represented by the varying capabilities of their students and their needs. Thus, a public school’s organization is bound to be complex and often subject to conflict. It also can appear incoherent because it tries to do so many things at once. Private schools are more likely to have a single objective, and select (and be selected by) the parents and their children who are in accord with that objective. A private school’s organization should therefore be less complex – more “aligned” around a coherent program. Sociologist Joan Talbert (1988) also argues for significant organizational differences between public, religious private, and nonreligious private schools on similar grounds.
But economist Byron Brown (1992) theorizes that private and public schools should not differ significantly. Because private and public schools and parents are all generally uncertain about children’s abilities and future employment prospects, he argues, they offer comprehensive curricula and similar teaching methods in order to reduce the uncertainty about schooling choices. If a school chooses to differ radically from other schools, it increases the risk to parents tha
t they have made the wrong choice. Whether by voice or exit, parents will push the school to behave much like other schools. Private schools tend to distinguish themselves, Brown concludes, not by innovating academically but by offering special “secondary services,” such as religious instruction, that do not affect labor market opportunities.
Much of the debate around differences between nonprofit public and private schools has revolved around statistical analyses that purport to control for student background characteristics, thus leading to the identification of a private school effect that demands explanation. Some scholars (e.g., Coleman, Hoffer, and Kilgore 1982; Peterson 1998) find that students in private schools score higher on achievement tests, even after student backgrounds are taken into account. Others (e.g., Witte 1996; Rouse 1997), however, find little or no private school effect after controlling for student background characteristics.
Because controls for background characteristics are so difficult to implement, it is unlikely that this debate will soon be resolved. Yet public policy does not wait for econometricians to achieve consensus. Controversy regarding the relative advantages of public or private schooling proceeds, and policies to shift resources from public to private schools, based on assumptions about private school superiority, gain increasing support. Ethnographic data may help shed light on this debate by giving us the opportunity to study in detail the principles and practices around which schools organize themselves.
In the pages that follow, we report on case studies of eight public and eight private elementary schools in California.1 One critique of a study that uses cases, rather than a large random sample of schools, is that it is not necessarily representative and hence cannot yield reliable conclusions regarding possible differences or similarities between private and public schools. It is true that our sample is neither random nor large, but the small number we chose to work with enabled us to observe instruction in classrooms and interview teachers, administrators, and parents. Time and resources did not permit such an observational and interview-based case study approach for more than 16 schools. And 16 is too few a number of schools to assure a representative sample had we selected it randomly. Schools also have too many varying characteristics to assure a representative sample of 16 if we had stratified all public and private schools prior to randomization.
Rather, we drew the sample to mirror the characteristics of important types of schools that might shed light on how nonprofit public and private schools differ. Private schools in our sample include both parochial and independent schools. Public schools include both neighborhood and “choice” (or charter) schools. We also targeted schools within each nonprofit sector (private or public) that served different social class populations (poor, working class, middle income, and affluent) and that varied by urbanicity and racial composition. We believe, despite the nonrepresentative character of the sample, that by visiting a reasonable number of public and private schools in two major metropolitan areas – schools catering to students of varied racial and ethnic backgrounds and from families both higher and lower in socioeconomic characteristics – we should be able to find differences. The widely repeated claim, after all, is that market-driven behavior is observably and significantly different from bureaucratically driven behavior. If that is the case, such differences should be observable in a sample of 16 schools.
Because interview respondents were assured anonymity, we use pseudonyms as school names, and we do not specify the California community in which each school is located. Appendix A provides a brief description of each school studied, along with a table that classifies each school both by the socioeconomic level of its clientele and by type, i.e., private-religious, private-nonsectarian, private-proprietary/for profit, public-neighborhood, or public-choice. The reader should refer to Appendix A when considering each of the sections below.
Our interviews were organized around a relatively simple working theory of school accountability. It assumes that schools actually have conceptions of accountability embedded in the patterns of their day-to-day operations, and that a school’s conception of accountability significantly influences how it delivers education. We expect that schools must solve problems of accountability in some way in order to function, and that how schools solve them is reflected in the way teachers, administrators, students, and parents talk about the fundamental issues of schooling. Schools are likely to have more “operative” internal accountability systems if their formal and informal mechanisms are aligned with individuals’ internalized notions of accountability (responsibility) and collective expectations of the school. In schools at the other extreme, there may be a high degree of incoherence among formal and informal mechanisms and individual notions, leading to a relatively weak or even dysfunctional internal accountability system.
We make no effort to draw conclusions from our case studies about the relative effectiveness of the schools we studied, and we did not select the schools to be statistically representative of broader school groupings. Rather, our purpose is descriptive and suggestive: we attempt to identify the practices in each of these schools that we believe may be typical of other schools; our findings could be tested in a larger sample in an effort to confirm a series of hypotheses about what public schools can learn from private schools. We did not attempt to quantify our observations or subject them to statistical analyses; our “tests” of these hypotheses in this very small sample, therefore, are based on interpretations of interviews and personal observations in classrooms and schools, not on quantitative data. This report, then, contributes to the discussion engendered by scholars like Chubb and Moe (1990) (who concluded from observational studies that private school effectiveness is the result of organizational factors – freedom from democratic/bureaucratic control) or Bryk, Lee, and Holland (1993) (who concluded from observational studies that the effectiveness of Catholic schools is the result of a shared sense of communal mission), without first proposing a statistical claim that public or private schools are superior.
Our most striking finding from these cases is that, to the extent patterns are generalizable, the most important distinctions among these schools did not separate public from private institutions. Rather, we found, for example, that private schools in inner-city communities were more similar in many respects to public schools in these communities than they were to private schools in suburban communities. Likewise, we found that the suburban public schools we observed had more in common in many respects with suburban private schools than they had with urban public schools. While these patterns were not uniform, they suggest that policy debates about education may be missing the most important issues if they focus on whether the quality of schools can be deduced from their public or private organization.
In sum, our answer to the question, “what can public schools learn from the private non-profit education sector?” is “not much.” But if phrased somewhat differently, i.e., “what can schools, public and private, learn from other schools, public and private,” our answer is “quite a bit,” depending on which schools we study and what we are attempting to learn.
We came to this conclusion by organizing our observations around the following commonly held generalizations about the superiority of nonprofit private schools:
Generalization No. 1: Private elementary school personnel tend to be more accountable to parents than are public elementary school personnel.
lization No. 2: Private school outputs and expectations for students tend to be more clearly defined than are public school outputs and expectations.
Generalization No. 3: Private elementary schools tend to produce higher nonachievement outputs – behavior and values, for example – than do public elementary schools. Moreover, private schools allocate a higher proportion of resources to these nonacademic objectives than do public schools.
Generalization No. 4: Teacher selection and retention practices at private schools are more efficient than are the selection and retention practices for teachers at public schools.
Generalization No. 5: Private schools’ academic success is accomplished with curricular materials that are not significantly different (in standard subjects) from curricular materials found in public schools.
Generalization No. 6: Private school innovations stimulate improved practices at the public schools with which they compete.
With respect to each of these generalizations about the superiority of private over public schools, our interviews and observations suggest differences between the schools we observed, but the differences did not arrange themselves along a public-private divide. Rather, we found both public and private schools that had greater or lesser parent accountability, more or less well-defined expectations and student outcome goals, more or less emphasis on nonacademic goals, traditional or less traditional curricular materials, more or less rigorous teacher selection and retention policies. We found evidence that schools learn from each other, whether they are public or private.
We discuss our observations with regard to each of these generalizations in turn.