The Ideology of Education
This page intentionally left blank.
The Ideology of Education The Commonwealth, the Marke...
7 downloads
760 Views
675KB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
The Ideology of Education
This page intentionally left blank.
The Ideology of Education The Commonwealth, the Market, and America’s Schools
Kevin B. Smith
State University of New York Press
Published by State University of New York Press, Albany © 2003 State University of New York All rights reserved Printed in the United States of America No part of this book may be used or reproduced in any manner whatsoever without written permission. No part of this book may be stored in a retrieval system or transmitted in any form or by any means including electronic, electrostatic, magnetic tape, mechanical, photocopying, recording, or otherwise without the prior permission in writing of the publisher. For information, address State University of New York Press, 90 State Street, Suite 700, Albany, NY 12207 Production by Christine L. Hamel Marketing by Anne Valentine Library of Congress Cataloging-in-Publication Data Smith, Kevin B., 1963– The ideology of education : the commonwealth, the market, and America’s schools / by Kevin B. Smith. p. cm. Includes bibliographical references and index. ISBN 0-7914-5645-5 (alk. paper) — ISBN 0-7914-5646-3 (pbk. : alk. paper) 1. Privatization in education—United States. 2. Education—Political aspects—United States. 3. Democracy—United States. I.Title. LB2806.36 .S63 2003 379.1'11—dc21 2002075761
10 9 8 7 6 5 4 3 2 1
For Catherine Susan Smith, who personifies all that a teacher should be. As long as individuals with her intelligence, professional dedication, and commitment to children are attracted to careers in public schools, public education remains something worth defending.
This page intentionally left blank.
Contents List of Tables and Figures
ix
Preface
xi
Acknowledgments
xv
Chapter One: Ideology and Education
1
Chapter Two: Education and the Economy
19
Chapter Three: Education and Equality of Opportunity
47
Chapter Four: Institutional Structure and Educational Goals
75
Chapter Five: Education and Civic Culture
101
Chapter Six: Education as Ideology
127
Notes
151
Methodological Appendix
157
References
171
Index
191
vii
This page intentionally left blank.
List of Tables and Figures Table 1.1 The Ideology of Education
14
Table 2.1 Education and Productivity in the States
36
Figure 2.1 Actual Versus Predicted Productivity
40
Table 2.2 SWAT Analysis of Education and Culture in High Productivity States
42
Table 3.1 Test Score Pass Rates In Texas School Districts
63
Table 3.2 Equity in Educational Outcomes in Texas School Districts
64
Table 3.3 SWAT Analysis of Test Score Pass Rates
69
Table 3.4 SWAT Analysis of Equity
69
Figure 3.1 Actual Versus Predicted TAAS Pass Rates
70
Figure 3.2 Actual Versus Predicted Equity
71
Table 4.1 Goals and School Type
87
Table 4.2 Marginal Effect of Institutions on Teaching Goals
93
Table 4.3 Estimated Institutional Impact on Adopting Religious/Moral Codes as Major Educational Objective
96
ix
x
List of Tables and Figures
Figure 5.1 Civic Culture and High School Graduates
107
Table 5.1 School Type and Civic Capital Correlations
114
Table 5.2 Race, Class and Type of School
116
Table 5.3 Determinants of Civic Capital
120
Figure 6.1 SAT Scores 1952–1995
130
Figure 6.2 SAT Scores 1952–1995
131
Table A.1 Estimated Institutional Impact on School Goals
164
Table A.2 Civic Capital Correlations
168
Preface Anyone reviewing the academic educational policy literature may be forgiven for concluding that, as Peter De Leon put it, the policy sciences have become “yet another vested tool for interest groups” (1997, x). These studies are routinely employed as partisan weapons in educational policy debates, and there is grumbling from the left and the right that academics are too often over-eager suppliers of political munitions letting ideological preference trump good scholarly practice (Glenn 2001, Tanner 1998, Peterson and Noyes 1997, Muir 1999). This troubling background supplied the motivation for writing this book. The scholarly policy sciences are reluctant to acknowledge that individual values and the political issue at hand are motivating forces strong enough to shape the direction of an analysis, let alone its conclusions (Myrdal 1969). There are good reasons for this firm grip on the positivist’s mantle of scientific neutrality. If academic policy analysis is really an empirical branch of ideology, it loses its role as a neutral authority in policy debates and its intellectual gyroscope. To borrow a phrase from the Nobel laureate Herbert Simon (1969), the policy sciences are sciences of the artificial because they deal not only with what is, but also what might be. This makes parsing the politics out of scientific inquiry difficult, while at the same time strengthening the case for a clear separation. The scientific method is simply a more reliable and systematic basis for understanding policy than the pressures of political expedience, the dogmas of ideology, or the inconsistent lessons of individual experience. It is to the former rather than any of the latter that academic policy analysts properly owe their first allegiance. While recognizing the normative idiosyncracies of the individual, and professing a decided preference for supporting liberal democratic values, in theory and method the policy sciences have good reasons to secure positivist foundations.
xi
xii
Preface
The problem is that in policy arenas like education reform, partisanship and ideology readily adopt the products of scientific inquiry when it suits their purpose. The arena of the debate leaves little neutral ground, even for the pure empiricist scrupulously applying best accepted theoretical and methodological practices. The awkward fact is that education reform debates are, and always have been, nakedly political struggles about who or what gets to define what we “ought” to do (Tyack and Cuban 1995). Anyone who ventures into the educational policy arena, even in the backwater confines of academic journals, rapidly learns that politics stick to education, and to publish any finding is to find oneself at least temporarily bonded—sometimes comfortably, sometimes with surprise and irritation—to a partisan cause. The more we seek to dissolve these attachments with the positivist’s solvent, claiming our conclusions simply reflect how our data reveal the world to be, the firmer the embrace of partisan agendas become. I believe that the increasingly precarious nature of nonpartisan claims in the policy sciences are, to a large extent, the fault of its practitioners. Some of this is may be due to those unscrupulously using a cloak of scholarly credibility to camouflage the advance of their political interests, as proclaimed by critics from all points on the ideological spectrum. Yet I think the larger problem is with scholars who are genuinely committed to systematically advancing knowledge. Mindful of our vulnerable positivist flanks, we maneuver around certain questions rather than assaulting them directly. Data driven, quantitative policy analysis is oriented towards isolating systematic relationships between variables, which orients policy studies towards to questions of “What works?” Such projects lend themselves to those with positivist/scientific sympathies. Yet while these questions are no doubt important, they cannot be practically separated from their political implications and the partisan plaudits and denunciations such implications unavoidably attract. I believe the more interesting questions in education policy, at least for those of us viewing schools through the disciplinary spectacles of political science, have little to do with the technical means for achieving what are (usually erroneously) assumed to be universally desirable ends. Instead of “What works?” it seems to me that the more interesting question is “Who wants this to work, and why?” This modest bit of rephrasing puts politics into the front of an inquiry rather than tolerating it as a not altogether welcome guest in the back. A fair and comprehensive reading of educational policy research during the past decade or two surely leads to the conclusion that there are consistent differences among scholars about “what works,” differences that are not going to be solved by the technical or theoretical merits of the next insight into test score
Preface
xiii
variation. Even a casual familiarity with battles over choice, charters, accountability, and other big education issues conveys the knowledge that these debates are as much about ideologically preferred ends as identifying the most pleasing set of instrumental means. Policy analysts tend to concentrate their efforts, at least nominally, on the latter issue. Yet directly addressing the former seems at least as important: What, politically, is at stake in education reform? This book reflects my attempt to grapple with this question. I wanted to scrape off the normative elements that varnish academic theory, see if they worked independently as a functional political vision of education, and assess the correlation between that vision and what could be reliably confirmed in the empirical world. While deliberately trying to lay bare the normative and ideological in the academic research, my goal here is not to denigrate positivism (nor the market, which is the main target of what follows), but to try and find its boundaries. I strongly believe the policy sciences have an appropriate role as a more or less neutral policy authority. Our job, however, should be to identify what is and what is not a political choice. It is not to pretend politics and values do not intrude upon our work or, even worse, to muddle advocacy with inference. I believe those of us who take the science in policy science seriously can directly and usefully address the key normative questions in educational policy without forfeiting our positivist sympathies. This book is a modest attempt to make that case by addressing what I see as the key political or normative question underlying most contemporary education policy controversies: Is the market a readily available solution to the problems of public education or a threat to its democratic ethos?
This page intentionally left blank.
Acknowledgments A book is rarely the product of a single person’s labor, and what follows is no exception. Numerous colleagues generously contributed their time and their thoughts to this project and rightfully share in any credit it accrues. I would like to offer special thanks to Jeff Gill at the University of Florida, who with little notice and even less compensation, made a number of key suggestions regarding methodology and presentation of empirical results. Thanks also to Ken Meier at Texas A&M for giving a very early version of the manuscript to a graduate seminar—it is invaluable to get the graduate student “tear down” before publication—and for periodically prodding me along with encouragement, threats, and a pleasant variety of stimulating libations. I am also grateful to the many who reviewed portions of the manuscript as they were presented as conference papers and offered valuable feedback. Most importantly, I would like to acknowledge the contribution of the anonymous reviewers engaged by SUNY Press. These referee reports made several key contributions to the final product. Finally, while the credit for this book is rightfully shared, I reserve for myself exclusive responsibility for any errors or criticisms.
xv
This page intentionally left blank.
1 Ideology and Education
P
ublic education and the market have both been described as secular religions. The market because it purports to possess divine attributes that, “are not always completely evident to mortals but must be trusted and affirmed by faith” (Cox 1999, 20). Public education because it embodies a “cosmological” belief among Americans that schooling offers a sure path to a better future (Cookson 1994, 83; Brogan 1962, 137). Though both represent broadly-held value systems, public education and the market are not particularly compatible creeds. Judging by the record of the past two decades, attempts to integrate the two tend to result in the sorts of controversies associated with the most bitter sectarian debates. At the heart of these controversies is a conflict between two camps. The first is a loose coalition of market theorists, business groups, religious groups, conservative think tanks, and policy advocates who see public education as seriously in trouble. They are united by a common perception that public education is rigid and bureaucratic, rule-bound and unaccountable, and mired in mediocrity. This camp views marketbased reforms of public education—policies such as choice, charters, vouchers, and outright privatization—as a ready set of solutions to clearly defined problems (Powers and Cookson 1999, Chubb and Moe 1990, Murphy and Schiller 1992, Ravitch 1997, Gerstner 1995, Hanushek et al. 1994, Moe 2001). The second is the public school “establishment,” an even looser coalition of teacher’s unions, school boards, and school administrators, progressive academics, and liberal foundations and policy advocates. They share a belief that market-based reforms misperceive the problems of education and threaten the democratic values that justify its existence (Engel 2000, Henig 1994, Cookson 1994, Smith and Meier 1995, Rebell 1998). These camps are held to be locked in a struggle for the “soul of education,” of whether schooling in America will be guided by a democratic metaphor that “leads to a belief
1
2
The Ideology of Education
in the primacy and efficacy of citizenship as a way of life,” or a market metaphor that “lead(s) to a belief in the primacy and efficacy of consumership as a way of life” (Cookson 1994, 9; Callan 1997). These differences center on the perceived motivations for pursuing market-based reforms and their expected consequences for public education. Advocates of market reforms largely reject the theological comparisons, arguing what they propose are the instrumental means to achieve universally desirable ends, not the imposition of a particular value system (Powers and Cookson 1999). This does not mean market reforms are value-neutral in specifics. Voucher systems, for example, might help some achieve value-based preferences such as a religious component to schooling. Yet these are individual choices; the values selected by one are not imposed on all. While values might drive some decisions in an education marketplace, the broader argument is that choice, competition, and deregulation promote universally beneficial characteristics like innovation, efficiency, and response to consumer preference, not a particular value agenda. If there is any attempt to institutionalize a value system in education, it exists in the status quo with its “one size fits all” bureaucracy. Market mechanisms are less insistent on uniformity and are presented as a largely apolitical management or organizational reform for the public sector that will promote more responsive, flexible, efficient, results-oriented schools (Chubb and Moe 1990, Hanushek et al. 1994, Moe 2001, Box et al. 2001). Opponents of market reforms are quicker to embrace the value-system metaphors. They see public schools as institutional manifestations of democratic values, a view anchored in the legal origins of public education: state constitutions authorize the existence of public schools in the United States and place their governance in the hands of state legislatures. State constitutions justify public education in terms of the “democratic imperative,” i.e., to serve society’s need for citizens capable of self-governance and committed to democratic values, and to uphold the collective commitment to those values by equitably distributing the educational opportunities that open the doors to social and economic opportunities (Rebell 1998). Public schools are thus seen as integral to the functioning of democratic society. Their job is to store and replicate the values of the polity, values defined broadly by state and national constitutions and in specifics by the outcomes of democratic processes. At the heart of this viewpoint, which I shall term the commonwealth perspective, is a belief that “public interest” is properly defined collectively through a democratic process, not individually through a market process. If one of these collectively determined values is church–state separation, public schools
Ideology and Education
3
should not engage in religious inculcation, regardless of the preferences of the individual. Using vouchers to support the individual preferences for religious indoctrination strips decision-making power about the appropriate role of religion in public education (and by extension, the larger sphere of public life) from democratic processes and passes it to market processes. This shift is not an apolitical decision with no repercussions beyond the individual; the market mechanism shrinks the role of representative institutions and democratic processes in shaping public life. In doing so the market threatens the broader value system that justifies public education (Callan 1997, Barber 1992). As such, the conflict is not a technical dispute about means; it is a normative and philosophical conflict about who—or what—gets to define the ends of public policy (Hartoonian 1999, Engel 2000; Witte 2000, 13). So the critical question at the heart of the conflict between the market and the commonwealth camps boils down to this: Is the market a utilitarian and instrumental response to a clearly defined set of problems, or is it an ideological agenda seeking to usurp democratic values with its own? This is a question with important implications for the purpose and operation of mass systems for public education, and raises critical issues for policy scholars struggling to separate analytical frameworks from ideological advocacy. It is the question that this book seeks to answer. The central research question posed here presents a difficult challenge for policy scholars because the central ideas of the market are formalized as positive theory (an explanation of how the world does work) and as normative theory (an explanation of how the world should work). As a positive theory, the market offers a powerful explanatory framework, an empirically verifiable means to explain and assess the cause and effect of policy. As a normative theory the market makes a persuasive case for accepting certain premises as “correct” or “just” that are validated by human fiat rather than empirically verified. In this guise it can quickly transform into an ideology, a consistent and interwoven set of values, attitudes, and beliefs about the political system and the appropriate role of government in society (Friedman 1982; Campbell et al. 1964; Schwartzmentel 1997, 2). The problem is the extent to which the positive theory needs the ideology in order to function. Disentangling one from the other, especially in the case of education, has proven to be extraordinarily difficult. A number of academics who use market frameworks at least nominally in a positive sense stand accused of pursuing ideological agendas, producing a lengthening list of studies that “more resemble manifestoes than policy memorandum” (Powers and Cookson 1999, 104; Tanner 1998; Witte 2000, 157–189; Muir 1999;
4
The Ideology of Education
Greene and Peterson 2000). This despite the fact that some studies using market frameworks support its normative dimensions (Chubb and Moe 1990) and some do not (Smith and Meier 1995). Clarifying the positive and normative dimensions of the market and sorting out its relative status as theory and ideology are prerequisites for a structured answer to the question posed here.
MARKET THEORY AND EDUCATION
As a positive theory of public-sector phenomena, the market is formalized by public choice, which is essentially the transformation of neoclassical economic theory into a theory of politics (Downs 1957, Buchanan and Tullock 1962, Olson 1965, Friedman 1982). This transformation is achieved by viewing the actions of citizens, politicians, and public servants as analogous to self-interested producers and consumers engaged in market exchange (Buchanan 1972). At the heart of publicchoice theory are two simple assumptions: (1) Individual utility maximization, the notion that an individual knows and can rank order their preferences and, given a choice, will maximize utility by taking the actions that fulfill these preferences at minimal cost. (2) Methodological individualism, the notion that only individuals make decisions. Collective decisions are seen as aggregations of individual choices, not a unique property of the group (Buchanan and Tullock 1962, 32). Put these assumptions into a market system of exchange—competition among producers, choice among consumers—and they comprise the essential characteristics of neoclassical economic thought. Public choice takes these intellectual tools and transfers them from private markets to the public sector. The justification for this transformation rests on one of the great insights of Adam Smith: self-interest can serve collective interests. In a market, businessmen might pursue nothing but profit and consumers nothing but their own satisfaction. Nonetheless, the results—cheap, widely available, high quality goods—are universally beneficial. No central authority dictates these outcomes; the collective benefit is produced by allowing self-interested actors to freely exchange goods and services as they see fit. Public-choice theory raises the possibility that market mechanisms can produce similar results for public goods and services (Tiebout 1956). Attempts to reform the public sector on this basis promote what is generically termed the “market model” of public administration. This advocates viewing citizens as consumers, treating government as a business within the public sector, and increasing efficiency
Ideology and Education
5
within that sector by having public goods and services delivered by competitive agencies (Box et al 2001, 611). Variations of public choice theory have been applied to education, most notably by Friedman (1955) and Chubb and Moe (1990; see also Schneider et al. 2001). Viewing education through the lens of public choice, these analyses conclude that public education is wasteful and autocratic because it is regulated through top-down systems of hierarchical control. Power is concentrated in bureaucracies (state education agencies, central administration in districts and schools) with monopolistic control over educational services. This translates into centralized, bureaucratic control over all fundamental aspects of the educational enterprise: enforcing the mandates of representative bodies (legislatures and school boards), evaluating educational needs, hiring and firing teachers, setting wages, planning schools, assigning students, and determining curricular content and standards (Vandenberghe 1999). According to public-choice theory, there are two key negative consequences to this institutional arrangement: (1) Inefficiency. As bureaucrats are also assumed to be self-interested, their power will be used to maximize their utility, not those of students and parents. Bureaucracies are assumed to maximize inputs, the resources (manpower, budgets, jurisdictional reach, etc.) that provide bureaucrats with income, prestige, and professional achievement (Niskanen 1994). Lacking competitive pressures, they have little incentive to focus on outputs, i.e., to deliver on the desired outcomes of the educational process. (2) Concentration of power in the bureaucracy. Parents and students have limited recourse with the educational bureaucracy. They are free to become involved in the democratic process that formally determines the goals and the budgets of public education, but get no guarantees that their preferences will be acted upon. They are politically weak compared to powerful organized interests (teachers’ unions, textbook publishers, etc.) who prefer the status quo. So students get trapped in poor schools, innovation and progress are discouraged, and high levels of inefficiency are tolerated because the institutional structure provides little incentive to change. Though the empirical evidence supporting these theoretical assumptions is mixed (Chubb and Moe 1990, Smith and Meier 1995, Schneider et al. 2001), the prescriptive solution they imply is politically popular. This solution is the imposition of the market model through policies that introduce market mechanisms (competition among producers and choice among consumers) to education via policies like vouchers, charter schools, school choice, or even outright privatization. The objective is to shift power from the bureaucracy to the consumer,
6
The Ideology of Education
and to force education service providers—schools—to respond to the demands of their primary clientele. Under conditions of competition and choice, schools are forced to respond to these preferences instead of the preferences of legislatures and their bureaucratic management mechanisms. This is expected to promote innovation and efficiency in education providers, and provide freedom of choice for consumers. The overall result is predicted to be a wide variety of high quality schools that are output, not just input, oriented. High quality services efficiently produced, in other words, collective benefits produced by harnessing the power of self-interested action through the power of market mechanisms. Even its advocates accept that the market model has limits. Left purely to market forces public education would cease to be public and be available only to those who could afford it. The market model retains government financing to combat this underinvestment (though in many guises allowing such support to flow to private schools), and recognizes the need for some central regulatory mechanisms to promote accountability and ensure health, safety, and minimal performance standards (Lamdin and Mintrom 1997). The goal is not to eliminate the government entirely from education, but to leverage the power of the market to produce better quality public goods and services, increase citizen/consumer satisfaction, and to do so at a reasonable cost. Given these objectives, public choice and the market model embrace what sociologists term a “functionalist” vision of schools. This is the idea that schools impart technical knowledge (a service to be consumed), and the form and specifics of this knowledge should be determined by the social and economic needs for particular skills (the preferences of clientele or customers; see Clark 1962). Functionalism argues the mission and organizational structure of schools should be driven by the task environment of education (contemporary needs for particular economic and social skills), not derived from ideological agendas. Public choice and the market model follow this instrumental and utilitarian argument in the sense that, whatever the general quality of public schooling is, there is a widely held perception that it is not as good as it should be. This broadly recognized problem has collective consequences, especially as a drag on economic advance (Rothman 1991, Verstegen and King 1998, Berliner and Biddle 1995, Loveless 1997). Public-choice theory offers an explanation for the underperformance of public education: its institutional arrangements promote inefficiency and offer no incentive to respond to demands to do better. The market model provides a well-developed solution to this problem— replace the system of hierarchical control with market mechanisms.
Ideology and Education
7
Thus theory and prescription can be claimed to be apolitical—the first explains a known and widely accepted problem, the second proposes a solution logically derived from this explanation.
MARKET IDEOLOGY AND EDUCATION
Public choice is not only employed as a positive theory but also as a normative theory. Compared to the orthodox model of public administration, which calls for public services to be delivered by bureaucracies in centralized jurisdictions, public-choice theory and the market model are argued to more closely realize the ideal of governance envisioned by writers such as James Madison (Ostrom 1973). Madison argued for a republican system where power was decentralized, and groups (factions in Madison’s terminology) were free to pursue their own interests within this fragmented political structure. Advocates of public choice as a normative theory see a strong parallel between these arguments justifying the American political system and the market model, which allows consumers to pursue their interests by choosing among competitive publicservice providers (Buchanan and Tullock 1962, 64). The market model is thus not only compatible with democratic ideals, it is also a much closer institutional realization of those ideals than the orthodox hierarchical control systems that characterize public education. These characterizations of public-choice theory and the market model are widely criticized. The primary thrust of these criticisms is that the market principles embodied in public choice are not only incompatible with democratic values and processes, but they are also fundamentally hostile to them. Markets favor efficiency and productivity over equality and representativeness, and by taking their cues from individual self-interest they can harm as well as advance collective interests. For example, compulsory universal education and special education programs are, from a typical market cost–benefit analysis, inherently inefficient propositions (Finn 1996, Hofstadter 1962). From a democratic perspective, the former is justified by the needs of the state, the latter on the rights of the individual and, because both are embodied in constitution and law, efficiency is a secondary concern. Schools are charged with upholding these values even though resources will be “wasted” in doing so. Imposing the central market criteria of “good” decisions and outcomes—efficient production and allocation—could conceivably contradict these fundamental purposes of public education. Racial segregation supplies the most well-known example of how individual interest can conflict with collective interest. After a series of
8
The Ideology of Education
political battles, de jure racial segregation in public education was ultimately banned by the Supreme Court in Brown v. Board of Education. Public education’s response to this ruling was initially uneven. Over the long term, it realigned itself with the ruling as the definitive interpretation of the Constitution, though achieving this realignment sometimes required the full coercive powers of the state and did little to address de facto segregation. One of the responses to Brown was the first systematic efforts to introduce market mechanisms into public education. This was undertaken by Southern whites frustrated with the public system’s increasing resistance to demands for segregation (Henig 1994, 102–105). They recognized that a market mechanism would respond to this demand even if it was not only incompatible with, but in direct contrast to, the constitutional ruling. This suggests that left to themselves, markets will efficiently allocate services in response to such demands, even if doing so is inequitable and abrogates collective democratic values (McCabe et al. 1999). The basic criticism is that while the market sees demand and customers, it is uncomfortable with the concepts of citizenship and community (Box et al. 2001). As such concepts are central to most notions of democracy, making market values appear comparable to, much less synonymous with, democratic values is misleading. On such grounds, the normative theory of the market is viewed as providing academic cover for an ideology hostile to democratic values at the heart of the ethos of public education (Engel 2000, McCabe et al. 1999). So, rather than an instrumental means to improve the underperformance of education, or as method to better align public education with democratic ideals, the ideology of the market is seen as a direct threat to the historical and legal justifications for education, i.e., storing and replicating the collective values that uphold a democratic society (Engel 2000, Callan 1997, Barber 1992, Kaestle 1983, Cremin 1980, Center on National Education Policy 1996, Pangle and Pangle 1993). Public education was historically rationalized in the United States by stressing the importance of putting schools under state (as opposed to church) control, and using them as institutions to promote civic virtue and a common set of democratic values (Kaestle 1983, Pangle and Pangle 1993, Vinovskis 1995, 92–103; Cremin 1980, 442). Thomas Jefferson’s “Report for the Commissioners for the University of Virginia,” for example, articulated six goals for public education. The first three goals dealt with the benefits of instilling basic literacy, the second three goals the importance of instilling communal responsibilities in the individual (Jefferson 1975, 332–346). Horace Mann, the father of the common school movement of the nineteenth century, emphasized the importance of a system of public education to achieve such goals. This meant
Ideology and Education
9
schools with some degree of centralization under state control and organized so they were oriented towards the polity as embodied by representative institutions and state and national constitutions (Cremin 1980, 148). For the commonwealth, public education is inextricably bound to the democratic ethos, and market mechanisms threaten rather than strengthen those beliefs. Given these origins, it is no surprise to find schools repeatedly shouldering an important role in processing value-based conflict within American society. The role of religion in public life; the range and limits of individual liberty; the articulation and promotion of core republican values; the need to socialize a polyglot influx of immigrants into Americans; providing the social cohesion and human capital necessary to deal with a depression, two world wars and a cold war; dealing with the painful legacy of race; striving to uphold the egalitarian commitment to equitably distribute social opportunities—for more than a century public education has been used for working through the difficulties that define the American experience (Tyack and Cuban 1995, Cremin 1980, Hofstadter 1962, Merelman 1980, Barber 1992). This explicitly political role is not only desirable; for the commonwealth it is the primary reason schools exist. For the commonwealth, if a conflict exists over the appropriate role of religion in public life then the scope of religious activities in public schools, at least over the long term, should reflect and replicate for the next generation the values constituting the common ground between competing factions in society. Schools serve as repositories for communal values, what can be accepted by most if not enthusiastically embraced by all. Commonwealth advocates argue the market represents a direct threat to this role by proposing to supplant the ideological structure justifying public schooling with an alternate value system. This argument gains support from market advocates who recognize the inherent tensions between market and democratic processes. Chubb and Moe (1990, 26–68), for example, explicitly hold democratic institutions and processes responsible for the “failure” of education. Others critique democracy obliquely, focusing the culpability of bureaucracy (e.g., Lieberman 1993, Ravitch 1997, Murphy 1996). As Chubb and Moe (1990, 38) point out, this argument is disingenuous because bureaucracy is a natural product of democratic control. Getting rid of bureaucracy means getting rid of the management mechanisms of representative institutions, which means reducing, if not eliminating, the role of the democratic process in controlling education. Others justify eliminating democratic/bureaucratic control as a way to eliminate, or at least ameliorate, the politics that surround public education. Hill et al. (1997, 36), for
10
The Ideology of Education
example, argue the problem with the governance structure of education is that it makes schools a focal point for conflict over values, civic responsibility, the rights of minorities, and a whole range of other disputes within communities. What these conflicts produce are a package of compromises that guide the socialization of children rather than producing graduates with solid academic foundations. A persistent theme of market arguments is that schools are too tied to the democratic process, too concerned with sorting out, storing, and replicating the common values of a community. To commonwealth advocates what is being rejected here is not just democratic control, but the entire justification for public educational systems to exist. It is not just a rejection of a technical means (bureaucracy) to achieving functionalist goals (e.g., higher test scores), but a rejection of education’s role in supporting the polity. For the commonwealth schools are primarily agents of socialization, not a means to transmit blocks of technical knowledge. While this includes acquiring the literacy and numeracy skills necessary for individual social and economic advance, it is the inculcation and replication of values rather than the particular levels of technical knowledge transmission that are important. For the commonwealth, the job of schools is to provide a “process of acquiring the norms to which all members of a society conform” (Arnstine 1995, 5), and they are expected to “express values more than achieve goals” (Noblit and Dempsey 1996, 3). Market advocates do not deny the importance of education’s role in securing and promoting civic values (e.g., Bennett 1992, 56–62). As Friedman (1982, 86) put it, “A stable and democratic society is impossible without a minimum degree of literacy and knowledge on the part of most citizens and without widespread acceptance of [a] common set of values. Education can contribute to both.” But they are largely silent on how these common values will be produced by a market system catering to individual preferences. The democratic values embedded in public education are taken as a given, a generic outcome of education itself rather than a specific political objective achieved imperfectly and at no small cost. Though its advocates presume the market model can produce a shared set of moral and civic values in education, there is little detail on how exactly markets can do this (Engel 2000, 71–74).1 This lack of detail concerns commonwealth advocates because a good deal of empirical evidence suggests markets atomize rather than build community. Market model education systems exist in a number of countries, and a central reason for adopting such systems was an inability to achieve common ground in value conflicts. The Belgian and Dutch market systems, for example, were a response to a struggle for control over educa-
Ideology and Education
11
tion between Protestant and Catholic churches, and the state. As no dominant force emerged from the theological and secular tussle to control the classroom, the education system was built around individual choice. In practice, this turned out to be a decision to divide education along sectarian lines. These divisions blurred considerably as these societies became more secular, but today these systems still seem to promote religious and social stratification (Vandenberghe 1999, Lutz 1996, Ritzen et al. 1997). In other nations, the introduction of market mechanisms into public education has favored particular groups, especially the better educated who quickly opt into schools with student bodies characterized by high socioeconomic status and better than average test scores (see Vandenberghe 1999, 278). The divisive aspect of the market is well documented in the United States. Social stratification, either by religion, class, or ethnicity is one of the few common themes to arise from research on market models of education (Powers and Cookson 1999). Choice or voucher systems seem to benefit the advantaged even among groups of lower socioeconomic status, where the better educated are more likely to reap the benefits (e.g., Schneider et al. 1997). Voucher programs in Milwaukee, Cleveland, Washington, D.C., and New York City have generated conflicts within and outside those communities with little consensus on whether they provide tangible or lasting benefits (see Witte 1996, 1998, 2000; Peterson and Noyes 1997). Given this track record, critics of the market model argue that it exacerbates the politics surrounding education without producing any broad civic contribution (e.g., Giroux 1998). These problems were not unforeseen by market advocates. Friedman (1955, fn. 1), for example, openly acknowledged that the market system he proposed could result in racially segregated schools. Nothing in Friedman’s argument counters the proposition that the segregation could also occur along religious or socioeconomic lines. Critics of the market model argue that its negative civic implications stem from its inability to focus on a core collective purpose. Tied to the presumption of methodological individualism, the theoretical basis of the market has a hard time articulating a core ideological commitment to democracy. Instead, the market’s functionalist perspective is that “schools have no immutable or transcendent purpose. What they are supposed to be doing depends on who controls them and what those controllers want them to do” (Chubb and Moe 1990, 30). The root concern for the commonwealth is that adopting the market model means that those controlling agents will cease to be institutions and processes that broadly accept the historical and legal justifications of public education. Responding to the individual rather than the democratic process,
12
The Ideology of Education
the market will exacerbate divisive social cleavages for the simple reason that the value agendas dividing social groups are saleable commodities. Instead of controlling the power of faction—the problem at the heart of the Madisonian conception of government—the market unleashes it, giving some groups better opportunities to advance their own interests at the expense of others. How public education promotes and preserves democratic values in practice is a controversial issue within the commonwealth camp even beyond market-related concerns. There are two basic theories on education’s role in reinforcing civic values. The first theory is embedded in consensus theory, reflecting the Jeffersonian vision of public schools providing the basic underpinning of a democratic society through inculcation of core civic values. Such visions are exemplified in the arguments of theorists and philosophers of education such as Dewey (1929), Guttman (1987), Barber (1992), and Callan (1997). There is, however, a darker side to this idealized portrayal and one that is often more empirically grounded. This view can be loosely referred to as “conflict theory,” the argument that schools are used to maintain social hierarchies for the benefit of a dominant class. In democratic systems, it is usually argued that this is achieved by socializing people to accept social inequalities as the product of merit rather than socially reproduced injustice (Karabel and Halsey 1977, 451). This view is often associated with Pierre Bourdieu’s (1973) notion of a “hidden curriculum,” the idea that below the surface of the real curriculum is a socialization process that benefits one set of people (in the United States, usually identified as white and middle- to upper class) by getting another set of people to accept their inferior lot (usually identified as minority and poor). For conflict theorists, schools are ideological tools but are not employed for civic benefit. They smuggle values for the dominant social class, who determine these values by using the democratic process to defend their social position and prerogatives. Commonwealth advocates recognize the dark side of public education’s cultural role, and fear market reforms on the grounds that its institutional prescriptions could exacerbate this problem. Hierarchical control hardly boasts a perfect track record, but it has forced the end of de jure racial segregation, struggled mightily for some form of religious neutrality, sought to ensure some minimal level of equality of opportunity, and overtly attempted to instill a set of civic values that include patriotism, commitment (at least in the abstract) to the democratic process, and tolerance (Kaestle 1983, Tyack and Cuban 1995, Engel 2000, Cremin and Borrowman 1956). Success in any of these areas is debatable, but so is the market model’s ability to correct any of the
Ideology and Education
13
shortcomings. As an example, consider inequity in resources, long a concern in public education. This inequity is a result of a heavy reliance on local property values, so well off neighborhoods are able to better support public schools than poor neighborhoods. The haves defend this resource advantage in the name of “local control,” often arguing that although they have the resource advantage and want to keep it, money does not matter (Berliner and Biddle 1995, 264–267; Kozol 1991). State attempts to equalize funding have been limited by the politics surrounding the funding mechanism, and have slowed but not eliminated the widening gap between rich and poor schools (Bundt 1997). Market advocates suggest something like a voucher system will help address these inequities, as vouchers will be equally funded and anyone can choose to go to a “rich” school. Yet, like any producer, schools with better facilities and better qualified teachers are likely to charge a premium. The better off are positioned to take advantage of these benefits with less interference from centralized government concerned about broad concerns of equity. Hierarchical control tied to the democratic process has not solved the problem of resource inequity, but does recognize there is a problem. The market simply sees and seeks to respond to demand. The existing politics of school funding clearly indicate a strong demand for funding inequities that is not being met by the existing system. Responding to such demands may not be fair in social-democratic terms, but the market is an efficient rather than a just allocator. Lacking regulatory constraints the market will respond to demand even when the socialdemocratic implications are negative. Such possibilities, commonwealth advocates suggest, means the market applies a starkly different set of values to education, values clearly at odds with democratic ideals.
THE IDEOLOGY OF EDUCATION
The two policy camps at odds over public education reform thus represent competing perspectives of education that can be viewed as separate and systematic value systems—ideologies of education (Engel 2000). Table 1.1 classifies the key values that separate the market and commonwealth perspectives of education (Halchin 1998, Witte 1998, McCabe et al. 1999). The first of these differences is the basic conception of education. The market views education as something that can be treated as a private good, i.e., a service provided to individuals that can be consumed in individual units (Lamdin and Mintrom 1997, 213). This suggests that, within broad limits, education can be treated much like the
14
The Ideology of Education TABLE 1.1 The Ideology of Education Contrasting Values of the Market and the Commonwealth
Value/Belief
Market
Commonwealth
Conception of education
Private good—service consumed by individual
Public good—process supporting socialdemocratic undertaking
Primary stakeholders of education
Individuals—parents/ students
Society/ community
Appropriate locus of control
Internal—schools respond to demands of clientele or “customers” (parents/students)
External—schools respond to democratic institutions, externally enforced rules/ regulations
Mechanism of control
Exit—dissatisfaction leads consumers to seek another service provider
Voice—dissatisfaction leads citizens to pursue change through democratic process
Primary mission
Output/Outcomes— transfer of economically useful knowledge and skills
Socialization—instilling values, beliefs, behavioral dispositions
Operating value (primary basis of judging mission success/failure)
Efficiency—maximum outputs with minimum inputs
Equity—provision equality of social/ economic opportunity to citizens
exchange of other goods and services that the market has demonstrated it can regulate efficiently (Friedman 1955). The commonwealth views education as a public good, a service that cannot be broken down into individual units. Education is like national defense in that the entire society has a stake in public schools, not just the individuals being educated. A citizen might not have children in school, but retains a vested interest in shaping future citizens and retains a right to participate in that future through the democratic process (Barber 1992). This difference in basic conception leads to different perceptions of the primary stakeholders in education. If a private good, the natural stakeholders are individual consumers. If a public good, the natural
Ideology and Education
15
stakeholder is the entire community or society. If a private good, education can and should be regulated by market mechanisms. If producers are allowed access to the market and consumers are free to choose among producers, a public education market will be highly responsive to its primary stakeholders (i.e., individuals). The primary mechanism to ensure this response should be exit, i.e., producers are oriented to consumer demand by the ever present threat that a dissatisfied customer will opt for a competitor’s services. If a public good, education can and should be regulated by representative institutions and their bureaucratic management mechanisms. This will ensure maximum response to education’s primary stakeholders (i.e., all citizens). The mechanism of control to ensure this response should be voice, i.e., the system is oriented to its stakeholders by an open and easily accessible democratic process (Hirschman 1970). The emphasis on “should” reflects the normative dimension of these arguments, they represent connected normative beliefs about the appropriate role of public education and government in society. The commonwealth perspective is predicated on the belief that a democratic polity embodies a set of values that exist independent of the individual, and education’s job is to ensure that individuals are socialized into those group values. The market views the polity simply as a social contract to provide services to the individual, and among these services are education (McCabe et al. 1999). These beliefs are reflected in the primary missions the market and commonwealth assign education and the criteria they use to assess in achieving those objectives. The market sees the primary job of schools to provide the individual with something useful, specifically the skills that provide social and economic opportunities. The outcome of the educational process should be the acquisition of these skills, and schools that maximize the transfer of these skills with the minimal resource inputs are viewed as the most successful, i.e., they are the most efficient. These may yield collective benefits, but outcomes that produce group advantages are byproducts rather than the primary mission or motivational force in a market system. For the commonwealth, the primary job of schools is to serve the collective, to produce citizens with the beliefs and behavioral dispositions that serve the continuation of democracy and to equitably distribute social and economic opportunities. Despite these normative foundations, this does not mean that the market cannot function as a positive theory. If the key normative assumptions correlate with what we know about the empirical world, the market functions as a useful explanatory framework—a logically ordered picture of how the world does work, not just a vision of how it
16
The Ideology of Education
should work. Even if the market is more a normative than a positive theory, this does not mean it should automatically be rejected as the basis for reforming education. The commonwealth’s claim to public education rests on a simple thesis: public education exists to serve public interests, and those interests are best served by building education on a system tied to the democratic process and democratic values. If the market is better at divining the public interest and responding to those interests, then it is defensible as a normative theory of education. If the market model can deliver on its own collective promises (efficient production of educational outputs) as well as those of the commonwealth (inculcation of civic values, equitable distribution of opportunity), it may better realize the historical and legal justifications for public education. If its key normative premises cannot be empirically confirmed, and it does not advance (or at least does not harm) the normative historical and legal justifications of public education, the market fails as a positive and a normative theory of education. It survives as an ideology but one that stands in direct contrast to the commonwealth—an ideology predicated on advancing the interests of others even as they harm collective interests by reforming public education so that it abandons its basic constitutional missions. This reasoning suggests a way to gain empirical leverage on the key research question posed by this book. As an instrumental means to achieve universally beneficial ends, the market claims that schools are failing to adequately respond to some functional demand that has broad social consequences. Most commonly, this is portrayed as the individual need for economically valuable skills (see Chapter 2). These skills not only support individual opportunity, they also contribute to the collective good in the form of human capital, the intellectual resources that fuel innovation and efficiency in the broader economy. Thus a key justification for market-based reforms of public education is that they will increase a critical contribution to the economy (Bishop 1989, Hanushek et al. 1994, Paris 1994). This suggests an empirical hypothesis: School outputs have a positive relationship with economic development. If this is the case, it supports the functionalist picture of schooling associated with the market as a positive theory. If school outputs do have such a determinative role, increasing the efficiency of schools—their ability to transform inputs into the skills that constitute outputs—is a utilitarian goal that market mechanisms may help achieve (Swanson and King 1991). Even if this aspect of the market argument is not confirmed, it does not impact the market’s status as a normative theory of education. Regardless of whether school outputs drive economic success, if market
Ideology and Education
17
mechanisms equitably distribute educational opportunities, promote democratic values, and further individual liberty in measures equal to or greater than the status quo, the result is a powerful and defensible normative vision of public education. This normative vision also suggests empirically testable hypotheses. If the market is a better way to realize the ideals of democratic governance in education, then institutional arrangements promoting consumer choice and competition among schools should result in the following: (1) Better educational opportunities than the status quo (see Chapter 3). (2) The promotion of civic goals and values (see Chapters 4 and 5). If, as its critics contend, market theory is a scholarly veneer covering an ideological agenda that threatens the commonwealth values underlying education, market mechanisms should have a wholly different observable impact. Competition and choice will correlate with greater inequities in educational opportunity and serve to advance narrow, sectarian interests. There are a number of conceptual and methodological challenges in testing these expectations, but these are not insurmountable. If the market is a utilitarian and instrumental response to clearly defined problems, the base presumption is that schools are failing to adequately respond to society’s needs for social and economic skills. This failure should manifest itself in a predictable relationship between what schools produce (human capital) and social and economic success. If it is a normative theory that will advance the value-based historical and legal justifications for public education, there should be an observable, positive relationship between market mechanisms and educational opportunity and the promotion of civic values. If market reforms represent a value system that threatens the democratic values public education is held to imbue, market mechanisms should have a negative relationship with educational opportunity and the promotion of civic values. The remainder of this book seeks to test each of these propositions.
CONCLUSION
The basic purpose of this book is an empirical investigation into whether market-based reforms of education represent a utilitarian response to known problems in public education or whether they represent a valuebased challenge to the public school ethos. The aim is to assess whether the deregulatory reforms represented by the market model are more likely to achieve functionalist ends (improved educational outputs and outcomes) or ideological ends (a reshaping of attitudes and beliefs and a drift away from democratic regulatory mechanisms). The starting point
18
The Ideology of Education
of the project is specifying two ideologies of education, systematic belief systems that provide prisms to judge education and to obtain a vision of what it should look like. One of these, the commonwealth, is explicitly ideological—it views public education as a primary means of storing and replicating the values of the polity. The second, the market, can be portrayed as positive theory, a normative theory compatible with the values embodied in the commonwealth, or as a rival ideology that seeks to replace the democratic values with market values. These contrasting perspectives center on assumptions about what public education contributes to society and how it should be institutionally structured to maximize that contribution. They also provide a relatively stable platform for comparative analysis: They generate expectations about causal relationships that can be, at least in theory, empirically confirmed. That confirmation, or the lack thereof, will be used as a basis to answer the question of whether the market offers an instrumental tool to help public education advance public interests, or whether it represents an ideological agenda that threatens the democratic values at the heart of the historical and legal justification for public education.
2 Education and the Economy
M
arket-based education reforms are frequently justified by the claim that schools inadequately respond to the demand for economically valuable knowledge. The belief that education’s primary function is to meet this demand by creating and transferring the intellectual skills that provide economic opportunity is at the heart of the market’s functionalist conception of schooling. As Hanushek (1994, 129) puts it, “For most students and their parents, the raison d’etre of schools is preparation for useful and rewarding careers.” Numerous studies claim that creating these economically valuable skills serves public as well as individual interests because they determine the economic success of the social aggregate. This was the central declaration of A Nation At Risk, the study that framed education reform as an economic imperative during the Reagan administration: “Knowledge, learning, information, and skilled intelligence are the new raw materials of international commerce. . . . If only to keep and improve on the slim competitive edge we still retain in the world markets, we must dedicate ourselves to the reform of our educational system” (National Commission on Excellence in Education 1984, 7). Emphasizing the link between educational performance and economic success in similar terms has been a rhetorical constant in market-based justifications of education reform ever since (Paris 1994, Apple 1986). The causal relationship between education and economic success is broadly accepted even by those with reservations about market control of public education (e.g., Reich 1990). Yet though widely accepted, this relationship linking education to economic success is only vaguely understood and frequently oversimplified. As Tyack and Cuban (1995, 34) observe, “Many policymakers have narrowed the currency of educational success to one main measure—test scores—and reduced schooling to a means of economic competitiveness.” Yet test scores have changed relatively little over the past 30 years as the economy has cycled
19
20
The Ideology of Education
through periods of recession and prosperity. The United States consistently enjoys the most productive labor force on the planet yet tends to be only slightly above average (or worse) in international education comparisons based on standardized tests. Despite the evidence cautioning that test score-based measures of education “production” have, at best, a modest relationship with individual or aggregate economic success, the notion that improvements on these yardsticks will yield economic benefits is deeply ingrained in economic and popular analysis. A central justification of the utilitarian market argument is that introducing choice and competition into public education will increase the acquisition of economically valuable intellectual skills, and that these skills can be measured by output indexes like test scores (Hanushek 1994, Chubb and Moe 1990, Bishop 1989, Gerstner 1995, Mayer 1999). The market’s claim to be a utilitarian response to a well defined problem thus rests heavily on the assumption that education outputs have important aggregate economic implications. The purpose of this chapter is to assess the theoretical and empirical case for this functionalist conception of schools.
HUMAN CAPITAL THEORY
The relationship between education and the economy underlying arguments for market-based education reforms rests largely on human capital theory, a direct outgrowth of neoclassical economic thinking. Though human capital theory’s antecedents predate Adam Smith’s major works by at least a century (Kiker 1966), its contemporary incarnation is largely the product of two University of Chicago economists, the Nobel laureates Theodore Schultz and Gary Becker. Schultz (1961) argued a labor pool’s intellectual skills are an economic input as valuable as physical capital, and Becker (1993) formalized and popularized the connection between these skills and economic productivity. Both stressed the importance of education as the primary formal producer of human capital. At the “hard core” of human capital theory “is the idea that people spend on themselves in diverse ways, not for the sake of present enjoyments, but for the sake of future pecuniary and nonpecuniary returns” (Blaug 1976, 829). Education represents one such expenditure; it is a rational individual investment made with the expectation of a return. The investment consists of lost time and forgone wages, sacrifices necessary to engage in the nonmarket activity of education. The return is the higher pay that increased productivity commands in the labor mar-
Education and the Economy
21
ket. Human capital is thus man’s “capacity to produce goods and services,” and the production of these intellectual skills is what connects education to the economy (Thurow 1970). Transferring the basic logic from individuals to the social aggregate is not hard. Societies investing more in the production of human capital will have decided economic advantages. Their labor pools will be more productive and characterized by a greater rate of economically desirable characteristics such as innovation. As the primary formal producers of human capital, schools represent both an aggregate social investment in human capital and the individual opportunity to do the same (Fosler 1991, 299–300; DeYoung 1989). The basic approach to modeling the economic impact of human capital is through an equation that formalizes the expected returns to education (Miller et al. 1995, Rouse 1999): W=I+E+F
(2.1)
Where W represents wages or earnings, I represents that individual’s innate capacities, E represents a measure of education, and F represents family background. A similar model (called an education production function) is employed to formalize the production of human capital by schools (see Hanushek 1997, Burtless 1996, Verstegen and King 1998 for variations):1 E=I+S+F
(2.2)
Where E represents a measure of education (the same sorts of measures that appear in equation 2.1), I represents innate individual ability, S represents school characteristics (these range from organizational variables, to per pupil spending, to specific pedagogic innovations or curriculum requirements), and F represents family background (especially socioeconomic factors), which are consistently reported to be the dominant predictors of outputs. Regardless of whether the unit of analysis is the individual or a social aggregate, when data are employed to operationalize these models, the basic goal is to assess the correlates of human capital. E is the critical measure of education’s response to its task environment, i.e., its production of economically valuable skills and knowledge. The difference is whether E is modeled as an outcome or an explanatory variable. Both models assume that E represents a formal objective of education, an objective justified on the basis of its ability to provide economic returns to the individual or to social aggregates (Hanushek 1994, Bishop 1989).
22
The Ideology of Education
The causal relationships portrayed by equations 2.1 and 2.2 are sometimes called the questions of internal and external efficiency (Swanson and King 1991). Internal efficiency studies (equation 2.2) seek to isolate the variables that maximize E. External efficiency studies (equation 2.1) seek to isolate E’s role in maximizing W. Together, these logical frameworks have provided much of the intellectual firepower behind calls for market-based reforms. Key are studies that claim private schools do a better job of producing outputs (higher test scores or graduation rates) and that private school graduates get better returns on their educational investment. Such findings support the argument that public schools should adopt the market characteristics of their private counterparts (Chubb and Moe 1990, Coleman and Hoffer 1987, Levin and Kelly 1994). Internal efficiency claims, especially the contested role of expenditures in determining educational outputs, have generated the most controversy in and outside academia (see Burtless 1996). Yet it is the external efficiency claim that forms the base justification for marketbased education reforms. Human capital theory as a basis for marketbased policy prescription stands and falls on education’s link to broader economic success, not simply on an education system’s ability to produce high test scores or graduation rates (Mayer 1999). The individual-level link between education and economic success is often assumed to be obvious, but the empirical record is actually quite mixed. The correlation between education and economic success at the individual level is firmly established, but it has proved all but impossible to disentangle education’s effects from those of innate ability and family background. The most promising studies seeking to isolate and quantify the relationship between education and economic success are those based on data from identical twins (for a survey see Bound and Solon 1999). Identical twins share the same genetic makeup and (usually) the same family background and socioeconomic environment, so a natural experiment occurs where the latter are held constant and the effects of education can be isolated. Yet even these studies have produced inconsistent results. Most of these analyses are based on log-linear models that estimate the returns to education on the basis of percent increase in income for each additional year of schooling. The average rate of return estimated by the 22 models from the seven studies surveyed by Bound and Solon (1999) was 7.7 percent, i.e., every year of education was associated with a 7.7 percent increase in earnings. The average is somewhat misleading because of the range of the results—from as low as 2 percent to as high as 12 percent. Such varied findings suggest that investment in education (at least in terms of time) yields a positive return on future wages, but the size of that return is up for debate. The paradoxical infer-
Education and the Economy
23
ence is that while the correlation between education (however measured) and individual earnings is well known and empirically established, these correlations are of limited use for policy prescription because the unique effect of schooling is so difficult to isolate from other sources of variation. Even the studies using data from identical twins have struggled to definitively isolate the relative impact of education on individual earnings. A similar problem plagues individual-level internal efficiency research. Education production functions studies have found it difficult to control for the effects of things like innate ability in determining test scores, and attempts to do so have yielded inconsistent results on what “really” drives education outputs (Burtless 1996, Verstegen and King 1998, Hanushek 1997). Measures of educational aptitude or achievement such as the SAT tend to be very highly correlated with measures of intelligence like IQ scores, and both are positively correlated with earnings. Separating out education’s unique contribution, if any, to achievement test scores or intellectual skills and abilities remains the subject of no small controversy. We are not sure what schools independently contribute to achievement or intellectual skills, and we are not sure what levels of achievement or intellectual skills independently contribute to income levels. The high degree of strength attributed to these causal relationships in policy debates requires a leap of faith above and beyond the existing empirical evidence. It is reasonably certain that education and income have a genuine relationship at the individual level. The correlation is too strong to dismiss as statistical artifact and resistant to any number of controls. Yet the inferential problem remains: Do measures of educational output represent something uniquely attributable to schooling? SAT scores and graduation rates clearly reflect socioeconomics (the most reliable predictor of educational achievement) as well as innate ability. How much of the remaining variation on these indexes can be confidently attributed to the educational process is not clear at all. Disentangling, rank ordering, and isolating the specific impacts of schooling for the purposes of policy prescription have never been done in the sense of a generalizable agreement. Research at the aggregate level uses the same input–output approach reflected in equations 2.1 and 2.2. Instead of individual wages, W translates to an aggregate economic measure such as productivity or per capita wealth, E to some population level of educational attainment (e.g., percent with a high school diploma, average test scores), and F to broader socioeconomic context (everything from crime rates to political environment). Research at this level has also struggled to produce general agreements on policy prescriptions to increase the production of
24
The Ideology of Education
human capital (no shortage of prescriptions, just a shortage of agreement on their validity). At least at the aggregate level things like innate ability are likely to wash out, that is, unless the blessings of genetics are unevenly distributed and concentrated in certain subpopulations or geographic locations.2 If such abilities are distributed randomly across groups of humans, it should be safe to assume innate intellectual ability is roughly comparable across aggregated populations such as schools, districts, states, and nations. At least in theory, this means the I in equation 2.1 is less likely to confound results. Despite the practical appeal of aggregate-level research, it, too, has a rocky history. Such studies have also had a difficult time successfully identifying policy prescriptions that will boost schooling’s production of human capital. Early external efficiency studies suggested aggregate investment through higher educational spending and compulsory attendance laws would yield significant economic payoffs. Those payoffs have been slow in materializing, or have not materialized at all (Salamon 1991, Fosler 1991, Denison 1962, Schultz 1963). Similarly, the education production functions literature have repeatedly addressed the internal efficiency question at the aggregate level, and mostly repeated the inconsistent and contradictory findings reported at the individual level (see Burtless 1996, Verstegen and King 1998). At the individual level and at the aggregate level, neither the internal nor the external efficiency arguments formally displayed in equations 2.1 and 2.2 have produced much that is definitive. We are reasonably certain at the individual level that education and income have a causal and not just a correlational relationship, but have little in the way of specifics and are unsure of schooling’s unique contribution to this relationship. At the aggregate level education and things like productivity and overall levels of wealth should be related given what we know from the individual level, but our explanation here is even more incomplete. Despite the gaps and contradictions, the basic logic of human capital theory is entrenched in the underlying assumptions for marketbased education reforms. Part of this rapid acceptance may be due to human capital theory’s basis in neoclassical economics and its natural fit with the market ideology. School objectives should be determined by the external (economic) task environment, and individuals should be free to invest whatever resources they consider necessary to gain the intellectual capacities to be economically productive. Human capital and education production function studies, and the popular market arguments that make use of them, assume or claim that individuals with higher levels of schooling or educational achievement have greater economic opportunity, and that societies with greater formal levels of
Education and the Economy
25
education or educational achievement within its population are more likely to be economically successful (Paris 1994, Apple 1986, Lieberman 1993, Hanushek 1994). These assumptions present a straightforward enough proposition: School outputs (human capital) should predict economic success. Why do we lack clear empirical confirmation for this proposition? The central problem is measurement. The concept of human capital is easy enough to grasp; it is the set of intellectual skills that contribute to the production of goods and services. But how is this measured? The measurement choice is critical because it can determine the outcome of internal and external efficiency research projects. There are three basic approaches to measuring human capital. The first is expenditure, which regardless of unit of analysis is usually an aggregate measure (commonly per pupil spending) and not a measure of individual investment. The problem here is immediate and obvious—expenditure measures a presumed investment in human capital, not human capital itself, and the investment is a result of a collective choice rather than an individual one. Picking expenditure also tends to muddy the causal water between the inputs and outputs of external and internal efficiency research, and the clean causal picture indicated by equations 2.1 and 2.2 (education causes human capital, human capital causes economic success) quickly starts to break down. In education production function models, expenditure is invariably an explanatory variable, something used to predict the production of human capital rather than human capital itself. Studies examining the relationship between education and economic success also employ education expenditures as an explanatory variable, even though this skips a necessary intervening variable—the actual educational product the investment is supposed to return (Salamon 1991, 6–10). S and E in equations 2.1 and 2.2 tend to get indiscriminately substituted for one another across 30 or 40 years of research, confusing the causal logic and tainting the validity of the conclusions. The second option, and probably the most common, is to measure human capital using the amount of formal education received (such as years in school or highest grade completed). This too has been castigated as a poor yardstick of human capital, one that further confuses what is an input and an output. Like expenditure, human capital theory views time in school as an individual investment or input, not the actual product or output of an education. Endurance in the education system is just as likely to be a product of state/family insistence or assistance than some independent indicator of economically useful skills. So using time in school as a measure of human capital suffers drawbacks similar to those in using expenditure and also confuses the
26
The Ideology of Education
roles of F and E in equations 2.1 and 2.2. As Welch (1975, 67) put it, “I find it hard to conceive of a poorer measure of the marketable skills a person acquires in school than the number of years he has been able to endure a classroom environment.” The third option is test scores. These have advantages over expenditures and levels of education. As indexes of concepts like educational aptitude or achievement, they offer at least some face validity to the claim that they isolate E in equations 2.1. and 2.2. Test scores have been enthusiastically adopted as yardsticks of educational success or failure on this sort of basis at both the individual and aggregate levels of analysis. Lower test scores are seen as threats to the economic success of both the individual and society because they are viewed as reasonable proxies for human capital (Bishop 1989, Henig 1994, Paris 1994, National Commission on Excellence in Education 1984, Hanushek 1994). The acceptance of test scores as proxies of human capital is controversial for several reasons. Tests like the SAT or the NAEP were designed to be measures of educational aptitude or achievement, not the capacity to be economically innovative or productive. It is not clear that these measures are even accurate yardsticks of what they were designed to measure (see Lemann 1995, Rothstein 1997, Carson et al. 1993). Even tests specifically designed to measure job skills and capacities boast only modest success. The General Aptitude Test Battery (GATB) is one such test that is widely used and studied, and it is a relatively poor predictor of job performance (for a survey see Jencks 1998, 75–76). The notion that SATs or NAEPs and their kind tap into some concept of economically important cognitive capacities is actually contradicted by a long list of empirical studies, and the faith vested in test scores as proxies for human capital comes in spite of, rather than because of, the available evidence (Henig 1994, Rothstein 1997, Levin and Kelly 1994). Cumulatively, that evidence suggests we do not know exactly what test scores measure, are unsure if they declined over time as market advocates frequently claim, and have no real broad empirical basis to treat them as correlates—let alone causal factors—of economic success at the national level (Rothstein 1998, Bracey 1997, Carson et al. 1993).
EDUCATION, CULTURE, AND THE ECONOMY
Despite its importance as an intellectual foundation of the market conception of schools, empirical studies of human capital theory have yet to definitively isolate and explain what education does to effect the capacity for innovation and productivity. Many studies solve the underlying
Education and the Economy
27
problem by simply assuming that education transfers economically valuable knowledge and skills. Such assumptions may make academic research more tractable, but do nothing to explain how the human capital produced by schools gets effectively deployed as an input into the broader economy. If the intellectual skills and the future productivity they represent are unobservable, it certainly leaves prospective employers with a problem: How can they make judgments about something they cannot see? If they cannot see this economically valuable fund of human capital, how do they harness it to their business operations and, by extension, to the larger economy? Economists who asked such questions were led to the proposition that education may function as a sorting and signaling device, not as an independent producer of human capital. Rather than creating human capital, education systems may simply rank order people on the basis of economically desirable characteristics. Even if education makes only modest independent contributions to the creation of these characteristics, it retains a useful economic function if the end result is to isolate and identify people with valuable intellectual skills. Educational credentials (high test scores, a degree from a good university) may not say much about the educational system’s independent capacity to produce human capital, but they will still provide important clues about the economic potential of the individual (Riley 1976). Solid credentials signal high future productivity, lower credentials (poor grades or test scores, a high school or college dropout) signal an individual with a set of economically undesirable characteristics such as absenteeism and irresponsibility (Weiss 1995). At first glance, this argument holds no particular threat for human capital theory. The basic notion of an individual foregoing current income in the expectation of a higher future return is intact, and so is the basic causal framework. Some critics recognized quickly, however, that sorting or signaling arguments pose a set of potentially devastating problems to human capital theory, and in doing so pose equally weighty problems for market-based justifications of education reform (Blaug 1976). Human capital theorists assume education signals economically valuable intellectual skills. But what if education signals something else? It is not clear what productive skills and capacities a high SAT score, graduation from an elite prep school, and an ivy league diploma represent. But that list of credentials conveys a high probability its bearer is at least middleclass (probably upper), and comes from a background that if not white, Anglo-Saxon, and Protestant, has at least bought into the values exemplified by this sociodemographic strata (Friedenburg 1974). Education may be signaling cultural values rather than economic skills.
28
The Ideology of Education
The empirical record backing this claim is extensive and consists of many threads. For example, gaps in educational achievement (regardless of the form of measurement) exist between certain groups in society. These gaps do not wholly explain the economic differences between ethnic groups in society (O’Neill 1990, Johnson 1998).3 According to human capital and broader economic theory, wage discrimination on the basis of class or race could not exist in a competitive labor market. Entrepreneurs will exploit the wage gap, pressuring not only their competitors but also imbalanced wage scales. Over the long run, wages should move towards an equilibrium, a natural market rate based on rewarding the productive capacity and economic skills of the worker, not their ethnic, religious, or social backgrounds. That this has not happened—the wage gap between blacks and whites seems particularly persistent—has led a number of researchers to suggest that more is involved in the explanation than education or training. Some argue that disparities in income levels and rates of return to education across ethnic or other sociodemographic groups may be the product of differing cultural values. Some groups promote values conducive to high productivity (e.g., Asians and Jews), even if they are the victims of open discrimination. Other groups foster values leading to low productivity (e.g., blacks), and it is these values rather than outright racial or class discrimination that produces wage gaps. Needless to say, this is a sensitive issue because it can lend academic credibility to offensive stereotypes, and critics have fought back with compelling cases that the culture argument has little empirical merit (e.g., Darity and Williams 1985). The argument that culture determines economic success, however, has a long empirical history. Max Weber (1958) argued that countries dominated by Protestant religions developed faster because their religious values supported hard work, innovation, and dedication to the business of business. Weber’s argument that Protestant religious teachings nurture the spirit of capitalism has some empirical merit, even though there are clearly exceptions to the general trend (e.g., Clark 1951; Wiarda 1992, 13–15). It is not just religion. Ethnic grouping, neighborhood, wealth, and family background combine into powerful predictors of educational attainment and economic success (e.g., Borjas 1995, Graham 1981, Darity and Williams 1985, O’Neill 1990). For example, consider the well documented cultural clash that sometimes occurs when inner-city minority students transfer into suburban and mostly white school districts. Minority students often report feeling isolated and marginalized, faced with a choice of “selling out” and buying into the predominant, white “achievement” value system, or remaining
Education and the Economy
29
true to their neighborhood social networks (Smith and Meier 1995, Wells 1996, Cook and Ludwig 1998). So rather than sorting and signaling the intellectual skills associated with innovation and productivity, education may signal socioeconomic groupings or particular value systems, i.e., culture and its associated social hierarchies. This is exactly the claim made by conflict theorists, that education reproduces and justifies existing social hierarchies and inequities by giving them the appearance of merit. Consider that test scores for black children raised in white families are dramatically higher than black children raised in black homes, and that these effects seem to wane in adolescence when they move into a social and cultural environment that becomes less white (Jencks and Phillips 1998, 3). This suggests that test scores are reflecting mainstream “white” culture rather than some fund of economically useful intellectual skills that are produced by schools or are innate qualities of the individual. According to human capital theory, everyone is a capitalist, everyone has resources to invest in the hope of securing future returns (Bowles and Gintis 1975). Economic inequities and social hierarchies are thus a product of merit, the result of individual choice and effort. Things like race, class, and socioeconomic status will be independent of levels of education because this is a function of an individual choice to forego existing wages in hopes of future returns, not the imposition or replication of particular value systems. This theoretical world is clearly a long way from empirical reality (Graham 1981). Such arguments have important implications for human capital theory and for the main thesis of this book. If value systems determine things like economic productivity, human capital theory is wrong, and market-based reforms take on a more ideological than utilitarian justification. Education institutions are no longer passing on blocks of technical knowledge or creating “intellectual skills.” They are smuggling values.
TESTING THE EDUCATIONAL AND CULTURAL DETERMINANTS OF PRODUCTIVITY
Tests of human capital theory share a basic logic and a set of assumptions. Key is that an individual rationally makes a choice to forego current wages in hope of future returns, and it is these micro decisions that drive macro-level behavior (DeYoung 1989, Becker 1993). Yet there is not much individual choice at all in primary and secondary education in the United States. All states have mandatory attendance laws, and
30
The Ideology of Education
collectives rather than individuals make the key choices about investments in education. It is the coercive powers of the state, not the rational decisions of the individual, that drive formal education. Even when individuals are free from the educational mandates of government (this varies from state to state, but it is not until beyond high school that the mandates disappear), their choices are constrained by their ability to pay for higher education and their ability to gain access to preferred educational institutions. That higher education is routinely subsidized by the government, either through loans, grants, or publically supported institutions, means that collective decisions are never really out of this equation. Rather than all being capitalists free to invest their resources on the basis of individual preference, most citizens invest in education at the outright insistence of the state, or their investments are heavily shaped by the policies of, and subsidies from, the state. The unavoidable role of the state in public education contrasts with the individualistic theoretical portrayal of human capital theory, and sets up the social aggregate as a unit of analysis above and beyond the accumulation of micro-level decisions. It is collective decisions— public policy on everything ranging from mandatory attendance, to school budgets, to standardized curriculums—that shape the individual patterns rather than vice versa. And if the aggregate level is where both individual and collective levels of human capital are at least partially determined, human capital theory runs into more empirical problems. While the correlation between education and measures such as income are firmly established at the individual level, the strength of this correlation tends to diminish at the aggregate level. Industrialized countries with “better” education systems (i.e., they have higher expenditures, average years of education, or test score performance) are not necessarily those who are going to rank higher on measures like gross domestic product per capita. A simple correlation between math scores reported in an international education comparison (National Center for Education Statistics 1997 and 1997b) and the per capita GDP of the participating countries and yields a coefficient of .12. In other words, the test scores often presented as economic barometers manage to explain 2 percent, or less, of the variation in per capita GDP in cross-sectional comparisons. Across time, the results are not much better. A simple correlation between per capita GDP and average SAT scores within the United States between 1959 and 1995 yields a negative coefficient, –.866 to be exact. Inferring causality from this correlation means concluding that increasing human capital results in lower economic growth. It is, of course, risky to base such inferences on the basis of simple bivariate correlations, but they show
Education and the Economy
31
that any assumption of a straightforward link between education and aggregate economic success is suspect. At a bare minimum such findings indicate that economic success at the aggregate level (as it certainly is at the individual level) is the result of a complex set of causal factors, and that education might only play a minor part in these relationships. A number of scholars have made persuasive cases that education’s broader economic role is overplayed for ideological reasons, i.e., the macro-level connection is exaggerated in order to make the case that market reforms will serve public and not just individual interests (e.g., Tanner 1998, Molnar 1994). The possibility that human capital theory might not be very good at explaining aggregate level outcomes because it has problems above and beyond measurement error—i.e., that the underlying assumptions might be wrong—is not particularly well investigated. Even on the individual level when the key assumptions seem to more or less hold, studies seem to raise as much doubt as confirmation (Fogel 1994). When individuals decide to send their children to private school, they definitely make a decision to forego current income because private schools require tuition. It can be reasonably assumed that at least part of that decision is driven by the expectation of future returns. A number of well known studies that have compared public and private schools on the basis of test scores follow this reasoning, though these differences yield very modest gains for private school students. In a study widely cited in support of market reforms, Coleman and Hoffer (1987) conclude that Catholic school students outperform public school students on standardized test scores, even after controlling for other factors such as race and socioeconomic status. Combining these results with data linking wages to performance on test scores, Levin and Kelly (1994) find that even if the basic conclusions of Coleman and Hoffer are correct, the higher test scores for Catholic students translate into a wage difference of 4 cents an hour over their public school counterparts. The return is small enough not only to question the wisdom of the initial individual investment, but also the underlying assumptions of human capital theory that help frame the market justifications for reform.
EDUCATION AND ECONOMIC SUCCESS—THE CASE OF THE STATES
The basic conclusions drawn from the assessment of human capital theory thus far are these: (1) Any empirical examination of the link between education and economic success in human capital terms is going to be at least partially determined by how education or human
32
The Ideology of Education
capital is measured. (2) The assumptions and premises of human capital are open to challenge even if these measurement issues are addressed—the basic educational indicators used as proxies for human capital may be reflecting cultural differences, or differences in social status, rather than the presence of economically useful skills or aptitudes. (3) There is a good argument for addressing the external efficiency question—the key question for justifying reform in terms of education’s economic task environment—at the aggregate level. Not only are the decisions on investment (in terms of time and money) generally made above the level of the individual, but in the aggregate some of the variation that is hard to isolate and account for in individual studies (e.g., innate abilities) is also likely to wash out—i.e., be held constant under the assumption of random distribution of such qualities across populations.4 Any empirical test of education’s link to economic success and of human capital’s ability to support systemic policy prescription should somehow account for all of these. This is a tall order. It requires a social aggregate that has an economy that can be isolated, a distinct education system whose outputs can be measured in multiple ways, and an identifiable value system or culture. It would also help if there was at least some general understanding of how such economies worked, so that any test of education’s relationship to the economy could be undertaken within the confines of a widely recognized set of controls, i.e., a set of variables with both a theoretical justification and an empirical track record of predicting variation in economic success. States come closest to meeting this list of requirements. They are sovereign governments and vested with the primary control over education through the X Amendment. Although much of this power is delegated to local education agencies, state governments bear the ultimate constitutional responsibility to provide for a public education system (Rebell 1998). States also shoulder a sizeable portion of the fiscal responsibility for public education, and by controlling the purse strings play a dominant policy role in education. While the case for states as distinct education systems is solid, it is less clear whether states have distinct economies. States have certainly become more professionalized and have assumed broader economic policy responsibilities during the past two or three decades, but has this translated into the capability to influence local economic conditions independent of national trends? States certainly act as they have some control over their economic destinies, and aggressively employ a broad array of policies designed to encourage business investment, employment, and revenue expansion (Jones 1990, Brace 1993, Dye 1980, Hendrick and Garand 1991, Brace and Dudley 1991, Lange and Garret 1985). Indeed, economic
Education and the Economy
33
development has emerged as the primary object of policymaking in the states. As Brace and Jewett (1995, 662) argue, it is the “Esperanto of state politics . . . (and) virtually every policy is now weighed by its anticipated impact on economic development.” A central part of these efforts to boost economic development are education policies, often with explicit human capital-based justifications (Brace 1993, Olson 1997, Smith and Rademacker 1998). For states, a poor education system is seen as the biggest stumbling block to economic development (Osborne 1988). So education policy has become one of the most important state-level responses to “the demand of business for highly skilled workers, and the pressures of a globally competitive economy” (Zinser 1994). Just as a number of studies conclude that state economies can be isolated for research purposes, considerable research suggests variation among states on measures such as test scores represents meaningful variation among education systems (e.g., Hanushek and Taylor 1990, Graham and Husted 1993, Dynarski and Gleason 1993). The states would thus seem to be a good laboratory to test the aggregate relationship between educational performance/investment and economic development. Much of the research on state economies incorporates a set of expectations about education’s economic role that is clearly anchored in human capital theory. Most commonly, education is seen as an investment that will increase the skills and productivity of a state’s labor pool, and high education outputs are seen as lures for businesses seeking to relocate or expand (Brace 1993, Zinser 1994, Smith and Rademacker 1999). Despite the theoretical importance of education in studies of state economies, however, empirical studies have produced remarkably little evidence corroborating the aggregate education–economy relationship. Jones (1990) reports that depending on the measure of economic development and time period studied, education is a positive influence, a negative influence, and/or (most commonly) an insignificant influence. Brace (1993) finds insignificant results. Smith and Rademacker (1999) find education investments exact a short-term economic cost, but confer a long-term benefit. Paul Brace (1993) undertook the most comprehensive examination of state economies to date, and as part of that work developed a base model of state-level economies that can be extended to education’s relationship to economic performance. Education plays a central theoretical role in studies that employ Brace’s framework, but this role is usually tested using only a single measure (test scores, percent of population with a high school diploma, or education expenditures). As discussed earlier, there are drawbacks to using any of these measures as proxies for stocks of human capital, and employing a single measure ignores the
34
The Ideology of Education
possibility that these indexes reflect different dimensions of education (at a minimum investment in education and educational output). Culture has had little theoretical and no empirical role in studies of statelevel economies. The theoretical justifications for accounting for culture in such studies were discussed earlier, and the practical barriers to including such a concept in an empirical test are relatively minor. States are known to have distinct political cultures, and quantitative measures of these cultures have repeatedly been found to have empirical validity (Elazar 1977, Sharkansy 1969, Lieske 1993, Johnson 1976). The best known and most widely used approach to political culture in the states is Elazar’s (1977) tri-fold classification scheme. Using traced migratory patterns of ethnic and religious groups across the United States to support his argument that settlement patterns resulted in distinct regional value systems, Elazar identified three distinct statelevel political cultures: individualistic, moralistic, and traditionalistic. Individualistic political culture is a value system that views government in utilitarian terms, as an institution to serve the needs of those who created it, and where politics is “based on a system of mutual obligations rooted in personal relationships” (1977, 95). Politics here is seen as something of a dirty business, while business itself—the reciprocal exchanges and relationships freely entered into by individuals—is seen as a central focus of social life. It is, obviously, a value system highly compatible with capitalistic economic activity, and in his description Elazar uses the term “business” to capture its view of public and political life. Individualistic states include Illinois, Indiana, and Ohio. In contrast, moralistic political cultures are more oriented towards the common good. Politics in this culture is considered a collective search for the good society and viewed as the means to advance public interests. Moralistic cultures use government to place constraints on individualism in the name of communal good. Moralistic states (examples include Wisconsin and Minnesota) tend to have progressive political histories and strong commitments to public education. While individualistic cultures view government as a necessary evil, and moralistic cultures view government as a necessary good, traditionalistic culture is ambivalent. In traditionalistic cultures, politics and government are viewed as the province of a political elite. Instead of a struggle for individual gain (individualistic), or a search for equality of opportunity (moralistic), social order in a traditionalistic culture is hierarchical and fixed. Traditionalistic cultures are concentrated in the South.This typology of culture includes a value system that obviously prizes economic success. Individualistic culture embodies market norms, so if culture drives economic success, individualistic states
Education and the Economy
35
should be more productive independent of the education measures commonly used as proxies for human capital. Thus we have the necessary ingredients for a test of how human capital theory contributes to economic success independent of a specific value system: An established model of economic performance, a set of commonly employed measures of human capital (expenditure, aggregate test scores, and years of formal education), and a long-established approach to identifying and measuring culture. If human capital theory is correct, at least some of the education indicators should show a consistent, positive relationship with economic performance when added to the basic model. More importantly, they should maintain in this relationship when culture is accounted for—if the education–economy relationship falters in the presence of a cultural control, it would indicate that education is reflecting culture rather than some set of independently transferred economic skills.
CULTURE, EDUCATION, AND PRODUCTIVITY
Table 2.1 presents the results of a regression analysis of productivity in the states for the years 1986 to 1994 (see methodological appendix for details). The outcome variable is productivity measured as gross state product (GSP) per worker, and all of the explanatory variables with the exception of the education and culture measures are taken from the model proposed by Smith and Rademacker (1999), which is a direct extension of the model formulated by Brace (1993).5 Smith and Rademacker model state-level productivity as a result of: (1) The political environment (political competition, unified partisan control of executive and legislature, federal support in the form of defense expenditures, the size of the public sector measured as tax collections as a percentage of total GSP, and policy activism measured as the number of tax incentives and business support packages approved by the state); (2) National productivity (GSP per worker for the entire nation); (3) Socioeconomics and labor pool characteristics (per capita income and level of unionization); (4) Education (lagged per pupil expenditures to measure long-term investment, concurrent percent of GSP devoted to education to measure short-term costs, variation in the distribution of within-state education resources to measure equity, and test scores). The test score measure was constructed from state-level SAT and ACT scores, and lagged five years to account for the timing of the economic impact of the human capital these scores purportedly represent.
TABLE 2.1 Education and Productivity in the States Coefficient (Standard Error)
Confidence Limits
Leamer Bounds
23* (10.6)
2.3:44
–1.43:7.27
Percent of GSP Devoted to Education
–2197* (380)
–2943:–1451
–4098:2197
Percent of Population with HS Diploma
–167 (142)
–447:112
–102:317
Equity of Education Spending
–23 (86)
–192:145
–253:31
Lagged Per Pupil Spending
.99* (42)
.14:1.83
–1.43:7.27
Political Culture
1569* (725)
147:2991
1569:1647
Defense Expenditures
–.40 (.50)
–1.39:.59
–40:3.79
Size of Public Sector
–1024* (253)
–1520:–527
–1024:–134
Policy Activism
–113* (34)
–180:–46
–435:–83
.59* (.12)
.32:.82
.54:.85
4637* (1533)
1631:7643
3315:8666
Political Competition
–62 (69)
–197:72
–62:36
Unified Control of State Government
224 (296)
–356:805
–1271:224
Per Capita Income
.43* (.14)
.15:.70
.22:.94
Variable Lagged Test Score Performance
National Productivity Unionization
Constant
14,945 (14,058)
N
405
R2
.62
Coefficients are GLS estimators. *p < .05 (see Methodological Appendix for details).
Education and the Economy
37
To this base model I simply add two other measures. The first is the percentage of a state’s adult population with at least a high school diploma. The second is a slightly modified version of the culture index proposed by Sharkansky (1969). The scale of the original index is 1 to 9, with 1 indicating a pure moralistic state, 5 a pure individualistic state, 9 a pure traditionalistic state, and mixes of the cultures scoring between these points. Although a commonly used measure of Elazar’s culture classifications, Sharkansky’s index has a serious drawback for this analysis. A key hypothesis is that individualistic states will have higher productivity than moralistic or individualistic states, and Sharkansky’s index scores individualistic states in the middle of the scale rather than the top or the bottom. This makes it hard to posit a direct linear relationship between productivity and the culture index. Accordingly, I “folded” the indexes: all scores from 1–5 were kept from the original scale, scores above 5 were subtracted from 10. This result is a 1–5 index where the closer a state is to a pure individualistic culture the higher its score. This allows a direct test of the culture hypothesis. If cultural values drive economic success, higher scores on the folded culture index should positively correlate with productivity (see Methodological Appendix for complete variable description and sources of data). The estimates of the resulting regression model presented in Table 2.1 generally confirm Smith and Rademacker’s conclusions: National productivity trends, the size of the public sector, socioeconomics, and rates of unionization are significant predictors of economic productivity; political competition and unified partisan control of state government are not (i.e., they are statistically insignificant). A minor difference with Smith and Rademacker findings is the policy activism variable, which they found to be insignificant but shows up here as a significant predictor. Of key importance for present purposes are the findings for education and culture. If human capital theory is correct, the education measures should positively correlate with productivity, and if the conflict theorists are correct, culture should positively correlate with productivity. The education findings are mixed. Expenditures have a long-term payoff (indicated by the positive coefficient for lagged per pupil spending) and a short-term cost (indicated by the negative coefficient for percent of GSP devoted to education). Test scores have a modest positive impact; equity in spending and percent of the population with a high school diploma have no impact (the coefficients are statistically insignificant). The mixed results hint at a certain instability in the relationship between education and economic productivity. Specifically, it raises that the possibility that how education is measured and which measures are
38
The Ideology of Education
included will determine how the regression model portrays the relationship between education and economics. This possibility is more clearly portrayed by the confidence limits and Leamer bounds reported for each coefficient. Confidence limits provide a range of values bounding the “real” coefficient with a 95 percent probability (i.e., there is only a 1 in 20 chance the true coefficient falls outside of this range). Leamer bounds report the upper and lower values for each coefficient from alternate model specifications and estimation techniques. Confidence limits and Leamer bounds provide information about the stability of an estimated relationship between the explanatory and outcome variable, information unaccounted for by traditional hypothesis tests (see Gill 1999). The latter are predicated on tests of statistical significance (a z or a t test) designed to assess the probability (called an Alpha level) that the true coefficient is zero. Alpha levels are traditionally set at .05, meaning if there is less than a 5 percent chance that the true coefficient is zero, the coefficient estimated by the regression model is considered a good description of the relationship between the explanatory and outcome variable. Alpha levels are purely arbitrary— there is no substantive or scientific argument for accepting a coefficient as “real” with an Alpha of .05 while rejecting one with an Alpha of .06. More importantly, the outcome of significance tests are heavily dependent upon sample size and model specification.6 As the number of data points available and the number of the variables included in the model change, so does the likelihood a coefficient will meet the traditional test of statistical significance. This is important because in any regression analysis, it is rare for a researcher to attempt only model specification. Typically model specification is an evolutionary process, with data availability, regression diagnostics, and other factors determining what ends up in the final model. The result is that what is portrayed as “real” by a regression analysis is dependent upon what variables and what measures a researcher uses to operationalize a model. Confidence levels and Leamer bounds show how sensitive the reported coefficients are to model specification, estimation technique, and sample size (see methodological appendix for more on the estimation techniques used to generate the results in Table 2.1). The key contribution of confidence limits and Leamer bounds is to show whether coefficients are bounded by zero, i.e., they show whether a coefficient’s values under differing testing scenarios are consistently positive or negative. If confidence levels or Leamer bounds cross zero, it means the coefficient switches from positive to negative depending on model specification, or that the range of probable values for the true coefficient includes both positive or negative numbers. Either case sug-
Education and the Economy
39
gests the possibility that the true coefficient is zero and the probability that there is no relationship cannot be discounted. Variables displaying these characteristics are suspect because their instability shows their relationship with the explanatory variable is dependent upon model specification and N-dependent statistics like standard errors. Regardless of their statistical significance in the reported model, they cannot be treated as substantively different from zero. Important for present purposes is that every single one of the education variables crosses zero in the confidence levels and/or Leamer bounds. This instability is not due to any obvious statistical culprits associated with regression or the specific technique used to generate the estimates (see methodological appendix). The estimated relationship between education and productivity shifts based on model specification, sample size, and variable measurement. This offers an explanation for the highly mixed record of education in studies examining the political economy of the states. Education’s relationship is unstable because it either does not exist except as a statistical artifact, or if it does, it is sporadically reflecting something else not being systematically accounted for in this research literature. What might that other something be? The obvious possibility is culture. The culture variable reported in Table 2.1 is uniformly positive within the range of values designated by the confidence limits and Leamer bounds. The range in Leamer bounds is tight, indicating a high degree of coefficient stability. In contrast to the education variables, it is safe to give this coefficient a substantive interpretation, to infer that it reflects a “real” relationship. The model coefficient estimates that for every one point increase on the culture scale—i.e., every time a state gets 1 point closer to an individualistic political culture—it has an associated increase in productivity of $1,569 of GSP per worker. To put that into context, consider that the overall productivity mean was $53,116 of GSP per worker, and the mean state culture score was 2.65 (i.e., a moderately individualistic culture). The model estimates a score of 5 on the culture index (a purely individualistic culture) would raise this average state’s productivity from $53,116 to $56,803, almost a 7 percent increase productivity. So what Table 2.1 reports is a base model of state-level economic performance that includes a variety of education measures used in human capital studies and a culture variable to reflect some of the long-stated rivals to human capital theory’s explanation of education’s relationship with economic success. The education variables confirm some expectations of human capital theory, but their underlying instability raises serious questions about the orthodox understanding of
40
The Ideology of Education
education’s relationship to the economy. Culture proved to be a robust positive estimator of productivity and, compared to education, is thus the more important factor. These findings, however, provide only limited information about education’s potential role in determining productivity. One of the limitations of regression analysis is that its results represent only the average case. Figure 2.1 shows a graph that plots the actual productivity values of the states against the productivity values estimated by the regression model reported in Table 2.1. The line through the data points represents the relationship between the observed and actual productivity values. Values above this line can be thought of as states that had higher levels
FIGURE 2.1 Actual Versus Predicted Productivity
Education and the Economy
41
of productivity given the inputs represented by the variables in Table 2.1 (Meier and Gill 2000). Clearly some states in some years were performing considerably above average. What were these high productivity states doing differently from the average state, and did it have anything to do with education? One way to get answers to such questions is through a statistically weighted analytical technique (SWAT). This is a statistical procedure designed to enhance program and policy analysis by extracting information about what makes high (or low) performing cases different from the average case (see Meier and Gill 2000, Meier and Keiser 1996). The basic idea is to use errors in estimation (i.e., the difference between the values that are observed and those estimated by a regression model) to identify high or low performers. The high (or low) performers are then iteratively assigned a set of weights that gradually give them greater influence in determining the results of the estimation (for technical details and a primer see the methodological appendix). As the high or low performing cases take on a bigger role in model estimation, the coefficients systematically shift to reflect any differences separating average from nonaverage cases. Applying SWAT to pooled data sets such as that employed here is a particularly good way to extract this information from a regression analysis, information that is not readily apparent in standard regression techniques focused on the average case (Smith, Meier, and Gill 2000). Table 2.2 reports the education and culture variable results taken from a SWAT analysis of the full regression model reported in Table 2.1 (see methodological appendix for statistical details). The numbers are simply proportions of the original regression coefficient and show how this proportion changes through the basic nine iterations of SWAT analysis (in the final iteration high productivity states have 10 times the influence on the estimates as the average performing states). So numbers less than 1 indicate a coefficient moving towards zero as high productivity states get more influence in the analysis. This indicates a variable has less of a role in separating high productivity from average productivity states. Numbers greater than 1 indicate a coefficient increasing as higher productivity states get more influence. This suggests that a variable plays a greater role in separating high productivity states from average states. On this basis, the SWAT analysis indicates that education does little to separate high productivity states from average states. Per pupil expenditure, percent of GSP devoted to education, and equity in education show little tendency to wander from the original regression coefficient—the proportions are close to 1 in all iterations of the analysis, indicating these variables do nothing to separate high performers
TABLE 2.2 SWAT Analysis of Education and Culture in High Productivity States Percent of GSP Devoted to Education
Equity of Education Spending
Percent of Population with HS Diploma
Culture
1
1.05
.98
1.02
.93
1.01
1.09
.95
1.04
1.06
.93
1.02
1.12
.92
1.07
4
1.08
.88
1.04
1.16
.92
1.10
5
1.09
.85
1.05
1.19
.80
1.13
6
1.11
.80
1.06
1.19
.70
1.18
7
1.11
.72
1.07
1.16
.54
1.24
8
1.09
.68
1.07
1.12
.25
1.32
9
.98
.59
1.06
1.16
CZ
1.46
Iteration
Per Pupil Expenditure
Test Scores
1
1.01
1.02
2
1.04
3
Cell entries are the SWAT coefficient divided by the original OLS coefficient for a given iteration of the analysis. Numbers close to 1 indicate the variable has little effect in separating high productivity states from average productivity states; numbers greater than 1 indicate the variable plays a more important role in distinguishing high productivity states from average productivity states; and numbers less than 1 indicate the variable’s role in predicting productivity diminishes as high productivity states are allowed to have more weight in the analysis. The ninth iteration reflects variable performance when high productivity states are allowed to have ten times more influence in the analysis than average and low productivity states. All estimates were generated using the full model reported in Table 2.1. See Methodological Appendix for complete details. CZ = Crosses zero (i.e., coefficient changes direction).
Education and the Economy
43
from average performers. The proportions for test scores and percent of population with a high school diploma shrink, indicating these factors are less important to high productivity states. In contrast, the impact of the culture variable steadily gets larger. When higher productivity states have 10 times the influence of average productivity states, the coefficients for the culture variable is 46 percent larger than in the original regression. These findings suggest that one thing clearly distinguishing high productivity states from average and low productivity states is an individualistic culture. Education’s role—regardless of the measurement—in making such distinctions runs from modest to nonexistent.
EDUCATION AND CULTURE
The analyses reported in Tables 2.1 and 2.2 show that compared to culture, education is a poor predictor of productivity in the states. Obviously, this finding is more supportive of human capital theory’s critics than its advocates. There remains a question of the relationship between culture and education. There is no overwhelming statistical relationship between the education and culture variables (bivariate correlations of about .3), but it is important to note that the folded culture variable measures only relative strength of individualistic culture. The transformation of the original Sharkansky scale was deliberately used to create a measure of a single value system, that hypothesized to be more supportive of market values and activity. While the folded scale highlights individualistic culture, it does a poor job of highlighting the other value systems, and these value systems may have independent relationships to education for purely noneconomic reasons. In particular, moralistic cultures are obvious candidates for very high support of public education, even if this does not directly translate into any economic benefits. When the culture scale is unfolded—i.e., using Sharkansky’s (1969) original 1 to 9, traditionalistic to moralistic scale—correlations with education measures become stronger. The cross sectional correlation between test scores and the original culture measure is approximately –.7. Between percentage of population with a high school diploma and culture, the correlation is even higher: –.8. Using the original scale, it is tough to statistically distinguish between culture and the test score/level of schooling approach most common in aggregate human capital studies. Moralistic states (Minnesota, Wisconsin, Iowa) have very high test scores and high percentages of high school completion; traditionalistic states (Louisiana, Mississippi) score low on these measures. This suggests that test scores and graduation rates are picking up the impacts of
44
The Ideology of Education
moralistic and traditionalistic cultures, essentially the presence or absence of commonwealth values.7 These value systems do not so readily or obviously transform into support of the economic sphere as the individualistic culture, so there is no particular reason to expect them to have strong relationships with variables like productivity. This also suggests an alternative causal view of education and its relationship to the economy than that commonly portrayed in human capital studies. Education is a reflection of culture, with moralistic states vesting the most in schooling and traditionalistic states the least. All else equal, individualistic states will be more likely to score higher on economic measures because the broad social value system prizes economic success. As a reflection of culture, education will have a highly uneven relationship with the economy, because it tends to highlight moralistic cultures (high on all education measures) and traditionalistic cultures (low on all education measures), value systems which are considerably less supportive of market activity than individualistic culture. When individualistic culture is brought out and allowed to assume a role independent of education, the education measures fall away as culture asserts its stronger and independent relationship with the economy. This is a fairly straightforward explanation of the instability in the education results reported in Table 2.1. Unlike arguments that ultimately come down to measurement, model specification, and estimation technique, the confidence limits and Leamer bounds indicate culture is going to provide much more consistent results when compared to the variety of available education measures.
CONCLUSION
One of the central claims of the market ideology of education is that schooling is important to economic success. While there is little doubt that people with more education tend to earn higher wages, the relationship between education and things like wealth and productivity is much more complex than the common “more equals more” assumption. Even if this simple causal proposition were correct, at the individual level it contains the seeds of its own demise—as more people hold a high school diploma that diploma loses its value, as more hold a college degree it too becomes worth less. Rather than more education translating into higher wages, more education ends up being required simply to prevent the returns to education from falling. Whatever the real returns to education are, they have eroded past two or three decades even as the traditional measures of human capital have risen (Levin and Kelly 1994).
Education and the Economy
45
The macro-relationship between education and economy seems weaker. A single analysis of productivity in the states is unlikely to prove definitive, but the results presented here indicate that higher test scores, or higher levels of educational attainment, are unlikely to have a dramatic impact on economic advance. They also suggest that any finding that education has a significant macro-level impact on economic success is going to be heavily dependent upon how an analyst decides to measure education, conceptualize the expected return to education, and specify the model. These choices will drive how a regression model portrays education’s relationship with the economy, and different choices are quite likely to result in different inferences. The analysis should not be read as a claim that education is unimportant to an individual’s economic fortunes. Perfect SAT scores and an Ivy League degree are no doubt valuable commodities that open up economic and social opportunities for their bearers. What is in dispute is whether such indexes in the aggregate represent human capital independently produced by educational institutions or embody a set of characteristics reflecting social status and value system. The analysis presented in Tables 2.1 and 2.2 seems to suggest the latter more than the former. This raises questions about education’s assumed relationship to its task environment, which this analysis suggests may be determined more by culture than by economics. If this is the case, the ideals embedded in the market ideology regarding education’s task environment are not as strongly embedded in empirical reality as is commonly assumed. Connecting education to the economy through the arguments of human capital theory turns out to be more of an ideological than an empirical claim.
This page intentionally left blank.
3 Education and Equality of Opportunity
A
s the previous chapter details, numerous studies raise questions about the external efficiency of schooling. Education outputs, especially test scores, have a tenuous claim to represent economically valuable cognitive skills, and raising these outputs seems unlikely to boost productivity in the aggregate (Levin and Kelly 1994). Yet these same outputs undoubtedly play an important role in apportioning economic and social roles. High school dropouts have fewer job prospects and are more likely to become social delinquents, and the test scores of graduates will determine access to higher education and the varied social and economic options it provides (Kolpin and Singell 1997). Whether education outputs objectively reflect human capital or provide meritocratic cover for the reproduction of cultural patterns and social inequity is a subject of intense controversy within academe, but this is mostly an abstract debate. In the practical world, test scores and the pedigree of a college degree matter regardless of their relationship to economic productivity or particular value systems. Irrespective of the reasons why, education still determines social and economic opportunity (Kozol 1991, Arnstine 1995, Jencks 1998). So while doubts about the external efficiency argument mean less justification for using market mechanisms to better align schooling with an economic task environment, they do not diminish the importance of the internal efficiency question. Market advocates use the internal efficiency argument to stake a claim to the commonwealth’s key operating value: equity. Public choice analyses of the internal efficiency question conclude that choice and competition will give schools the incentive to maximize education outputs while broadening access to schools with the highest graduation rates and test scores (Chubb and Moe 1990, Witte 2000, Peterson and Noyes 1997; Greene, Giammo, and Mellow 1999;
47
48
The Ideology of Education
Hanshek 1994). For practical purposes it matters little whether these educational indicators are objective indexes of human capital, subjective reflections of value systems, or something else. Schools that maximize these indicators for the greatest number of people, in effect, are argued to still serve egalitarian ends. Maximizing efficiency, internal efficiency, is thus equated with maximizing equality of opportunity. A market system gives students trapped in poor schools, students consigned to limited social and economic opportunity by the public education monopoly, the option to go elsewhere. The same system will pressure schools to ensure those options include the opportunity to gain the educational credentials that promote success. In staking a claim to promote equality of educational opportunity, the market seeks to incorporate within its institutional prescriptions the ideological core of the commonwealth and a key justification for the existence of public schools. The purpose of this chapter to examine the theoretical and empirical veracity of this claim.
DEFINING EQUALITY OF EDUCATIONAL OPPORTUNITY
The market and commonwealth ideologies both claim to embrace and support the near universal perspective that public education should play a key role in distributing social and economic opportunity (Friedman 1982, Brighouse 1994, Pangle and Pangle 1993, Callan 1997, Guttman 1986). Neither ideology, nor the universal vision, is particularly clear about what “equality of educational opportunity” means. As Howe (1997, 15) puts it, “equality of educational opportunity is now recognized as an essential requirement of equality of opportunity more generally. Beyond this shared commitment common to all forms of liberal democratic theory, however, vast differences exist among its various strands regarding what equality of educational opportunity requires.” Assessing the market’s claims regarding equality of educational opportunity first requires an understanding of what “equality of educational opportunity” means. Though equality of educational opportunity has had different meanings in different historical eras, the contemporary understanding of the term is a product of the Coleman report (Coleman et al. 1966, Coleman 1974). The Coleman report was undertaken in response to Section 402 of the Civil Rights Act of 1964, which ordered the U.S. Commissioner of Education (forerunner of the Secretary of Education) to conduct a survey on the “lack of availability of equal educational opportunities for individuals by reason of race, color, religion, or national origin in public educational institutions at all levels in the United States” (Coleman et al.
Education and Equality of Opportunity
49
1966, iii). The immediate challenge involved in responding to this Congressional mandate was tackling the problem of definition. Historically, equality of educational opportunity was judged on fairly straightforward criteria: all children in a given locality have access to the same schools and roughly the same curriculum at no cost. Using this as a yardstick, the school and the state’s role in providing equal opportunities was passive, their obligation discharged by ensuring that educational institutions were available and providing more or less equal services at no charge. The individual parent or student had to take the active role by taking advantage of the opportunities offered by a public education system. It was clear from the general tone of the reform debate following the Supreme Court’s 1954 desegregation ruling in Brown v. Board of Education, that equality of educational opportunity had evolved into something more than free and freely available public schooling. Coleman identified not one concept important to defining equality of educational opportunity in the modern era, but five. Three were centered on inputs: equality of educational resources (money and the things it is used for— facilities, teachers, smaller class sizes, curricular innovations); the educational or socioeconomic backgrounds of a given student population; and intangibles such as teacher expectations, morale, and commitment that the distribution and interaction of the other inputs produced. Two were centered on what effect these inputs had on the effects or results of schooling, essentially the relative equality of outcomes for people with equal inputs, or its conceptual converse, the inequality of outcomes for people with unequal inputs (Coleman 1974, 10–12). From these raw materials, three central elements emerged to define equality of educational opportunity. First, equality of opportunity was to be judged by looking at what educational outputs were produced for a given set of inputs. The central concern was on what accounted for differences in educational outcomes between racial or class groups—specifically, in the case of the Coleman report, between blacks and whites. These outcomes were generally operationalized as levels of educational attainment, educational aspirations, and, especially, standardized test scores. Second, equality of educational opportunity was to be judged by a school’s ability to produce the same average outcomes for different population groups. This is a critical point—the operational definition refers to aggregate outcomes, not individual ones, a distinction Coleman (1974, 15) made explicit. Third, schools and the state were no longer allowed a passive role—their responsibility was to bring the average educational outcomes of different racial groups into rough parity. Equality of educational opportunity was thus operationally defined into a question of internal efficiency, and the research challenge became to isolate the
50
The Ideology of Education
inputs that had the most impact on outputs. After the Coleman report “it became increasingly the practice, even the demand that equality be measured by school outputs; that is to say, by results of tests of academic achievement” (Mosteller and Moynihan 1972). Inputs mattered only to the extent that they contributed higher outputs. This definition set the context for the most widely known findings of the Coleman report: school resources matter little to outputs; the family and socioeconomic background of students matters a great deal. These findings put schools on the defensive. First, test score gaps between ethnic groups showed they were failing to provide equal opportunities. Second, the common lament of public education—insufficient resources—was removed as a justification for failure. The institutions themselves, not their levels of support, came to be seen as at fault. Rather than changing institutional inputs, the focus of reform shifted towards changing institutions. The Coleman report’s claim that family rather than school characteristics determined educational achievement had existing empirical support. David et al. (1961), for example, found that the primary predictors of educational attainment were, in order of precedence, levels of parental education, the occupational status of the father, family income, and family size. Interestingly, both the Coleman and David studies can be interpreted as supporting the claims of conflict theorists: children of higher status parents do better in school, and because they do better in school, they go on to become people of higher status themselves. In short, schools replicate social hierarchies and inequities. Such interpretations were overshadowed by the subsequent political controversies generated by the debate over the role, or lack thereof, of money in equalizing opportunities between racial and class groups. Following the Coleman report, education production functions—the internal efficiency relationships formalized in equation 2.2—became the dominant form of assessing how schools distribute social and economic opportunities. Across four decades and hundreds of studies, this research has repeatedly replicated the Coleman report’s central claim that school resources have little impact on outcomes (for surveys see Burtless 1996, Hanushek 1997, Verstegen and King 1998).
MONEY AND EDUCATION: THE POLITICS AND POLICY OF EDUCATION PRODUCTION FUNCTIONS RESEARCH
The general policy conclusion extracted from education production functions literature is exemplified by Hanushek’s (1981, 1997) claim
Education and Equality of Opportunity
51
that educational outcomes cannot be improved by “throwing money at schools.” Rather than money, the way to improve student outcomes (and thus equality of educational opportunity) was through institutional reform. Researchers began trying to isolate the organizational practices of educational institutions that consistently produced high test scores with the minimal inputs. Efficiency, the central operating value of the market, thus emerged as a primary criteria for judging equality of educational opportunity. Compared to their private counterparts, public schools—especially large urban systems with high concentrations of minorities—faired poorly in education production functions research. Efficiency’s importance to equality and the seemingly superior performance of private sector schools cleared the way for the popular resurrection of a marketbased education system based along the lines first suggested by Friedman (1955, 1982). Scholars such as Hanushek (1981,1994) and Chubb and Moe (1990) pushed these ideas into the mainstream, where they found favor in part because of their explicit promises of better educational opportunities for those in sub-par public schools. The appeal crossed traditional political divisions, with voucher and charter programs attracting the support of (mostly) liberal Democratic inner-city minority interests, Republican governors, religious private schools, and conservative think tanks (Witte 1996, 1998). The market’s promise of better educational opportunities simultaneously appeals to the practical, sometimes desperate, needs of one group, while furthering the political and ideological preferences of the other. These promises were given academic credence by education production functions research which also solidified the mainstream concept for judging educational opportunity established by the Coleman report: high test scores translated into good opportunities; low test scores translated in poor opportunities. Reforms that led to the former, or at least mitigated the latter, could claim to advance the egalitarian ideal of equality of educational opportunity. As a central claim of the education production functions research is that private—i.e., more market like— schools have higher outputs, making public education more market-like should increase outputs and more evenly distribute educational opportunity (Hanushek 1994, Chubb and Moe 1990, Friedman 1982). The market’s claim to better advance the commonwealth’s core operating value invited a strong reaction. The education production functions literature is anchored in economic rather than democratic theory, and some argue this intellectual foundation hopelessly entangles the research enterprise in the ideological preferences of the market. While the results may mimic good scholarly practice, the underlying
52
The Ideology of Education
intellectual framework and its associated methodological practices tilt the conclusions toward market preferences. The net result is to give academic cover to a blatantly ideological, anti-public school agenda that seeks to institutionalize rather than reduce inequality (Tanner 1998, Verstegen and King 1998). These commonwealth counterarguments attack the key education production function claims on empirical, methodological, theoretical, and normative grounds. The central target is the finding that educational resources are not a significant correlate of school outputs. This is because in addition to justifying the pursuit of market-based institutional reforms, the claim that money does not matter simultaneously insulates such reforms against criticisms that they are rooted in a political agenda that wants to preserve existing inequities in resource distribution to secure the advantages it confers to certain social groups (Alexander 1998a). If such inequities have no impact on the performance or quality of schools, then gross inequities in the distribution of resources has little role to play in addressing equality of educational opportunity. Commonwealth advocates draw on two primary arguments to counter the claim that money does not matter. First, an increasing number of empirical studies undertaken within the education production functions framework find money is positively correlated with school outcomes (Hedges, Laine, and Greenwald 1994; Verstegen and King 1998). Second, the underlying epistemology of education production functions research (drawn largely from economic productivity analysis) is portrayed as having an inherent normative bias for supporting market preferences (Jondrow et al. 1982, Taylor 1994). Economic productivity analysis is theoretically oriented towards for-profit firms. Such organizations ordinarily have one clearly defined overarching goal (profit), one easily quantifiable measure of output (the product), and a series of easily identifiable inputs (the costs of labor and the physical means of production). The overarching goal (profit) is maximized when outputs are maximized for a given level of inputs—the input–output ratio defines efficiency and gives a clear picture of the firm relative to its overarching goal. In contrast, schools pursue multiple goals, and there are multiple ways to measure their outputs. The relative efficiency of schools—how their inputs relate to their outputs—is dependent upon the choice of output measure. The selection of output measure is a normative choice because no objective universal index of education “production” exists. Test scores have long been established as the dominant measure of output in the education production functions literature (Prather and Gibson 1977; Hanushek 1997; Eccles 1999, 6). Yet
Education and Equality of Opportunity
53
replace these with other output measures—especially in time series analysis—and resources take on a much stronger, positive, and significant relationship with outputs. SAT scores over the past 40 years may have fallen as expenditures have risen, but general levels of educational attainment have gone up (Simon and Boggs 1997). A different output measure can thus yield a different inference on the role of educational resources. This means the money does not matter argument is as much a political as an empirical claim; it rests on the subjective judgment that test scores are better indexes of educational “production” than, say, graduation or dropout rates (Eccles 1999). Methodologically, the education production functions research is tied to regression analysis, a statistical technique where inference is wholly dependent upon the underlying quality of the data. As seen in the previous chapter, there are a wide set of options to measure education outputs; these options are not particularly well correlated with each other, and may be picking up different concepts. The inferences taken from regression analysis are also based upon the null hypothesis criterion; it is assumed that no relationship exists between inputs and outputs and the statistical evidence has to be overwhelming to reject this assumption. Given the problems of measurement and the less than clear parallels between corporate and education production function analysis, amassing this overwhelming evidence is going to be hard (Spencer and Wiley 1981). In education production function studies the null hypothesis is that money does not matter, so probabilistically money is more likely to be found to have no relationship with educational outputs. Even when money is a positive predictor, the relationship can be rejected on the grounds of statistical insignificance when a coefficient does not cross an arbitrary Alpha level threshold in a hypothesis test. These are not mere technical matters; they represent the entire case for the claim that money does not matter. For example, Hanushek’s (1997) review of 377 education production functions studies is widely cited in support of the resources-do-not-matter argument. Of these 377 studies, the vast majority reported positive relationships, and a plurality had positive and statistically significant results. The claim that there is no clear or consistent relationship between resources and school outputs is propped up by the roughly third of the studies reporting insignificant results. The argument that money does not matter is thus based almost entirely on what is technically a series of nonfindings (Verstegen and King 1998; Hedges, Laine, and Greenwald 1994; Unnever et al. 2000). The distribution of resources is also a matter of normative concern independent of measurement concerns. Literally valuing some children’s education more than others stands in contrast to social democratic
54
The Ideology of Education
notions of equity (Molnar 1994, 1995). Alexander (1998a, 237) argues the question “does money matter?” should be properly be stated as “does money matter when expended by public schools?” Why? Because “it is only in the context of public education that such a banal question could be posed and given any credibility.” Money might not buy a higher SAT score, but it certainly can buy other things. Inequities in funding affect everything from class size to physical infrastructure to the quality and available quantity of instructional materials, and these sorts of differences translate into inequities in educational opportunities (Peevely and Ray 1989, Necochea and Cline 1996). Money may not be a silver bullet for all of education’s problems, but providing more resources is not the policy equivalent of shooting blanks (Slavin 1999). Even if educational resources have no relationship to test scores, they seem to have important implications for future social economic success—students who are the beneficiaries of greater educational investments tend to do better in life (e.g., Card and Krueger 1992, Rizzuto and Wachtel 1980). Kozol (1991) argues disparities in financial resources represent a set of “savage inequalities” institutionalizing an underclass that is sociodemographically characterized as largely urban, poor, and minority. Given education’s central role in social, economic, and political opportunities, the disparities, in effect, deny certain classes the opportunities given others. If this is the case, we have a “democratic imperative” to pursue equalizing educational funding on the basis of need, because to fail to do so means not just accepting, but supporting, a system that abandons the core democratic value of equality (Rebell 1998). Because the primary method of financing public education is local property taxes, funding inequities are tied to geography, and many of the school systems with the most pressing needs have the least ability to deal with them. Large urban school systems, typically with large minority enrollments, often serve a population dealing with a breakdown in community and family systems, and their ability to compensate is constrained by a weak property tax base. The net effect is that many minorities, especially African Americans, do not have the same educational opportunities as whites that typically attend well-funded suburban schools (Wilson 1988). This is not to say that urban districts necessarily spend less than other school systems, but that, “simply put, the problems confronting urban school districts are bigger, costlier, more numerous and tougher to overcome than those facing most rural and suburban systems” (Olson and Jerald 1998, 9). Market advocates acknowledge these disparities in educational opportunity, and that minority, inner-city residents bear the brunt of
Education and Equality of Opportunity
55
such inequalities. Yet they tend not to link such inequities with resources, instead claiming the organizational arrangements of large urban school districts (monopolistic, rule-bound, bureaucratic) are the source of the problem (e.g., Ravitch 1997). It is not money that provides a solution to such inequities but market forms of regulation. And even if money did play a role in perpetuating these inequities, market advocates can argue their proposed reforms could provide a base-level of financial equity. A voucher system along the lines proposed by Friedman (1955) would provide educational funding directly to parents rather than to the educational institutions themselves, and parents could use these vouchers in the educational marketplace, picking the service provider that best matched their own desires. As the vouchers would be of equal value, the financial disparities that currently exist would be eliminated. As service providers jockeyed for customers, they could supply the desired educational opportunities on a more customized and individualized basis. This would provide the populations suffering the sharp end of any “savage inequalities” a set of options and opportunities they do not currently have (Chubb and Moe 1997). In theory, such a system could be constructed to preserve equitable distribution of educational resources without promoting other generally recognized threats to equality of educational opportunity such as segregation based on class, race, or religion (Brighouse 1994). Such arguments help the market restake a theoretical claim to commonwealth values, but the empirical support of the theory is mixed. The limited voucher programs that exist in the United States have produced an enormous amount of harshly worded claim and counterclaim, but no consensus on whether they have fulfilled the promises of their backers or the fears of their critics (Witte 1996, Peterson and Noyes 1997). Market-based systems in other nations confirm commonwealth fears of promoting greater segregation and inequity rather than overall increases in performance or educational opportunities (e.g., Ritzen et al. 1997, Parry 1996). In the United States, there is no way to empirically demonstrate a market system will advance or constrain educational opportunities because no such system exists in a general sense at the elementary and secondary level (Smith and Meier 1995). In higher education, however, we do have the structural elements of such a system and at least a 40–year history of their impact on public colleges and universities. The best known educational voucher system in U.S. history was created by the “Serviceman’s Readjustment Act,” better known as the G.I. Bill. This federal legislation provided veterans direct financial aid that they could take to the educational institution of their choice—public or private. A number of public colleges and universities
56
The Ideology of Education
objected to this system when it was created, proposing instead that financial aid flow to institutions rather than to individuals. The argument was that if the money went to individuals, private (many religious) institutions with higher tuition rates would disproportionately benefit because the program would, in effect, subsidize a class system. Those who could afford to add to their government-backed stipend would leave behind the public system of universities for the more exclusive private system (Chambers 1968). Despite such concerns, the federal government, followed by many state governments, accelerated programs that sent financial aid directly to students rather than to institutions. These direct aid programs distributed resources on the basis of need, not just veteran status. The justification for this was same equity argument made by those advocating voucher systems at the K–12 level—it would boost educational opportunities for those who most needed them. By the mid-1970s large numbers of students were able to take tax dollars pretty much to the college of their choice, and by the late 1970s “the federal voucher scheme for funding higher education was fully operational” (Alexander 1998b, 167). The result is a rough version of the sort of system advocated by Friedman (1955) and Chubb and Moe (1990). It is, admittedly, a watered down form of such proposals because most public institutions still get direct subsidies from the state (often as a primary source of revenue). Still, there is consumer choice, and government funding increasingly has gone to the individual rather than to the institution. In two separate analyses Alexander (1998a, 1998b) argues private institutions are the clear winners of these voucher programs. In 1996 private institutions enrolled 24 percent of the nation’s full-time students and got 45 percent of the dollars from state-level direct student aid programs. Private institutions received $747 more per student in taxpayer-funded aid distributions than public institutions. These financial disparities are not as troubling as the shift in enrollment demographics that accompany them. The percentage of low income students in private institutions declined between 1975 and 1996, while the percentage of lower income students in public institutions has increased. Alexander concludes that the net effect of vouchers in higher education has been to funnel considerable amounts of taxpayer assistance to private (many religious-affiliated) institutions, and to subsidize a middle-class exodus from public to private institutions. The evidence from the one broad-scale, government-backed, quasimarket educational system currently operating in the United States suggests the “savage inequalities” portrayed by Kozol (1991) will increase rather than diminish.
Education and Equality of Opportunity
57
Though Alexander’s studies suggest public education institutions have good reason to be concerned about a voucher system’s impact on their bottom line, market advocates are skeptical. A frequent criticism of public education—especially large urban systems—is that it concentrates on inputs at the expense of outputs, an imbalance that is argued to be pathological to education systems with large centralized bureaucracies (e.g., Ravitch 1997, Hanushek 1994). Commonwealth advocates respond that this criticism misses the point: the focus on inputs is more than self-interested bureaucrats seeking to maximize their budgets; it reflects a genuine concern about equity. If school districts with larger proportions of low-income, minority, and low English-proficient student populations seek to provide equality of opportunity, they have to set up programs to deal with the challenges presented by their clientele. Bureaucracy is an effect of these challenges, not a cause of performance problems (Smith and Meier 1995). The much criticized bureaucracies of large urban school systems do seem to try to maximize their inputs from nonlocal sources, but they are funneling most of these resources into core instructional programs that reflect the needs of their clientele (Parrish, Matsumoto, and Fowler 1995; Meier, Wrinkle, and Polinard 1997; Necochea and Cline 1996; Smith and Meier 1995). While bureaucracy is not blameless for the problems in education, its culpability may be exaggerated, and its positive contributions to providing equity overlooked.
THE USE AND MISUSE OF TEST SCORES
Education production function research leans heavily on test scores to operationalize education outputs, and commonwealth advocates question whether this emphasis tilts findings towards market preferences (Eccles 1999, 6). In one sense, it is obvious why test scores are employed as the predominant method to measure outputs. Regardless of their substantive pros and cons, standardized tests help determine social and economic opportunities inside and outside education. Tests are often mandated as graduation requirements, are used to allocate higher educational opportunities, and sometimes shape initial opportunities in the labor market. They may not be reliable indicators of human capital or to what extent schools can improve cognitive capacity, but this does not reduce their practical importance. Maximizing test scores, or minimizing test score gaps between social groups, is an obvious way to get empirical leverage on equality of educational opportunity as conceptualized post-Coleman.
58
The Ideology of Education
Given the vagueness surrounding what test scores actually mean, the real question for commonwealth advocates is whether they should matter, or at least, whether they should matter so much. From the market perspective, test scores have a number of important characteristics. They are quantifiable measures of school output that serve as a basis to judge institutional performance. They represent information that can be used to make individual, self-interested choices among service providers. They are also presumed to be a proxy for human capital. This all assumes that test scores provide an objective, valid measure of individual intellectual skills or capacities, and also reveal something about a given educational institution’s ability to produce those skills. Yet the validity of these underlying assumptions has been attacked on both empirical and normative grounds. Educational measurement in the form of standardized testing has two generally recognized functions: (1) to assess individual competence on some dimension of cognitive capacity or skill; and (2) to assign people, on the basis of their test scores, into ideologically acceptable social categories. The first function is the general presumption of the education production functions research, though it has long been argued the second function is the more important, i.e., test scores help justify social inequities by presenting them as the product of individual merit. As Friedenberg (1974, 31) put it, testing “serves chiefly the ideological function of convincing the young that the American social system recognizes and rewards individual competitive achievement.” The claim of test scores to assess individual cognitive competence and the performance of schools in producing intellectual skills and abilities has been repeatedly challenged. For example, the SAT is one of the best known and most widely employed standardized tests. Tests like the SAT (and similar indexes such as the ACT, NAEP, GRE, MCAT, and LSAT) trace their origins to the early work in psychometrics, the same scholarly endeavor that produced the IQ test (SAT and IQ scores tend to be highly correlated). The SAT was primarily designed, and is still used, to measure an individual’s potential for academic performance, not his or her level of education. On this basis it is employed as a way for colleges to sort applicants—i.e., as a direct way to distribute opportunities in higher education. The SAT undoubtedly provides some useful information for such purposes, but it is also clear the SAT is a highly imperfect yardstick of academic potential. One of the first studies to examine the relationship between SAT scores and college performance found that the scores “could predict about a tenth of the variance in grades among freshmen” (Lemann 1995, 87). This modest performance indicates significant limits to the information the SAT conveys about the concept it was designed to measure.1
Education and Equality of Opportunity
59
In fact, it is not clear what standardized tests actually do convey. Perhaps they do tell us something about schools, although exactly what that is has never been identified in a general sense. Even their ability to indicate simple levels of information acquisition—presumably a necessary component of basic literacy and something schools can be held accountable for—is suspect. Arnstine (1995, 51) reported that after an exhaustive review covering a 60-year period, he found only two studies addressing whether these tests are tapping into the concept of individual acquisition and retention of information, and concluded there was no basis to contradict the claim that the, “vast majority of students in schools . . . forget the bulk of the subject matter knowledge they were able to retain long enough to pass end-of-semester examinations.” Standardized tests, in other words, say a lot about how well someone takes tests but not much about what they are likely to know. Adding to questions about the validity of standardized tests as objective indicators of cognitive capacities or intellectual competence are claims that these tests are racially or culturally biased. Racial gaps in test scores are held to be indicative of a bias towards whites and the middleto upper-class. While these claims are difficult to evaluate, the history of standardized tests in providing “scientific” evidence to justify racist beliefs about the intellectual capacities of different ethnic groups is long and disturbing (Gould 1996). Such charges have been leveled in the contemporary era against scholars like Jensen (1969), and Murray and Herrnstein (1994), who used IQ scores and achievement tests to support controversial arguments that genetically-based cognitive capacities were unevenly distributed on the basis of race and class. Such claims rest heavily on the presumption that these tests measure some fixed quantity of aptitude or intelligence, a presumption that seems increasingly contradicted by empirical evidence. That evidence, accumulated from hundreds of studies, indicates test scores are not valid indicators of such properties as fixed intelligence or aptitude. This growing research literature does support the following conclusions: social environment has a large impact on test results; environmental differences rather than genetics are more probable explanations for test score gaps across race and class; test score gaps between blacks and whites virtually disappear when family background is properly accounted for; racial test score gaps correlate with familiarity of white or middle-class culture. These studies, undertaken across a variety of academic disciplines, provide strong, repeated, and consistent evidence that test scores reflect cultural and class differences between racial groups rather than differences of aptitude, achievement, or intelligence (for surveys see Jencks and Phillips 1998, Gould 1996; for a counter view see Murray and Herrnstein 1994).
60
The Ideology of Education
Compounding the problems of racial or cultural bias in the tests themselves are criticisms that tests have been improperly employed by academics. Coleman (1974) operationally defined equality of opportunity as the same average outcomes for different population groups, i.e., when there was no difference between the mean test scores of whites and blacks, equality of educational opportunity would be achieved. This implies equality of opportunity assessments should be based on explanations of variation between racial groups. Instead, virtually all of the studies used to make such assessments—the bulk of the education production functions literature—seeks to explain variation within racial groups. The difference is critical because it can lead to opposite results and inferences. When the same data used by the Coleman report are used to examine group rather than individual differences, the results and conclusions reverse themselves. The reanalysis by McPartland and Sprehe (1973), for example, found considerable effects for educational resources, estimating that if whites attended typical black schools their achievement test scores would fall between a fifth and two-fifths of a standard deviation. The differences between Coleman’s, and McPartland and Sprehe’s analysis of the same data is due to the fact that they are using it to answer different questions: What causes individual achievement? (Coleman); and, What causes achievement differences between races? (McPartland and Sprehe). The answers to these questions derive from different pieces of variation in the data, and only the latter is oriented to equality of educational opportunity as defined by Coleman (1974). The Coleman report, in short, mismatched measurement and methodology with its conceptual target. The education production functions literature methodologically followed Coleman’s influential lead, in the process repeating Coleman’s mistake of failing to properly operationalize equality of educational opportunity (Miller 1981). The argument that money does not matter to equality of educational opportunity, and its corollary that market-based institutional reforms will matter, is built to a considerable extent on studies that have the “wrong” explanatory variable. There are not only doubts about how the education production functions literature operationalizes equality of educational opportunity. Others argue much of the research in this tradition downgrades the potential role of resources (and upgrades market-based institutional variations) on the basis of little more than normative preferences. For example, in a typical education production function study, all correlations between individual achievement and school resources that are also related to race and class are assumed to be spurious. This means a correlation between socioeconomic context (e.g., levels of poverty in a
Education and Equality of Opportunity
61
school district) and level of school resources (e.g., per pupil expenditure) can be used to dismiss a correlation between resources and educational performance. This is because the latter relationship is assumed to be a product of socioeconomic context’s relationship to resources and outcomes (e.g., Hanushek et al. 1994, Unnever et al. 2000). As Miller (1981) points out, these are value-laden assumptions. Any disputed variance in achievement differences is being automatically attributed to racial background rather than to school resources. This downgrades by assumption the argument that class or social background determines the level of resources available to a school, and that these resource levels in turn determine student outcomes. A number of prominent critics of education production functions research point to such arguments to support their claims that it is as much an ideological agenda as good social science (Alexander 1998a). The economic-based theoretical foundation, the epistemology and methodology, the measurement and conceptualization problems all serve common ends: take the focus off resource inequities, shift policymaking towards institutional reforms that will insulate the haves from the claims of the have-nots, and allow education systems to go on replicating existing social hierarchies (Alexander 1998a, Rebell 1998, Miller 1981). Equality of opportunity may be a universal goal in the abstract, but in reality the game is rigged, and the “uneven playing fields reflect a dark unspoken sense that other people’s children are of less inherent value than our own. Now and then, in private, affluent suburbanites concede that certain aspects of the game may be trifle rigged to their advantage” (Kozol 1991, 177). The academic studies concluding that money does not matter helps remove the elitist sting of such admissions, and the result for such people is “uncontaminated satisfaction in their victories. Their children learn to shut from mind the possibility that they are winners in an unfair race, and they seldom let themselves lose sleep about the losers” (Kozol 1991, 177). These arguments suggest that the conflict theorists are right about the real purposes of education, that the market ideology wants to preserve those purposes, and that market regulatory mechanisms will achieve the goal of preserving the inequalities in the status quo.
EQUALITY OF EDUCATIONAL OPPORTUNITY IN TEXAS
The market’s claim to the core commonwealth value thus boils down to the argument that equity and efficiency are compatible, if not synonymous. Market mechanisms will make schools more output focused,
62
The Ideology of Education
competitive pressures will drive up the overall level of those outputs, choice will allow broad access to schools that produce high outputs, and the result will be better opportunity for all (Chubb and Moe 1990, Hanushek 1986, 1994, 1997, Peterson and Noyes 1997; Teske et al. 1993, Maranto, Milliman, and Stevens 2000). Criticisms of this claim suggest it rests on market-biased economic theories and is supported by empirical studies that have institutionalized these biases in their epistemological, methodological, and conceptual frameworks. Recognize these biases and account for them, suggest the critics, and not only does money matter, but market mechanisms are also revealed as indifferent to equality of educational opportunity, and quite possibly hostile to it (Alexander 1998a, Tanner 1998, Verstegen and King 1998, Kozol 1991, Henig 1994). Tables 3.1 and 3.2 report analyses of educational outcomes in Texas school districts that attempt to embody the empirical case for both sides of this argument. For several reasons, Texas makes a good statistical laboratory for this exercise. First, it provides enormous variation at the district level. Texas is a large and diverse state with more than a thousand school districts and almost 4 million students. The districts range from one-school operations with less than 100 students (more than a fifth of Texas school districts have only one school), to large urban districts with hundreds of schools and 100,000s of students. The socioeconomic and racial mix of student populations across these districts varies and is equally diverse. Second, and more importantly, there is variation in the structure of local governance. Like most states, Texas delegates much of the regulation of public education to local education agencies (LEAs) that have broad policy powers. By far the most dominant institutional model for an LEA is the Independent School District (ISD). ISDs are governed by a locally elected board of trustees. Typically the key role of the school board is to make broad policy decisions and hire a superintendent to head a centralized administration that clarifies and enforces these policies. ISDs are thus a prototypical example of the “one best system” of school organization that market advocates find objectionable: a monopoly public service provider run by bureaucrats (Tyack 1974, Chubb and Moe 1990, Hanushek 1994, Texas Education Agency 2000a). There are, however, alternative LEA institutional structures, including charter schools. Charter schools were authorized by statute beginning in 1995. Under Texas law, charter schools are public schools substantially released from state education regulations and are separate from local school districts. The number of charters is limited by law, and a requirement to be granted a charter is sponsorship to be an institution of higher
Education and Equality of Opportunity
63
TABLE 3.1 Test Score Pass Rates in Texas School Districts Variable
Coefficient
Confidence Limit
Leamer Bounds
% Economically disadvantaged
–.11* (.02)
–14:–.07
–.11:–.08
% White students
.07* (.02)
.03:.11
.07:.10
Instructional expenditures per student (× 1,000)
.50 (.04)
.001:1.0
.20:.50
Students per teacher
.05 (.14)
–.22:.33
–.10:.20
% White teachers
.04* (.02)
.003:.08
.01:.04
Teacher experience
.39* (.11)
.17:.60
.19:.39
Bureaucracy
–.34* (.15)
–.64:–.03
–.34:–.29
Charter school
–19.5* (1.5)
–22.6:–16.5
–19.5:2.4
Constant
72.6 (4.07)
N
1,104
Adj. R-2
.47
Unstandardized coefficients (standard errors) reported. *p < .05.
education, a nonprofit organization, or a governmental entity. In the 1999–2000 academic year data used for the studies reported in Tables 3.1 and 3.2, there were 142 charters operating 176 schools and serving more than 25,000 students. Charter schools have no guaranteed clientele and essentially function as independent units of governance that inject an element of entrepreneurship into education regulation. Groups of parents, for example, can band together to form a nonprofit and petition for a charter, and a handful of charters operate more than one
64
The Ideology of Education TABLE 3.2 Equity in Educational Outcomes in Texas School Districts
Variable
Coefficient
Confidence Limit
Leamer Bounds
% Economically disadvantaged
.005 (.03)
–05:.06
–.0007:.007
% White students
.15* (.03)
.10:.20
.07:.15
Instructional expenditures per student (× 1,000)
2.9* (1.0)
2.0:4.0
.001:3.0
Students per teacher
.54* (.22)
.10:.98
.005:.54
% White teachers
–.13* (.03)
–.18:.–08
–.14:.001
Teacher experience
–.06 (.17)
–.40:.27
–.13:.19
Bureaucracy
.27 (.28)
–.28:.82
–.29:.36
Charter school
–5.2* (2.3)
–9.9:–6.3
–739:3.8
Constant
68.9* (6.5)
N
830
Adj. R-2
.07
Unstandardized coefficients (standard errors) reported. *p < .05.
school. They are LEAs with market-like characteristics, and as such they are useful to make comparisons between the traditional “one best system” institutional structure represented by ISDs, and the more autonomous clientele-driven service providers market supporters advocate. Charter schools in Texas no doubt fall considerably short of a market ideal—they are far from free of state regulation or from interference of the entity that sponsors them. Nonetheless, they are a viable, marketlike alternative that provides an exit option from the local school dis-
Education and Equality of Opportunity
65
trict. As such they provide a path to empirically assess if more marketlike institutions can provide a better distribution of educational opportunities. Texas charter schools may be a particularly good test of this because they are roughly evenly split between charters serving areas where student populations are seen as “at risk” or “not-at-risk.” In other words, they are providing educational services for a substantial number of students whose educational opportunities in the traditional public school system, for various reasons, are not considered particularly good.2 Table 3.1 reports the results from a regression model that is typical of the education production function literature. The outcome variable is the percentage of students in the district who passed the Texas Assessment of Academic Skills (TAAS), a standardized test designed to measure skills and knowledge associated with the statewide curriculum. The test is mandated across several grades and also serves as an exit exam that is a prerequisite for a high school diploma. The explanatory variables are typical of those employed to represent the likely socioeconomic and school-specific determinants of educational outcomes (i.e., they represent F and S in equation 2.2, and the I, or innate capacities of students, are assumed not to be a major concern because of the aggregate unit of analysis). The three variables of particular concern here are the role of money (measured here as instructional expenditures per student), the size of the LEA’s bureaucracy (percentage of staff employed in central administration), and whether or not the LEA is a charter school. These serve as the basic test of the market argument that institutional structure matters to educational outcomes, while money does not (see methodological appendix for complete variable descriptions and other statistical details). The results in Table 3.1 confirm the market arguments, with one significant exception. Like the bulk of other educational production function studies, the money variable in this analysis positive but statistically insignificant (it does not meet the .05 Alpha criterion). The size of the bureaucracy is negatively related to outcomes, i.e., larger bureaucracies are associated with fewer students passing the TAAS exams. Both findings fit market expectations. Institutional structure matters (large centralized bureaucracies are shown to be a drag on performance), money does not. The significant exception is the variable for charter schools, which is negatively related to pass rates. The size of the coefficient for charter schools is substantial, estimating that if all else is equal, TAAS pass rates are about 20 percentage points lower in charter schools compared to ISDs. Some of this may be attributable to charter schools that are oriented to at-risk students, though much of the
66
The Ideology of Education
impacts of such student body characteristics should be accounted for by the control variables included in the model (specifically the percentage of students that are economically disadvantaged and the percentage of students that are white). It is also possible that missing data ended up creating a biased sample of charter schools (74 are included in the analysis). This seems unlikely, however, as data selection was deliberately structured to put charter schools in the best possible light in terms of the TAAS pass rate measure (see methodological appendix). Confined to the traditional parameters of education production function analysis (the first column of Table 3.1), the results confirm the oft-made claim that money does not matter, and that bureaucracy negatively affects school performance, but fails to confirm the argument that market-based institutional prescriptions will improve performance. When the analysis steps beyond the traditional confines of education production function analysis, the market argument faces more problems. Table 3.2 uses the same variables and the same data as Table 3.1. The only difference is the outcome variable, which is an index of equity formulated by Kenneth Meier and other scholars connected to the Texas Excellence in Education Project (see Meier et al. 2000, Meier et al. 1997).3 The equity measure “Q” formulated in TEEP-related scholarship is based on TAAS pass rates, and its basic objective is to capture in a single number the comparative performance between the three primary racial groups in the Texas public school student population (whites, blacks, and Latinos). It is calculated thus: Q = (Pblack/Latino)/PAnglo This is a ratio of black and Latino pass rates weighted by their population and divided by white pass rates.4 When Q takes on the value 1, it indicates that minority pass rates on the competency exam are exactly the same as white pass rates. Scores below 1 indicate that minority groups in a given district are passing these exams at lower rates than whites, and scores over 1 (these are quite rare) indicate the opposite. Q scores are susceptible to being distorted by small numbers in any given racial group (especially whites—a school with one white student has a 100 in the denominator if the student passes, and if the student fails, Q cannot be calculated because the denominator is zero). Accordingly, I include only Q scores that fall between .5 and 1.5, and for ease of interpreting regression coefficients multiplied Q by 100 (see methodological appendix for a detailed justification of the data selection procedure). The end result being a 50 to 150 index where higher scores indicate minority students are doing better relative to white students. As the vast
Education and Equality of Opportunity
67
majority of scores fall between 50 and 100, higher scores can be practically interpreted as higher levels of equity between the three racial groups in terms of TAAS pass rates. Using Q as the outcome variable, the analysis reported in Table 3.2 is thus an education production function that attempts to operationalize in the outcome variable the concept of equality of educational opportunity as originally defined by Coleman. The results for the key variables of interest, again with one exception, reverse the findings reported in Table 3.1. First, money matters. The variable is significant by the traditional Alpha .05 criterion, and estimates that for every $1,000 increase in per pupil instructional expenditures the equity index increases by 2.9. Second, institutional structure does not matter, at least in terms of the size of the bureaucracy—the bureaucracy variable is statistically insignificant. Third, the charter school variable stays negative and significant. All else equal, charter schools are associated with 5.2 fewer points on the equity measure compared to ISDs. The findings clearly indicate that when equality of educational opportunity is operationalized as differences between ethnic groups, the best known claims of the production function literature can empirically be contradicted. The claim that institutional structure rather than money is the key input variable is weakened further when the traditional confines of regression analysis are expanded to include information about coefficient stability and the role of the key variables in separating high performing from average performing school districts. The expenditure variable is not statistically significant in Table 3.1, but it is consistently positive in both confidence limits and Leamer bounds (the same is true in Table 3.2). This suggests that the positive coefficient is a reliable indicator of a real relationship; it is simple too consistent a positive predictor to dismiss as indistinguishable from zero on the basis of an arbitrary Alpha level. In contrast, the variables designed to capture institutional structure cross zero in confidence limits, Leamer bounds, or both. This suggests the findings on institutional structure are dependent upon model specification and data selection, i.e., there is a reasonable likelihood they are a statistical artifact rather than an accurate estimate of a true relationship. The only exception is the bureaucracy variable in Table 3.1, which is consistently negative. Large bureaucracies are consistently negatively related to TAAS pass rates. This result can be used to support market arguments, but as Smith and Meier (1995, 57–60) argue, large bureaucracies are also a sign of student populations facing a broad variety of problems—not just poverty, but limited English proficiency, family dysfunction, etc. The bureaucracy variable may be picking up this impact rather than the corrosive effect of administration on
68
The Ideology of Education
performance. The analysis does not provide an empirical answer to which inference is more accurate. A SWAT analysis, however, does little to bolster market claims. Tables 3.3 and 3.4 show the impact on the three key variables as LEAs with higher pass rates/equity scores gradually allowed to have more weight in the analyses reported in Tables 3.1 and 3.2. Both for overall pass rates and equity scores, expenditures play a greater role in separating high performing districts from average performing districts (in the equity model the expenditure variables increases up to 240 percent of the original coefficient). This suggests higher expenditures—more money—plays a role in high performing school districts. By looking at what separates high and average performers, the SWAT analysis points decisively towards the perspective that school resources matter. The SWAT results for the two institutional variables are mixed. The charter school coefficient rapidly decays toward zero, and in the equity model actually crosses zero and becomes a positive predictor in the last two iterations. Overall, these results tend to support the inference that charter schools have no real impact on equity or performance. Whatever the pros or cons of charter schools, all else equal, this analysis indicates they are not distinguishable from public schools in terms of TAAS pass rates or the Q index. The results for the bureaucracy variable tell two different stories. In the SWAT analysis of TAAS pass rates (Table 3.3), the bureaucracy coefficient steadily declines toward zero, indicating the size of the bureaucracy does little to enhance or detract from high performance. In the SWAT analysis of equity (Table 3.4), the bureaucracy variable increases fairly dramatically as high equity school districts are allowed to have more weight in the analysis. The bureaucracy coefficient becomes as much as 440 percent larger than reported in the original analysis reported in Table 3.2. This is interesting because the original coefficient is positive, i.e., it indicates all else equal, larger bureaucracies are associated with more equity. The SWAT analysis suggests that high equity districts tend to have larger centralized bureaucracies than average equity districts. This finding fits with Smith and Meier’s argument (1995, 57–60) that bureaucracy is, at least partially, an undeserving scapegoat in education. The programs that large bureaucracies are established to run do increase equality of educational opportunity between racial groups, regardless of their impact on overall pass rates. While these findings contradict market claims, a word of caution is appropriate. Figures 3.1 and 3.2 show the actual pass rates/equity scores plotted against those predicted by the models reported in Tables 3.1 and 3.2. The SWAT analysis is based on treating observations well above the lines drawn through these data points as high performers.
TABLE 3.3 SWAT Analysis of Test Score Pass Rates
Iteration
Intructional Expenditures Per Student
Bureaucracy
1
1.0
.91
2
1.2
.82
.95
3
1.4
.73
.92
4
1.4
.58
.89
5
1.6
.47
.89
6
1.6
.35
.80
7
1.6
.20
.74
8
1.6
.11
.66
9
1.6
.14
.54
Charter School 1.0
Cells are proportions of the coefficients reported in Table 3.1.
TABLE 3.4 SWAT Analysis of Equity
Iteration
Intructional Expenditures Per Student
Bureaucracy
Charter School
1
1.03
1.18
.94
2
1.07
1.37
.84
3
1.13
1.59
.75
4
1.17
1.80
.63
5
1.27
2.1
.50
6
1.34
2.4
.34
7
1.51
2.8
.17
8
2.4
4.4
.CZ
9
1.9
3.1
.CZ
Cells are proportions of the coefficients reported in Table 3.2. CZ = crossed zero (slope coefficient changes direction).
70
The Ideology of Education
The caution comes from the scatter around the lines. In both cases there is a good deal of spread around the line, suggesting the model is producing estimates that are not a tight fit with the observed data. Slight alterations in model specification (adding or dropping a variable) could change the picture presented by these scatterplots, and conceivably alter what is being considered a high performer, and thus also alter the results of the SWAT analysis. Even here, though, the impressions given by Figures 3.1 and 3.2 diverge from the traditional conclusions of education production functions research. Overall, the model produces estimates that are a better fit with the observed data
FIGURE 3.1 Actual Versus Predicted TAAS Pass Rates
Education and Equality of Opportunity
71
when TAAS pass rates are the outcome variable rather than the equity index (though not directly comparable, this is also reflected in the adjusted R-squares reported in Tables 3.1 and 3.2). The typical education production function model shown in Tables 3.1 and 3.2 does a reasonable job of predicting overall performance using a standardized test measure, but it is less accurate when the outcome variable is fashioned to reflect Coleman’s conception of equality of educational opportunity. Whatever education production functions tell us about overall performance, this does not necessarily apply to equality of educational opportunity. This suggests that the efficiency so central to the
FIGURE 3.2 Actual Versus Predicted Equity
72
The Ideology of Education
education production functions literature is not the same thing as equality of educational opportunity as defined by Coleman. Annual evaluations mandated by the Texas State Board of Education add some anecdotal support to the inferences taken from Tables 3.1 through 3.4. These evaluations indicate that students in charter schools serving at risk populations graded them more highly than the public schools they attended in no small part because charter schools tended to provide smaller class sizes than public schools. Students in charter schools serving not-at-risk populations were less satisfied, reporting that their schools were not significantly better than the traditional public schools they had previously attended. In particular, they expressed frustration with the limited choices of class offerings at charter schools. Common to both the positive and negative assessments of charter schools is their basis in resource-related inputs (smaller class sizes and broader curriculum offerings do not come cheap). Students believe they are getting a better educational opportunity from a charter school when its resources—in the form of smaller classes and more teachers—are seen as significantly better than the locally available alternative. Students tend to downgrade the improvement when these differences are smaller, or in the form of a less diverse curriculum, worse than the public schools they previously attended (Texas Education Agency 2000b). In so many words, students also say that money matters. Not too much should be made of anecdotal evidence, and I make no claims that the more general analysis reported in Tables 3.1 through 3.4 is definitive. The study is of a single state (though a large and diverse one), the institutional variation from the traditional organizational form is small and of one type, and while the usual disagreements over model specification and inference may be minimized by adopting confidence limits and Leamer bounds as co-equals with null hypothesis test, they are unlikely to be eliminated. Nonetheless, the study does present itself as representative of what is likely to happen in a production function study when the outcome variable reflects the concept of equality of opportunity conceived by Coleman, and the normative and epistemological criticisms of the education production functions approach are explicitly accounted for. When this is done, market-like institutional arrangements are more likely to be found harmful to equality of educational opportunities rather than promoting them, and school resources—money—matters. If both findings are generalizable, the internal efficiency argument as conceived within the market belief system loses its key empirical supports and with it the claim to the commonwealth’s core operating value.
Education and Equality of Opportunity
73
CONCLUSION
Even if the external efficiency claim (school outputs determine aggregate economic success) is suspect, there still remains a powerful justification for pursuing institutional reforms based on the internal efficiency claims of market advocates. The central claim is that market-like educational institutions will promote equity, not just efficiency. If market-based regulatory forms can better distribute educational opportunities, they align themselves with the almost universally accepted republican aims of public education. This claim has received considerable support from education production functions research. This research, however, has been criticized for having in-built market biases in its underlying theory, and in the conceptual and epistemological approach at the heart of the production function research enterprise. Picking through these issues reveals a wide set of disagreements and criticisms on research choices on measurement and methodology, and also of the normative underpinnings involved in these choices. Taking these criticisms into account, a general empirical test using Texas school districts does not support market claims. This test suggests the foundations of the internal efficiency justifications for market-based reforms of public education, like the external efficiency justifications addressed in the previous chapter, lean on normative beliefs as much as empirical evidence.
This page intentionally left blank.
4 Institutional Structure and Educational Goals
T
he arguments and analyses in the preceding chapters empirically challenge the market’s arguments in several ways. Value systems seem to do a better job of predicting aggregate productivity than test scores and the useful knowledge and skills they are held to represent. These test score measures in turn seem to be at least partially reflecting social inequities related to class and race. Much of the epistemology underlying research supporting the market case has normative roots, and when this is recognized and accounted for, the empirical findings that support calls for market-based institutional reform moderate or are even reversed. At a minimum, the functionalist perspective of education—the notion that schools are in the business of transferring technical blocks of knowledge that are socially and (especially) economically useful—is highly incomplete. Yet in raising these challenges to the market, the preceding analyses have said more about what schools do not do, rather than what they do. At least in an aggregate sense, schools do not seem to be in the business of creating stocks of human capital that lead to appreciable gains in economic productivity, and market-based organizational structures do not help schools more equitably distribute educational opportunities. This leaves open the possibility that the mission of schools is more cultural than functionalist and that market reforms may be more valueladen than utilitarian; but if schools are smuggling values rather than transferring technical knowledge and skills, little thus far illuminates specifically what values are being smuggled, or how market-based institutional reform might affect those values. Such an investigation requires more focus on what schools are doing rather than what they are not doing. A question central to the purpose of this book is still
75
76
The Ideology of Education
unanswered: What are the goals of schools and how are market reforms of education likely shape those goals? This is the question that this chapter seeks to answer.
THE “ARE” AND “OUGHT” OF SCHOOL GOALS
The debate between market advocates and their opponents over what schools ought to do is often framed as a set of differences about means rather than ends. The Jeffersonian ideals most visible in consensus theory are, in the abstract, universally embraced. These universals include education’s importance to underwriting the democratic process and equitably distributing social and economic opportunity. While such goals are easily embraced as abstract “oughts,” they become highly controversial as they are translated into policy specifics. These controversies make clear that the conflict between the market and the commonwealth is not a disagreement about the best technical means to achieve a commonly desired end but a fundamental philosophical difference about the primary purposes of public education. Chubb and Moe (1990, 53–55 and 78–79), for example, point out that public schools tend to have a lot of goals, many of them vague, contradictory, or mutually exclusive, and argue there is nothing particularly sacrosanct about any of these goals, that whatever objectives schools pursue is simply a product of their controlling agents. The central problem for public education is the number of controlling agents, which indirectly includes any constituency able to get the democratic process to respond to its policy preferences. As public schools must respond to the preferences of various powers in democratic bodies and their bureaucratic management mechanisms, they end up trying to conform to a bewildering tangle of regulations and top-down mandates. An advantage claimed for more market-like schools is an ability to cut through this clutter and provide a clearer and more focused mission (Chubb and Moe 1990). This mission clarity is obtained adopting the demands of education’s clientele, i.e., parents and students, as the primary organizational objective. Institutionalizing this connection between clientele demand and school mission is thus justified by a standard functionalist conception of education (schools have no immutable goals, their job to respond to whatever demands are generated by a dynamic task environment) and the market conception of who should ultimately regulate an organization’s goals and behavior (the customer). Yet this approach to defining the mission of schools by consumer preference stands in stark contrast with a long philosophical tradition
Institutional Structure and Educational Goals
77
surrounding public education. At least since the Founding generation, public schools have been justified with a commonwealth vision—they are repositories and replicators of democratic values, and as such should be oriented towards the polity and not the individual student or parent (Pangle and Pangle 1993, Kaestle 1983). Their primary mission is to advance the democratic interests of the state, not the self-interest of the student. Certainly, this sort of mission seems to inform and drive the arguments of theorists and philosophers of education such as Dewey (1929), Guttman (1986), and Callan (1997). And as articulated by such advocates, this mission is immutable—it is the central justification for the existence of a mass system of public education, and its formulation and execution at a given time lies properly in the hands of democratic institutions and processes. In contrast, the latter are what Chubb and Moe (1990) and other market advocates (e.g., Lieberman 1993, Gerstner 1995, Osborne and Gaebler 1993) explicitly or implicitly charge with being the primary source of problems with the public education system. In short, different ideological beginnings—normative conceptions of what schools ought to do—point towards different determinants of public school objectives and different preferences on how schools should be organized to achieve those objectives. As organizational structure is built around the ideologically preferred source of organizational mission, this raises the possibility that different institutional arrangements will produce different educational objectives. If schools pursue goals that reflect the preferences of their controlling agents, any change in the relative influence of those controlling agents means that goals can also change. All that is required for such change are controlling agents with heterogeneous preferences—shifts in power among these agents will translate into different preferences being reflected in school goals. Commonwealth and market advocates generally agree on this logic but disagree about who are the appropriate controlling agents. Commonwealth advocates object to reorganizing public education to respond largely to small homogeneous clienteles because the goals such consumers impose on schools have a very mixed historical track record. Most obviously, education consumers have wanted (and in some cases received) segregation by race, class, religious indoctrination, or other goals that stand in direct contrast to the immutable republican mission described by scholars such as Callan (1997) and Guttman (1986). The answer to what objectives schools ought to be pursuing is thus clearly linked not only to ideology, but also directly to the differing organizational structures that reflect and institutionalize ideological values. Indeed, this seems to be a direct justification employed by Chubb and
78
The Ideology of Education
Moe (1990, 53–55) for adopting market-based institutional reforms, and a primary argument employed by those who resist them (e.g., Cookson 1994, Henig 1994). The former sees more market-like arrangements as more flexible and better able to serve a variety of educational goals by focusing on a core, client-driven mission. The implication is that the goals of schools ought to be clear, focused, and springing from local market demand, not universally imposed from above. Institutional arrangements that shun a multiplicity of goals in favor of more focused response to a variety of market demands fit the bill. Those with commonwealth sympathies fear that some of those client-driven missions may result in segregation by race, class, or religious preference, threatening the common school ethos that justifies the school system’s existence as a vital buttress of democracy. The implication is that the goals of schools ought to be the product of a more communal, democratic process, not narrowly tailored to a niche of preferences that may or may not fit with the liberal democratic ethos. But while these sorts of differences are at the heart of the contemporary education debate, the consideration of school objectives has remained resolutely in the realm of “ought” rather than “is.” There has been remarkably little attention given to the empirical examination of whether shifting institutional arrangements also shifts the primary objectives schools pursue. Yet both market and commonwealth advocates argue that this should be the case, that institutional arrangements should affect not just how schools pursue their mission but help determine what that mission is. Given the potential implications for the social-democratic justifications used to construct and defend public education systems, this seems to be at least as important a question as the impact of institutional arrangements on the achievement outcomes that seem to be the central concern of most education policy research.
THE IMPORTANCE OF INSTITUTIONAL STRUCTURE— GETTING GOALS FROM THE TOP OR THE BOTTOM
There are two basic views on how institutional structure shapes the specific educational objectives pursued by schools. One view argues that external controls can be effective in shaping the goals schools pursue through law (e.g., prohibition of racial segregation) or regulatory policies such as teacher certification standards, curriculum requirements and the like (Smith and Meier 1995, 93–106). This perspective fits well with commonwealth preferences. As agents of the state serving common ends, schools should be organized to respond to the law and to the pref-
Institutional Structure and Educational Goals
79
erences of the broader citizenry, implying a top-down, loosely hierarchical, organizational arrangement is most appropriate for public education. At the top of this hierarchy are state constitutions, which provide the general outline of public education’s mission and set up its basic system of governance. Generally speaking, state constitutions grant state governments the central power over public education, and instruct them to exercise this power to help achieve common ends. For example, Article XIII, Section 1 of the Minnesota Constitution specifies that, “The stability of a republican form of government depending mainly upon the intelligence of the people, it is the duty of the legislature to establish a general and uniform system of public schools. The legislature shall make such provisions by taxation or otherwise as will secure a thorough and efficient system of public schools throughout the state.” Similarly, Article IX of the California Constitution states that, “A general diffusion of knowledge and intelligence being essential to the preservation of the rights and liberties of the people, the Legislature shall encourage by all suitable means the promotion of intellectual, scientific, moral, and agricultural improvement.” Thus state constitutions justify the existence of public education in explicitly republican terms, imply that any specific educational objective serve broad social-democratic purposes, and identify state government as the key controlling agent (Rebell 1998). The top-down approach that typically finds support among commonwealth advocates thus has its roots in the broad (often vague) constitutional provisions that justify public education and place its control under the institutions and processes of representative democracy. In institutional specifics, this suggests the goals of public education are properly established by representative bodies and then externally imposed on schools. Accordingly, what schools should do is properly an issue for state legislatures or local proxies, such as school boards, that have been specifically delegated policy powers by law or constitutional provision. The management mechanisms of these controlling agents (laws, rules, regulations) are assumed to be strong enough to exert influence over the day-to-day operation of schools, over what actually goes on when the classroom door closes. Either by instilling particular values through the process of becoming professionally qualified to administer or teach in a public school, or more directly in mandates or constraints on what and what not to do in the classroom, the top-down approach shapes the mission of education. The second perspective is associated with the market and argues that regardless of their constitutional roots, top-down external controls are going be of limited use in public education. To begin with, democratic processes and institutions are not particularly good at formulating
80
The Ideology of Education
clear goals, often issuing vague and contradictory missions to public agencies and leaving state and local administrative agents a good deal of autonomy to determine what objectives will take priority. Given such an area of gray, self-interested education bureaucracies will determine what schools do, not the democratic process nor laudable, but highly abstract, constitutional principles. Even if clear, specific missions were thrown up by the democratic process, it is unclear they could be achieved. Every layer in the educational hierarchy has a limited ability to monitor or control the core purposes and actions of the layer below it, a situation that diffuses responsibility and invites shirking. This perspective suggests that a more bottom-up organizational structure is preferable because it will avoid concentrating the power to determine school goals into the hands of whatever groups manage to effectively mobilize in the political arena. Institutionalizing a direct connection between customer demand and service provision passes power to individual interests while also allowing homogenous interests to aggregate and deliver clear messages to a local service provider. This clarifies mission priorities for the education supplier and, with an exit option available to customers, immediate accountability for failing to achieve them. The net effect is to undercut the “bureaucracy’s” or the “system’s” ability to shape educational mission on the basis of its own self-interest and to turn this power over to parent and student consumers. This second perspective involves an element of the unknown. If the consumer is granted the power to shape the mission of education, what will that mission be?
MARKETS AND GOALS
Scholars across several disciplines have proposed explanations of how institutional structure can determine organizational goals. One of the more common arguments is that organizational environment shapes organizational culture, which in turn creates norms that regulate behavior, structures intraorganizational relationships, and affects individual commitment to the organization and its objectives. Private and public school teachers, for example, operate in different organizational environments and have been found to have systematically different value orientations towards education (e.g., Reyes and Pounder 1993). While the cultural approach to understanding what organizations do and why they do it is an important focus of organizational theory scholars, it has had a mixed impact on market arguments and the broader education debate. While borrowing bits and pieces from orga-
Institutional Structure and Educational Goals
81
nizational culture scholarship, justifications for market-based reforms in education rarely stray far from the basic principles of public choice. Chubb and Moe (1988, 1990), for example, authored what is probably still the best known argument on the theoretical link between institutional characteristics and school goals. They begin with one of the more consistent observations of the organizational theory literature, i.e., that complex environments tend to manifest themselves in complex organizational structures. Public schools have highly complex environments characterized by a large and heterogenous constituency stretching far beyond parents and students. Particular interests within that constituency vie with each other to get the democratic institutions controlling educational policy to respond to their preferences. Everyone from textbook manufacturers to teachers’ unions to religious groups seeks to get their agenda reflected in education policy decisions, and in many cases the democratic process is at least partly responsive to these varied and often competing interests. This environmental complexity is reflected in the broad range of regulations, standards, methods, and mandates issued by state legislatures and school boards. Clarifying and prioritizing these into clear mission objectives is difficult because for schools externally controlled by representative bodies voice is used as the primary signal of dissatisfaction (Hirschman 1970). This is the right to participate in collective decision making, the fundamental component of a classical conception of democratic citizenship. Any citizen has a right to weigh in at a public meeting, and any group of common interests is free to mobilize in support of particular policies or elected policymakers. While the democratic process guarantees and protects the use of voice, it makes no promises on how (or even if) it will be reflected in policy decisions. Those who are most likely to gain effective leverage over education through the voice option are well organized interest groups (e.g., teachers’ unions) who can deploy consistent and sustained pressure on the political system through collective action. These groups are not necessarily seeking the same educational objectives as parents and students. Chubb and Moe argue the basic institutional framework of external control by democratic bodies and their bureaucratic management mechanisms are common to virtually all public schools, and so are the goalrelated consequences of being organizationally wedded to a broad and heterogeneous constituency. As a consequence of the democratic process, the mission of public education will reflect the mish-mash of preferences that have had some success in using voice to elicit a response from democratic institutions. On these grounds, the comparison between differing institutional frameworks needed to assess the impact
82
The Ideology of Education
on organizational mission has to be between public and private schools.1 In contrast to public schools, private schools are argued to have much less complex environments. They have small and homogenous constituencies—their clientele—with an exit option. In contrast to voice, exit signals dissatisfaction according to the principle that “if we don’t like it, we will leave,” and presupposes a business-like conception of government (McCabe et al. 1999, 368). In theory, exercise of the exit option translates into immediate and clear signals on organizational mission and performance—if a school pursues objectives at odds with the preferences of its clientele, it soon finds itself without a clientele. The less complex environment and the more direct mechanism used to connect controlling agents to education results in a clearer and more focused core mission. These contrasting pictures of institutional structure—public education system as top-down and private schools as bottom-up—lead to the hypothesis that organizational mission priorities will differ between public and private schools as a result of this institutional variation. Simply by their existence, Chubb and Moe argue, private schools are demonstrating that they are better at clarifying and achieving some goals for some people better than public schools. The big question is, what goals are they pursuing, and how different are they from the goals of public schools? As Chubb and Moe (1988, 1990) put it, “A church school that attracts students on the basis of religious and moral training almost surely outperforms the local public school in this dimension. But this says nothing about their relative effectiveness in transmitting democratic values or an appreciation of cultural values.” If the real purpose of schools is to “smuggle values,” what values are private schools smuggling? Are they just doing a better job of the sorts of things public schools universally are expected to strive for? Or does the shift in institutional form carry with it a change in goals that will ideologically reshape the mission of public education? The general assumption is that a shift towards market-based institutional forms will promote the goal of academic excellence (Chubb and Moe 1990, 82). Because they are controlled by a process dealing with ambiguity, internal inconsistency, and aimed towards compromise, public schools are going to be oriented towards the lowest common denominator and will, “ordinarily find it politically and organizationally difficult to place high priority on academic excellence” (Chubb and Moe 1988). Thus the argument is that, as far as goals go, a shift towards market-based educational institutions will promote a higher priority for academic excellence, which is presented as a universally desirable goal. Academic excellence—again, generally judged by the use of test scores—is
Institutional Structure and Educational Goals
83
seen as desirable largely because of the functionalist context of the reform debate. Academic excellence (i.e., higher test scores) is seen as evidence that higher levels of economically and socially useful intellectual knowledge and skills are being transferred (e.g., Lieberman 1993, Coleman and Hoffer 1987). Independent of the questions surrounding such functionalist assumptions, Chubb and Moe’s institutional theory is challenged by critics on normative and empirical grounds. One of the key criticisms is that freed from the external controls of the democratic process, schools may pursue goals less desirable than promoting academic excellence. That religious preference, geographical proximity, racial or ethnic segregation may rival or even surpass academic excellence as an organizational goal in response to market demand is something critics of market-based reforms have frequently argued (Smith and Meier 1995, Henig 1994, Kozol 1991). Evidence from nations where choice or voucher systems are the norm provide considerable empirical backing for such claims (e.g., Lutz 1996, Ritzen et al. 1997). Even Chubb and Moe’s own data analysis (from a national survey of teachers and administrators) support some of these criticisms. In looking at responses from teachers and administrators on educational goals, Chubb and Moe (1988, 1080) report that private schools are mission-oriented towards academic excellence and personal growth. In contrast, public schools were goal-oriented towards basic literacy, citizenship, and good work habits. These results were presented as evidence that institutional structure did indeed help determine organizational mission and interpreted as favorable to the argument for pursuing market-based institutional reforms in public education. Yet these findings also suggest that, regardless of their focus on academic excellence, public schools are more oriented to the mission of public education spelled out in state constitutions. Few quibble with the empirical finding that private schools are more likely to prize academic excellence. The inference that this helps justify more market-like reforms in public education, however, requires a normative leap that is open to challenge. The goals of public schools embedded in state constitutions—the legal justification and authority for their existence—generally have little to do with academic excellence. Instead, they are much more closely related to goals such as instilling basic literacy and developing the skills associated with productive democratic citizenship: a willingness to get along with others, tolerance of differences, and good work habits (Rebell 1998). Chubb and Moe’s (1988, 1080) results show private schools do a poorer job of adopting and pursuing these sorts of goals. Perhaps some who use private educational services do prize academic excellence, and
84
The Ideology of Education
private schools respond to this by prioritizing it in organizational mission. It also can be argued that the goals public schools were found to prioritize reflect that they are more responsive to the state—its laws, regulations, and ultimately the constitution—than private schools. From a social-democratic standpoint, such findings can just as easily be normatively interpreted as an argument to preserve, rather than change, the institutional status quo in education. If institutional structure does make a difference to school goals, there is evidence even from studies favoring market reforms that those differences put public schools closer to the republican missions that justify their existence. Part of the tradeoff in achieving this may well be making academic excellence a lower priority.
MEASURING SCHOOL GOALS
The basic hypothesis that institutional structure determines school goals, then, is one that both market and commonwealth perspectives agree on, even though there is considerable disagreement on exactly what goals will emerge as priorities in a bottom-up system. The contrasting expectations on goals can be empirically tested by examining the differing mission priorities of public and private schools. If the market is correct, private schools will place a higher priority on academic excellence. If the commonwealth is correct, they will place a higher priority on promoting values at odds with the democratic mission of public education. The problem in undertaking such a test is isolating school objectives. How can school goals be measured? It is important to emphasize that this is a question seeking an empirical rather than a philosophical answer, an “is” rather than an “ought.” There is no doubt that schools in the United States persue a lengthy list of goals and objectives, a number of which seem mutually exclusive and contradictory. Schools are expected to socialize children to be obedient and to be independent critical thinkers; to store and pass on the knowledge of the past while teaching skills that are marketable today; to instill cooperation as a basic value while promoting competition as a basic expectation of life; to focus on basic academic skills while simultaneously offering a broad curriculum and allowing students to choose from its offerings (Tyack and Cuban 1995, 43). All of these are expected to serve republican ends—to serve the collective needs of a democratic system—though this is often more of an assumption than a rational causal argument or a well documented empirical correlation (see Arnstine 1995).
Institutional Structure and Educational Goals
85
Extracting any universal consensus on what schools “ought” to do in specifics from this complex and often conflicting set of missions seems unlikely. However, it may be possible to empirically establish an “is,” even if a universal philosophical “ought” is out of reach. Numerous education scholars argue that what schools do is ultimately dependent upon what teachers do (Tyack and Cuban 1995, Nuthall 1999, Chubb and Moe 1988). Whatever schools ought to be doing, what teachers do in the classroom represents what they are doing. Despite the differences on how to formulate goals and hold schools accountable for these objectives, both the top-down and bottom-up perspectives recognize that teachers play the key implementation role. Teachers retain a good deal of autonomy when they close the classroom door, and if institutional arrangements determine school goals, it logically follows they must shape teacher attitudes and behavior (Hess 1999, 37). Both sides of this debate agree that whatever goals schools seek to achieve, they are ultimately dependent upon teachers adopting and pursuing them. This fits with a common argument in organizational theory that “operators,” or the street-level bureaucrats who actually perform the work of fulfilling organizational objectives, ultimately determine what those objectives are (Simon 1997, Wilson 1989, Lipsky 1980). Thus it is possible to empirically determine the core goals of schools by determining the primary objectives teachers pursue in the classroom. If the arguments of reformers and defenders of public education have any validity, these goals should vary in a predictable fashion according to differing institutional arrangements. Both would expect public education systems to differ from, say, private parochial schools. The latter would cater to a set of specific preferences—i.e., religious preferences— of a particular segment of education service consumers. Public schools are prohibited de jure from incorporating any element of religious indoctrination into their curriculum. This is, in effect, a goal that has been specifically denied to public school systems, and the machinery of external democratic control (including the courts charged with interpreting the laws of democratic institutions) goes to some lengths to enforce this ban. Beyond this external control, private parochial systems are free to make some form of religious indoctrination a specific educational objective. More generally, public schools are expected to pursue a broader set of objectives than private schools, reflecting the difference between responding to the complex and conflicting set of priorities that are thrown up by the broadly inclusive democratic process, and responding to a more homogenous set of preferences expressed by a particular group of consumers in an educational market. Private schools will have
86
The Ideology of Education
more narrowly focused missions, focusing on academic excellence, and, in the case of sectarian schools, promoting religious values. As teachers represent the key actors in pursuing or achieving school objectives, the basic test of the hypothesis posed here is to examine whether the educational objectives teachers pursue in the classroom vary by institutional structure.
PRELIMINARY EVIDENCE
Table 4.1 provides preliminary and mixed empirical confirmation of these expectations. This lists the results of a national teacher survey conducted by the National Center for Education Statistics (1998) that included questions about the educational goals teachers consider the most important, and the type of school where they were employed (public, Catholic, other religious affiliated, and nonsectarian private schools).2 The cell entries indicate the percentage of teachers in a given type of school who rank the indicated goal as one of their three most important educational objectives. These rankings do show some variation in mission priorities by institutional type, although perhaps not as much as one might expect. Clearly the dominant educational goal, regardless of institutional type, is building basic literacy skills. Basic literacy is less a concern in the private sector, but at least 60 percent of private school teachers rank this as one of their most important educational goals. Private schools do make academic excellence a higher priority than public schools, confirming market expectations. Yet these differences are mostly marginal. Catholic schools, by far the dominant segment of private sector education, are virtually indistinguishable from public schools in prioritizing academic excellence. Just under 40 percent of public and Catholic school teachers rank this as one of their most important goals. Just under 50 percent of teachers in other sorts of private schools make academic excellence a mission priority. Similar sorts of marginal differences occur for other goals, with public school teachers placing more emphasis on promoting good work habits and self-discipline, and instilling vocational–technical skills than teachers in the private sector. Dramatic separation between public and private schools is evident only in the last column: private sector school teachers are much more likely to make the promotion of specific moral or religious codes a primary educational goal. A solid majority of Catholic school teachers rank this as one of their primary goals, though a smaller majority than that for basic literacy or promoting personal growth. For Catholic school teachers, promoting specific moral or religious codes ranks roughly
TABLE 4.1 Goals and School Type Building basic literacy skills
Encouraging academic excellence
Vocational/ technical skills
Promoting good work habits and self-discipline
Promoting personal growth
Promoting human relations skills
Promoting specific moral or religious values
Public School (N = 4,761)
77.6
38.7
17.2
69.9
62.3
24.6
8.9
Catholic School (N = 770)
64.2
38.3
5.2
55.5
63.1
13.2
55.2
Other religiousaffiliated (N = 708)
62.7
47.4
4.9
58.5
43.2
10.6
71.2
Non-sectarian (N = 323)
60.4
47.7
9
63.5
79.6
25.4
21.4
School Type
Cells indicate percentage of teachers at a given school type (row) who rank the goal listed in the associated column as one of their three most important educational goals.
88
The Ideology of Education
equal with promoting good work habits and self-discipline. For other religious-affiliated schools, moral and religious and indoctrination jumps out as the primary organizational mission, with more than 70 percent of teachers ranking it as a primary goal. Even in nonsectarian schools, a fifth of teachers rank it as a top priority, more than twice the number of public school teachers. The results thus do hint at clear and systematic variation in the educational goals pursued by teachers according to type of institution, though this variation has little to do with academic excellence. Public schools are primarily oriented towards instilling basic literacy (specified in the survey question as basic reading, mathematical, and speaking skills), promoting good work habits and self-discipline, and promoting personal growth (specified in the survey question as promoting “selfesteem and self-knowledge”). Other types of schools place somewhat less emphasis on basic literacy, a little more on academic excellence, and considerably more on instilling specific moral codes or promoting religious development. With the large exception of promoting specific moral or religious values, the most striking impression of this survey is not how much variation there is between different types of schools, but how little. With the notable (and, in the case of schools with religious affiliations, entirely expected) exception of moral and religious values, private school teachers seem to be pursing more or less the same educational objectives as public school teachers. This is especially the case with Catholic schools where, in terms of ranking goals, the Catholic school teachers largely mirror the responses of public school teachers. This may reflect the historical parallels between the Catholic and public schools, and the impact of a more secular society that has mitigated the deep religious differences that prompted a clear and continued separation between these systems. Certainly by the late 1960s or early 1970s, Catholic schools had essentially embraced the liberal democratic ethos and many of its secular implications (Macedo 1998). Religion still plays an important role in Catholic education, but in contrast to other religious affiliated schools it is not its core mission—basic literacy, personal growth, and promoting good work habits and selfdiscipline both outrank it in terms of goal priorities among Catholic school teachers. It would not be correct to describe Catholic schools as simply a second “public” system, but in terms of goals neither would it be entirely inaccurate. Like their religious counterparts, nonsectarian private schools also diverge from public schools in the goals they pursue, but again in rather marginal fashion. They have relatively less focus on basic literacy, and more on personal growth and academic excellence, which fits the pat-
Institutional Structure and Educational Goals
89
tern hypothesized by scholars such as Chubb and Moe (1988, 1990). Again, they also place a higher value on instilling religious or moral codes than public schools, but this is a secondary focus compared to its emphasis in schools affiliated with religious organizations. Yet in looking at the core objectives prioritized by private, nonsectarian schools, these seem to largely consist of instilling the basics, promoting personal growth, and promoting good work habits and self-discipline—essentially the same core objectives of public schools. The hypothesized pattern of greater scatter in public school objectives does not appear in Table 4.1. The three core objectives of public school teachers are well defined, and no other goal listed even gets close to majority acceptance. If anything, private school objectives seem to be less well defined. The most important goals in private schools tend to have smaller majorities of teachers embracing them, the exceptions being promoting moral and religious values for religious-affiliated schools, and promoting personal growth for nonsectarian schools. If institutional structure shapes mission priorities, Table 4.1 suggests it is private schools that have a harder time clarifying their core objectives. The figures in Table 4.1, of course, present a fairly unsophisticated picture of the variation among educational objectives. While eyeballing these figures provides some initial confirmation that goals vary by institutional arrangement, it does so in a fairly crude manner. The goals of teachers are almost certainly determined by variables other than the type of school they are employed by, and the table only presents variation highlights between institutional types. As others have pointed out, institutional arrangements vary within these types as well as between them (Chubb and Moe 1990, Smith and Meier 1995, Brown 1992). Rather than a rigorous test, Table 4.1 provides a justification for pursuing the hypothesis that institutional arrangements help determine educational objectives.
OTHER DETERMINANTS OF EDUCATIONAL OBJECTIVES
Besides the type of institution, there are at least three other likely sources of variation in educational goals: (1) the power arrangements within the institution (i.e., the degree of autonomy of the teachers); (2) the characteristics of the students in the classroom; and (3) the background and personal characteristics of the teachers themselves (Chubb and Moe 1988, Smith and Eccles 1998, Nuthall 1999; for similar arguments applied to public agencies in general see Wilson 1989 and Lipsky 1980).
90
The Ideology of Education
Power arrangements are central to the arguments of market reformers, who generally claim that in public education too much power is concentrated in the administrative and bureaucratic hierarchy beyond the classroom and the individual school (e.g., Chubb and Moe 1990, Friedman 1955, Hanushek 1994). The top-down administrative structure characterized by the “one best system” (Tyack 1974) in effect ties schools and teachers to a set of externally determined goals they may or may not agree with, and may or may not fit the demands and wishes of students and parents. Freed from these constraints, teachers have more leeway to pursue objectives that correlate with those of their core clientele. If academic excellence is the key unmet demand in education’s task environment, we should expect higher levels of teacher autonomy to positively correlate with academic excellence as an important educational objective. Student characteristics also seem a logical determinant of educational objectives. A teacher at an elite prep school has a different set of possibilities in terms of educational goals than a special education teacher in an inner-city school. Student attributes such as apathy/enthusiasm, academic background, or demonstrated cognitive capacity obviously may play an important role in determining the educational objectives a teacher will choose to pursue. Objectives may also be constrained by the socioeconomic backgrounds of the students. Such characteristics are argued to set the boundaries on educational goals as well as educational opportunity (Alexander 1993, Rebell 1998). The individual characteristics of teachers are also likely to play an important role in determining their educational objectives. It seems reasonable to suggest that teachers may construct their own ideology of education, and there is a substantial research literature indicating that the individual characteristics of teachers determine a classroom’s educational environment and direction (e.g., Nuthall 1999). Such individual impacts should be considered independent of institutional constraints and student characteristics. Taking all of these into consideration, it should be possible to construct a model capable of controlling for and highlighting the separate contributions of teacher and student characteristics, institutional type and power sharing to determining the primary goal teachers pursue in the classroom. These three general concepts—institutional power sharing and autonomy, student characteristics, and teacher characteristics— should provide a good set of controls for assessing the general impact of institutional type upon goal selection in a more rigorous, multivariate analysis.
Institutional Structure and Educational Goals
91
DATA AND METHODS
The data for undertaking this analysis come from the same source used to generate the results presented in Table 4.1, the second Teacher Followup Survey (National Center for Education Statistics 1998), a nationally representative survey of K–12 teachers undertaken by the U.S. Department of Education in 1991–1992. One of the specific objectives of the survey was to gather data on attitudes towards professional objectives, and the survey includes a broad enough set of questions to provide a rich resource to construct the sort of multivariate model required for present purposes. The outcome variables were the educational objectives of teachers as listed in Table 4.1. This means the analysis consists of using the same basic model to predict seven outcome variables (one for each educational goal). Each outcome variable is a dichotomous measure, with 1 indicating that a teacher ranked that particular goal as one of their three most important educational objectives. The explanatory variables include a serious of dichotomous measures to indicate whether a teacher taught at a public school, Catholic school, a conservative Christian school, some other form of religious-affiliated school, a nonsectarian school, or a private special education school. These are finer measures for institutional type than those used in Table 4.1, the aim being to gain as much empirical leverage on the underlying hypothesis as possible in the multivariate analysis. The public school measure was excluded as the reference category. Probably the most important control variable included in the model is the relative power or autonomy of teachers. Market advocates argue private sector schools and teachers enjoy more autonomy, and this allows them to help shape more explicit missions as a result of signals from their primary clientele (Chubb and Moe 1988, 1990). The Teacher Followup Survey asked a series of questions relating to teacher autonomy and influence that cut along a number of conceptually distinct dimensions. Rather than pick one or two variables as representative of a multidimensional concept like power or influence, I used factor analysis to collapse this broad set of questions into three separate indexes (see methodological appendix for details). These three separate indexes reflect relative levels of control over what is taught (whether teachers are free to determine the content of the classes and the instructional materials they use to teach it), control over how things are taught (whether teachers can choose pedagogical techniques, grading systems, homework, and have disciplinary enforcement powers), and control over
92
The Ideology of Education
school policies (input on curriculum, disciplinary standards, professional development, and grouping or tracking of students). I have labeled these indexes educational content, pedagogy, and school policy, and higher numbers on each reflect greater levels of teacher autonomy or influence within a given area. To round out the specification of the model with a set of strong controls, I used factor analysis to create four indexes of student and parental characteristics. These reflect student and parental educational attitudes and support of school objectives (student academic involvement, parental involvement and support, student apathy), levels of student violence (possession of weapons, physical and verbal abuse of other students and teachers), levels of student drug or alcohol abuse, and levels of socioeconomic stress (poverty, drug/alcohol abuse by parents, racial tension, and cultural conflict). Details on all these variables can be found in the methodological appendix. As a final set of controls I included a set of “usual suspects” individual characteristics—teacher age, sex, and race. These admittedly do not constitute a direct measure of a concept such as “education ideology,” but given the limitations of the data seemed the best options for controlling individual-level teacher characteristics. Because the outcome variables are dichotomous, I employed a logistic regression procedure to generate estimates, an approach more suited to nominal dependent variables than OLS and, with some minor mathematical manipulation of the base coefficients, capable of generating intuitively interpretable results (see Long 1997).
RESULTS
Table 4.2 presents the impacts of institutional type on the odds of selecting a given goal as one of the three most important educational objectives (these estimates are taken from the fully specified model described above). The marginal effect on the odds are reported because they are substantively easier to interpret than unstandardized logit coefficients (Long 1997). The cell entries can be viewed as the estimated change in odds of picking a given educational goal compared to public school teachers, while controlling for all the other variables in the model. For example, in the first column, first row entry indicates that, all else equal, being at a Catholic school rather than a public school multiplies the predicted odds of picking basic literacy as a primary educational goal by .44, all else held constant. The multiplicative nature of the coefficients means a number greater than 1 indicates a positive impact on the odds
Institutional Structure and Educational Goals
93
TABLE 4.2 Marginal Effect of Institutions on Teaching Goals Teaching Goal Basic Literacy
Catholic
Conservative Other Christian Religious
Special Nonsectarian Education
.44
.35
.44
.43
.47
1.06
1.15
1.06
1.66
.86
Occupational/ Vocational Skills
.37
.52
.31
.38
1.77
Good Work Habits/ Self-Discipline
.50
.55
.57
.75
.64
Personal Growth
.80
.19
.43
1.9
2.3
Human Relations
.50
.26
.45
1.04
1.22
1.87
.84
Academic Excellence
Moral/ Religious Codes
14.5
63.4
22.7
Cell entries are the change in odds relative to public school teachers in ranking the listed goals as primary educational objectives. The numbers are multiplicative, i.e., numbers less than 1 indicate a decrease in the odds of adopting a given goal as a primary educational goal relative to public school teachers, while numbers greater than 1 indicate an increase in the odds of adopting a given goal relative to public school teachers.
of selecting a given goal as an educational priority, and a number less than 1 has a negative impact on the odds (Long 1997, 82).3 Likewise, the closer a coefficient is to 1, the smaller its impact on the odds of choosing one of the listed goals over basic literacy. The results presented in Table 4.2 do indicate some systematic institutional effects on the goals teachers pursue in the classroom. Some of this systematic variation is undoubtedly being picked up as a result of the finer institutional distinctions than those used in Table 4.1. Special education schools in the private sector, for example, clearly seem to
94
The Ideology of Education
have a different organizational mission—they are more focused on vocational skills and personal growth than public schools. This finding intuitively fits with market claims that private sector schools will shape their missions in response to the needs and preferences of their primary clientele because there is an obvious case for special education schools prioritizing these sorts of goals to better serve their clientele. But by far the most significant differences between public and private schools are the those relating to instilling specific religious or moral codes of behavior—the same inference drawn from the less sophisticated portrait presented in Table 4.1. Even when controlling for other likely determinants of goal selection, Catholic school teachers are 14.5 times more likely to pick instilling moral or religious codes compared to public school teachers, conservative Christian school teachers a whopping 63.4 times more likely, and other religious-affiliated schools 22.7 times more likely. In contrast to these huge effects, institutional type seems to have much less impact on prioritizing academic excellence. Catholic, conservative Christian, and other religious-affiliated schools are moderately more likely to prioritize academic excellence, but the differences with public schools are insignificant (see below). Non-sectarian schools are considerably more likely to emphasize academic excellence, although it is worth noting that the change indicated here is not as large as it is for instilling moral and religious codes. All else equal, the models estimate that non-sectarian school teachers are 66 percent more likely that public school teachers to pick academic excellence as an educational priority, but 87 percent more likely to prioritize moral or religious codes. By the standards of traditional hypothesis tests, all inferences of a positive relationship between institutional type and selection of academic excellence are, with one exception, unjustified—only the variable indicating non-sectarian private school crosses meets the .05 Alpha criterion. If what teachers try to achieve in the classroom captures the empirical reality of what schools do, and (as Chubb and Moe argue) public-private sector comparisons are the best way to assess the impact of institutional structure on education, then this analysis suggests dropping the top-down, external controls typical in public education is more likely to make specific value systems a higher educational priority rather than to promote academic excellence. Schools, it appears, are serious about smuggling values, and institutional structure plays a key role in achieving this aim. This inference is bolstered by looking at the results of the full models (all of these are reported in the methodological appendix in Table A.1), most of which did little to improve the ability to predict a teacher’s educational priorities over and above picking the modal category of the
Institutional Structure and Educational Goals
95
dependent variable. Simply speaking, neither institutional type nor any of the other variables evidenced anything but a marginal capability in discriminating between the goal choices of teachers. Absent any other information, the best way to estimate the probability of any given teacher prioritizing a particular goal is to simply to pick the model category in the outcome variable. The multivariate models did little to increase the predictive power of this blunt approach to probability estimation. Compared to simply picking the model category, for example, the academic excellence model managed to reduce errors by about 5 percent. This weak performance is not too surprising—as Table 4.1 indicates, a bit less than half of teachers prioritize this, and a bit more than half do not, meaning there is not much meaningful variation by institutional type. Explanations of why some teachers pick academic excellence and some do not lies not only beyond institutional type but also beyond the other control variables included in the model. Many of the variables were statistically significant and were bounded by zero in both confidence limits and Leamer bounds, but their predictive power was at the margins (see Table A1 in the methodological appendix). The big exception to this generally weak performance was the model predicting religious and moral codes. The full model predicting this particular dependent variable is reported in Table 4.3. This model managed to reduce predication errors by about 30 percent over the model category, and the institutional variables are clearly playing a dominant role in providing this explanatory horsepower. In this one model, the clear divisions illustrated in Table 4.1 hold up in the presence of the controls. To make interpretation of the model easier, Table 4.3 reports the marginal impact on the odds (Exp[B]) as well as the unstandardized coefficient, and reports confidence limits and Leamer bounds for the latter rather than the former. The huge effects indicated by marginal impacts on the odds calculated from the unstandardized coefficients are confirmed by the confidence limits and the Leamer bounds. Neither differing model specifications, nor the range of reasonably possible values of these effects, does anything to change the basic inference—private schools (with the exception of special education schools) are much more likely to make a specific type of religious or moral indoctrination a primary educational goal compared to public schools. This is true even for nonsectarian school teachers, who are estimated to be 87 percent more likely than their public school counterparts to adopt instilling specific religious or moral codes as a primary educational objective. The results in Table 4.3 also typify the mixed findings in regard to teacher power and autonomy. The variables designed to tap into
TABLE 4.3 Estimated Institutional Impact on Adopting Religious/Moral Codes As Major Educational Objective
Variable
Coefficient (Standard Error)
Exp(B)
Confidence Limits of Exp(B)
Leamer Bounds for Exp(B)
11.9:17
13.5:14.4
42:95
58:64
18:28.8
21:22.9
Catholic
2.67 (.09)*
14.4
Conservative Christian
4.14 (.20)*
63
Other Religious
3.12 (.12)*
22.7
Nonsectarian
.62 (.18)*
1.87
1.29:2.7
1.8:1.87
Special Education
–.17 (.52)
.84
.3:2.3
.81:.91
Control over educational content
–.09 (.04)*
.91
.82:1.0
.94:.93
Control over pedagogy
.04 (.04)
1.04
.95:1.14
1:1.04
Control over school policies
.15 (.05)*
1.16
1.05:1.27
1.16:1.16
Parental/student support for school
–.04 (.06)
.96
.84:1.09
.92:.97
Student violence
–.15 (.06)*
.85
.76:.95
.85:.86
Student drug/ alcohol abuse
.02 (.05)
1.02
.91:1.14
1:1.02
Race and socioeconomic problems
.04 (.05)
1.04
.94:1.16
1.04:1.06
Sex
.33 (.08)*
.79
.61:1.02
.33:1.38
Age
.05 (.03)*
1.05
.98:1.12
.05:1.06
(continued on next page)
Institutional Structure and Educational Goals
97
TABLE 4.3 (continued)
Variable
Coefficient (Standard Error)
Race
–.23 (.13)*
Constant
–2.37 (.15)*
N
6562
–2 Log likelihood
4951
Pseudo R-2 (Cox and Snell)
Exp(B)
Confidence Limits of Exp(B)
Leamer Bounds for Exp(B)
1.4
1.18:1.65
–.23:/.77
.21
Estimators generated by logistic regression (dependent variable is categorical). *p < .05.
the concepts of teacher control over educational content, pedagogy and school policies, point towards no consistent conclusions. The variables are frequently in different directions, and are sometimes statistically insignificant or not bounded away from zero in confidence limits or Leamer bounds (using the multiplicative Exp B coefficients, this means they cross 1 in Table 4.3).4 With the partial exception of religious and moral indoctrination, neither institutional type nor the power arrangements within them seem to make a large difference on separating the organizational mission of the private sector from the public sector.
CONCLUSION
The primary purpose of this chapter was to empirically assess the primary mission of schools and determine whether these goals varied according to institutional arrangement. The basis for this analysis builds from the assumption that in practice a school’s mission is inseparable from how teachers prioritize and pursue educational goals, and that assessing the impact of institutional structure on these goals is best done by comparing public schools to private schools. The analysis here suggests there is some systematic variation in teacher goals by institutional
98
The Ideology of Education
arrangement, but this variation—with the exception of instilling moral and religious codes—is not particularly notable. The most significant inference from this analysis is that the impact of weakening the external constraints on public schools and making them more like the private sector will be to shift their primary mission away from promoting basic literacy and towards instilling specific religious and moral codes—an ideological (even theological) rather than a utilitarian change. There may be a slightly larger priority accorded to academic excellence, but certainly not to the extent suggested in research backing market-based reforms (see esp. Chubb and Moe 1988, 1990). All else equal, aside from the moral and religious goals, private and public schools pursue educational objectives that are more alike than dissimilar. Whatever they ought to be doing, private and public schools are primarily oriented towards building basic literacy, promoting good work habits and self-discipline, and promoting personal growth. The private and public sector differ primarily in the fact that the latter is banned de jure from instilling religious values, and that most of its teachers accord instilling specific moral codes a much lower priority than private teachers. In one sense, this makes the findings trivial and the inferences an exercise in confirming the obvious. Religious-affiliated schools rarely seek to hide their religious origins or sanitize their curriculums from its influence. To the contrary, religion is often an explicit selling point for such schools, and even nonsectarian schools are free to promote particular moral codes of conduct.5 In contrast, public schools break the law when they promote specific religious values, and their moral teachings are limited to qualities—work hard, be tolerant of others—that can be affirmed universally within the polity. Yet, if public–private sector comparisons provide useful insight into the potential impacts of institutional reform on educational objectives, the findings raise more substantive concerns. As the bulk of private sector primary and secondary education has some religious affiliation, it seems reasonable to assume that an increasingly decentralized, marketbased system of public education that funnels tax dollars and students towards the private sector could significantly expand the enrollments of such schools in much the same way that de facto voucher programs have already done in higher education (Alexander 1998b; see discussion in previous chapter). If so, the findings reported here suggest one of the results of such a move will be to significantly shift the ethos of “public” education by increasing the probability that separate sectarian concerns will gain a larger foothold and a higher priority within the education system.
Institutional Structure and Educational Goals
99
In a normative sense, whether this is good or bad for the democratic purposes of schooling is up for debate. Religion and moral codes are certainly not incompatible with a good education, and Catholic schools provide ample evidence they need not be incompatible with a broad set of liberal democratic values (Macedo 1998). Still, such a development clearly seems at odds with the arguments of philosophers such as Guttman (1987), Dewey (1929), and Callan (1997). As scholars such as Barber (1992) have argued, the traditional purpose of public education has a strong emphasis on public; its job is to produce citizens who can effectively participate in the communal life of a democracy. Using schools to promote sectarian doctrines clearly raises some problems for this goal. The findings also contradict a set of market- and commonwealthbased expectations on institutional structure. First, institutional variation between public and private schools does not correlate with a large increase in the priority of academic excellence. Though the functionalist perspective of many market advocates sees academic excellence as an important goal to tie schools to their external task environments, this only minimally seems to be the case. If the teacher preferences reported in Table 4.1 are used to rank the goals of education, academic excellence comes in fourth for public schools, religious-affiliated schools and nonsectarian schools, and fifth for Catholic schools. In the private sector, personal growth and/or perpetuating value systems are prized well above the technical knowledge represented by academic excellence. Second, goals in the public sector do not seem to be more diffuse than in the private sector, an expectation shared by market and commonwealth advocates. According to all the results presented here, public schools are primarily focused on building basic literacy, promoting self-discipline and good work habits, and creating a sense of selfesteem and self-knowledge—all compatible with the general republican objective of creating productive citizens capable of self-government and providing them with the tools needed to make and take advantage of their talents and opportunities. They are heavily tilted towards instilling behavioral predispositions and attitudes—work hard, know yourself, get along with others—that most would accept as compatible with and supporting the democratic ethos, so these findings certainly can be interpreted to support the argument that democratic culture and values seem to be a large part of the public school mission. Yet private schools pursue these same missions to a more or less similar degree, and, at least empirically, it is not clear what threat (if any) the greater promotion of specific religious or moral value systems presents to the republican justifications for public education.
100
The Ideology of Education
The bottom line is that if the goals teachers pursue in the classroom truly reflect “is” rather than the “ought” of school missions, then these findings suggest that, outside of promoting basic literacy and numeracy skills, schools seem to be more interested in smuggling values than transferring technical blocks of knowledge. The institutional variation that does exist between goals and school type indicates that having public schools adopt market-like institutional arrangements is more likely to alter what values are being smuggled than to place a higher premium on academic skills in the public sector.
5 Education and Civic Culture
A
ccording to any number of commonwealth-oriented scholars and analysts, the primary purpose of public education is to “indoctrinate the coming generation with the basic outlooks and values of the political order” (Key 1961, p. 316). In this perspective schools are repositories and replicators of value systems that manifest themselves in norms and behavioral dispositions (Arnstine 1995). Their job is not to respond to their external task environments but to shape that external environment by storing and passing on the norms and values that constitute political or civic culture. This notion is deeply embedded within the commonwealth ideology, shaping its conception of education and its primary objectives, preferences on who should control education, and how that control should be administratively structured. These beliefs support the claim that education has an important role in supporting a democratic polity, a claim with a long history in the United States. A number of revolutionary-era politicians argued a system of schools under centralized government control would be critical to sustaining and expanding the success of the American experiment in democracy. The best known proponent of this argument was Thomas Jefferson, who tried—with very mixed success—to establish a centralized system of public education in Virginia.1 Although Jefferson and other revolutionary-era notables such as John Adams were unsuccessful in their attempts to persuade their fellow citizens to support a mass system of public education, their argument that public schooling could make a critical contribution to the civic health of the nation never lost currency. A half-century after Jefferson introduced a bill calling for a statewide system of public schools into the Virginia House of Delegates, Horace Mann and others like him managed to transform the original idea of a mass system of public education into bricks and mortar reality on a state by state basis. Because of his successful efforts to proselytize the virtues of public education, Mann, a
101
102
The Ideology of Education
Massachusetts lawyer and politician, is remembered as the father of public schooling. At the center of his proselytizing efforts was the argument that nothing less than the survival of the republic depended on the creation of centralized systems of public education. Schools would support democracy by teaching the values of tolerance and public spiritedness, and by promoting a willingness to participate in social and political life. While Mann also promoted the economic value of education, his main arguments favored those “rooted in moral principles or civic virtues” (see Vinovskis 1995, 92–103). So the primary reasons mass systems of public education exist in the United States, and the reason governments run them, is because education is seen as an effective means of instilling civic values into the citizenry. Such republican sentiments are often explicitly articulated in the state constitutions that legally mandate the existence of public education (Rebell 1998). From at least Jefferson’s time, the importance of education to the nation’s civic health was considered so critical that commonwealth advocates argued that schools should be under the direct control of the state and free of sectarian agendas.2 Although Jefferson’s attempts to establish public grade schools were largely unsuccessful, educational historians argue they are particularly notable because of their general insistence that they be free from church interference, and served to deliberately shift the “prestige and resources of the state from the church to the school as the most appropriate agency for carrying on . . . extramural education” (Cremin 1980, 442). The previous chapter suggested that institutional structure is capable of shifting the mission priorities of education away from the communal ends of the democratic process and towards more sectarian value systems. Market-based education systems create this potential by allowing the indoctrination of narrow value agendas favored by small, homogenous groups of citizens (what Madison would call factions) to become primary school objectives. Such possibilities raise obvious problems for the civic role of education traditionally used to justify its existence. Looking at the variation in the educational objectives pursued by teachers, however, provides only a preliminary glimpse into the potential cultural impact of structural reforms in education. The larger issue is how more market-like systems shape the civic attitudes and behavioral predispositions of students, the “coming generation” who will ultimately determine how the nation’s political culture evolves. While commonwealth supporters from both consensus and conflict theory vantage points argue that schools play an important role in determining civic attitudes and behavior, this argument is also adopted by market advocates as the functionalist justifications for reform run into
Education and Civic Culture
103
empirical challenges. Expanding on these functionalist foundations, market backers have in recent years begun to argue that private schools not only do a better job of promoting academic excellence but are also more effective at creating civic attitudes and beliefs in their students, and have presented this as an additional reason for pursuing educational reform (e.g., Peterson et al. 1998, Greene et al. 1999). If such claims are correct, market-based reform of education has a defensible commonwealth justification independent of the traditional functionalist claims that have been challenged here and elsewhere. The socialization of future citizens into productive members of a civic culture is at the core of both the intellectual and legal justification of the commonwealth conception of public education. If it can be demonstrated that the market is more effective at this task than traditional public schools, it has a powerful cultural argument justifying its reform prescriptions, even if its more traditional functionalist claims are suspect. This issue can be collapsed into two basic empirical questions: (1) Do schools actually affect civic attitudes and beliefs? The general presumption that schools are effective smugglers of democratic values is widely held in the abstract, yet the evidence supporting this proposition is mixed. (2) Are certain types of schools more effective at promoting such values? If more market-like institutional structures do a better job of promoting things like tolerance, participation, and general public spiritedness, the market institutional reforms—if not the whole underlying market ideology—has to be accepted on commonwealth terms. Answering these questions, and exploring their importance, is the focus of this chapter.
DEMOCRACY AND SOCIAL CAPITAL
Political scientists have long posited that effective democratic governance (as well as economic growth) are at least partially dependent upon a society or community’s dominant political values and belief systems. The general argument is that political culture is identifiable by a distinct and coherent pattern of attitudes held by most citizens within a polity, and that these support and are in turn replicated by the social, political, and economic institutions that order society (Almond and Verba 1963, 1980, Inglehart 1988, Granato et al. 1996). Political cultures that are more civic, i.e., cultures characterized by a broad commitment to participating and contributing to public life, are held to be more successful in supporting stable democratic regimes and promoting economic development than other cultures. The antecedents of such propositions run
104
The Ideology of Education
deep into the history of political philosophy, where the notion of civic virtue has long been held to be an important ingredient of sustainable democratic governance. From its beginnings, observers of social life in the United States have identified particular characteristics that contribute to a unique American political culture. These often center on the polity’s core attachment to individualism, the belief that the individual should be free to make choices without government interference. This attachment to individualism, though, presents accepted risks to the polity—taken to extremes it makes collective action difficult and limits the ability of the polity to guarantee the liberties of the individual. De Tocqueville was one of the earliest observers to note the paradox this presented for American society, and to articulate the need for balancing the atomizing social effects of individualism with the necessity of communal cooperation that was a prerequisite for preserving individual liberties. He proposed a theory of associations as the solution to this problem, a solution that was already culturally rooted in American society. Simply put, De Tocqueville observed that Americans have a “penchant and ability to form voluntary associations” and that this served the dual purpose of maintaining individual liberty while supporting effective democratic institutions (Elazar 1999). This penchant for voluntary associations, a broad willingness on the part of citizens to get involved in public life, and the psychological characteristics that support such norms of participation (e.g., feelings of efficacy, a tolerance of opposing viewpoints) is argued to be at the heart of American civic culture (Putnam 1995). The notion that such social connections underpin effective democratic governance generally has been greatly expanded by contemporary scholars studying the concept of social capital. Social capital, like human capital, is an intuitively easy to grasp construct that defies universal definition or measurement. Social capital has been variously defined as “a set of institutionalized expectations that other social actors will reciprocate co-operative overtures” (Boix and Posner 1998), or “features of social life—networks, norms and trust—that enable participants to act together more effectively to pursue shared objectives” (Putnam 1995a, 664–665). Social capital, in other words, is a more general form of De Tocqueville’s theory of voluntary associations. Both concepts have at their core a set of norms involving reciprocity, trust, and a willingness to participate in public life. As such, social capital can also confer economic as well as social advantages—effective cooperation makes it easier to achieve economic objectives—and thus bears more than a passing relationship to human capital (Fukuyama 1995). Social capital, however, is more broadly oriented
Education and Civic Culture
105
towards the polity than to individual interests. It encompasses and supports a more extensive view of the common good than the economic focus that accompanies scholarship on human capital. As Putnam (1995a, 1993) points out, social capital does not necessarily lead to positive social contributions. Some people band together and cooperate in ways that have a negative effect on social life and a stable democratic order. Organized crime, militias, and gangs are (mostly) voluntary associations that often involve norms of trust and reciprocity, but few consider them beneficial to the running of a liberal democracy. The positive side of social capital is expressed in terms like “civic capital” or “civic culture,” which encompass the propensity for a given community or society to support and engage in civic acts. Unlike the dynamism of human capital and the economy, social capital and civic culture are mutually reinforcing, semi-permanent characteristics of social aggregates. The norms and beliefs that make up this culture are embedded in and replicated by key communal institutions that are highly biased towards preserving the social status quo. Putnam (1993) found civic culture in the regional governments of Italy to be remarkably tenacious, somehow reproducing itself in a highly stable manner over the course of centuries. The implications of having this social characteristic can hardly be overstated—the rough summation of Putnam’s work in Italy is that regions with highly civic cultures fostered good governance and economic prosperity to a greater extent than those with lesser civic cultures. This particular relationship between civic culture and effective democratic governance and economic performance has been studied in literally dozens of countries (including the United States) with good empirical results (Almond and Verba 1980, Rice and Sumberg 1997, Granato et al. 1996; for a dissenting view see Jackman and Miller 1996). Generally speaking, social capital and the civic culture it creates is argued to promote good governance by making citizens more sophisticated consumers of politics, reducing the likelihood of political free riders (e.g., tax cheats), increasing commitments to collective concerns over purely individualistic ones, promoting cooperation among and reducing shirking within public agencies, and by fostering compromise among antagonistic elites and organized interest groups (Boix and Posner 1998). Thus the question of how a civic culture is established and reproduced is of enormous importance to scholars interested in the success of democratic regimes, and in a more applied sense to those who desire policies to promote and preserve social order and economic development. Civic culture is almost certainly a product of a multitude of factors, emerging from a set of political and social traditions and characteristics
106
The Ideology of Education
that combine and interact in ways not yet fully understood. For example, the religious history of a nation has been argued to be a powerful contributor to the existence of a given political culture. The most famous example of the latter is Weber’s (1958) claim for a causal link between the Protestant work ethic and the successful emergence of capitalistic systems in Europe. It is a claim that has an extensive empirical backing— religious tradition has proven to be a remarkably good proxy for the factors historically associated with the emergence of modern economic development (DeYoung 1988, 1148). Regardless, the important point for present purposes is that education systems are argued to play a key role in storing and reproducing the norms and values that constitute a civic culture. This is exactly the argument employed by Jefferson and Adams employed in their mostly unsuccessful attempts to establish mass public education systems under centralized government control. Just as market advocates argue that schools are a major supplier of human capital, education systems are also argued to be major suppliers of social capital. Public education produces this social capital by bringing together proto-citizens of different backgrounds onto a more or less even playing field. This early mixing of the demographic pot is expected to increase tolerance of differences, promote norms of cooperation, and generally encourage an ethic of civic responsibility (McClosky 1967; Sullivan, Piereson, and Marcus 1982; Wolfinger and Rosenstone 1980, 102; Brehm and Rahn 1997). Researchers interested in questions related to civic culture thus tend to be more interested in how long people spend in school rather than their performance on achievement tests. If schools are institutional repositories of the norms of tolerance and cooperation, the longer individuals are exposed to such norms the more likely they are to make positive contributions to social capital. There is a fairly extensive empirical record to back such claims. Putnam (1995a, 667–668) finds the more time people spend in institutions of formal education, the higher that population’s level of social capital. Across a broad set of studies, similar education measures are consistently a strong correlate of participation in civic life and the norms that support a vigorous public realm (Brehm and Rahn 1997). This process of creating social capital is essentially the reproduction of the prevailing political culture, whether it is civic in nature or not, and lies at the heart of the claimed cultural mission of public education. A sense of how important a foundation education is to civic culture can be gained from looking at Figure 5.1. The Y-axis in this chart is Rice and Sumberg’s (1997) index of civic culture for the American states. This is an explicit attempt to replicate Putnam’s (1993) concept of civic culture in a single empirical measure that places states along an interval-
Education and Civic Culture
107
FIGURE 5.1 Civic Culture and High School Graduates
level scale (high numbers are associated with more civic cultures).3 The X-axis is the percent of a state’s population that has at least a high school diploma. Little in the way of sophisticated statistical analysis is needed to make the point: as the number of high school graduates increases, so does the level of civic culture.4 The relationship portrayed in Figure 5.1 is representative of a wide variety of studies (particularly at the aggregate level) across a number of different disciplines that examine the relationship between education and the norms and behavioral dispositions held to support democratic governance. High levels of schooling, and the systems of education that make them possible, are repeatedly confirmed as primary supports of a civic culture.
108
The Ideology of Education EDUCATION AND CIVIC CAPITAL
The rosy picture of education’s positive relationship with civic culture, however, is far from uniform, especially when the analysis moves from the aggregate to the individual level. Public education tends to retain an important role in shaping civic norms and attitudes in such studies, though that role is more complex than the straightforward relationship portrayed in Figure 5.1. A number of scholars studying the sociology and psychology of education argue that schools are conservative institutions oriented toward reproducing status quo social hierarchies (Katsillis and Rubinson 1990, Bourdieu 1973). These analyses confirm the important role of education in shaping culture but suggest the positive picture conveyed by Figure 5.1 tells only half the story. In conflict theory perspectives, the social haves coopt the cultural role of public education for self-interested purposes, using schools to justify and retain their favorable social position. Schools teach a “hidden curriculum,” where commonwealth values are rhetorically supported but subtly subverted, inequitably distributing social capital towards the upper-end of the social hierarchy. Those at the top of this hierarchy end up with the highest levels of political efficacy, espouse a greater commitment to the public sphere, are more likely to participate in public life, and possess the necessary social and organizational sophistication to overcome the obstacles to collective action. Paradoxically, these civic attributes and the greater political effectiveness they bestow upon their bearers then end up being employed to defend and maintain their favored social position rather than for less self-interested public ends. Conflict theory thus suggests schools help maintain the status quo by ensuring the resources needed to effectively participate in public life—norms like efficacy and tolerance, and the civic skills that support participation—are inequitably distributed (Litt 1963). Thus underneath the positive picture portrayed by Figure 5.1 is a “hidden reality” where education reinforces the advantages of class and race. Probably the best known study along these lines is Edgar Litt’s 1963 article comparing three different public school districts, one working-class, one lower middle-class, and one upper middle-class. Litt found differences between these districts in the political themes of the civic education texts, the attitudes of community leaders, and the effects of course work on the political attitudes of students. Most important, he found that only upper middle-class students were being oriented towards a “realistic” view of the political process, one that emphasized participation and political conflict. In contrast, students in the working-class district were oriented toward a more idealistic view
Education and Civic Culture
109
of democracy, emphasizing harmony and de-emphasizing the need for participation to achieve political aims. Litt concluded students in the three communities were being trained to play different political roles in society, and to orient themselves towards politics in different ways. Working class students were being socialized to accept idealistic views of politics and to be passive in their political roles. Upper middle-class students were being socialized to participate and influence the dynamics of public decisionmaking. Although decades old, Litt’s study remains notable not only because of its empirical support for some of the starker claims of conflict theory, but also because it highlights the difficulties of trying to assess the effect of school type on civic attitudes or the production of civic capital. All three districts in Litt’s study were public and presumably had the same rough organizational structure (i.e., the “one best system” with a school board responsible for broad policy decisions and a centralized bureaucracy headed by a superintendent for day-to-day management). The variation in civic socialization examined in Litt’s study was explained by class differences, not by institutional structure. In essence, Litt suggested the upper middle-class was effectively using schools to perpetuate a set of democratically attractive civic attitudes and behavioral dispositions that could be used to help perpetuate their social position. Upper middle-class students were being trained to shape important public decisions, working-class students to accept them, or at least accept there was not much they could do about them. The general solution commonwealth-oriented scholars propose for this situation, i.e., one where schools reproduce unfair or unjust social hierarchies, is to force the entire education system into a tighter embrace of the commonwealth values public education is supposed to embody (e.g., Arnstine 1995). The problem pointed out by Litt’s study (and by the previous chapter) is that schools are embracing such values. It is the broader environmental context (largely socioeconomic, but also curricular) that drives how these values are transmitted to students and employed as agents of socialization. It is not the particular organizational or administrative arrangement that counts, but the controlling agents. Since Litt’s study, an extensive research literature confirms that education serves primarily to reinforce the political socialization desired by family or the broader “social milieu” (race, class, peers) of an individual (for a survey see Paulsen 1991). This presents a much more complex picture of “creating citizens” than intimated by abstract appeals to Jeffersonian ideals. If conflict theorists are correct, schools cannot escape the cumulative socialization effects of their external and internal environments. Schools simply reproduce or reinforce the value systems of these
110
The Ideology of Education
environments, which creates different socialization experiences for those with different backgrounds regardless of institutional structure. Wells (1996), for example, found that black students who transferred from inner-city to suburban schools in St. Louis felt pressure to adopt the “achievement” ideology of the white middle-class, i.e., the belief that education provides equal social opportunities and that any broader social inequities are the result of individual failure rather than systematic injustices. Adopting this value system, however, isolated some of those students from the cultural context of their families, neighborhood peers, and from their own personal experiences. The key factor in education’s role in the socialization process, in other words, was in amplifying the values of the localized culture, not the mechanics of school organization or governance. Adding to this general skepticism about the ability of institutional reform to produce socialization processes independent of community or culture is persuasive evidence that the positive civic attitudes attributed to education, especially those expressed by social haves, are superficial. When holding these civic attitudes threatens social status or position, or at least no longer can be employed to defend that status or position, the well educated tend to abandon them. For example, Jackman (1978) found that higher status, well educated whites did indeed have high scores on abstract indexes of tolerance (asked if they favored desegregation, they tended to respond in the positive). Yet when asked if they would support government action to uphold the abstract principles they approved of, commitment shrunk considerably, and the difference between the highly educated and the poorly educated in terms of support for government action became “trivial to nonexistent.” Seeking to explain this finding, Jackman concluded that education serves to familiarize people with the “correct” democratic response to controversial issues such as racial integration, but does not strengthen their commitment to the democratic principles they putatively espouse. Following up on this argument, Jackman and Muha (1984) replicated the findings of superficial commitment, concluding that the “dominant social groups routinely develop ideologies that legitimize and justify the status quo, and the well-educated members of these dominant groups are the most sophisticated practitioners of the group’s ideology” (p. 751). In other words, education teaches those at the top of the social hierarchy how to use civic values universally accepted in the abstract to retain and justify their position. The commitment to these principles nosedives once they threaten the self-interested preference for the status quo. Despite these complications, there is also an extensive individuallevel research literature arguing that education promotes the entire
Education and Civic Culture
111
gamut of attitudes and behavioral predispositions associated with civic capital and a strong civic culture. Education—variously measured as years received, test score performance, or participation in school activities and extracurriculars—is reported to independently promote tolerance, trust in others, support of democratic principles and norms, and participation in social and political life through membership in voluntary associations, voting or engagement in other dimensions of civic life (see Wolfinger and Rosenstone 1980;Bobo and Licari 1989; Paulsen 1991; McNeal 1999; Verba, Schlozman and Brady 1996). Despite the doubts raised by conflict theory, there remains considerable empirical support for education’s claim to make a universally positive contribution to the public good.
ORGANIZATIONAL STRUCTURE AND CIVIC CAPITAL
Given the complex and sometimes contradictory research findings, it should come as little surprise to discover there is significant disagreement on how particular institutional arrangements shape the production of civic capital. A number of studies report significant differences in how public and private educational institutions shape the cultural attitudes and orientations of their charges. Despite the routine claims from market advocates that private schools provide a better quality product than public schools, there is a long history of empirical studies finding that public school graduates outperform private school graduates in college academic work (e.g., Harris 1937, Seltzer 1948, McArthur 1954, 1960). Some explanations for why this occurs, at least for those who attend the most elite universities, lean heavily on the differing value systems promoted by public and private educational systems: public school graduates seem to rely more on merit and evidence a willingness to succeed on the reciprocal basis that hard work and performance will be rewarded, private school graduates (especially those of elite prep institutions) rely more on social prestige and connections for their success (Zweigenhaft 1992, 1993). Such findings support the argument that different types of schools pass on different types of value systems to their students. More impressionistically, Hutton (1995, 310–328) came to just such a conclusion regarding the role of elite private (confusingly called “public”) schools in the United Kingdom in shaping attitudes towards government. He argued that the private school system there replicates a particularly pure version of the market ideology, one that denigrates the worth and performance of public agents generally. Other studies indicate that market reforms in education may increase the production of social capital, but that it will be inequitably
112
The Ideology of Education
distributed. Schneider et al. (1997), for example, looked at the information networks of parents in choice and nonchoice public school districts to test the hypothesis that choice systems will stimulate citizen interest in acquiring information on the quality of public services. The quality of these information networks—in and of themselves a form of social capital—varied most notably by socioeconomics and race. The institutional effect of choice versus nonchoice districts on these networks tended to be very marginal, but when it did show up it tended to highlight segregation of such networks by class and race. While all of these studies are far from conclusive, cumulatively they do provide some empirical evidence supporting the argument that public schools promote the virtues associated with the egalitarian aspects of social capital to a greater extent than private schools, and that more market-like arrangements in schools may exacerbate stratification by race and class (Smith and Meir 1995, Henig 1994). Such findings support arguments that the socioeconomically advantaged are better positioned to reap the benefits of vouchers, choice, and privatization efforts, and the widespread use of an exit option will split school communities into those of haves and have nots. The end result being a more extreme version of the socialization process and resulting social inequity that Litt documented in “one best system” public schools. Such findings, however, are not universal. Though the evidence here is even less extensive, some studies do suggest that more market-like schools do produce greater civic norms in their students compared to private schools. Peterson et al.’s (1998) analysis of school choice in Washington, D.C., found that students in public schools report higher levels of cheating, violence, destruction of property, racial conflict, and general socially deviant behavior than students in private schools. The study also indicated that private schools may do a better job of promoting social capital than public schools, because private school parents are more actively engaged in school and community. In a methodologically more sophisticated analysis, Greene et al. (1999) examine the civic attitudes of Latinos and find a significant positive effect for private schools; they spell out a specific set of explanations for why private schools would do a better job of promoting civically desirable characteristics such as political engagement and tolerance for others: (1) Private schools are voluntary associations with a common purpose, and to join such a group means to adopt pursuit of that collective goal in a way that transcends race and other differences. (2) Private schools, especially Catholic schools, devote considerable pedagogical and curricular attention to promoting civic values, something often given lower priority in public schools. (3) The governance system of private schools may promote
Education and Civic Culture
113
democratic values because parents who choose their school (and sometimes pay heavily for the privilege) have a greater stake in the institution and as a result tend to participate more, a pattern that may extend beyond the school. Although the positive coefficients associated with years of private school education in this study support these arguments, socioeconomics also played a role (income was a significant factor), and because the analysis was focused on a particular ethnic group it did not account for differences in race. These mixed findings exacerbate the problems illuminated by the research looking at the more general impact of education on democratic norms and values. Education is widely expected to have a positive effect on civic capital, and this expectation is repeatedly empirically confirmed. Yet what role institutional structure plays in this, and whether it is independent of the effects of race and class is very much in dispute. Scholars who have specifically examined the link between market-based institutional reforms and democratic norms and values frequently argue the former is likely to have a deleterious effect on the latter. Cookson (1994), Henig (1994), and Kozol (1991) argue to various degrees that a system based on rejecting or shrinking regulation by the democratic process can hardly be expected to promote the communal commitments that characterize civic culture. The best can be hoped for is a superficial resemblance of such commitments, with a strong commitment to using such norms to preserve class, race, or religious prerogatives, but much less in the way of a genuine commitment to the communal whole.
RACE, CLASS, CIVIC CAPITAL, AND SCHOOLS
The arguments presented thus far indicate that several challenges present themselves to any empirical assessment of how (or even if) market-based education reforms shape the norms and behavioral predispositions that constitute civic culture faces. The first of these is a measurement problem. The social or civic capital that is argued to promote civic culture is a multidimensional concept, and to measure it most scholars combine a broad variety of indicators into a common index (e.g., Inglehart 1990, Jackman and Miller 1996, Putnam 1993). These usually combine attitudinal variables (e.g., tolerance, feelings of efficacy) with behavioral variables (e.g., reported participation or membership in voluntary associations). To be defended as valid, there must be some evidence that the variables constituting such an index reflect a single underlying dimension that can reasonably be viewed as representing “civic capital” or “civic culture.” This can then be used
114
The Ideology of Education
to assess whether there is variation on this conceptual dimension across differing institutional arrangements. Table 5.1 presents a set of candidates for such a measure taken from the characteristics of American grade school students: tolerance, efficacy, participation, and political knowledge, all concepts found in previous research to be central components of the values and behaviors that constitute civic capital (e.g., McNeal 1999, Katsillis and Rubison 1990, Flathman 1996, Bobo and Licari 1989, Jackman and Muha 1984, Paulsen 1991, Greene et al. 1999). The data for these variables come from the 1996 Youth Civic Involvement (YCI) survey, a national poll aimed at examining the civic involvement of youth in grades 6 through 12 (National Center for Education Statistics 1996). The variables are simple additive indexes constructed using sets of survey questions that tap into the underlying concepts. The tolerance index, for example, is TABLE 5.1 School Type and Civic Capital Correlations
Type of School
Political Political Knowledge Knowledge (1) (2)
Tolerance
Efficacy
Participation
Public Assigned
1.45 (.62) N = 3230
2.9 (.99) N = 3230
2.04* (1.16) N = 5040
2.03* (1.6) N = 1610
1.75 (1.52) N = 1620
Public Chosen
1.47 (.62) N = 578
3.02 (.94) N = 578
2.03 (1.17) N = 832
1.92 (1.61) N = 271
1.58* (1.48) N = 307
Private Religious Affiliated
1.47 (.60) N = 303
3.19* (.92) N = 303
2.64* (.98) N = 466
2.75* (1.61) N = 148
2.34* (1.58) N = 155
Private Non Religious Affiliated
1.68* (.56) N = 102
3.38* (.84) N = 102
2.46* (1.10) N = 156
2.85* (1.5) N = 47
2.47* (1.54) N = 55
Overall
1.46 (.62) N = 4213
2.94 (.98) N = 4213
2.09 (1.16) N = 6494
2.09 (1.51) 2076
1.78 (1.53) N = 2137
0–2
0–4
0–4
0–5
0–5
Scale
Means (standard deviations) reported. *Significantly different from overall mean at p < .05in a two-tailed t-test.
Education and Civic Culture
115
built from two questions asking whether students believed that people should have a right to speak out against religion and whether they believed certain books should be kept out of libraries. So a zero on this index indicates a student believes people should be prevented from speaking out against religion and that some books should be censored from library shelves, and a 2 indicates they support the freedom to speak against religion and oppose censorship of books. Indexes for efficacy (built from questions on whether a student understood politics and government, believed their family had a say in what government does and were able to contact a public official and to speak at a public meeting), and participation (whether a student participated in student government, in-school activities, out-of-school activities, and community service) were constructed in the same fashion. High scores indicate higher rates levels of efficacy and participation. The political knowledge variables consist of two separate five-question “quizzes” on political knowledge given to separate sub-samples of the survey, with the indexes reporting the number of correct answers (see methodological appendix for details). Overall scores on these indexes are reported, and are also broken down by a four-category school typology used in the survey. An initial glance at these figures indicates the average private school student— especially in nonsectarian schools—scores higher on these indexes than the average public school student. To get a basic reference point for which of these differences are meaningful, Table 5.1 also indicates (with an asterisk) which of the separate school means are significantly different from the overall means in a two-tailed t-test. A couple of interesting findings emerge from this basic analysis. First, choosing a public school does not seem to make much difference in the norms and behaviors associated with civic culture. The only significant difference in public schools chosen by parents and students was slightly lower levels of political knowledge. Second, scores from public schools essentially represent the middle and bottom of the distribution (though the range for all indexes for all school types ranged from the minimums to the maximums). Scores from private schools consistently place in the upper third of the distribution. This is especially the case with the efficacy and political knowledge measures. This initial analysis then, seems to indicate that private schools do a better job of promoting civic capital than public schools, though choice in public schools does little to replicate this advantage. What is left out here, of course, is any control for race and class. And this is the second big challenge for any attempt to assess the effect of institutional structure on civic capital; that institutional impact has to be separated in
116
The Ideology of Education
some fashion from the effects of race and class. In practice this is hard to do because school populations reflect the demographic diversity—or lack thereof—of their communities or clientele. Table 5.2 provides a glance at the ethnic and socioeconomic differences across school type, and the variation here is more dramatic than it is for civic capital.5 Students who are assigned to public schools are more ethnically mixed, are much less wealthy, and are much less likely to have parents who are college graduates compared with private schools. These all serve to create an a priori expectation that public school students will score lower on civic capital measures (see McClosky and Brill 1983). Some of the more interesting results presented in Tables 5.1 and 5.2 are the comparisons between assigned and chosen public schools. Students going to schools specifically selected by their parents are less likely to be white, are poorer, but come from families where the parents are slightly better educated than the typical public school-assigned student. While such results cannot be considered conclusive—and there is little in the survey to pursue the issue further—they do suggest that given an exit option in the public system, those most likely to exercise it are comparatively well-educated minority families seeking to leave schools they perceive as somehow failing. This is backed by the fact that students who TABLE 5.2 Race, Class, and Type of School
Type of School
Percent White
Percent with household incomes over $50,000
Percent receiving food stamps in last 12 months
Percent with parents who are college graduates
Public school— assigned
73
35.6
10.4
31
Public school— chosen
59
32.5
13.2
33.7
Private school— church related
77
53
3
51.7
Private school— not church related
84
62.8
6
68.8
Overall
72.6
36.1
10.1
34.3
Education and Civic Culture
117
are attending public schools by choice seem to be concentrated in large cities (44 percent versus 23 percent of assigned public school students). Backers of school choice have frequently argued that it can especially benefit inner-city minorities trapped in sub-standard schools, and conservative advocacy groups have found common ground with inner-city leaders of liberal leanings on just such grounds. The benefits of exercising this choice, however, may be limited. As Table 5.1 indicates, students in a public school by choice are little different across the civic capital measures from those in assigned public schools, and actually fare significantly worse in the measure of political knowledge. Whatever the benefits of public school choice, it does not seem to increase overall levels of civic capital and, at least when it comes to political knowledge, may actually reduce it. While extrapolating such inferences from descriptive statistics has to be considered a largely speculative exercise, these tables do indicate the variation of civic capital across race, class, and school type is consistent with the findings and arguments by researchers such as Jackman and Muha (1984), and Litt (1963). Private schools are largely (though certainly not exclusively) the province of the white middle- and upperclasses, and finding that this group is more supportive of civic norms and behaviors is not surprising: across the existing research literature, high socioeconomic status is probably the most consistent predictor of civic attitudes. These data, however, are not really equipped to answer the important question of whether the high scores reflect some genuine commitment to the underlying values of a democratic civic culture, or a more shallow commitment that consists of a learned response to the “right” values that can be used to help justify and maintain elite status. Nonetheless, the data are a good basis to attack the question of whether there is an independent institutional contribution to such measures of civic capital. This is at least as important as the commitment question because the civic culture argument for market-based reforms rests on two key assumptions: (1) If public schools were more like private schools, the civic capital of their students would increase. (2) If public schools were more like private schools, race and class would not be so important to civic capital, thus making its distribution more equitable. If these assumptions are correct, market reforms of public education can claim a commonwealth justification regardless of how deep or shallow the resulting attitudes turn out to be. As a first step towards a multivariate analysis aimed at such objectives, I collapsed the tolerance, efficacy, participation, and political knowledge variables into a single civic culture index using factor analysis, a technique designed to extract the shared variance underlying a set
118
The Ideology of Education
of variables to create a smaller number of measures (called factors) designed to represent the common dimensions underlying the variables. In this case, the factor analysis generally confirmed the assumption that these variables were tapping into a single concept, creating a single reliable factor that represented 40 percent of the variance in the original variables. The correlations between the four original variables and the single factor index were all significant and positive, ranging from .45 for tolerance to .8 for political knowledge (see methodological appendix for details). This index has at least face validity as representing the level of civic capital held by students. The key assumptions referred to above now become empirical questions: how good predictors of this variable are to race, class, and institutional structure? If private schools are consistent independent and positive predictors, market reforms may indeed have a commonwealth justification.
A MULTIVARIATE ANALYSIS
Using data from the CYI survey, a multivariate regression model can be specified to assess the independent effects of institution, race, and class, and how these interact with each other. The explanatory variables of primary interest in this model will be dummy measures indicating the type of school a student attends (public schools will be excluded as the reference category), and the key control variables will be those indicating race (a dummy variable where 1 = white) and socioeconomic status (household income, and dummy variables indicating whether the student’s family has received food stamps in the past 12 months, and whether at least one parent is a college graduate). These variables constitute the basic vehicle to assess the independent impacts of race, class, and institutional structure on civic capital. Some of that impact, however, is likely to come in a combination of these elements rather than independently because, as considerable previous research indicates, those of high socioeconomic status use schools to better equip themselves to more effectively participate in public life as a means of defending their social status (Jackman 1978, Jackman and Muha 1984, Litt 1963, Bobo and Licari 1989). The upper classes, in other words, may get a “value added” effect from schools, especially private schools, by using them to more effectively socialize students into taking on leadership roles in public life. To account for such possibilities, I also included two interaction terms that combined class and institutional type. These in essence were dummies where 1 indicated a student was from a family where the parents were college graduates, the
Education and Civic Culture
119
household income was above $50,000, and they attended private school (one interaction for religious-affiliated private schools, one for secular private schools).6 While these variables are of primary interest to this analysis, a number of other studies indicate they do not constitute a fully specified model to predict a concept such as civic capital. Most of what is important and omitted is encapsulated in the notion of an individual student’s “social milieu” (see Paulsen 1991). This includes a broader set of characteristics in family, social, and educational environment than is captured in measures of income, race, and type of school. To round out controls for the impact of family background, I include measures indicating whether the family allows the student to have a say in setting family rules, whether other members of the family participate in community service, and whether the mother works at least 35 hours a week. The rationale for selecting these variables was to assess whether civic capital can be generated by family example through parental civic or work involvement, or through allowing a child to participate in family decisions that have an important effect on a student’s life (these sorts of variables have been argued to be important source of civic skills and values, see Verba, Schlozman, and Brady 1996). To capture the effects of broader social setting, I included a series of dummy variables indicating whether the student lived in a small city (smaller than 50,000), a large city (greater than 50,000), or a rural area (suburb was excluded as the reference category). I accounted for school environment in two ways. The social environment of the school was represented by school size, and ethnic composition (the latter a dummy variable where 1 = at least 75 percent of students are of the same ethnicity as the respondent). I also wanted to assess whether school curriculum could have an impact on civic capital, and thus included a variable indicating if the student’s school required a service activity, and another indicating if the student had given a speech or an oral report as part of a required class assignment (again, both dummy variables). Finally, I also included a series of variables to account for individual student characteristics: age, sex, whether they worked for pay, and how often they watched the national news (see methodological appendix for more details on the model). Table 5.3 presents the results of the multivariate analysis. The findings on institutional impact are mixed. While all the coefficients are positive, their ability to support reliable inference is suspect in two cases. Public schools attended by choice and private schools with no religious affiliation are not discernible from public-assigned schools—the coefficients for these variables do not meet the p < .05 criterion, and are not
TABLE 5.3 Determinants of Civic Capital Confidence Limit
Leamer Bounds
.03 (.06)
–.06:.16
.01:.04
.366 (.117)*
.138:.595
.36:.38
Private Not Religious Affiliated
.186 (.24)
–.285:.657
.18:.71
Parents College Grads
.42 (.05)*
.32:.51
.42:.44
High SES* Private Religious Affiliated
.001 (.157)
–.29:.32
.001:.02
High SES* Private Not Religious Affiliated
.118 (.295)
–.45:.69
.11:.18
Household Income
.04 (.009)*
.03:.07
.04:.05
White
.20 (.05)*
.10:.31
.19:.22
Age
.11 (.01)*
.08:.15
.11:.13
Sex
.07
–.005:.16
.06:.07
School Size
–.01 (.02)
–.05:.03
–.06:–.01
Ethnic Composition of School
.002 (.04)
–.06:.10
.001:.002
Family Lets Child Have Say in Rules
.133 (.04)*
.05:.21
.11:.15
Works for Pay
.11 (.04)*
.03:.20
.10:.11
School Requires Service Activity
–.09 (.05)
–.20:.008
–.09:.07
Variable Public Choice Private Religious Affiliated
Coefficient
(continued on next page)
Education and Civic Culture
121
TABLE 5.3 (continued)
Variable
Coefficient
Confidence Limit
Leamer Bounds
Family Participates in Community Service
.22 (.04)*
.14:.30
.22:.25
Frequently Watches National News
–.15 (.02)
–.19:–.11
–.15:–.16
Gave Speech/Oral Report in Class
.38 (.05)*
.28:.48
.33:.38
Family Receives Food Stamps
–.13 (.08)
–.30:.04
–.12:/–.13
Rural residence
–.07 (.06)
–.19:.05
–.07:–.08
Small City Residence
–.08 (.06)
–.20:.04
–.08:–.10
Large City Residence
–.06 (.06)
–.18:.06
–.08:–.09
Mother Works
–.04 (.04)
–.13:.04
–.04:.11
Constant
–.27
N
.28
Adj. R-2
.27
*p < .05. Unstandardized coefficients (standard errors) reported.
bounded away from zero in the confidence limits. The coefficient for private schools with a religious affiliation, however, has a positive coefficient that is assessed valid by statistical significance, confidence limits, and Leamer bounds. Cumulatively, the findings weaken the initial impression conveyed by Table 5.1 but do not rule out the possibility of a systematic relationship between certain types of schools and variation in civic capital. Students at public schools by choice, and in nonsectarian private schools, have no discernible differences in levels of civic capital when compared with students at assigned public schools. All else equal, students at private, religious-affiliated schools, however, do have
122
The Ideology of Education
somewhat higher levels of civic capital. The difference is modest, but consistently positive (the coefficient is small, but more substantive interpretation is hard as there is no intuitive scale in a factor index; this issue is discussed more indepth below). The findings on race and class are generally more clear cut. Higher household incomes, having parents that are college graduates, and being white are positively associated with civic capital. The food stamp variable is in the expected direction, but is neither statistically significant nor bounded away from zero in confidence limits. Overall, these findings tend to reinforce the picture presented in Table 5.2. Simply, race and class are important determinants of civic capital. The interaction terms designed to address the possibility of a “value-added” effect between class and private schools do little to alter the inferences on institutional type or race and class. Neither of them cross the thresholds established to be judged reliably different from zero. Of the control variables, the most interesting findings are those associated with family background and school curriculum. Letting a child participate in family rule making, and parental involvement in community service are reliable positive predictors of the civic capital index. Family, in other words, seems to have a very important role in producing civic capital. Classes that require participation through oral reports or speeches have a similar impact. While institutional arrangements show modest to nonexistent effects in this analysis, schools do seem to be capable of independently supporting civic capital, though this seems to be done through curriculum and pedagogy rather than the differing organizational structures attributed to the public and private sectors.7 There are two surprises in the variables that controlled for individual characteristics. Watching the national television news decreases civic capital, a somewhat unexpected finding on the presumption that this would tend to increase knowledge and information that would support democratic values and effective participation. What may be going on here is that this variable is simply serving as a proxy for television viewing in general, and television has been argued by some to be a central culprit in declines of civic capital (Putnam 1995). The other is that working for pay is positively associated with civic capital. Although employment may put some pressure on academic performance (e.g., as homework is put off to go to work) the combination of social interaction with others and responsibilities it entails may serve to boost civic capital. Verba, Schlozman, and Brady (1996) make just this sort of argument. One of the problems with this analysis is the difficulty in substantively interpreting it—the dependent variable is a factor with no intu-
Education and Civic Culture
123
itive interpretation of its scale, so it is hard to gauge the substantive meaning of the coefficient reflecting a positive relationship between religious-affiliated schools and civic capital. One way to address this is to examine standardized coefficients, i.e., coefficients that present the effect of a given independent variable on the dependent variable in terms of standard deviations. This allows a rough comparison of the magnitude or importance of the independent variables by putting them all on a common scale of “standard deviation” units. There are 11 variables in the model that meet all three criteria to be considered reliable (i.e., statistical significance and bounded by zero in confidence limits and Leamer bounds). Of these 11 variables, the most important is having parents who are college graduates, which has a standardized coefficient of .2, indicating a change of one standard deviation in this variable is associated with a change of .2 standard deviation in the civic capital index. Rounding out the five most important variables as ranked by this method are watching television news (–.16), required oral reports or speeches in class (.15), a student’s age (.14), and household income (.13). In other words, of the five most important variables in the model, those with the greatest impact on civic culture, two are representative of class (income and parents who are college graduates), one is a curricular requirement independent of school type, and two are individual characteristics. The only institutional variable meeting the thresholds for reliable inference, attendance at a private religious-affiliated school, was ranked seventh, with a standardized coefficient of .09. Even when detected, the effect of institutional structure on civic capital is fairly modest. These findings bolster the cumulative inference that the presumptions underlying the market’s claim to commonwealth justifications for reform have little in the way of an empirical foundation. Private schools (or public schools attended by choice) produce little improvement in civic capital compared to public schools. Instead, class and curricular requirements are the key factors. When it comes to civic capital, the “one best system” seems to be at least as good as its private counterparts and makes its contributions with a much more socioeconomically and ethnically diverse student population.8
CONCLUSION
The two primary goals of this chapter were to examine whether schools positively affect the civic capital of students, and if so, whether certain types of schools were more effective at doing this than others. The
124
The Ideology of Education
answer to the first question is a qualified yes. Not only is there a lengthy empirical record finding that education is positively associated with the norms and behavioral predispositions that support a civic culture, the empirical analysis here also indicates that schools are capable of independently contributing to that culture. Requiring students to interact and defend differing points of view, as required in class assignments involving speeches and oral reports, seems to strengthen civic norms and behavioral predispositions. The qualification to this positive assessment comes from a research that raises some questions about how deep abstract professions of tolerance are, and whether participation and the political efficacy that motivates and supports it are universally beneficial for the communal good, or employed for more self-interested advantage. Regardless, the results of this analysis, and those of an extensive research literature, confirm the cultural mission of schools. The debate comes over whether this mission represents an egalitarian contribution to supporting the democratic good (consensus theory), or a more elitist contribution that helps reinforce the social status quo (conflict theory). The answer to the second question has fewer caveats. It is clear that levels of civic capital vary considerably by type of school, with public school students, on average, scoring lower on such indices than private school students. It is equally clear that this variation in civic capital is paralleled by differences in socioeconomic status and race. A multivariate analysis strongly indicates what explains the variation in civic capital is the latter (socioeconomics) rather than the former (type of school). There is some evidence that private religious-affiliated schools do have some positive effect on promoting civic capital. As school attendance and church membership have been argued to be important sources of civic skills and attitudes (Verba, Schlozman, and Brady 1996), perhaps the combination of the two is where any “value added” effect shows up. Given the dominance of Catholic schools in this institutional type, this is probably more accurately interpreted as a Catholic school effect rather than a general religious school effect. Regardless, even this positive finding turns out to be modest, with church schools only a marginal improvement over public schools. A marginal improvement, of course, is still an improvement. Yet the analysis in the last chapter suggets a tradeoff for this improvement. Religious affiliated schools are much more likely to prioritize instilling religious values or particular moral codes as a primary educational objective. The marginal improvement in civic capital—the greater political effectiveness bestowed by these schools—may thus be employed to advance sectarian interests in the polity in much the same way conflict theorists argue the socioeconomically better off use these advantages to sustain their own social positions.
Education and Civic Culture
125
Basically, two premises must anchor the claim that more market-like schools can independently increase the fund of communally-beneficial levels of civic capital in society: (1) More market-like arrangements in education will result in civic capital production that reflects the private rather than the public school performance in Table 5.1. (2) The increase in civic capital will not contribute to or result from the inequities across socioeconomic categories as reflected in Table 5.2. The analysis presented here indicates both premises are empirically untenable. Bluntly, the claim that institutional reform can independently contribute to civic capital gains above that produced by the existing system seems weak. This does not mean market reforms will not make some positive contributions to civic culture. Those contributions, however, must rest to a larger extent on assumptions that such reforms, (1) will not exacerbate existing segregation by race and class, and (2) that the higher civic capacities of the middle- and upper-classes will be used for universal benefit rather than to defend their social position. Given the existing research, both assumptions are questionable. At a minimum, the empirical evidence that market-like reforms will allow self-segregation by groups (not all necessarily by class, though this seems to be a common factor) and give them a greater ability to defend their social position and preferences is at least as strong as the counterevidence (see Wells and Lopez 1999). In commonwealth terms, the risks for market-based reform of education are fairly high, while the payoff remains uncertain.
This page intentionally left blank.
6 Education as Ideology
T
he central purpose of this book was to assess whether marketbased education reforms were more likely to produce positive, utilitarian change (improved economic prospects through higher levels of human capital, more equitable distribution of social and economic opportunity), or cultural change (a reshaping of normative institutional goals and how they shape individual attitudes and behavioral predispositions). The starting point to understanding these contrasting claims was to specify two value systems that defined and isolated the competing policy camps, and to claim these value systems constituted separate ideologies of education. These ideologies form differing pictures of the world and education’s role within and to it, establish different priorities in educational goals, and favor differing institutional models and education policies to reflect these visions and priorities. Although not formal deductive or inductive causal theories, these contrasting ideologies provided a reasonably stable and systematic platform for comparative empirical analysis. Their assumptions and beliefs about how the world should work generated a set of empirical propositions to test the assumptions and likely effects of the prescriptions that spring from these competing ideological frameworks. The overall inference from those empirical tests is that, although not without some merit, the functionalist claims of the market ideology are suspect in a number of key areas, and they seriously underestimate the negative social-democratic effects of market reforms on educational goals and civic values. The claims supporting market reforms rest to a considerable extent on normative beliefs about how the world should work, rather than empirical evidence about how the world is working. The analyses undertaken here tend to lend credence to a conflict theory perspective regarding market-based education reforms, i.e., they are as much about “smuggling values” as they are about boosting productivity or equalizing social and economic opportunities.
127
128
The Ideology of Education
Such findings have little impact on the market’s status as a value system, and they are unlikely to change any normative beliefs about the pros or the cons of market-based regulation. These analyses do raise a number of questions about market-based analytical constructs, such as public choice. As employed by academic policy analysts, these frameworks frequently support the argument that the market is a ready response to known failures in public education. Yet market-based theoretical frameworks are far from value neutral. They import into education studies normative preferences about the goals of education and reinforce assumptions about a wide range of relationships whose empirical validity is debatable. There is a common normative foundation to the market as a value system and as an analytic construct, and market-based analyses thus end up pushing values as well as promoting instrumental responses to problems in education. It is important to specify those normative assumptions and their implications for policy analysis and education reform.
ASSUMPTIONS
Advocates of market-based reform commonly make a set of assumptions regarding the goals and performance of public schools, and a further set of behavioral assumptions about individuals. Generally, the purpose of education is assumed to be largely economic in nature, the individual is assumed to bear the primary costs and benefits of schooling, and public schools are assumed have a poor performance record (for variations, critiques, and surveys of such arguments see Friedman 1982, Chubb and Moe 1990, McCabe and Vinzant 1999, Schneider et al. 2001, Apple 1986, Paris 1994, Hanushek 1994, Berliner and Biddle 1995, Mayer and Peterson 1999, Tyack and Cuban 1995, 141). Given this, it is assumed a rational individual seeking to satisfy his or her own self-interest will seek out high quality educational services that confer economic benefits, and that if competitive service providers are free to respond to such demands they will provide those high quality educational services (e.g., Becker 1993, Chubb and Moe 1990). These assumptions serve to define the basic problem of education and also provide the justification for its reform. Yet, as the analyses presented here indicate, the empirical veracity of these assumptions is open to challenge, and a case can be made that they are grounded to a significant extent in ideology, in a perspective of what the world should look like through the lens of a particular set of values. The goals of education and the presumption that individuals are its primary stakeholders have been discussed at some length in various
Education as Ideology
129
places throughout this book. While market advocacy and analysis tilts towards a functionalist perspective focused on the individual, there is a strong counterargument that public education’s primary goals are cultural and oriented towards the social aggregate. Both are normative arguments, though it is the latter that tends to have a stronger grounding in the constitutional language authorizing the existence of mass systems of public education. The assumption of education’s strong connection to economic success at the individual and aggregate level, while almost certainly valid on some level, simplifies what is a complex and only vaguely understood relationship (see Chapter two). When market frameworks are turned towards social-democratic concerns beyond economics, the assumption that individuals are the appropriate unit of analysis can lead to significant measurement and methodological issues. These tend to be resolved in favor of market preferences, and when the normative implications of these approaches are made explicit and brought into empirical analysis, the market case tends to weaken (see Chapter three). This is especially the case when school goals are addressed as an empirical “is” rather than a philosophical “ought.” For example, the primary difference between private and public schools in terms of educational goals has less to do with academic excellence—the assumed consequence of freeing supply to meet demand—and more to do with moral and religious indoctrination, and it requires something of a “leap of faith” to assume that by responding to individual selfinterest schools can promote a common set of civic values (see Chapters four and five). The empirical challenge to the presumption of public education failure has not been a primary focus within this book, but it is the subject of a fairly broad research literature, much of it focused on the role of test scores and what they say about school performance (for surveys see Henig 1994, Berliner and Biddle 1995, Rothstein 1997). Although considerably oversimplified, the basic empirical debate can be exemplified by Figures 6.1 and 6.2. Figure 6.1 uses national averages on SAT scores to provide a visual picture of a precipitous decline in SAT scores between the early 1960s and the mid-1990s. Specifically, Figure 6.1 replicates a graph popularized by William Bennett (1994) to provide direct evidence of a decline in educational performance (I use a longer time period than Bennett—1952 to 1995—to deliberately emphasize the point Bennett is making about the historical differences in SAT averages). Note that there are two periods of stability. The first is between 1952 and 1960, where means fluctuate around 970, and the second is after 1980, when scores stabilize around 900 after a two-decade decline. This conveys an impression of a clear and quantifiable drop in school
130
The Ideology of Education
FIGURE 6.1 SAT Scores 1952–1995
“output.” The numbers are indeed factual, but as Stone (1997, 167–168) argues, “numbers make normative leaps,” and decisions on how to measure and present something are often tied to a desire to change something else. Consider the scale used in Figure 6.1. The Y-axis recording SAT scores runs from 880 to 980. That is, it covers a 100point range of a 1600 point test (the SAT’s math and verbal sections each range from 200 to 800 possible points). Chopping off the range of scores beyond the measured decline has the visual effect of exaggerating the drop. Average SAT scores have stayed within a 100-point range for 40 years, but individual scores have varied well beyond this. What happens if this broader range of variation is taken into account? Figure 6.2 uses the same information as Figure 6.1. The only difference is the scale of the Y-axis, which in Figure 6.2 reflects a more representative range of 600 (representing students at the lower range of variation) and 1600 (representing students at the uppermost range of variation). This adjust-
Education as Ideology
131
FIGURE 6.2 SAT Scores 1952–1995
ment flattens out the downward trend considerably and tends to highlight stability rather than decline (no cohort average represented in the graph falls below 90 percent of highest average). As this relatively stable performance occurred during a massive demographic shift in the testtaking pool as a result of expanding educational opportunity, Figure 6.2 can easily be interpreted as a positive reflection on school performance (Carson et al. 1993). The point is not to make a case that the numbers support one interpretation or the other, but to make the case that it is the interpretation, not the numbers themselves, that are important. The choice between the contrasting messages of educational performance presented by these two pictures is a normative choice, a choice easily guided by an ideological preference. As Tyack and Cuban (1995, 112) observe, every education reform movement of the Twentieth century followed an identical two-step plan: (1) Attempt to convince the public that the existing system is “inefficient,
132
The Ideology of Education
anachronistic, and irrational.” (2) Claim a sure-fire solution. The reason for this endless tactical repetition is that regardless of the particular reform agenda, public school success robs the proposals of their primary justification. A case for public school failure, in other words, is a political necessity for those seeking a change in the educational status quo. If this is indeed the case, then advocates of market-based reform have a significant stake in public education’s failure, and thus have an incentive to interpret numbers, such as test score averages, a particular way. Opponents of market reforms, of course, have just as much of an incentive to filter numbers through an ideological lens, but the importance of doing so is more important to the success of market reform efforts. Public school failure is not simply a convenient justification for market reform but an important theoretical component for any quasimarket solution: dissatisfaction with current performance provides the basic source of unmet demand to motivate the responses of alternate service providers. Market reforms are at least implicitly predicated not only on the assumption that public schools are performing below par and that educational consumers know this, but also that educational consumers know what they do want and will rationally seek to satisfy those goals. This assumes there exists a large pool of unmet demand for high quality educational services, and that given an opportunity consumers and producers will seek to satisfy that demand. If such a demand does not exist, market-based reforms are unlikely to produce any significant improvements over public school “failure.” There are at least three empirical challenges to the demand assumption. One is that while there is evidence of a broad perception that public schools could be doing better, the claim of widespread dissatisfaction with public education is problematical. Attitudes towards education seem to exhibit a clear division, with local schools held in relatively high regard and public education as a whole rated considerably lower (see Loveless 1997). When the focus is limited to local or neighborhood schools, there appears to be little in the way of wide dissatisfaction with public schools (e.g., Goodlad 1984, 36; Rose and Gallup 2000, 43). Undeniably, there are population pockets where levels of dissatisfaction with local schools are high, and these tend to be concentrated in large, urban, heavily minority school districts where a case for failing institutions can be persuasively made. Overall, however, Americans tend to be fairly happy with their local schools, even if they have deeper concerns about public education in general. Indeed, the relatively high regard with which parents regard their local schools prompts some market advocates to grumble about the gullibility of this group of service consumers (see Berliner and Biddle 1995, 113; Witte 2000, 5–7).1
Education as Ideology
133
The level and quality of information held by educational consumers raises the second empirical challenge. For market mechanisms to work as advertised in education, there must be clear demands for suppliers to respond to and consumers must make informed, rational decisions among service providers to satisfy their demands. If they are easily fooled, as even some prominent market advocates suggest, they make easy prey for unscrupulous suppliers and reinforce the case for educational governance to tip toward hierarchical control. Most academic research indicates that people have relatively low levels of information about local schools, though there is some evidence that marginal consumers—e.g., people who have recently moved and thus possess an immediate incentive to gather information on local educational services—may be enough to make some form of quasi-market regulation work (see Teske et al. 1993, Schneider et al. 1997). Marginal consumers are going to represent a small fraction of a given community at any given point in time (probably less than 10 percent), but this pool may be constant enough to force response from suppliers. The marginal consumer argument, however, raises a normative democratic concern: Should education be institutionally reformed so that a good deal of regulatory power is shifted to such a small and localized minority? If the demands of the marginal consumer reflect the broader desires of the community, such concerns are lessened. If the marginal consumer does not reflect these concerns (e.g., if some marginal consumers want to be isolated from certain groups within the community), if their self-interests are not also communal interests, the negative implications are considerable. The third empirical challenge comes from research looking at what service providers are offering rather than what consumers are demanding. Private schools differ from public schools not by emphasizing academic excellence, but mostly by offering religious or value-based curricular elements not available in public schools (see Chapter four). What demands would drive a highly deregulated educational marketplace are the subject of a good deal of dispute. What current private sector service providers are undeniably providing that public schools do not are religious services and, to a marginally lesser extent, insulation from student populations in socioeconomic distress.2 In a context currently defined by high levels of satisfaction with the status quo, low levels of information, and religious services and social segregation serving as the primary selling point for the private sector, the assumption that a new market will be driven by rational consumers seeking academic excellence (i.e., greater individual acquisition of human capital) is propped up by ideological visions of how the world should work rather than the available empirical evidence.
134
The Ideology of Education
The cumulative point here is that a number of value-based assumptions or claims within the market ideology clearly extend into the theoretical frameworks of academic policy analysts. This is equally the case with the commonwealth, whose values also reach deep into the work of academics. However, this presents a greater problem for the market by weakening the argument that deregulatory reforms are ready solutions to known programmatic problems. If market reforms represent political preferences about the appropriate goals of education, then they face formidable commonwealth counterarguments that tend to be more comfortable (and constitutionally and historically supported) on such grounds. If this is the case, that education reform is an ideologicallygrounded dispute over normative agendas, one of the immediate cautions for academic policy analysts is the repetition of an old lesson— analytic frameworks, like the analyst, cannot be value free. Regardless of what education policy analysts produce, it will be judged at least in part on how it fits within certain ideological frameworks, and ignoring or ducking the normative elements of this enterprise is a disservice if the policy sciences are to live up to their democratic responsibilities to public education (see Lasswell 1951). Assuming these normative concerns can be safely ignored as long as research rigorously applies accepted scientific methods is not only fallacious but also, as Stephen Jay Gould, put it, silly: “Objectivity must be operationally defined as fair treatment of data, not absence of preference. . . . One needs to understand and acknowledge inevitable preferences in order know their influence. . . . No conceit could be worse than a belief in one’s own intrinsic objectivity, no prescription more suited to the exposure of fools” (1996, 36). Social scientists who study education routinely uphold the “irrational taboo” of forthright discussions of the values underpinning their analytic frameworks (Tanner 1998, 345). Given the strong argument for a set of value preferences underlying market-based (and, to be fair, commonwealth-based) analysis, this seems to be a particular disservice visited by academic policy analysts upon the important questions of education reform. Our responsibility is not to camouflage the ideological nature of the questions we pose with scholarly trappings, but to make clear how those foundations relate to the implications of our findings and to the political choices they prescribe.
THE BLACK BOX OF FUNCTIONALISM
The basic problem with using analytical frameworks grounded in market values and applying them to education is that they cover-up a basic
Education as Ideology
135
“black box” problem. In a functionalist sense, theoretical constructs such as public choice have never satisfactorily addressed how demand is actually translated into things like intellectual skills or cognitive capacity. Outside of religious orientations, there is little in the way of pedagogy, curriculum, staff competencies, or facilities within private schools that cannot also be found in the public sector. Indeed, institutional reforms that replace hierarchical control with a quasi-market system are in some ways more notable for what they do not change. For example, the easing of external controls may allow suppliers to respond to demand, but if existing market experiments and the private education industry are any guide, the internal changes within schools are likely to be minimal. Structurally speaking, the vast majority of schools, regardless of their governance structure, tend to look more alike than different in terms of curriculum and organization (Brown 1992). Organizationally all schools tend to exhibit the same basic bureaucratic characteristics (administrative hierarchy, division of labor between teachers and principals, etc.), and differences in primary curricular focus, pedagogy, and the like tend to be in degree rather than kind. In terms of pedagogy and curriculum, public schools, if anything, tend to be more innovative than private schools, at least in the sense that they are willing to respond rapidly to educational trends and community concerns (Hess 1999). And while the public school system offers less choice among service providers, internally it tends to offer a much greater set of curricular choices than private schools. Catholic schools, the dominant player in private sector education, tend to offer, as one Catholic school superintendent put it, “academic programs (that) are skin and bone compared to public schools. There aren’t a lot of frills” (Smith and Meier 1995, p. 5). And while there are undoubtedly many gifted and committed teachers in private schools, the same is true of public schools. The big difference between public schools and the market or quasi-market alternatives again seems to come down to students: private schools or publically funded charters tend to be taken advantage of by the better off, even when targeted at lower socioeconomic groups (Powers and Cookson 1999, Vandenberghe 1999, Smith and Meier 1995, Hoxby 1996, Bowe and Ball 1992, Lutz 1996, Marschall 1998, Hoxby 1994, Chapter 5). Rather than specify a particular set of processes that explain how deregulation can be more effectively translated into cognitive skills, market advocates tend to lean heavily on appeals to the “invisible hand,” Adam Smith’s term for the force that produces communal benefits from self-interested behavior (Smith 1986, Mueller 1993, Kuttner 1993). The general form of this argument is that the competitive pressures of the market will unleash the creative energy currently stifled by
136
The Ideology of Education
external control, resulting in new methods, new choices, and new opportunities that will benefit everyone (e.g., Gerstner 1995, Lieberman 1993). Besides its vagueness on specifics, such arguments largely ignore that markets have losers as well as winners, and that these down sides are unlikely to be avoided in a quasi-market system of education. As Mintrom (1998) observes, “arguments favoring market-like reform often betray a naive faith in markets. Frequently, they are informed by idealized conceptions of the market process rather than by careful observation of how markets work in practice, and how they interact with social institutions, such as government and family.” The market promise that runs through even much of scholarly literature on market reforms on education tends to lean heavily on pictures of wily entrepreneurs moving quickly to take advantage of competitive opportunities to provide informed and demanding consumers the educational services they want. Under such idealized conditions perhaps market-based reforms can be, as they have been described, a “panacea” (Chubb and Moe 1990, 217). Yet these idealized conceptions often have a hard time translating into reality, and when problems arise, their backers tend to rediscover the advantages of hierarchical control. When problems of amateurism, parochialism, inequality, or outright corruption arise under quasi-market systems, the same reformers who seek to tear down public bureaucracies on the basis of their institutionalized incompetence often suggest that well-crafted regulations, conscientiously enforced, will solve the problems (Henig 1998). In short, they rediscover the utility of traditional bureaucratic management mechanisms. A good example of the black box problem, and its reliance on faith in market values to eliminate a set of extraordinarily complex and difficult issues, is the record of for profit education companies. The notion of public authorities contracting out with private companies to run public schools or educational programs has a documented track record stretching back at least to the 1970s. Such contractual arrangements frequently retain a larger role for public authorities than a pure voucher system but remain premised on the belief that introducing market forces into the production of public services will increase efficiency and efficacy (Sclar 2000, 3). Those market forces are introduced by public authorities who, rather than provide services directly, issue a set of service criteria and call for bids from private sector companies to fulfill those requirements. The advantage of such arrangements is that public authorities can specify the results they want and what they are willing to pay for them, and in accepting a bid contractually bind a for profit company to meet this standard of performance. By continuing to elect the authorities who issue the contracts, all citizens retain indirect influence
Education as Ideology
137
over public education service providers and, disciplined by the competitive pressures of the market, for profit companies will be forced to focus on results or risk losing business. With missions contractually clarified, their survival dependent upon achieving those missions and satisfying their customers, these companies must perform or risk being replaced by a competitor. At least in theory, this makes them less likely to get bogged down in messy political conflicts, more likely to adopt “whatever works” innovations, and provides them with a large incentive to keep costs reasonable without harming quality (see Hill et al. 1997). The history of for profit involvement companies in public education during the past few decades demonstrates the idealistic nature of this theoretical picture (see Ascher, Fruchter, and Berne 1996). The two best known for profit education ventures of the 1990s were Education Alternatives Inc. (EAI) and the Edison Project. EAI started off the 1990s as the exemplar of privatizing education. It won a number of high profile contracts, including agreements to run significant segments of the public education systems in Baltimore, Maryland, and Hartford, Connecticut, on promises of improving quality without increasing costs. EAI’s fortunes, however, took a turn for the worse after winning these contracts. Teachers’ unions objected to EAI’s approach to schooling (especially its use of inexperienced or minimally trained personnel), the results on performance improvement were mixed, and the company ran into a series of financial problems. After sparking a series of high profile political battles within the communities it served, it lost its Baltimore contract in 1995 and the Hartford contract a year later. Changing its name to Tesseract, the company shifted its headquarters to Arizona, a state that in the late 1990s began pursuing deregulation in a large way through liberal charter school laws, creating a market that Tesseract sought to take advantage of. The company also underwent a change of focus. Initially geared towards taking over struggling public schools in low-income urban areas, it began focusing on providing a “Harvard” for the children of the well off. In 2000 it operated 37 school sites serving approximately 8,000 students, with the majority of its operations concentrated in Arizona. That same year it underwent a financial meltdown—losing millions of dollars, its stock plunged to an all time low of 9 cents a share, it laid off much of its staff, was delisted from the Nasdaq stock exchange, and was so broke it could not afford the postage to mail report cards.3 This was an startling implosion for a company that during the 1990s had been a highly effective fund raiser. John Golle, chief operating officer for much of the company’s history, made millions cashing in stock options in the early 1990s when its stock price was relatively high, and EAI’s top officers earned incomes ranging from
138
The Ideology of Education
$100,000 to $350,000 per year (Miner 2000). The company’s effectiveness in demonstrating the market’s ability to produce a viable, long-term alternative to traditional public schools never matched the ability to secure generous salaries for its executives. The Edison Project began in 1991 when its founder, media entrepreneur Christopher Whittle, announced a plan to build 1,000 new forprofit schools. The project immediately attracted $44 million in private investment and gained instant credibility by hiring people like former Yale University president Benno Schmidt to translate the vision into reality. By 1993, the scale of the Edison Project was reduced, and it focused on managing public schools rather than operating private ones. It began operating its first public schools in 1995, and in the next five years raised millions more in private investment and expanded its operations to running 79 schools in 16 states and the District of Columbia. In 1999 it changed its name to “Edison Schools, Inc.,” and like EAI/Tesseract before it, went public. Like EAI/Tesseract it also struggled to show a profit, losing hundreds of millions of dollars during its existence, while paying its top executives generous salaries. Unlike EAI/Tesseract, Edison has a solid track record when it comes to academic performance. Most notably, Edison elementary schools have a strong focus on Success for All (SFA), a reading program designed to raise student achievement in low-performing schools. Edison uses a longer school day and a longer school year than most public schools, something that many parents find attractive, and it runs some charter schools that, most observers agree, are doing well. The performance record in terms of raising achievement test scores is mixed and contested (Walsh 1999) .4 Edison’s critics argue it does little that is unique, obscures its true educational expenditures for public relations purposes, cuts costs with large class sizes, has high staff turnover, relies too heavily on inexperienced (and therefore cheaper) teachers, and that its test score record is exaggerated and benefits from running schools that attract students from high socioeconomic backgrounds (American Federation of Teachers 2000). At a minimum, it is probably fair to say that Edison’s has provided little pedagogical and curricular innovation, choosing instead to adopt successful approaches largely developed elsewhere and that are widely employed by other public schools (this is the case with SFA). It probably can spend more money on its schools thanks to its focus and success in attracting private investment. As a viable quasi-market alternative to hierarchical control, it is also probably fair to say that Edison and the other private education management companies that have followed in its wake have yet to prove themselves.
Education as Ideology
139
While their localized successes or failures can be debated, there is no doubt the idealized pictures of private ventures producing a broad scale, universally attractive alternative to traditional forms of public education has, thus far, failed to materialize. Private education companies have consistently run into problems with teachers’ unions, often failed to deliver on their high promises, and raised questions about their credibility as worthy investments of public trust by providing their executives with lavish salaries while seeking to cut instructional labor costs. Rather than innovate, they have tended to adopt successful approaches already employed in public schools, and their record on purely market criteria is dismal—collectively they have lost enormous amounts of money over extended periods of time (see Molnar 1994, 1995, Miner 2000). Public bureaucracies with a similar records—being flashpoints for political conflict, having mixed performance records, failing to innovate, and spending large sums of money for debatable returns—would likely be (and in education, frequently are) excoriated. Backers of these ventures argue they are learning from their mistakes and that the promises of the market can still be realized in public education (Chubb 1997). Yet from all of this, there is still no clear explanation or evidence of how deregulating supply transforms demand into cognitive skills. The results of companies like EAI/Tesseract and Edison are perhaps most remarkable for the growing realization that for profit schools were in many respects the same as public schools. Some of these schools are good, and some of them are not so good. What seems to separate the successful from the not-so-successful are adequate funding, good teachers, good leadership, a vision and the autonomy to act on it, and broad-based support within the community (Nifong 1996). These are exactly the same characteristics associated with successful public schools, and none seem exclusive to a particular institutional arrangement except from an ideological standpoint, i.e., a value-based belief that one option will generate more of these positives than the other.
THE CULTURAL BLACK BOX
The market actually has a much clearer explanation of how a deregulated education system can link and translate demand into values, though the black box reappears when these values are extended to social aggregates. This difference turns on the extent of external control by democratic institutions. Under their traditional forms of governance, public schools frequently emphasize the less pleasant sides of public life, the conflict and need for compromise when dealing with heterogenous
140
The Ideology of Education
groups and opinions. This serves to reinforce a communal commitment to democratic values: “Through their involvement in democratic structures, people in schools not only learn how hard it is to be democratic, but also come to understand the new challenges presented by large and diverse groups to democratic structures” (Mosher, Kenny, and Garrod 1994). These very visible conflicts and the disagreements they generate are one of the problems market analysis identifies with democratic governance (Hill et al. 1997, Chubb and Moe 1990). They can be eliminated, or at least diminished, by freeing up the supply of educational service providers and allowing schools to respond to different sets of preferences. The market promises to get the politics out of education on the theory that if you get what you want, the cause of political conflict disappears. The problem with this, at least from a commonwealth perspective, is that it contracts the public sphere and diminishes the importance of democracy to the polity. If you do not have to deal with those who are different in terms of background, personal characteristics, and opinions, there is little incentive to search for common ground and little need for democratic processes that support such quests. Political scientists, in particular, have paid increasing attention to market-based reforms with an eye towards education’s democratic implications. Those analyzing vouchers, school choice, and charter schools have offered some preliminary evidence that, carefully adopted, some institutional features of the market might contribute to increasing civic engagement and encourage places for “deliberative democracy” (e.g., Marschall 1998, Hoxby 1996, Schneider et al. 2000, Mintrom 1998). The key positive social contribution identified by such reforms is that coupling an exit option to voice option, i.e., promoting access to decision makers to voice concerns and complaints and giving such communications weight with the threat of exit, can improve a school’s response to its primary clientele, increase levels of participation, and drive up levels of parental satisfaction. These positive reforms are generally attributed to the extra degree of autonomy given schools under such reforms rather than the result of any competitive pressure placed upon them. The exact institutional characteristics that can help promote direct citizen engagement in decisions that affect their own lives are still a matter of some debate, but some increase in local autonomy (which the market seeks to guarantee for all service providers) is the common factor. While there are some positive signs here, overall the evidence on the potential for market reforms to make positive social-democratic contributions is mixed. The possibility of these reforms contributing to “skimming” or social stratification on the basis of class or race are noted by
Education as Ideology
141
some of the same researchers who identify the positive social contributions of such reforms (e.g., Scheider et al. 2001, Hoxby 1996). And an overreliance on the exit option seems to undermine the potential to promote civic engagement, as high rates of turnover can maintain rather than address organizational effectiveness (Mintrom 1998). And while decentralization and local autonomy in education systems may promote greater participation and increase levels of satisfaction, historically such reforms have also been shown capable of promoting dissension, confusion, and outright corruption. A good example of this are some of the high profile experiments using publically funded vouchers to allow students to attend private schools. In Milwaukee, where probably the best known voucher system was established in the early 1990s, several private schools failed (at least one is under charges of embezzlement and fraud), there has been enormous turnover of participants in the plan, and the effects of the quasi-market experiment are a matter of extreme controversy, prompting a number of high profile and quite bitter exchanges among those studying the program or with a stake in its success or failure (see Witte 1996; Peterson and Noyes 1997; Greene, Peterson, and Du 1996). The program has promoted a divisive legal battle, an influx of outside money and influence into school board races, and accusation and counteraccusation on all sides. According to former Milwaukee school Superintendent Howard Fuller (a supporter of the choice program), “It’s like a war in Milwaukee. It’s an issue that divides people politically, philosophically, and ideologically” (quoted in Koch 1999, 32). Regardless of whether the experiment has increased outputs—and even granting the positive case to the market advocates, these seem to have been minimal—there is little doubt this policy experiment has promoted dissension, tarnished professional reputations, undermined the legitimacy of public institutions at the state and local level, and shrunk the possibilities of finding common ground on education policy reform outside of specific partisan or ideological agendas. In commonwealth terms, whatever has or is to be achieved, the price for doing so has already been high. Even accepting the positive findings on this issue and ignoring the negative, there remains a problem with how market-based reform will affect public education’s relationship with the larger social aggregate. Although definitely contestable, assume that carefully crafted versions of voucher, choice, charter, or contracting out arrangements are capable of promoting greater levels of participation within the school community. This does not necessarily mean that these internally cohesive, civically engaged, and flush with social capital groups export these these same things into the general community. Such arrangements may be used as a
142
The Ideology of Education
vehicle to benefit certain groups who are better able to serve their own interests even if it is at the expense of the broader community. One of the longest-running and most extensive examinations of charter schools has paid particular attention to this issue, to the potential for marketlike reforms to make positive contributions to social-democratic values and to its potential to exacerbate social divisions, and found them capable of doing both simultaneously (Wells and Lopez 1999). Charter schools are sometimes used as a way to fashion institutions that serve to build community and foster participation in public life. Charter schools, for example, can provide a sense of emancipation and freedom to traditionally disenfranchised groups or communities long frustrated by the inability of local school systems to provide high quality educational opportunities. The negative companion to this positive portrait is that the advantages of such reforms tend to accrue to particular groups of people. Privileged communities use charter schools to break off from the traditional public school system and in doing so help protect their social advantages and help replicate existing social hierarchies. In less privileged communities, charter schools tend to benefit students whose parents are most committed to and engaged in their children’s education. This provides a significant benefit to the charter school and those connected with it, but it also deprives the broader public school system of desperately needed social capital. Charter schools, in other words, might well help to support the shared political project at some very localized level. Yet they may also serve to exacerbate the existing unequal distribution of social, economic, and political opportunities between and within communities. What determines the extent of the contribution to either of these potential outcomes seems to rest to a large degree on who is the driving force behind the school (activist, teacher, parent, entrepreneur) and what their primary motivations are for doing so (religious or ideological indoctrination, community building, academic or pedagogic excellence, defending favored social position, profit, or some combination of these). The central concern regarding the market’s potential adverse socialdemocratic effects can also be exemplified by a thought experiment proposed by Eamon Callan (1997, 1–2): imagine a world where the rights associated with a liberal democracy—freedoms of expression, political participation, and the like—have the force of law. Imagine a society where wealth is distributed according to whatever criteria you think is best, and where peace and prosperity are the norm for all citizens. Then introduce to this utopian picture a set of worrisome social-democratic characteristics. When elections are held, hardly anyone votes. The media all but ignore politics, because their consumers do not find it appealing.
Education as Ideology
143
Political parties are controlled by a common set of elites, and little separates them. Citizens are largely indifferent to questions of good and evil because they see “the point of their lives simply as the satisfaction of their desires, or else they commit themselves so rigidly to a particular doctrine that dialogue with those who are not like-minded is a thought to be repellent or futile.” The society contains a diverse number of groups defined by race, culture, class, or demographics, but they shun each other. Freedom of speech has a largely vestigial role because although the groups respect each other’s legal rights, their dislike for one another means speech is largely used for condemnation, and only rarely to defend distinct visions of the common good. This scenario, which Callan refers to as The Brave New World, seems quite clearly to embody much of what the market ideology promises to embed in schools and other political institutions: a focus on individual goals and interests rather than collective ones, a significant reduction of the importance of public institutions and a narrowing of their purpose to efficient service provision, and a redefinition of democratic freedom as the right to consume what you want when you want to, and an institutionalized denigration of the democratic process. The empirical picture reported by Wells and Lopez (1999) provides a positive a side to The Brave New World conceptualized by Callan, but it does little to contradict the potential down side illuminated in such abstract exercises. The contrast between the arguments of market advocates and the empirical and conceptual portraits by Wells, Lopez, and Callan reflects a critical difference in the conceptualization and provision of common values. Macedo (1998) argues that while the liberty of the individual and of private communities must be protected, the support these same individuals and communities provide to the shared political project is of enormous concern. A liberal democracy requires a certain type of civic culture and certain types of public institutions. Public education systems are central to this project because they constitute a critical mechanism to store and replicate that culture; they represent the commitment to the shared political project. Commonwealth advocates view the creation and maintenance of this civic culture as a specific political objective that needs to be reflected in the institutional specifics of education systems. Market advocates certainly do not ignore the common good nor view it as irrelevant. But they also tend to assume it is a natural product of education, and that a deregulated system of education based upon response to individual self-interest will produce a common good. This assumption rests heavily on appeals to the market’s “invisible hand,” not on a clear explanation of how individual self-interest is transformed into communal good by a set of institutional reforms.
144
The Ideology of Education
These differences can be exemplified by the role of religion in education. The market finds little objectionable about encompassing schools that promote particular religious beliefs or practices as the central source of “truth” about all realms of social life under the “public” umbrella. If people take tax money to such institutions to satisfy their self-interest, then beyond a set of minimal regulatory matters, all is well from a purely market perspective. The high degree of satisfaction within this school community will promote participation and internal unity in the pursuit of a common goal. But what of this community’s relationship with the larger polity? Commonwealth advocates argue that part of the cost of assimilation into a liberal democratic culture is the recognition that religious views are, in terms of the social aggregate, politically irrelevant (Macedo 1998; see also Herberg 1960). Religiously affiliated education institutions that teach differently are accepted in a liberal democratic culture that takes seriously its commitment to individual liberty. But from a commonwealth standpoint, such institutions cannot be termed “public” regardless of where their funding comes from. Their broader relationship to the polity is deliberately limited, and may even be openly antagonistic. This simply does not fit with the commonwealth conception of public education, and to argue otherwise requires a redefinition of “public.”5 It is not just the separatist potential of religious affiliation on the common weal that concerns commonwealth advocates. More general and wholly secular market values are, if anything, seen as more of a threat to core democratic values. Historically, education has been conceived of as a public good and a fundamental right, and as such, democratic values have taken precedence over commercial values in public schools (Giroux 1998). The market’s penetration of public education threatens that hierarchy of values by institutionalizing a set of political preferences that erode democracy and redefine it in market terms. The effect of such cultural pressures extends well beyond school: increasing numbers of young people define democracy as the freedom to buy and consume what they wish (Wright 1997, 182). This is democracy cast largely in market terms, a concept barely recognizable as democratic from the commonwealth perspective. Institutional reforms that emphasize individual self-interest and take the public contributions of education as givens rather than specific political objectives only serve to weaken the commonwealth social-democratic vision. The potential for market values to create something along the lines of Callan’s “Brave New World” is supported by a series of studies that examines the behavior and attitudes of those who are exposed to market values in the classroom. Students of economics, for example, fre-
Education as Ideology
145
quently behave more self-interestedly and are less committed to the provision of public goods than those trained in other disciplines (Carter and Irons 1991, Marwell and Ames 1981). Frank et al. (1993, 1996) conclude that at least some of the differences in the willingness to cooperate for communal benefit between economists and noneconomists are attributable to the former’s academic training. Blais and Young (1999) report that students being taught rational choice theory—which has selfinterested utility maximization as a basic theoretical assumption—were less likely to vote. In short, the assumptions and values at the center of public choice theory, the same values and assumptions that underpin market-based reform of education, have been shown to be capable of socializing people into behavioral predispositions that are not particularly conducive to a strong civic culture. Are these same values and assumptions likely to produce similar results if they are institutionally embedded in the regulatory framework of public education? From the commonwealth perspective, this a question of no small significance, especially if conflict theorists are correct. Democracy, as Deborah Meier (1997, 61) argues, “is not a ‘natural’ form of human organization. It depends . . . on certain habits of mind and heart that have very insubstantial foundations in most societies.” It is not particularly clear how values and assumptions seem to have a negative effect on civic engagement when an explicit part of the curriculum can reverse this effect when they are used as the basis for a regulatory framework. Again, a “leap of faith” can traverse the black box, though this does little to assuage commonwealth concerns.
POLICY ANALYSIS AND POLITICAL VALUES
Political scientists have long recognized that industrialized democracies tend to support two cultural conceptions of government that produce tension between market and democratic belief systems. The ideologies of education are in one sense simply extensions of a broader cleavage in political culture. Generically speaking, the market is a predisposition to view bargaining among self-interested groups and individuals as the appropriate path to political order. The commonwealth prefers political order to be a product of cooperative effort among citizens to promote the common interest (Elazar 1977). These represent two value systems that push government in different directions and often resist easy synthesis. The source of much of the historical tension in the American political process, they offer competing views not just on what government should or should not do, but what government is. The market and
146
The Ideology of Education
the commonwealth encapsulate opposing sets of values that opposing groups have long struggled to embed in public institutions. The conservative renaissance that rippled through the political process in the last two decades of the twentieth century pushed the marketplace conception of government to the fore. Virtually everywhere progressive notions of participatory democracy lie dormant or in retreat. Far from being viewed as the social mechanism of conflict resolution and solution, the institutions of representative democracy are vilified as the source of society’s problems. Legislatures are perceived as corrupt, unresponsive, and overbearing. Public agencies are viewed as slow and sclerotic, promoting the interests of a favored few while choking the general social process with rules and red tape. The proffered remedy is to cut government back to a minimum and have what remains mimic the organizational framework of business. Transform the citizen into a consumer, eliminate monopolistic governmental control of a broad swath of public programs, link the supply of public service to demand, and the invisible hand will transform individual self-interest into the communal good. Government is to be thus “reinvented” as a benevolent corporation, responsive to the demands of its customers and sensitive to competitive pressures (for variations of these arguments and their likely impacts see Osborne and Gaebler 1993, Kuttner 1997, Hibbing and Theiss-Morse 1995, Plank and Boyd 1994). This is the political context in which education reform debates are currently embedded, and this context tends to favor market values. As the public sphere has come to be less and less valued, the purpose of education has increasingly narrowed to economic concerns, and the basis for judging educational success has increasingly narrowed to test scores. Under such circumstances it becomes logical to promote market-based alternatives to common schools, and academic policy analysts have done exactly that (Tanner 1998). Yet such alternatives ignore public education’s historical purpose to provide “comprehensive social and political goods more than narrow, instrumental ends” (Tyack and Cuban 1995, 141). And presenting them as logical policy prescriptions on the basis of scholarly analysis ignores the value-based foundations of the whole reform enterprise. Academic policy analysis should seek to clarify political choices, not camouflage the fact that they are political. Ultimately, as a practical matter, it is pointless to seek an ultimate victory for one ideology’s prescriptions. Whether the education system in the United States undergoes radical transformation in the coming decades or fossilizes the status quo, the competing value systems will have to accommodate each other in some fashion. The pragmatic question is not necessarily which ideology has the “best” or “right” vision of
Education as Ideology
147
education, but how one can seek and find common ground with the other. It is here, where education reform can find political acceptance, if not enthusiastic support, and where reforms are likely to have lasting and useful impacts. The market seeks a loosening of supply to promote efficiency, performance, and response to individual demand. The commonwealth seeks a system that preserves citizen involvement and participation, equity, and accountability to the common weal. How can all of these preferences be accommodated? How can the civic goals of education be preserved in the face of broad social, economic, and political changes that make market values ascendent over commonwealth values? How can public education be institutionally preserved on commonwealth terms when there is growing pressure to deliberately dismantle this institutional core and distribute the fragments to those who want them? Despite their importance, academic policy analysts have nothing approaching a consensus on an answer to such questions. One reason for academics’ inability to provide much in the way of solutions is obviously the critical role of values in determining how such questions are answered. The disagreements generated by such questions have less to do with facts and data, the “stuff” of empirical social science analysis, and more to do with normative preferences, assumptions about what “stuff” is important and how it should be treated. Nonetheless, the policy disciplines, as Herb Simon argued (1969), are sciences of the artificial: they are concerned not just with what is, but what might be. In that spirit, it is perhaps fitting to conclude with a set of policyrelated suggestions designed to bring the normative foundations of market-based education reforms to the fore and explicitly include them in political decision making. The first of these suggestions is to public school advocates in the commonwealth tradition: recognize the market has a powerful normative case because it appeals to the value preferences of what has become the political mainstream, and it is unlikely empirical analysis is going to do much to diminish that appeal, at least in the short term. Perhaps unwittingly, defenders of the commonwealth have actually made the market case much easier. As Tyack and Cuban (1995, 140) argue, public school advocates are also frequently its most persistent critics, and some of their criticisms are now being used to justify eliminating the common school that defines the core of their philosophy. If these criticisms are valid, alternative responses to them should be given serious consideration, and this means not rejecting all market-based reform simply on principle. The second suggestion is to market advocates: recognize that over the long term that commonwealth concerns about the market’s potential
148
The Ideology of Education
for negative social-democratic impacts will have to be addressed. The market case is not going to be substantially bolstered by another education production function study claiming a quasi-market institutional characteristic marginally improves test scores. Not only have the empirical veracity of such claims come under increasing pressure in recent years, these studies are also vulnerable to normative criticisms regarding methodology, measurement, and model specification that are capable of severely mitigating their validity (see Chapter three). More of these studies are unlikely to “prove” anything, at least in the sense of shifting existing preferences.6 Scholars who are paying serious attention to the impact of quasi-market reforms on social networks, civic engagement, and deliberative democracy are probably doing the most important work in terms of effecting the long-term sustainability of market preferences in education (e.g., Mintrom 1998, Schneider et al. 2001). These sorts of studies are not producing uniformly positive confirmation of market preferences, but they have provided at least some preliminary evidence that, under limited circumstance, some market elements can be incorporated into education with positive social-democratic results. The third suggestion is for policy makers generally, rather than to any particular camp: keep market experiments small, at least initially. Current experiments with charter schools, vouchers, and contracting out have produced no definitive results that conclusively support the claims of their advocates or the worst predictions of their advocates (see Witte 2000). Coupled with the quasi-market’s very mixed record in education elsewhere, caution is warranted. The political environment will probably brake the speed and scope of market reforms anyway, but the caution bears telling regardless. More than a century of constantly searching for some mechanism to introduce radical improvement to public schools through altering governance or organizational arrangements has produced little except for marginal improvement—or marginal failure— compared to the status quo (see Tyack and Cuban 1995). There may be some schools, perhaps even some school districts, that are so bad there is little to lose in rolling the dice and swapping hierarchical control for a quasi-market system. Instituting that system, however, is unlikely to remove some familiar issues: lack of fiscal resources, difficulty in defining parental roles, establishing appropriate school climate, culture and decision-making mechanisms. As a number of charter schools have discovered, ridding yourself of traditional forms of governance creates the need for a new form of administration, and under such conditions old fashioned bureaucracy sometimes can suddenly seem attractive. Even for profit companies acknowledge that their administrative challenges and approaches are very similar to that of public schools, recognize that pub-
Education as Ideology
149
lic schools run a relatively tight administrative ship, and reject the common conception of American education under a bloated and inefficient bureaucracy as unfair and inaccurate (Whittle 1997). Conditions of uncertainty are common to most schools, and in uncertain environments it is important to preserve the things that are working reasonably well, and this describes a large portion of public education. The process of public education seems to be such a pervasive and central part of the nation’s social life that we often only notice its failures, are unable to see its successes, and are sometimes curiously incapable of distinguishing between the two. An attempt to make that distinction explicit should be made before engaging in any sweeping reform. The fourth suggestion is that any market reform embodied in public policy—and this certainly includes vouchers, contracting out, and charters—should include value-based constraints. Specifically, any private or nonprofit institution that agrees to take public money to provide grade school educational services: (1) Must agree to make any religious instruction optional for the student. This does not constitute a ban on religion in the curriculum, it simply means you cannot take public money and use it to coerce someone into religious indoctrination. Students at a school on the taxpayer’s dollar should be allowed to opt out of such instruction if they so choose. (2) Must agree to have an open admissions policy. This means no selection criteria on the basis of academic performance or aptitude, special needs, family background, or characteristics such as race and religion. Anyone who has a right to a public school education (in other words, just about anyone who is a resident and within set age limitations) should also be able to exercise that right from an educational service provider who is receiving taxpayer money. (3) Must meet all the same certification standards (teacher competence as well as health and safety standards), and all basic statewide curriculum mandates that public schools must abide by. (4) Must make all records—financial and governance as well as academic—open to public inspection. This is simply an accountability demand, not in the sense of a performance standard, but in an insistence that the public has a right to know where its money is going and what policies and decisions it is being used to support. These constraints are specifically aimed at forcing the underlying values discussed throughout this book to the forefront of policy decisions. Those who oppose these constraints are, in effect, forced to argue that alternatives to traditional public education should be able to use public money to promote religious agendas even against a student’s will, that they should be able to use public money to create elite student
150
The Ideology of Education
bodies, that they should be held to different standards of teacher competence and health and safety than public schools, and that the public should have no right to a paper trail that documents how taxpayer money is spent and what it is used to support. Obviously, these are normative issues, and, equally obviously, they are tilted towards forcing reforms to confront contradictions between market and commonwealth values. Under these conditions, i.e., under explicit commonwealth-based value constraints, participation by private and nonprofit schools is likely to be considerably lower than if no such constraints are present. A 1997 Department of Education survey of private schools assessing their willingness to accept public school students (and public money) found that two-thirds would be likely unwilling to do so if it entailed permitting exceptions from religious instruction or activities, and three-quarters would be unwilling if it required them to accept special needs students (Koch 1999, 26). The arguments such constraints are likely to promote will have less to do with efficiency, individual choice, human capital, and test score performance and much more to do with ideological preferences. And when the argument is on those grounds, the commonwealth ideology can likely hold its own in the political arena with the market ideology.
Notes CHAPTER ONE: IDEOLOGY AND EDUCATION
1. There is some research that has looked at the effect of market-based reforms on social capital and civic virtues, with some positive results (e.g., Greene et al. 1999). However, whether this is due to an actual effect of market-like institutional arrangements or a reflection of socioeconomic stratification is very much in question (see Schneider et al. 2001, and Chapters four and five in this book).
CHAPTER TWO: EDUCATION AND THE ECONOMY
1. Education production functions treat education systems as analogous to business firms, as organizations that transform a given set of inputs into a set of outputs. This approach does not rest entirely on human capital arguments, though the most significant controversies in this literature—especially the role of money and its relationship to outputs—are in essence disagreements over a given set of returns for a given investment (see Burtless 1996). The parallels with human capital premises seem clear. 2. This is exactly the claim that Murray and Herrnstein made in The Bell Curve (1994). The methodology and the conclusions of this work have been soundly criticized, and the claim that intellectual ability is going to be randomly distributed across populations and geographic locations seems more empirically supportable given the historical evidence (Gould 1996). 3. Jencks and Phillips (1998) argue that closing the gap could go a long way to evening out economic and social opportunities. 4. This is likely to encounter some objections because it goes not only against the individualistic anchors of human capital theory, but also the microeconomic preferences of the market. Nonetheless, there is no getting around the fact that for most people investing in K–12 education—in terms of money, time, or even particular institution—is decidedly not an individual decision. These are
151
152
Notes
made at the level of the collective, either by school boards or state legislatures. And even under most market reform schemes they will mostly remain collective decisions. Vouchers might offer some choice of institution, but they will not negate compulsory attendance laws, and the amount of the voucher will be decided at much higher levels than the individual. If this reasoning is not persuasive enough, it is also the case that to serve as a basis for policy prescription for public education, human capital must be shown to work in the aggregate— i.e., that whatever individual benefits it confers collectively contribute to the common good. If it does not work at the aggregate level, the justification for reforming education on the basis of its collective importance to the economic task environment is weakened considerably. 5. A big difference between what is presented in Table 2.1 and the Brace (1993) and Smith and Rademacker (1999) models is that they use dynamic models, i.e., they include a lag of the dependent variable as an independent variable. I did not do this for a simple reason. Models with lagged dependent variables essentially are models of differences, with the coefficients indicating a change in the first year of the study, and subsequent impacts being accounted for in the lag (the lag essentially acts as a “super control” variable, accounting for all causes of the dependent variable at t-1). This is not a particularly useful approach here because I am interested in levels rather than differences—in particular there is no across time variation of the culture measure. Culture is held to be a stable characteristic of states that change very slowly, and there is no measure that I know of that would pick up changes on a year-to-year basis. In other words, a lag is probably going to pick up culture’s historical effect, leaving its independent indicator to account for what happens when culture changes. As the culture measure only changes from state to state, rather than year to year, it is not likely to pick up much at all. In essence, I am more focused on the between state variation in an effort to pick up long-term relationships, while the earlier models were more focused on exploiting the within state variation to pick up short-term relationships. 6. There is also an argument that such tests are not only technically suspect, but in this case also irrelevant. The N of this study essentially represents the population of the states and as long as inference is confined to the time period studied, there is no need to infer from a sample to a population. Measures of variation (e.g., the standard error) still provide important information, but the t-test is arguably pointless. 7. The causal argument here virtually has to run from culture to education rather than vice versa. The subcultures Elazar identified not only existed long before things like SAT scores, but also in some states, even before identifiable mass systems of public education.
CHAPTER THREE: EDUCATION AND EQUALITY OF OPPORTUNITY
1. There is controversy over even what the SAT is supposed to measure. Virtually from its inception critics have argued that it is an exceedingly poor
Notes
153
measure of academic aptitude. By changing the name from scholastic aptitude test to scholastic assessment test in 1994, even the test’s manufacturer seems to have at least implicitly acknowledged the validity of some of these claims. The semantic fallback to “assessment”—despite a 50–year claim to be testing aptitude—does little to quell the surrounding controversies. If the SAT is assessing scholastic achievement, it is doing so in a highly imperfect manner. A number of companies make respectable profits by offering effective cram courses for the SAT. 2. There are two other governance models besides ISDs and charter schools in Texas. These are Common School Districts (CSDs) where county commissioners govern the school district, and Municipal School Districts (MSDs) where the city council governs schools. Only a handful of CSDs and MSDs exist and they are not included as independent variables in the analysis. 3. The Texas Excellence in Education Project (TEEP) is a joint effort by Texas A&M and the University of Texas Pan-American to undertake a sustained scholarly analysis of educational performance in the State of Texas. Numerous studies are conducted under its umbrella and made available (often with the accompanying data) on the TEEP web site: . A number have also been published in academic journals and books. 4. Thus if a school district’s student population was 50 percent white, 30 percent black, and 20 percent Latino, and the white pass rate was 80 percent, the black pass rate 60 percent, and the Latino pass rate 70 percent, Q’s numerator would be calculated as [(30*60) + (20*70)]/(30 + 20), or 64. Q would be a ratio of this number to the white pass rate—64/80 or .80.
CHAPTER FOUR: INSTITUTIONAL STRUCTURE AND EDUCATIONAL GOALS
1. Chubb and Moe’s (1988, 1990) argument, of course, extended far beyond an examination of how differing institutional frameworks affected organizational mission. Their primary, and best known, argument is that institutional framework helped determine outcomes, a topic covered in the previous chapters. 2. In this survey public school teachers and private school teachers were given slightly different response options than those listed in Table 4.1. All teachers were given eight possible response categories; the first seven of these categories were the same for both private and public schools. Public school teachers were not given an option of selecting instilling religious values as their primary religious educational goal (religious indoctrination is banned de jure in public schools). I collapsed promoting specific moral codes (public and private school teachers were given this option) into the same category as promoting religious beliefs to allow comparison across private and public schools on a single valuebased dimension. See methodological appendix for details.
154
Notes
3. Thus a coefficient of .44 indicates the odds of a Catholic school teacher picking basic literacy as a primary educational goal decrease by 66 percent compared to a public school teacher. A coefficient of 1.44 would indicate the odds increase by 44 percent increase. 4. Collinearity is one possible explanation for such results, but it does not seem to be a particular problem. Correlations between these three variables run between .4 and .55. The factors seem to be picking up genuinely different dimensions of the power/autonomy concept. 5. One example of nonsectarian schools promoting particular moral systems are the honor codes of military academies.
CHAPTER FIVE: EDUCATION AND CIVIC CULTURE
1. Jefferson introduced a Bill for the More General Diffusion of Knowledge into the Virginia House of Delegates in 1778. It laid down the basics of organizational structure and administration for a statewide network of schools, and went as far as to lay down curriculum mandates. It failed to pass the House of Delegates in 1778 and 1780. It passed the House in 1785, but was defeated in the Senate. In 1796 the legislature finally approved a watered down version of the original proposal. Critically, the version that became law allowed county courts to have the final say in the establishment of local schools. As this meant taxes on the local populace, the final say was without exception negative (Cremin 1980, 11). 2. This was only fitfully accomplished. In the nineteenth century, public schools essentially promoted an ecumenical Protestantism, something that evolved over the course of a century into a strict secularism. Catholic schools were established as separate educational systems largely out of concerns about defending the faith, and sometimes all too real fears of persecution (Cremin 1980, Macedo 1998). 3. Rice and Sumberg (1997) created the scale by taking a broad variety of indicators selected to represent four basic categories argued to be at the core of Putnam’s (1993) concept of civic culture: civic engagement; political equality; solidarity, trust and tolerance; social structures of cooperation. The scale itself was created using factor analysis to extract the common underlying dimension (i.e., civic culture) of the variables representing these categories. 4. In a regression analysis, the percent of population with a high school diploma is capable of explaining more than half of the variance in civic culture by itself. Adding a mixed and varied menu of controls to such a model—per capita income, residential mobility, ethnic composition of the population, crime rates, etc.—does little to detract from the strength of this relationship. 5. Using a basic difference of means test similar to that used in Table 5.1, just about all school specific cells in Table 5.2 show up as significantly different from the overall mean.
Notes
155
6. I experimented with numerous forms of these sorts of interactions, all aimed at trying to control for the “value-added” effect of being from the middle- to upper-classes and attending private school. Many of the resulting variables, however, created severe cases of multicollinearity in the final model. The variables described in the text achieved the basic aim of including something to account for the possibility of “value-added” without creating this problem. 7. Other versions of the regression model included variables related to curriculum (e.g., whether a class had stimulated interest in government), and at least one of these variables always showed up as significant. 8. The findings in Table 5.3, of course, are indicative of the average case, it is obviously of more than passing interest to attempt to find out what is particularly different about students who score at the upper end of the civic capital index. In previous chapters I employed a SWAT analysis to examine this sort of question. A SWAT analysis on the full model reported in Table 5.3, however, has little bearing on the substantive inferences made for the average case. The coefficients for chosen public schools and private secular schools do trend up, but never become significant. The coefficient for private religious-affiliated schools never wanders far from its original estimate. The most interesting findings from the SWAT analysis were for the interaction terms, the variables that combine class and private school. Both switched signs and indicated progressively negative effects, a trend that indicates private schools may act as a drag on the civic development of students in higher socioeconomic categories. Again, however, neither crossed the acceptable thresholds for reliable inference. Overall, and in contrast to the results presented in previous chapters, a SWAT analysis had little to add here and is thus not reported.
CHAPTER SIX: EDUCATION AS IDEOLOGY
1. Many of these advocates can also point out, quite correctly, that at least a significant minority (about 40 percent) of the public supports experimenting with market-based reforms such as vouchers. This can reasonably be interpreted as at least a tacit acknowledgment that public schools are performing considerably below expectations. As is so often the case with education, however, public attitudes turn out to be complex and resist easy interpretation. Not only does support for sending students to private schools at public expense in public opinion polls seem to be waning, this support also collapses if it would result in money being diverted from public schools. Instead, overwhelming majorities of the public—as high as 75 percent—strongly support the existing public school system and prefer to see it improved and strengthened rather than replaced with voucher or other market systems (see Rose and Gallup 2000, 49–50; Center on Policy Attitudes 2000). The argument that support for market-based reforms has, at best, fairly broad but shallow support is supported by the fate of statewide ballot initiatives on voucher issues, which have generally been decisively
156
Notes
defeated at the polls. In the 2000 election for example, proposals for voucher systems were overwhelmingly rejected in Michigan and California (about 70 percent of voters in both states were opposed). 2. More than three-quarters of private schools have religious affiliations (Koch 1999, 22). 3. Stock and salary information for both EAI/Tesseract and Edison was taken from Yahoo’s market guide: http://biz.yahoo.com/p/t/tsst.ob.html and http://biz/yahoo.com/p/e/edsn.html. Accessed 10/10/2000. 4. At least some of the differences of opinion regarding Edison’s effect on test scores stem from differences in measurement and methodology. In its reports Edison looks at whether test scores improve from year to year in its schools. Other analyses compare test score performance in Edison schools with performance in traditional public schools with similar student bodies. This can lead to different inferences on the overall impact of Edison on academic performance (Wyatt 2001). 5. It can be justifiably argued that Catholic schools, the largest religiously affiliated school system in the nation, do fit the commonwealth conception of “public.” The Catholic K–12 education system openly embraces liberal democratic ideals. However, in historical terms this is a relatively recent development, and is a least partially a product of commonwealth pressures created by the public education system (Macedo 1998). 6. From a purely academic perspective, future education production function analysis that pays particular attention to methodology, measurement, and specification issues remains fruitful and important work. The argument here is that it is unlikely to be decisive in terms of “making” or “proving” the market case.
Methodological Appendix All data used in the analyses for this book are publically available, or can be obtained from the author. I used three basic statistics packages to perform the analyses: Stata, SPSS, and NCSS. All OLS models were estimated using NCSS and/or SPSS; all pooled and GLS models were estimated using Stata. Technical details and data sources are presented by table, which hopefully should make it easier for the statistically inclined to judge the validity of the results without an undue departure from the main text.
TABLE 2.1
Table 2.1 reports the results of a pooled time series analysis predicting state-level productivity measured as GSP per worker. Although a powerful technique that combines the advantages of both cross-sectional and time series analysis, pooled models also import and compound the problems of both approaches (Stimson 1985, Sayrs 1989). Such models often generate complicated error structures that violate the basic assumptions of ordinary least squares (OLS). A number of substitute estimators have been suggested to deal with the effects of these violations. In this particular case, a series of diagnostics indicated a problem with unit effects— in essence there were state level impacts not being accounted for in the model. One of the most common approaches to dealing with these problems are to generate estimators using generalized least squares (GLS), and the GLS estimators are reported in the table. An alternate estimation technique is a fixed effects (FE), or least squares with dummy variables (LSDV) approach, which addresses the unit effects problem without the assumption of randomness required of GLS. There were two drawbacks to employing FE in this case: (1) High degrees of multicollinearity and the unit dummies (measures of “specific ignorance”) accounting for a good
157
158
Methodological Appendix
deal of variance. (2) FE cannot estimate time invariant variables, which are zeroed out of the model. This drops four variables (percent with high school diploma, equity in educational spending, unionization and political competition) from the model. These were theoretically important to retain. Accordingly, I opted to report the GLS coefficients, though the coefficients from both OLS and FE estimation techniques are included in the range of values indicated by the Leamer bounds in Table 2.1. The data were stacked traditionally (see Sayrs 1989), and the time period and states used in the analysis were largely dictated by data availability: Sharkansky’s (1969) culture scale does not include Alaska and Hawaii; no score is available for Louisiana in the political competition measure; Washington does not report aggregated test score results; and South Dakota was dropped because of missing data in some years. Statelevel ACT scores—necessary to calculate a test score index that avoids the selection bias problem—were not available for all years after 1990. This left data on 45 states for nine years. Sources of Data Productivity: GSP per worker in constant (1992) dollars. U.S. Department of Commerce, Bureau of Economic Analysis, Regional Economic Information System (REIS) CD-ROM. Bureau of the Census. Various Years. Statistical Abstracts of the United States. Washington, D.C., U.S. Government Printing Office. Policy Activism: Total number of tax incentives and other support packages listed in Tables 1, 2, and 3 of the Site Selection Handbook. Industrial Developments Site Selection Handbook. Annual political issue. Atlanta, GA: Conway Publications. National Productivity: GSP per worker in constant (1996) dollars. REIS CD-ROM, and Statistical Abstracts of the United States, various years. Unionization: A dummy variable where 1 = 15 percent or more of workers are labor union members, and 0 = less than 15 percent are members. Calculated from data reported in the Statistical Abstracts, various years. Political Competition: Index of state-level political competition calculated by Holbrook and Van Dunk (1993). Unified Control: A dummy variable where 1 = unified partisan control of legislature and governer’s office, 0 = divided partisan control. Calculated from data reported in the Statistical Abstracts, various years. Size of the Public Sector: Total state tax collections measured as a percentage of GSP. REIS CD-ROM. Bureau of the Census, Various Years.
Methodological Appendix
159
Federal Expenditures by State. Washington, D.C.: U.S. Government Printing Office. Defense Expenditures: Total Defense Department expenditures within a state in constant (1990 Texas) dollars. Office of Management and Budget. Annual. Federal Expenditures by State. Washington, D.C.: U.S. Government Printing Office. Percent of GSP Devoted to Education: Total state and local government expenditure on education as a percent of total GSP. GSP data taken from the REIS CD-ROM and the Statistical Abstracts of the United States, various years. Test Scores: This is either an SAT score or an estimated SAT score based on a state’s ACT score. This measure is produced using the so-called Wainer Transformation to merge SAT and ACT scores onto a common metric that avoids the selection bias inherent in each individual measure (see Smith and Meier 1995, Lehnen 1992). SAT data taken from various years of the Digest of Education Statistics. ACT data taken from Lehnen (1992). ACT scores were not available for every year—in missing years the ACT was estimated by averaging the two nearest data points. Equity in Educational Resources: A 0–100 index with 100 indicating complete equity (i.e., all districts spend exactly the same per pupil). From Quality Counts: A Report Card on the Condition of Public Education in the 50 States. 1997. Washington, D.C.: Education Week/Pew Charitable Trusts. Percent of Population with HS Diploma. Percent of population with at least a high school diploma. Figures taken from 1990 decennial census data and reported in the Statistical Abstracts of the United States. Per Pupil Spending: Spending per pupil in constant (1992) dollars. State Comparisons of Education Statistics: 1969–70 to 1993–94. Political Culture: Sharkansky’s (1969) political culture index, folded as described in text. Per Capita Income: Per capita income in constant (1996) dollars. REIS CD-ROM and Statistical Abstracts of the United States, various years.
TABLE 2.2
Table 2.2 reports the impact of the education and culture variables in a statistically weighted analytical technique (SWAT) analysis of the model
160
Methodological Appendix
reported in Table 2.1. The basic philosophy of SWAT is to use the residuals of an OLS regression to extract information about what separates high or low performing cases from the average case (for a complete exposition and examples see Meier and Gill 2000, Meier and Keiser 1996). To get the results reported in Table 2.2, the regression model in Table 2.1 was run using OLS. All observations with a studentized residual greater than .7 in this analysis (the basic rule of thumb for identifying high performers in SWAT analysis) were classified as high productivity states. These observations were assigned a weight of 1 and all other observations were assigned a weight of .9, and the OLS analysis rerun. The ratio of the original OLS estimate to this new weighted estimate is reported in the first row of Table 2.2. The analysis was iteratively rerun with the weight of the high productivity states set at 1, and the weights of all other states reduced in .10 increments (so the final weight was 1 for high productivity states and .10 for all others). In this fashion, higher performers (those assigned a weight of 1) are gradually allowed to have more weight in the analysis, and the resulting change in coefficient magnitude is used to assess what separates these from the average case typified by the standard OLS estimates. Sources of data for this analysis are the same as those reported in Table 2.1.
FIGURE 2.1
This is a scatterplot of the observed productivity values to the values estimated by the OLS regression used to generate the results in Table 2.1. The line representing the average case is a regression line through origin (slope coefficient = 1.0).
TABLES 3.1 AND 3.2
Tables 3.1 and 3.2 are OLS regression estimates and report the same model except for the outcome variable. In attempting to specify a typical production function model, I draw heavily from the arguments of Smith and Rademacker (2001). The outcome variable in Table 3.1 is the percentage of students passing the TAAS exams. The outcome variable in Table 3.2 is the equity index Q, multiplied by 100 (details are discussed in the text; see also Meier et al. 1997). The unit of analysis is Local Education Agency (LEA) for the 1999–2000 academic year. All data were downloaded from the Texas Education Agency (TEA) at: . There
Methodological Appendix
161
were two problems with these data. First, there was a good deal of missing data for charter schools, especially TAAS scores by ethnic grouping. It was possible to estimate some of the missing data. To avoid identifying students the TEA does not report TAAS pass rates when at least one, but less than five students pass the test. In a detailed report on charter schools, the TEA indicates these circumstances with “LT5” (less than 5). Where possible, I estimated missing TAAS data by assuming five students passed the test and calculated the resulting percentages on this basis. This probably over-estimates pass rates in some charter schools, but given the controversies surrounding market-based institutional reform that Chapter 3 addresses, I deliberately wanted to put charter schools in the best possible light. The second problem with these data were extreme outliers in TAAS pass rates. This was a particular concern because outliers at the low end of the scale were dominated by charter schools (mostly those dealing with at risk students), and where the number of students in particular ethnic group were small, the resulting TAAS pass rates pushed equity scores to extremes in a handful of cases. These outliers severely affect the estimates generated by the models reported in Tables 3.1 and 3.2, especially on the charter school variable (the findings for the bureaucracy and expenditure variables were fairly robust across all diagnostic versions of the model, but beginning with all available data points it was possible to change the direction of the charter school coefficient by dropping or adding as few as three observations). Various attempts to control for these effects through alternate model specifications were unsuccessful (these impacts are reported in Leamer bounds). I also attempted to deal with the problem through various data transformations, but with similar results. Dummying out the influential observations (defined as having a studentized residual of greater than 3) required adding as many as 25 variables to the regression model. I considered using a robust regression technique, but the statistical properties of combining this approach with SWAT are not particularly clear (this was also a consideration in deciding not to dummy out the influential observations). Given this, I handled the problem of influential outliers by dropping all cases where fewer than 30 percent of students passed the TAAS exams for the model reported in Table 3.1, and included only cases with an equity score of 50 to 150 in Table 3.2. This eliminated 41 observations from the analysis in Table 3.1 (all charter schools with low TAAS pass rates) and 28 observations from the analysis reported in Table 3.2 (a majority are charter schools with low equity scores). The total N of LEA was 1,183, so the dropped cases represented, at most,
162
Methodological Appendix
about 3.4 percent of data set. Other cases were excluded from the analysis because of missing data on other variables (mostly unreported overall TAAS scores and unreported TAAS scores for particular ethnic groups). By dropping the observations I managed to reduce the problem of outliers biasing regression line, mitigate the severe instability in the charter school coefficient, retain the maximum number of charter schools given the problems of outliers and missing data (74 of the 142 charters are included in the analysis reported in Table 3.1 and 40 in Table 3.2), while also putting charter schools in the best possible light (most low performing charters were eliminated from the final analysis). Even so, given both the reported results and the diagnostic findings that preceded them it is hard to vest much confidence in any inference about charter schools except for recognizing the high probability that they have no systematic relationship to TAAS pass rates or the Q index. Accordingly, this is the conclusion about charter schools that I stress in the text. Variable measures are as follows: Percent Economically Disadvantaged: Percent of students eligible for reduced price or free lunch programs. Percent White Students: Percent of the total student body that is white. Per Pupil Instructional Expenditures: Budgeted instructional expenditures divided by total students. Students Per Teacher: The total number of students divided by the total teacher FTE count. Percent White Teachers: The FTE count of teachers reported as white, expressed as a percent of the total teacher FTE count. Teacher Experience: A weighted average obtained by multiplying each teacher’s FTE count by his or her years of experience, summing for all weighted counts, and then dividing by total teacher FTEs. Bureaucracy: The FTE count of personnel identified as working in central administration divided by the total FTE count. Charter School: A dummy variable with 1 indicating a charter school. Percent of all Students Passing Test: The total number of students who passed all the TAAS tests they attempted expressed as a percentage of the total number of students who took one or more tests. The performance of students tested in grades 3–8 and 10 in reading and mathematics, and grades 4, 8, and 10 in writing are included. TAAS science and social studies results are not included.
Methodological Appendix
163
Equity: The Q index calculated from TAAS pass rates, as explained in the text.
TABLES 3.3 AND 3.4
These tables report the impact of the charter school, instructional expenditure, and bureaucracy in a SWAT analysis of the models reported in Tables 3.1 and 3.2. The methods used to generate these estimators are the same as those described under Table 2.2.
FIGURES 3.1 AND 3.2
These are scatterplots with observed data on the Y axis and the estimated values from the models reported in Tables 3.1 and 3.2 on the X axis. The line representing the average case is a regression through the origin (the slope coefficient is 1.0).
TABLES 4.1, 4.2, AND 4.3
All data used in the analyses reported in these tables were taken from the second Teacher Followup Survey (National Center for Education Statistics 1998). As part of this survey, both private and public school teachers were asked to select their most important educational goals from an eight-response survey item. The response categories were identical except for the final option, which was “promoting multicultural awareness” for public school teachers and “fostering religious/spiritual development” for private school teachers. The latter goal is essentially barred de jure for public school teachers. To merge these two survey items into a single variable for analysis, I collapsed “promoting specific moral values” and “fostering religious/spiritual development” into a single category, reasoning they were clearly related items; and the former makes the most likely (and more legally acceptable) substitute for the latter in the public sector. I also collapsed “promoting multicultural awareness” into “promoting human relations skills” for public school teachers (the former was chosen by only 1 percent or so of the public school teachers surveyed). The result was the seven-category list of goals listed in Table 4.1 and used as a dependent variable for the regression analyses that generated the estimates reported in Tables 4.2 and 4.3.
TABLE A.1 Estimated Institutional Impact on School Goals
Basic Literacy
Academic Excellence
Vocational/ Technical Skills
Good work habits/ self-discipline
Personal growth
Human relations skills
Moral/ religious values
Catholic
–.80 (.08)*
.06 (.08)
–.98 (.17)*
–.67 (.08)*
–.21 (.08)*
–.68 (.11)*
2.67 (.09)*
Conservative Christian
–1.04 (.15)*
.14 (.15)
–.65 (.29)*
–.58 (.15)*
–1.6 (.16)*
–1.34 (.28)*
4.14 (.20)*
Other Religions
–.80 (.10)*
.06 (.10)
–1.15 (.22)*
–.55 (.10)*
–.83 (.10)*
.78* (.14)*
3.12 (.12)*
Non-sectarian
–.82 (.13)*
.50 (.13)*
–.95 (.25)*
–.28 (.13)*
.65 (.15)*
.04 (.15)
.62 (.18)*
Special Education
–.73 (.29)*
–.14 (.30)
.57 (.33)
–.44 (.28)
.85 (.37)*
.20 (.30)
–.17 (.52)
Control over educational content
–.15 (.04)*
–.02 (.03)
.30 (.05)*
–.02 (.03)
–.006 (.03)
.09 (.04)*
–.09 (.05)
Variable
(continued on next page)
TABLE A.1 (continued)
Basic Literacy
Academic Excellence
Vocational/ Technical Skills
Good work habits/ self-discipline
Personal growth
Human relations skills
Moral/ religious values
Control over pedagogy
.01 (.03)
.02 (.03)
–.15 (.04)*
.03 (.03)
.01 (.03)
–.01 (.03)
.04 (.04)
Control over school policies
–.07 (.03)*
.01 (.03)
–.08 (.05)
–.06 (.03)*
.03 (.03)
.04 (.04)
.15 (.04)*
Parental/student support for school
–.02 (.05)
–.09 (.04)*
–.21 (.06)*
–.06 (.04)
.22 (.04)*
.09 (.05)
–.03 (.06)
Student violence
.02 (.04)
.09 (.04)*
–.04 (.05)
–.06 (.04)
.02 (.03)
.008 (.04)
–.15 (.05)*
Student drug/alcohol abuse
.18 (.04)*
–.15 (.04)
–.23 (.05)*
.11 (.04)*
.06 (.04)
–.02 (.04)
.02 (.05)
Race and socioeconomic problems
–.009 (.04)
.10 (.04)*
.11 (.05)*
.08 (.04)*
–.11 (.04)*
–.19 (.04)*
.04 (.05)
Sex
–.31 (.06)*
.23 (.05)*
.23 (.07)*
.006 (.06)
–.35 (.06)*
.12 (.06)*
.33 (.08)*
Variable
(continued on next page)
TABLE A.1 (continued)
Basic Literacy
Academic Excellence
Vocational/ Technical Skills
Good work habits/ self-discipline
Personal growth
Human relations skills
Moral/ religious values
Age
–.03 (.02)
.13 (.02)*
.08 (.03)*
.07 (.02)
–.17 (.02)*
–.09 (.02)*
.05 (.03)
Race
.17 (.10)
–.4 (.09)*
–.18 (.12)
.57 (.09)*
.12 (.09)
–.25 (.10)
–.23 (.13)
N
6,562
6,562
6,562
6,562
6,562
6,562
6,562
–2 Log likelihood
7,305
8,638
4,928
8,200
8.380
6,692
4,951
.04
.02
.06
.02
.06
.03
.25
Variable
Pseudo R-2 (Cox and Snell)
Estimators generated by logistic regression (dependent variable is dichotomous). Unstandardized coefficients (standard errors) reported. All models were estimated with a constant (unreported). *p < .05.
Methodological Appendix
167
The indexes created by factor analysis for the regression models were generated by first doing an exploratory factor analysis to examine the underlying structure in response to items relating to student/parent attitudes and characteristics. On the basis of these results, separate sets of questions were factor analyzed to actually generate the indexes. This approach resulted in the maximum extraction of variance from the variables used. All of the factor analysis-based indexes in the analysis represent well over 50 percent of the variance in the underlying response items and are the only factors produced in each analysis with an eigenvalue greater than 1. For replication purposes, the survey code numbers of the questions in each factor analysis are as follows: Control over educational content: TSC 248, TSC 249 (Eigenvalue 1.6, 78.8 percent of variance). Control over pedagogy: TSC 250, TSC 251, TSC 252, TSC 253 (Eigenvalue 2.41, 60.2 percent of variance). Control over school policies: TSC 244, TSC 245, TSC 246, TSC 247, TSC 248 (Eigenvalue 2.5, 50.2 percent of variance). Student/Parent attitudes towards school: TSC 254, TSC 255, TSC 257, TSC 268, TSC 270, TSC 271, TSC 269 (Eigenvalue 3.9, 55.7 percent of variance). Student violence: TSC 261, TSC 262, TSC 263 (Eigenvalue 2.5, 84 percent of variance). Student drug/alcohol abuse: TSC 261, TSC 262, TSC 263 (Eigenvalue 2.5, 84 percent of variance). Race and socioeconomic problems: TSC 272, TSC 273, TSC 274, TSC 275 (Eigenvalue 2.65, 65.8 percent of variance). The full logistic regression models referred to in the text are reported in Table A.1. I have reported psuedo R-2s to provide a readily accessible idea of how poorly these models performed (with the exception of the model predicting moral codes/religious values). Proportionate reduction in error (PRE) statistics emphasize the point—the only models that indicated a substantial reduction in error over just picking the model category of the dependent variable were the moral codes/religious values model, with a PRE of .30, and the basic literacy model with a PRE of .64. Other PREs ranged from zero (vocational–technical skills), i.e., no reduction in error, to .005 (academic excellence).
TABLES 5.1, 5.2, AND 5.3
All data for the analyses in Tables 5.1, 5.2 and 5.3 come from the Civic Youth Involvement (CYI) survey (National Center for Education Statistics
168
Methodological Appendix TABLE A.2 Civic Capital Correlations Variable
Pearson Correlation
Tolerance
.44
Efficacy
.69
Participation
.60
Political Knowledge (scale 2)
.75
1996), and all the independent variables in Table 5.2 were taken directly from survey questions and coded as reported in the text. The construction of the tolerance, efficacy, participation, and political knowledge scales, and code numbers for the specific questions are as follows: Political Tolerance: A 0–2 scale constructed from questions YD5C and YD5D. A 0 indicates respondent believed some books should be kept out of public libraries, and that people should be prevented from speaking out against religion. A 2 indicates respondent opposed keeping certain books out of libraries and was in favor of allowing people to speak out against religion. Efficacy: A 0–4 scale constructed from questions YD5A, YD5B, YD6, and YD7. A 0 on this scale indicated a respondent felt they could not understand politics, believed their family had no say in what the government does, did not feel like they could write a letter to a government official, and did not feel like they could make a statement at a public meeting. A 4 indicates a respondent felt like they understood politics, that their family did have a say in what government does, and that they were capable of writing a letter to a government official and speaking at a public meeting. Participation: A 0–4 scale constructed from questions YB2, YB3, YB4, and YC1. A 0 on this scale indicates a respondent did not participate in student government, school activities, out of school activities, or community service. A 4 indicates a respondent participated in all of these. Political Knowledge: Two 0–5 scales constructed from YD8A, YD8B,YD8C, YD8D, and YD8E (first scale), and YD9A, YD9B, YD9C, YD9D, and YD9E (second scale). The questions were basic tests of polit-
Methodological Appendix
169
ical knowledge (e.g., Who nominates federal court judges? What is the job held by Al Gore?) that were coded into correct/incorrect (coded 1 or 0) and summed. The second political scale was used in the dependent variable for the model reported in Table 5.3. Using the first scale had no substantive impact on the results. Civic Capital: The civic capital index used as the dependent variable for the model reported in Table 5.3 was created from a factor analysis of the tolerance, efficacy, participation, and political knowledge (second scale) variables described above. A single factor with an Eigenvalue of greater than 1 was created from this analysis (actual Eigenvalue = 1.59). It extracted 40 percent of the variance from the four variables. The bivariate correlations (all significant at the p < .01 level using a two-tailed ttest) between the original four variables are reported in Table A.2. The central concern with this analysis was the reduction in N—there were a little less than 8,000 respondents in the entire survey, but because only subsamples were asked the political knowledge question, this reduced the N to 1,762 for the reported regression analysis. A second regression analysis using the first political knowledge scale was also performed, with no substantive difference in results. Given that the two analyses created similar results and between them included close to 50 percent of the sample, they seem a reasonable basis for inference.
FIGURE 5.1
The data used to create the graph in Figure 5.1 come from Rice and Sumberg’s (1997) index of civic culture in the states, and the Statistical Abstracts of the United States.
This page intentionally left blank.
References Alexander, F. King. 1998a. “Voucher in American Education: Hard Legal and Policy Lessons from Higher Education.” Journal of Educational Finance. 24: 153–178. Alexander, F. King. 1998b. “Private Institutions and Public Dollars: An Analysis of the Effects of Federal Direct Student Aid on Public and Private Institutions of Higher Education.” Journal of Education Finance. 23: 390–416. Alexander, Kern. 1993. “Money Matters: Commentary and Analyses.” Journal of Education Finance. 24: 237–242. Allen, Robert E. 1996. “The Anxiety Epidemic.” Newsweek. April 8, p. 15. Almond, Gabriel and Sidney Verba. 1963. The Civic Culture: Political Attitudes and Democracy in Five Nations. Princeton: Princeton University Press. Almond, Gabriel and Sidney Verba. 1980. The Civic Culture Revisited. Boston: Little, Brown. American Federation of Teachers. 2000. “Trends in Student in Student Achievement for Edison Schools, Inc.: An Emerging Track Record.” Http://www. aft.org/research/edisonschools/Index.html Accessed Jan. 14, 2002. Apple, Michael. 1986. “National Reports and the Construction of Inequality.” British Journal of Sociology of Education. 7(2): 171–190. Arnstine, Donald. 1995. Democracy and the Arts of Schooling. New York: State University Press of New York. Ascher, Carol, Norm Fruchter, and Robert Berne. 1996. Hard Lessons: Public School and Privatization. New York: Twentieth Century Fund. Barber, Benjamin. 1992. An Aristocracy of Everyone. New York: Oxford University Press. Becker, Gary S. 1993. Human Capital, 3rd ed. Chicago: University of Chicago Press.
171
172
References
Bennett, William. 1992. The De-Valuing of America. New York: Touchstone Books. Bennett, William. 1994. The Index of Leading Cultural Indicators: Facts and Figures on the State of American Society. New York: Simon and Schuster. Berliner, David C. and Bruce J. Biddle. 1995. The Manufactured Crisis: Myths, Fraud, and the Attack on America’s Public Schools. New York: AddisonWesley. Bishop, John H. 1989. “Is the Test Score Decline Responsible for the Productivity Growth Decline?” American Economic Review. 79: 178–197. Blais, Andre and Robert Young. 1999. “Why Do People Vote? An Experiment in Rationality.” Public Choice. 99: 33–35. Blaug, Mark. 1976. “The Empirical Status of Human Capital Theory: A Slightly Jaundiced View.” Journal of Economic Literature. 14: 827–855. Bobo, Lawrence and Frederick C. Licari. 1989. “Education and Political Tolerance: Testing the Effects of Cognitive Sophistication and Target Group Affect.” Public Opinion Quarterly. 53: 285–308. Boix, Charles and Daniel Posner. 1998. “Social Capital: Explaining its Origins and Effects on Government Performance.” British Journal of Political Science. 28: 686–692. Borjas, George. 1995. “Ethnicity, Neighborhoods, and Human-Capital Externalities.” The American Economic Review. 85: 365–390. Bound, John and Gary Solon. 1999. “Double Trouble: On the Value of Twinsbased Estimation of the Return to Schooling.” Economics of Education Review. 18: 169–182. Bourdieu, Pierre. 1973. “Cultural Reproduction and Social Reproduction.” In Knowledge, Education and Cultural Change, Richard Brown, ed. London: Tavistock. Bowe, Richard and Stephen J. Ball (with Anne Gold). 1992. Reforming Education & Changing Schools. London: Routledge. Bowles, Samuel and Herbert Gintis. 1975. “The Problem with Human Capital Theory—A Marxian Critique.” The American Economic Review. 65: 74–82. Box, Richard C., Gary S. Marshall, B.J. Reed, and Christine M. Reed. 2001. “New Public Management and Substantive Democracy.” Public Administration Review. 61: 608–619. Brace, Paul and Robert Dudley. 1991. “Military Expenditures and State Economic Growth.” In Public Policy and Economic Institutions. Mel Dubnick and Alan Gitelson, eds. Greenwood, CT: JAI Press.
References
173
Brace, Paul. 1993. State Government and Economic Performance. Baltimore: The Johns Hopkins University Press. Brace, Paul and Aubrey Jewett. 1995. “The State of State Politics Research.” Political Research Quarterly. 48: 643–682. Bracey, Gerald. 1997. Setting the Record Straight. Washington, D.C.: Association for Supervision and Curriculum Development. Brehm, John and Wendy Rahn. 1997. “Individual-Level Evidence for the Causes and Consequences of Social Capital.” American Journal of Political Science. 41: 999–1023. Brighouse, Harry. 1994. “The Egalitarian Virtues of Educational Vouchers.” Journal of Philosophy of Education. 28: 211–219. Brogan, Dennis. 1962. The American Character. New York: Time, Inc. Brown, Byron. 1992. “Why Governments Run Schools.” Economics of Education Review. 11: 287–300. Buchanan, James M. 1972. “Towards Analysis of Closed Behavioral Systems.” In Theory of Public Choice: Political Applications of Economics. Ann Arbor: University of Michigan Press. Buchanan, James M. and Gordon Tullock. 1962. The Calculus of Consent. Ann Arbor: University of Michigan Press. Bundt, Julie L. 1997. “School Finance Policy: Ameliorating or Aggravating Inequalities.” Paper presented at the 1997 annual meeting of the Midwest Political Science Association. Burtless, Gary, ed. 1996. Does Money Matter?: The Effect of School Resources on Student Achievement and Adult Success. Washington, D.C.: The Brookings Institution Press. Callan, Eamon. 1997. Creating Citizens: Political Education and Liberal Democracy. New York: Oxford University Press. Campbell, Angus, Philip E. Converse, Warren E. Miller, and Donald E. Stokes. 1964. The American Voter. New York: John Wiley & Sons. Card, David and Alan B. Krueger. 1992. “Does School Quality Matter? Returns to Education and the Characteristics of Public Schools in the United States.” Journal of Political Economy. 100: 1–40. Carson, C.C., R.M. Huelskamp, and T.D. Woodall. 1993. “Perspectives on Education in America: An Annotated Briefing.” The Journal of Educational Research. 86: 259–310. Carter, John R. and Michael D. Irons. 1991. “Are Economists Different, And If So, Why?” Journal of Economic Perspectives. 5: 171–177.
174
References
Center on National Education Policy. 1996. “Do We Still Need Public Schools?” Policy Paper. Washington, D.C.: Phi Delta Kappa. Center on Policy Attitudes. 2000. “Public Wants Federal Government to Play Active, But Not Punitive, Role In Strengthening Public Schools.” Http: //www.policyattitudes.org/OnlineReports/Education/edu_brief.html. Accessed Jan. 14, 2002. Chambers, M.M. 1968. Higher Education: Who Pays? Who Gains? Danville, IL: The Interstate Printers. Chubb, John. 1997. “Lessons in School Reform from the Edison Project.” In New Schools for a New Century. Diane Ravitch and Joseph Viteritti, eds. New Haven: Yale University Press. Chubb, John and Terry Moe. 1988. “Politics, Markets and the Organization of Schools.” American Political Science Review. 82: 1065–1087. Chubb, John and Terry Moe. 1990. Politics, Markets and America’s Schools. Washington, D.C.: Brookings Institution. Chubb, John and Terry Moe. 1997. “Politics, Markets and Equality in Schools.” In Autonomy and Choice in Context: An International Perspective. Rina Shapira and Peter Cookson, Jr., eds. Oxford: Pergamon. Clark, Burton R. 1962. Educating the Expert Society. San Francisco: Chandler. Clark, S.D. 1951. “Religion and Economic Backward Areas.” The American Economic Review. 41: 258–265. Coleman, James et al. 1966. Equality of Educational Opportunity. Washington, D.C.: U.S. Government Printing Office. Coleman, James. 1974. “The Meaning of Equal Educational Opportunity.” In Equality of Educational Opportunity, LaMar P. Miller and Edmund W. Gordon, eds. New York: AMS Press, Inc. Coleman, James and Thomas Hoffer. 1987. Public and Private High Schools. New York: Basic Books. Collins, Randall. 1971. “Functional and Conflict Theories of Educational Stratification.” American Sociological Review. 36: 1002–1019. Cook, Philip J. and Jens Ludwig. 1998. “The Burden of ‘Acting White’: Do Black Adolescents Disparage Economic Achievement?” In The Black-White Test Score Gap, Christopher Jencks and Meridith Phillips, eds. Washington, D.C.: The Brookings Institute. Cookson, Peter, 1994. School Choice. New Haven: Yale University Press. Cox, Harvey. 1999. “The Market as God.” The Atlantic Monthly. 283 (3): 18–23.
References
175
Cremin, Lawrence A. and Merle Borrowman. 1956. Public Schools in Our Democracy. New York: Macmillan. Cremin, Lawrence A. 1980. American Education, the National Experiences. New York: Harper and Row. Cubberly, Elwood. 1905. School Funds and Their Apportionment. New York: Columbia University Teachers College. Darity, William A. and Rhonda M. Williams. 1985. “Peddlers Forever?: Culture, Competition, and Discrimination.” The American Economic Review. 75: 256–261. David, Martin et al. 1961. Educational Attainment, Its Causes and Effects. Ann Arbor, MI: Survey Research Center, University of Michigan. DeLeon, Peter. 1997. Democracy and the Policy Sciences. Albany, NY: SUNY Press. De Long, J. Bradford. 1988. “Productivity Growth, Convergence, and Welfare: Comment.” American Economic Review. 78: 1138–1154. Denison, Edward. 1962. The Sources of Economic Growth and the Alternatives Before Us. Supplementary Paper No. 13. New York: Committee for Economic Development. De Tocqueville, Alexis. 1945. Democracy in America. New York: Alfred Knopf, Inc. Dewey, John. 1929. My Pedagogic Creed. Washington: Progressive Education Association. DeYoung, Alan J. 1989. Economics and American Education. New York: Longman. Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper. Dye, Thomas. 1980. “Taxing, Spending, and Economic Growth in the American States.” The Journal of Politics. 42: 1085–1107. Dynarski, Mark and Philip Gleason. 1993. “Using Scholastic Aptitude Test Scores as Indicators of State Educational Performance.” Economics of Education Review. 12: 203–211. Eccles, Jeremy. 1999. “Testing the Tests: The Limits of Contemporary Education Policy.” Ph.D. dissertation. University of Nebraska–Lincoln. Elazar, Daniel. 1977. American Federalism: A View From the States. 2nd ed. New York: Cromwell. Elazar, Daniel. 1999. “Tocqueville and the Cultural Basis of American Democracy.” PS: Political Science & Politics. 32: 207–210.
176
References
Engel, Michael. 2000. The Struggle for Control of Public Education: Market Ideology versus Democratic Values. Philadelphia, PA: Temple University Press. Finn, Chester, 1996. “Corrupted Intentions: Reforming Special Education.” National Review. 48: 46–48. Flathman, Richard E. 1996. “Liberal Versus Civic, Republican, Democratic, and Other Vocational Educations.” Political Theory. 24: 4–32. Fogel, Robert W. 1994. “Economic Growth, Population Theory, and Physiology: The Bearing of Long-Term Processes on the Making of Economic Policy.” The American Economic Review. 84: 369–382. Fosler, R. Scott. 1991. “Human Capital and Federalism.” In Human Capital and America’s Future. David Hornbeck and Lester Salamon, eds. Baltimore: The Johns Hopkins University Press. Frank, Robert H., Thomas Gilovich, and Dennis T. Regan. 1993. “Does Studying Economics Inhibit Cooperation.” Journal of Economic Perspectives. 7: 159–171. Frank, Robert H., Thomas Gilovich, and Dennis T. Regan. 1996. “Do Economists Make Bad Citizens?” Journal of Economic Perspectives. 10: 187–192. Friedenberg, Edgar Z. 1974. “Social Consequences of Educational Measurement.” In Equality of Educational Opportunity. LaMar P. Miller and Edmund W. Gordon, eds. New York: AMS Press, Inc. Friedman, Milton. 1982. Capitalism and Freedom. Chicago: University of Chicago Press. Friedman, Milton. 1955. “The Role of Government in Education.” In Economics and the Public Interest. Robert A. Solo, ed. New Brunswick, NJ: Rutgers University Press. Fukuyama, Francis. 1995. Trust: The Social Virtues and the Creation of Prosperity. New York: The Free Press. Gerstner, Louis V. 1995. Reinventing Education. New York: Plume Giroux, Henry A. 1998. “Education Incorporated?” Educational Leadership. 56 (October): 12–17. Gill, Jeff. 1999. “The Insignificance of Null Hypothesis Significance Testing.” Political Research Quarterly. 52: 647–674. Glenn, David. 2001. “The Voucher Vortex.” Lingua Franca. 11(4) 32–41. Goodlad, John. 1984. A Place Called School. New York: McGraw-Hill. Goodsell, Charles. 1994. The Case for Bureaucracy. Chatham, NJ: Chatham House.
References
177
Gould, Stephen Jay. 1996. The Mismeasure of Man. New York: W.W. Norton. Graham, Amy E. and Thomas A. Husted. 1993. “Understanding State Variations in SAT Scores.” Economics of Education Review. 12: 197–202. Graham, John. 1981. “An Explanation for the Correlation of Stocks of Nonhuman Capital with Investment in Human Capital.” The American Economic Review. 71: 248–255. Granato, Jim, Ronald Inglehart, and David Leblang. 1996. “The Effect of Cultural Values on Economic Development: Theory Hypotheses, and Some Empirical Tests.” American Journal of Political Science. 40: 607–631. Greene, Jay, Paul E. Peterson, and Jiangtao Du. 1996. “The Effectiveness of School Choice in Milwaukee: A Secondary Analysis of Data from the Program’s Evaluation.” Paper presented at the annual meeting of the American Political Science Association, San Francisco, September 1996. Greene, Jay and Paul E. Peterson. 2000. “Should Public Discussion of Political Science Research Be Controlled? Why Interest Group Recommendations on the Proper Procedures for Reporting Research Should Be Treated With Skepticism.” PS: Political Science & Politics. 33: 221–223. Greene, Jay P., Joseph Giammo, and Nicole Mellow. 1999. “The Effect of Private Education on Political Participation, Social Capital and Tolerance: An Examination of the Latino National Political Survey.” The Georgetown Public Policy Review. 5: 53–71. Gujarati, Damodar N. 1995. Basic Econometrics. New York: McGraw-Hill. Guttman, Amy. 1987. Democratic Education. Princeton, NJ: Princeton University Press. Halchin, L. Elaine. 1998. “. . . And This Parent Went to Market.” Paper presented at the American Political Science Association annual meeting, Boston. Hanushek, Eric. 1981. “Throwing Money At Schools.” Journal of Policy Analysis and Management. 1: 19–41. Hanushek, Eric. 1986. “The Economics of Schooling: Production and Efficiency in Public Schools.” Journal of Economic Literature. 24: 1141–1177. Hanushek, Eric and Lori L. Taylor. 1990. “Alternative Assessments of the Performance of Schools: Measurement of State Variations in Achievement.” Journal of Human Resources. 25: 179–201. Hanushek, Eric et al. 1994. Making Schools Work: Improving Performance and Controlling Costs. Washington, D.C.: Brookings Institute. Hanushek, Eric A. 1997. “Assessing the Effects of School Resources on Student Performance: An Update.” Educational Evaluation and Policy Analysis. 19: 141–164.
178
References
Harris, D. 1937. “Factors Affecting College grades: A Review of the Literature: 1930–1937.” Psychological Bulletin. 37: 125–151. Hartoonian, Michael. 1999. “Education For Sale: What’s a (Democratic) Principle Worth?” Social Education. 63: 241–243. Hedges, L.V., R.D. Laine, and R. Greenwald. 1994. “Does Money Matter? A Meta-Analysis of Studies of the Effects of Differential School Inputs on Student Outcomes.” Educational Researcher. 23: 5–14. Hendrick, Rebecca M., and James C. Garand. 1991. “Variation in State Economic Growth: Decomposing State, Regional, and National Effects.” Journal of Politics. 53: 1093–1110. Henig, Jeffrey. 1994. Rethinking School Choice: Limits of the Market Metaphor. Princeton, NJ: Princeton University Press. Henig, Jeffrey. 1998. “Market and Community-based Proposals for School Reform.” Policy Currents. 8:(1): 7–8. Herberg, Will. 1960. Protestant, Catholic, Jew; An Essay in American Religious Sociology. Garden City, NY: Anchor Books. Hess, Frederick. 1999. Spinning Wheels: The Politics of Urban School Reform. Washington, D.C.: Brookings Institution Press. Hibbing, John R. and Elizabeth Theiss-Morse. 1995. Congress as Public Enemy: Political Attitudes Toward American Political Institutions. New York: Cambridge University Press. Hill, Paul, Lawrence Pierce, and James Guthrie. 1997. Reinventing Public Education: How Contracting Can Transform America’s Schools. Chicago: University of Chicago Press. Hirschman, Albert. 1970. Exit, Voice , and Loyalty: Responses to Decline in Firms, Organizations and States. Cambridge, MA: Harvard University Press. Hochschild, Jennifer and Bridget Scott. 1998. “Governance and Reform of Public Education in the United States.” Public Opinion Quarterly. 62: 79–120. Hofstadter, Richard. 1962. Anti-Intellectualism in American Life. New York: Vintage Books. Holbrook, Thomas and Emily Van Dunk. 1993. “Electoral Competition in the American States.” American Political Science Review. 87: 955–962. Howe, Kenneth R. 1997. Understanding Equal Educational Opportunity. New York: Teachers College Press. Hoxby, Caroline M. 1994. “Do Private Schools Provide Competition for Public Schools?” NBER Working Paper No. 4978.
References
179
Hoxby, Caroline M. 1996. “Evidence on Private School Vouchers: Effects On Schools and Students.” In Performance-Based Approaches to School Reform. Washington, D.C.: Brookings Institution. Hutton, Will. 1995. The State We’re In. London: Vintage. Inglehart, Ronald. 1988. “The Renaissance of Political Culture.” American Political Science Review. 82: 1203–1230. Inglehart, Ronald. 1990. Culture Shift In Advanced Industrial Society. Princeton: Princeton University Press. Jackman, Mary. 1978. “General and Applied Tolerance: Doe Education Increase Commitment to Racial Integration?” American Journal of Political Science. 22: 302–324. Jackman, Mary and Michael Muha 1984. “Education and Intergroup Attitudes: Moral Enlightenment, Superficial Democratic Commitment, or Ideological Refinement?” American Sociological Review. 49: 751–769. Jackman, Robert W. and Ross A. Miller. 1996. “A Renaissance of Political Culture?” American Journal of Political Science. 40: 632–659. Jefferson, Thomas. 1975. The Portable Thomas Jefferson. New York: Viking Press. Jencks, Christopher and Meredith Phillips. 1998. “The Black-White Test Score Gap: An Introduction.” In The Black-White Test Score Gap, Christopher Jencks and Meridith Phillips, eds. Washington, D.C.: The Brookings Institute. Jencks, Christopher. 1998. “Racial Bias in Testing.” In The Black-White Test Score Gap, Christopher Jencks and Meridith Phillips, eds. Washington, D.C.: The Brookings Institute. Jensen, Arthur R. 1969. “How Much Can We Boost I.Q. and Scholastic Achievement?” Harvard Educational Review. 39: 1–123. Johnson, Charles A. 1976. “Political Culture in American States: Elazar’s Formulation Examined.” American Journal of Political Science. 20: 491–509. Johnson, William. 1998. “Basic Skills and the Black-White Earning Gap.” In The Black-White Test Score Gap, Christopher Jencks and Meridith Phillips, eds. Washington, D.C.: The Brookings Institute. Jondrow J., C.A.K. Lovell, I.S. Materov, and P. Schmidt. 1982. “On the Estimation of Technical Inefficiency in the Stochastic Frontier Production Model.” Journal of Econometrics. 19: 233–238. Jones, Bryan. 1990. “Public Policies and Economic Growth in the American States.” The Journal of Politics. 52: 219–233. Kaestle, Carl. 1983. Pillars of the Republic. New York: Hill and Wang.
180
References
Karabel, Jerome and A.H. Halsey, eds. 1977. Power and Ideology in Education. New York: Oxford University Press. Katsillis, John and Richard Rubinson. 1990. “Cultural Capital, Student Achievement, and Educational Reproduction: The Case of Greece.” American Sociological Review. 55: 270–279. Key, V.O. 1961. Public Opinion and American Democracy. New York: Knopf. Kiker, B.F. 1966. “The Historical Roots of the Concept of Human Capital.” The Journal of Political Economy. 74: 481–499. Koch, Kather. 1999. “School Vouchers.” CQ Researcher. 9(13): 283–303. Kolpin, V. and L.D. Singell. 1997. “Asymmetric Information, Strategic Behavior and Discrimination in the Labor Market.” Economic Theory. 12: 175–184. Kozol, Jonathan. 1991. Savage Inequalities. New York: HarperCollins. Kuttner, Robert. 1993. “Training Programs Alone Can’t Produce $20–an-Hour Workers.” Business Week. March 8, p. 16. Kuttner, Robert. 1997. Everything for Sale: The Virtue and Limits of Markets. New York: Alfred A. Knopf. Lamdin, Douglas J. and Michael Mintrom. 1997. “School Choice in Theory and Practice: Taking Stock and Looking Ahead.” Education Economics. 5 (3): 211–244. Lange, Peter and Geoffrey Garrett. 1985. “The Politics of Growth.” The Journal of Politics. 49: 792–827. Lasswell, Harold D. 1951. “The Policy Orientation.” In The Policy Science, Daniel Lerner and Harold D. Lasswell, eds. Palo Alto, CA: Stanford University Press. Lawrence, Curtis. 1996. “Choice Backers Defend Expansion.” Milwaukee Journal Sentinel. Feb. 22, p. 1B. Lehnen, Robert G. 1992. “Constructing State Education Performance Indicators from ACT and SAT Scores.” Policy Studies Journal. 20: 22–40. Lemann, Nicholas. 1995. “The Great Sorting.” The Atlantic Monthly. 276(3): 84–97. Levin, Henry M. and Carolyn Kelly. 1994. “Can Education Do It Alone?” Economics of Education Review. 12: 97–108. Lieberman, Myron. 1993. Public Education: An Autopsy. Cambridge, MA: Harvard University Press. Lieske, Joel. 1993. “Regional Subcultures of the United States.” Journal of Politics. 55: 888–913.
References
181
Lipsky, Michael. 1980. Street-Level Bureaucracy: Dilemmas of the Individual in Public Services. New York: Russell Sage Foundation. Litt, Edgar. 1963. “Civic Education, Community Norms and Political Indoctrination.” American Sociological Review. 28: 69–75. Long, J. Scott. 1997. Regression Models for Categorical and Limited Dependent Variables. Thousand Oaks, CA: Sage. Loveless, Tom. 1997. “The Structure of Public Confidence in Education.” American Journal of Education. 105: 127–159. Lutz, Sabrina. 1996. “The Impact of School Choice in the United States and the Netherlands on Ethnic Segregation and Equal Educational Opportunity.” Equity & Excellence in Education. 29: 48–54. Macedo, Stephen. 1998. “Transformative Constitutionalism and the Case of Religion.” Political Theory. 26: 56–80. Maranto, Robert, Scott Milliman, and Scott Stevens. 2000. “Does Private School Competition Harm Public Schools? Revisiting Smith and Meier’s ‘The Case Against School Choice.’” Political Research Quarterly. 53: 177–192. Marschall, Melissa J. 1998. “Institutional Reform, Citizen Involvement and Social Capital: The Effects of Coproduction Inner-City Schools and Communities.” Paper presented at the Annual Meeting of the Midwest Political Science Association, Chicago, IL, April 23–28. Marwell, Gerald and Ruth Ames. 1981. “Economists Free Ride, Does Anyone Else?” Journal of Public Economics. 15: 295–310. Mayer, Susan and Paul E. Peterson, eds. 1999. Learning and Earning: How Schools Matter. Washington, D.C.: Brookings Institution Press. Mayer, Susan. 1999. “From Learning to Earning.” In Learning and Earning: How Schools Matter, Susan Mayer and Paul E. Peterson, eds. Washington, D.C.: Brookings Institution Press. McArthur, C.C. 1954. “Personalities of Public and Private School Boys.” Harvard Educational Review. 24: 256–262. McArthur, C.C. 1960. “Sub-Culture and Personality During the College Years.” Journal of Educational Sociology. 33: 260–268. McCabe, Barbara Coyle and Janet Coble Vinzant. 1999. “Governance Lessons: The Case of Charter Schools.” Administration & Society. 31: 361–377. McClosky, Herbert. 1967. “Personality and Attitude Correlates of Foreign Policy Orientation.” In Domestic Sources of Foreign Policy. James N. Rosenau, ed. New York: Free Press.
182
References
McClosky, Herbert and Alida Brill. 1983. Dimensions of Tolerance. New York: Russell Sage Foundation. McNeal, Ralph B., Jr. 1999. “Participation in High School Extracurricular Activities: Investigating School Effects.” Social Science Quarterly. 80: 291–309. McPartland, James and J. Timothy Sprehe. 1973. “Racial and Regional Inequalities in School Resources Relative to Their Educational Outcomes.” Social Science Research. 2: 321–332. Meier, Deborah. 1997. “Habits of Mind: Democratic Values and the Creation of Effective of Learning Communities.” In Common Schools, Uncommon Futures. Barry S. Kogan, ed. New York: Teachers College Press. Meier, Kenneth J. and Jeff Gill. 2000. What Works: A New Approach to Program and Policy Analysis. Boulder, CO: Westview Press. Meier, Kenneth J., Jeff Gill, Robert D. Wrinkle, and J.L. Polinard. 2000. “Weighting with an Exogenous Variable or with Two Exogenous Variables: Equity Versus Excellence in Organizations.” In What Works: A New Approach to Program and Policy Analysis. Kenneth J. Meier and Jeff Gill, eds. Boulder, CO: Westview Press. Meier, Kenneth J. and Lael R. Keiser. 1996. “Public Administration as a Science of the Artificial: A Methodology for Prescription.” Public Administration Review. 56: 459–466. Meier, Kenneth J., Robert D. Wrinkle, and J.L. Polinard. 1997. “Equity and Excellence in Education: A Substantively Weighted Least Squares Analysis.” Paper presented at the annual meeting of the Midwest Political Science Association. April 12–14, Chicago, IL. Merelman, Richard. 1980. “Democratic Politics and the Culture of American Education.” American Political Science Review. 74: 319–332. Meyer, Philip. 1993. “SAT Scores, Journalism and Public Policy.” Economics of Education Review. 12: 195–196. Miller, Paul, Charles Mulvey, and Nick Martin. 1995. “What Do Twins Studies Reveal About the Economic Returns to Education? A Comparison of Australian and U.S. Findings.” The American Economic Review. 85: 586–599. Miller, Trudi C. 1981. “Political and Mathematical Perspectives on Educational Equity.” American Political Science Review. 75: 319–333. Miner, Barbara. 2000. “For-Profit Firm On the Ropes.” Rethinking Schools Online. Http://www.rethinkingschools.org/Archives/14_03/eai43.htm. Accessed Jan. 14, 2002. Mintrom, Micahel. 1998. “Local Organizations As Sites for Deliberative Democracy: Learning From Charter Schools.” Paper presented at the
References
183
Annual Meeting of the American Political Science Association, Boston, MA, September 3–6. Moe, Terry. 2001. Schools, Vouchers, and the American Public. Washington, D.C.: Brookings Institution Press. Molnar, Alex. 1994. Giving Kids the Business: The Commercialization of America’s Schools. Boulder, CO: Westview Press. Molnar, Alex. 1995. “School Funding: The Right Issue, The Wrong Logic.” Educational Leadership. 53: 58–59. Mosher, Ralph, Robert A. Kenny, Jr., and Andrew Garrod. 1994. Preparing for Citizenship: Teaching Youth to Live Democratically. Westport, CT: Praeger. Mosteller, Frederick and Daniel P. Moynihan, eds. 1972. On Equality of Educational Opportunity. New York: Vintage Books. Mueller, Jerry Z. 1993. Adam Smith in His Time and Ours: Designing the Decent Society. New York: The Free Press. Muir, Ed. 1999. “They Blinded Me With Political Science: On the Use of NonPeer Reviewed Research in Education Policy.” PS: Political Science and Politics. 32(December): 762–764. Murphy, John and Jeffrey Schiller. 1992. Transforming America’s Schools: An Administrator’s Call for Action. LaSalle, IL: Open Court Publishing. Murphy, Joseph. 1996. The Privatization of Schooling: Problems and Possibilities. New York: Corwin Press. Murray, Charles and Richard J. Herrnstein. 1994. The Bell Curve. New York: Free Press. Myrdal, Gunnar. 1969. Objectivity in Social Research. New York: Pantheon Books. National Center for Education Statistics. 1996. National Household Education Survey. CD-ROM. National Center for Education Statistics. 1997a. Education in States and Nations. Http://www.ed.gov/NCES/pubs/esn/n25a.html. Accessed June 14, 1997. National Center for Education Statistics. 1997b. Education and the Economy: An Indicators Report. Washington, D.C.: U.S. Government Printing Office. National Center for Education Statistics. 1998. Teacher Followup Survey. U.S. Department of Education. CD-ROM. NCES 98–312. National Commission on Excellence in Education. 1984. A Nation at Risk: The Full Account. Portland, OR: USA Research, Inc.
184
References
Necochea, Juan and Zulmar Cline. 1996. “A Case Study of Within District Funding Inequities.” Equity & Excellence in Education. 29: 69–72. Nifong, Christina. 1996. “School Privatization Heads to Remedial-Education Class.” Christian Science Monitor. Feb. 2, p. 12. Niskanen, William. 1994. Bureaucracy and Economics. Brookfield, VT: Edward Elgar Publishing. Noblit, George W. and Van O. Dempsey. 1996. The Social Construction of Virtue: The Moral Life of Schools. Albany, NY: State University of New York Press. Nuthall, Graham. 1999. “Introduction and Background.” International Journal of Educational Research. 31: 141–256. Olson, Lyn. 1997. “Keeping Tabs on Quality.” Quality Counts. Washington, D.C.: Education Week/Pew Charitable Trusts. Olson, Lynn and Craig D. Jerald. 1998. “Barriers to Success.” In Quality Counts ’98: The Urban Challange. Washington, D.C.: Education Week/Pew Charitable Trusts. Olson, Mancur. 1965. The Logic of Collective Action. Cambridge, MA: Harvard University Press. O’Neill, June. 1990. “The Role of Human Capital in Earnings Differences Between Black and White Men.” The Journal of Economic Perspectives. 4: 25–45. Osborne, David. 1988. Laboratories of Democracy. Boston: Harvard Business School Press. Osborne, David and Ted Gaebler. 1993. Reinventing Government: How the Entrepreneurial Spirit is Transforming the Public Sector. New York: Plume. Ostrom, Vincent. 1973 (1989). The Intellectual Crisis in American Public Administration. Tuscaloosa, AL: The University of Alabama Press. Pangle, Lorraine Smith and Thomas L. Pangle. 1993. The Learning of Liberty. Lawrence, KS: University Press of Kansas. Parrish, T.B., C.S. Matsumoto, and W.J. Fowler, Jr. 1995. Disparities in Public School District Spending 1989–90: A Multivariate, Student-Weighted Analysis, Adjusted for Differences in Geographic Cost of Living and Student Need. Washington, DC: National Center for Education Statistics, U.S. Department of Education. Paris, David. 1994. “Schools, Scapegoats, and Skills: Educational Reform and the Economy.” Policy Studies Journal. 22: 10–24. Parry, Taryn Rounds. 1996. “Will Pursuit of Higher Quality Sacrifice Equal Opportunity in Education? An Analysis of the Education Voucher System in Satiago.” Social Science Quarterly. 77: 821–841.
References
185
Paulsen, Ronnelle. 1991. “Education, Social Class, and Participation in Collective Action.” Sociology of Education. 64: 96–110. Peevely, G.L and J.R. Ray. 1989. “Equity as Determined by Locally Funded Teaching Positions in Tennessee Schools.” Journal of Educational Finance. 15: 189–204. Peterson, Paul E. and Chad Noyes. 1997. “School Choice in Milwaukee.” In New Schools for a New Century. Diane Ravitch and Joseph Viteritti, eds. New Haven: Yale University Press. Peterson, Paul E., Jay P. Greene, William G. Howell, and William McGready. 1998. “Initial Findings of School Choice Programs in Washington, D.C.” Paper presented at the annual meeting of the American Political Science Association. Boston, September 1998. Plank, David N. and William Lowe Boyd. 1994. “Antipolitics, Education, and Institutional Choice: The Flight From Democracy.” American Educational Research Journal. 31: 263–281. Powers, Jeanne M. and Peter W. Cookson, Jr. 1999. “The Politics of School Choice Research: Fact, Fiction, and Statistics.” Educational Policy. 13: 104–122. Prather, James E. and Frank Gibson. 1977. “The Failure of Social Programs.” Public Administration Review. 37: 556–564. Putnam, Robert D. 1993. Making Democracy Work. Princeton: Princeton University Press. Putnam, Robert D. 1995a. “Tuning In, Turning Out: The Strange Disappearance of Social Capital in America.” PS: Political Science and Politics. 28: 664–683. Putnam, Robert D. 1995b. “Bowling Alone, Revisited.” The Responsive Community. Spring: 18–33. Ravitch, Diane. 1997. “New York the Obsolete Factory.” In New Schools for a New Century. Diane Ravitch and Joseph Viteritte, eds. New Haven: Yale University Press. Rebell, Michael. 1998. “Fiscal Equity Litigation and the Democratic Imperative.” Journal of Education Finance. 24: 23–50. Reich, Robert B. 1990. “Education and the Next Economy.” In Education Reform: Making Sense of It All. Samuel B. Bacharach, ed. Boston: Allyn & Bacon. Reyes, Pedro and Diana G. Pounder. 1993. “Organizational Orientation in Public and Private Elementary Schools.” Journal of Educational Research. 87: 86–93.
186
References
Rice, Tom W. and Alexander F. Sumberg. 1997. “Civic Culture and Government Performance in the American States.” Publius: The Journal of American Federalism. 27: 99–114. Riley, John G. 1976. “Information, Screening and Human Capital.” The American Economic Review. 66: 254–260. Ritzen, Jozef, Jan van Dommelen, and Frans J. De Vijlder. 1997. “School Finance and Choice in the Netherlands.” Economics of Education Review. 16: 329–336. Rizzuto, R. and P. Wachtel. 1980. “Further Evidence on the Returns to School Quality.” Journal of Human Resources. 15: 240–254. Rose, Lowell C. and Alec M. Gallup. 2000. “The 32nd Annual Phi Delta Kappan/Gallup Poll of the Public’s Attitudes Toward the Public Schools.” Phi Delta Kappan. 82(1): 41–66. Rothman, Robert. 1991. “Revisionists Take Aim at Gloomy View of Schools.” Education Week. November. Rothstein, Richard. 1997. What Do We Know About Declining (or Rising) Student Achievement? Arlington, VA: Educational Research Service. Rothstein, Richard. 1998. The Way We Were? New York: Twentieth Century Fund Press. Rouse, C.E. 1999. “Further Estimates of the Economic Return to Schooling from a New Sample of Twins.” Economics of Education Review. 18: 1–22. Salamon, Lester M. 1991. “Why Human Capital? Why Now?” In Human Capital and America’s Future. David Hornbeck and Lester Salamon, eds. Baltimore: The Johns Hopkins University Press. Samuelson, Robert J. “Hollow School Reform.” Washington Post. April 28, 1993, p. 19. Sayrs, Lois. 1989. Pooled Time Series Analysis. Newbury Park, CA: Sage. Schneider, Mark, Paul Teske, Christine Roch, and Melissa Marschall. 1997. “Networks to Nowhere: Segregation and Stratification in Networks of Information About Schools.” American Journal of Political Science. 41: 1201–1223. Schneider, Mark, Paul Teske, and Melissa Marschall. 2001. Choosing Schools: Consumer Choice and the Quality of American Schools. Princeton, NJ: Princeton University Press. Schultz, Theodore. 1961. “Investment in Human Capital.” American Economic Review. 51: 1–17. Schultz, Theodore. 1963. The Economic Values of Education. New York: Columbia University Press.
References
187
Schwarzmantel, John. 1997. The Age of Ideology: Political Ideologies from the American Revolution to Postmodern Times. Washington Square, NY: New York University Press. Sclar, Elliott D. 2000. You Don’t Always Get What You Pay For: The Economics of Privatization. Ithaca, NY: Cornell University Press. Seltzer, C.C. 1948. “Academic Success in College of Public and Private School Students: Freshman Year at Harvard.” Journal of Psychology. 25: 419–431. Sharkansky, Ira. 1969. “The Utility of Elazar’s Political Culture: A Research Note.” Polity. 2: 66–83. Simon, Herbert. 1969. The Sciences of the Artificial. Cambridge, MA: MIT Press. Simon, Herbert. 1997. Administrative Behavior, 4th ed. New York: The Free Press. Simon, Julian L. and Rebecca Boggs. 1997. “Trends in the Quantities of Education: A Pictorial Essay.” Economics of Education Review. 16: 69–80. Slavin, Robert E. 1999. “How Can Funding Equity Ensure Enhanced Achievement?” Journal of Education Finance. 24: 519–528. Smith, Adam. 1986. The Essential Adam Smith. Robert L. Heilbroner, ed. New York: W.W. Norton. Smith, Kevin B. and Jeremy Eccles. 1998. “Buying a Better SAT Score: A Renewed Search for the Elusive Link between Education Expenditures and Outcomes.” State and Local Government Review. 30: 42–51. Smith, Kevin B. and Kenneth J. Meier. 1995. The Case Against School Choice. New York: M.E. Sharpe. Smith, Kevin B., Kenneth J. Meier, and Jeff Gill. 2000. “SWAT In Pooled Analysis.” In What Works: A New Approach to Program and Policy Analysis, Kenneth J. Meir and Jeff Gill, eds. Boulder, CO: Westview Press. Smith, Kevin B. and J. Scott Rademacker. 1998. “Out With the Input: The Misplaced Focus of Educational Policymaking.” Paper presented at the Midwest Political Science Association annual meeting, Chicago, April 1998. Smith, Kevin B. and J. Scott Rademacker. 1999. “Expensive Lessons: Education and the Political Economy of the American States.” Political Research Quarterly. 52: 709–727. Smith, Kevin B. and J. Scott Rademacker. 2001. “Money Only Matters If You Want It To?” Paper presented at the Midwest Political Science Association annual meeting, Chicago, April 2001. Spencer, Bruce and David Wiley. 1981. “The Sense and the Nonsense of School Effectiveness.” Journal of Policy Analysis and Management. 1: 43–52.
188
References
Stimson, James. 1985. “Regression in Space and Time: A Statistical Essay.” American Journal of Political Science. 29: 914–927. Stone, Deborah. 1997. Policy Paradox: The Art of Political Decision Making. New York: W.W. Norton. Sullivan, John L., James Piereson, and George Marcus. 1982. Political Tolerance and American Democracy. Chicago: University of Chicago Press. Swanson, Austin and Richard King. 1991. School Finance: Its Economics and Policies. New York: Longman. Tanner, Daniel. 1998. “The Social Consequences of Bad Research.” Phi Delta Kappan. 79(5): 344–350. Taylor, Lori L. Economic Review. Third Quarter 1994: 1–13. Teske, Paul, Mark Schneider, Michael Mintrom, and Samuel Best. 1993. “Establishing the Micro Foundations of a Macro Theory: Information, Movers, and the Competitive Local Market for Public Goods.” American Political Science Review. 87: 702–713. Texas Education Agency. 2000a. Snapshot 99: School District Profiles. http://www.tea.state.tx.us/perfreport/snapshot/. Accessed March 3, 2000. Texas Education Agency. 2000b. Texas Open-Enrollment Charter Schools: Third Year Evaluation. http://www.tea.state.tx.us/charter/eval.html. Accessed April 24, 2000. Thurow, Lester C. 1970. Investment in Human Capital. Belmont, CA: Wadsworth. Tiebout, Charles. 1956. “A Pure Theory of Local Expenditures.” Journal of Political Economy. 64: 416–424. Tyack, David. 1974. The One Best System: A History of Urban Education. Cambridge, MA: Harvard University Press. Tyack, David and Larry Cuban. 1995. Tinkering Toward Utopia: A Century of Public School Reform. Cambridge: Harvard University Press. Tyler, William B. 1985. “The Organizational Structure of the School.” American Review of Sociology. 11: 49–73. Unnever, James D., Allan C. Kerckhoff, and Timoth J. Robinson. 2000. “District Variations in Educational Resources and Student Outcomes.” Economics of Education Review. 19: 245–259. Vandenberghe, Vincent. 1999. “Combining Market and Bureaucratic Control in Education: An Answer to Market and Bureaucratic Failure?” Comparative Education. 35: 271–282. Verba, Sidney, Kay Lehman Schlozman, and Henry Brady. 1996. Voice and Equality. New York: Belknap Press.
References
189
Verstegen, Deborah A. and Richard A. King. 1998. “The Relationship Between School Spending and Student Achievement: A Review and Analysis of 35 Years of Production Function Research.” Journal of Education Economics. 24: 243–262. Vinovskis, Maris A. 1995. Education, Society, and Economic Opportunity: A Historical Perspective on Persistent Issues. New Haven: Yale University Press. Walsh, Mark. 1999. “Most Edison Schools Report Rise in Test Scores.” Education Week. 18 (31): 9–12. Weber, Max. 1958. The Protestant Work Ethic and the Spirit of Capitalism. New York: Scribner’s. Weiss, Andrew. 1995. “Human Capital vs. Signalling Explanations of Wages.” Journal of Economic Perspectives. 9: 133–154. Welch, Finis. 1975. “Human Capital Theory: Education, Discrimination, and Life Cycles.” The American Economic Review. 65: 63–73. Wells, Amy Stuart. 1996. “African American Students’ View of School Choice.” In Who Chooses, Who Loses? Bruce Fuller and Richard Elmore, eds. New York: Teachers College Press. Wells, Amy Stuart and Alejandra Lopez. 1999. “Charter Schools as Postmodern Paradox: Rethinking Social Stratification in an Age of Deregulated Education.” Harvard Educational Review. 69(Summer): 172–205. Whittle, Christopher. 1997. “The Edison Project Founder’s Musings on American Schooling.” The School Administrator. 1(54): 6–9. Wiarda, Howard J., ed. 1992. Politics and Social Change in Latin America. Westview Press: Boulder, CO. Wilson, C. Dwayne. 1988. “Equity-Based Education: The Elusive but Attainable Dream.” Equity & Excellence in Education. 24: 22–25. Wilson, James Q. 1989. Bureaucracy: What Government Agencies Do And Why They Do It. New York: Basic Books. Witte, John F. 1996. “Reply to Greene, Peterson and Du: ‘The Effectiveness of School Choice in Milwaukee: A Secondary Analysis of Data from the Program’s Evaluation.’” Paper presented at the American Political Science Association Meeting, San Francisco, August 1996. Witte, John F. 1996. “Politics, Markets, or Money? The Political-Economy of School Choice.” Paper presented at the American Political Science Association Annual Meeting, San Francisco, August 1996. Witte, John F. 2000. The Market Approach to Education: An Analysis of America’s First Voucher Program. Princeton, NJ: Princeton University Press.
190
References
Wofinger, Raymond E. and Steven J. Rosenstone. 1980. Who Votes? New Haven: Yale University Press. Wright, R.G. 1997. Selling Words: Free Speech in a Commercial Culture. New York: New York University Press. Wrinkle, Robert D., Joseph Stewart, Jr., and J.L. Polinard. 1999. “Public School Quality, Private Schools and Race.” American Journal of Political Science. 43: 1248–1253. Wyatt, Edward. 2001. “Mixed School Record for Edison Foes.” New York Times. April 25, p. B8. Zelinsky, Wilbur. 1973. The Cultural Geography of the United States. Englewood Cliffs, NJ: Prentice-Hall. Zinser, Jana. 1994. Reinventing Education. Issue Paper No. 1, Investing in People Project. Denver, CO: National Conference of State Legislatures. Zweigenhaft, Richard L. 1992. “The Application of Social and Cultural Capital: A Study of the 25th Reunion Entries of Prep School and Public School Graduates of Yale College.” Higher Education. 23: 311–320. Zweigenhaft, Richard L. 1993. “Prep School and Public School Graduates of Harvard: A Longitudinal Study of the Accumulation of Social and Cultural Capital.” Journal of Higher Education. 64: 211–225.
Index ACT. See Test scores Adams, John, 106 Alexander, F. King, 52, 54, 56, 61, 62, 84, 90, 98 Almond, Gabriel, 103, 106 American Federation of Teachers, 138 Apple, Michael, 19, 25, 128 Arnstine, Donald, 10, 47, 59, 101, 108 Ascher, Carol, 137 Ball, Stephen, 135 Barber, Benjamin, 3, 8, 9, 12, 14, 99 Becker, Gary, 20, 29, 128 Bennett, William, 10, 129 Berliner, David, 6, 13, 128, 129, 132 Biddle, Bruce, 6, 13, 128, 129, 132 Bishop, John, 16, 20, 21, 26 Blais, Andre, 145 Blaug, Mark, 27 Bobo, Lawrence, 111, 114, 118 Boggs, Rebecca, 53 Boix, Charles, 104, 105 Borjas, George, 28 Borrowman, Merle, 12 Bound, John, 22 Bourdieu, Pierre, 12, 108 Bowe, Richard, 135 Bowles, Samuel, 29 Boyd, William, 146 Box, Richard, 2, 5 Brace, Paul, 32, 33, 152n
Bracey, Gerald, 26 Brehm, John, 106 Brighouse, Harry, 48, 55 Brill, Alida, 116 Brown v. Board of Education, 8, 49 Brown, Byron, 135 Buchanan, James, 4, 7 Burtless, Gary, 21, 22, 23, 24, 151n Callan, Eamon, 2, 3, 8, 12, 48, 77, 99, 142 Campbell, Angus, 3 Card, David, 54 Carson, C. C., 26, 131 Carter, John, 145 Center on National Education Policy, 8 Center on Policy Attitudes, 155n Chambers, M.M., 56 Chubb, John, 1, 2, 4, 5, 9, 11, 20, 22, 47, 51, 54, 56, 62, 76, 77, 81, 82, 83, 85, 89, 90, 91, 98, 128, 136, 139, 140, 153n Civic capital: Catholic schools and, 124–125; defined, 105; determinants of, 120–121; education and, 108–113, 124; measure of, 154n, 167–169; private schools and, 115–116, 121–122, 123; race, class and, 114–116, 122, 123; schools and, 112–115. See also Social capital, Education
191
192
Index
Clark, Burton, 6 Clark, S.D., 28 Cline, Zulmar, 54, 57 Coleman, James, 22, 31, 48, 49, 60, 67, 83 Coleman Report, 48–49 Commonwealth: civic culture and, 143–145; democracy and, 9; perspective on public education, 2, 102–103; school goals and, 77–80, 101–102; values of, 14, 47, 127 145–146. See also Education, Market, Public education Conflict theory, 12, 29, 102, 108, 109, 127 Consensus theory, 12, 76, 102 Cook, Philip, 29 Cookson, Peter, 1, 2, 3, 11, 78, 113 Cox, Harvey, 1 Cremin, Lawrence, 8, 9, 12, 102, 154n Cuban, Larry, xi, 9, 12, 19, 84, 85, 128, 131, 146, 147, 148 Darity, William, 28 David, Martin, 50 Dempsey, Van, 10 Denison, Edward, 24 De Leon, Peter, xi De Tocqueville, Alexander, 104 De Young, Alan, 21, 29, 106 Dewey, John, 12, 77, 99 Downs, Anthony, 4 Dudley, Robert, 32 Dye, Thomas, 32 Eccles, Jeremy, 52, 53, 57, 89 Economic productivity: cultural and educational determinants of, 29–31, 34, 35–45. See also Education, Market Edison Project, 137, 138, 139, 156n Education: as ideology, 13–18, 127, 150; culture and, 43–44 105–108, 143; economic purposes of, 19–20, 102, 129; economic
returns of, 22–24, 31–32; efficiency and equity in, 51, 61–62, 73; equality of educational opportunity and, 60, 67, 71; external efficiency of, 22, 24, 47, 48; for profit, 136–139, 148; goals of, 75–78, 101–102, 103; goals, determinants of, 89–98, 163, 164–166; goals, institutional structure and, 78–80, 81–84; goals, measurement of, 84–86; goals, school type and, 86–89; inequities in, 52, 54–55; information on, 133; internal efficiency of, 22, 23, 24, 47, 48, 72; money and, 50–57, 60–61; outcomes of, 51; production functions, 51–53, 73, 151n; public opinion and, 155n; religious, 98, 99, 144, 149. See also Public education, Market, Commonwealth, Human capital, Test scores, Texas Education Alternatives Inc., 137–138, 139, 156n Elazar, Daniel, 34, 104, 145, 152n Engel, Michael, 1, 3, 8, 10, 12, 13 Finn, Chester, 7 Flathman, Richard, 114 Fogel, Robert, 31 Fosler, R. Scott, 21, 24 Friedenberg, Edgar, 27, 58 Friedman, Milton, 3, 4, 5, 10, 11, 14, 48, 51, 55, 56, 90, 128 Fukuyama, Francis, 104 Fuller, Howard, 141 Functionalism, 6, 19, 75, 103, 129, 134 Gaebler, Ted, 77, 146 Gallup, Alec, 132, 154n Garand, James, 32 Garrett, Geoffrey, 32 Gerstner, Louis, 1, 20, 77, 136 Gibson, Frank, 52 Gill, Jeff, 38, 41, 160 Gintis, Herbert, 29
Index Glenn, David, xi Goodlad, John, 132 Gould, Stephen Jay, 59, 134, 151n Graham, John, 29 Granato, Jim, 103, 105 Greene, Jay, 4, 47, 103, 112, 114, 141, 151n Guttman, Amy, 12, 48, 77, 99 Halchin, Elaine, 13 Halsey, A.H., 12 Hanushek, Eric, 1, 2, 16, 19, 21, 23, 25, 26, 33, 47, 50, 51, 52, 53, 57, 61, 62, 90, 128 Harris, D., 111 Hartoonian, Michael, 3 Hedges, L.V., 52, 53 Hendrick, Rebecca, 32 Henig, Jeffrey, 1, 8, 26, 62, 78, 83, 112, 113, 129, 136 Herrnstein, Richard, 59, 151n Hess, Frederick, 85 Hibbing, John, 146 Hill, Paul, 9, 137, 140 Hirschman, Albert, 15, 81 Hoffer, Thomas, 22, 30, 83 Hofstadter, Richard, 7, 9 Holbrook, Thomas, 158 Howe, Kenneth, 48 Hoxby, Caroline, 135, 140, 141 Human capital: culture and, 27–29; economic success and, 26–27; empirical tests of, 29–30, 33–34; measurement of, 25–26, 47, 48; signaling and, 27, 29; state economic success and, 33; theory of, 20–21, 24, 31–32, 151n Hutton, Will, 111 Hypothesis tests, 38, 53, 94, 152n Inglehart, Ronald, 103, 113 IQ. See Test scores Irons, Michael, 145 Jackman, Mary, 110, 114, 117, 118 Jackman, Robert, 105, 113
193
Jefferson, Thomas, 8, 102, 106, 109, 154n Jencks, Christopher, 26, 29, 47, 59, 151n Jensen, Arthur, 59 Jerald, Craig, 54 Jewett, Aubrey, 33 Johnson, Charles, 33 Johnson, William, 28 Jondrow, J., 52 Jones, Bryan, 32, 33 Kaestle, Carl, 8, 12, 77 Karabel, Jerome, 12 Katsillis, John, 108, 114 Keiser, Lael, 41, 160 Kelly, Carolyn, 22, 26, 31, 44, 47 Key, V.O., 101 Kiker, B.F., 20 King, Richard, 6, 16, 21, 22, 23, 24, 52, 53, 62 Koch, Kather, 141, 150, 156n Kolpin, V., 47 Kozol, Jonathan, 13, 47, 56, 61, 62, 83, 113 Krueger, Alan, 54 Kuttner, Robert, 135, 146 Lamdin, Douglas, 6 Lange, Peter, 32 Lasswell, Harold, 134 Lehnen, Robert, 159 Lemann, Nicholas, 26, 58 Levin, Henry, 22, 26, 31, 44, 47 Licari, Frederick, 111, 114, 118 Lieberman, Myron, 9, 25, 77, 83, 136 Lieske, Joel, 34 Lipsky, Michael, 85, 89 Litt, Edgar, 108, 109, 117, 118 Long, J. Scott, 92, 93 Lopez, Alejandra, 125, 142, 143 Loveless, Tom, 132 Ludwig, Jens, 29 Lutz, Sabrina, 11, 83, 135 Loveless, Tom, 6
194
Index
Macedo, Stephen, 88, 99, 143, 154n, 156n Madison, James, 7, 102 Mann, Horace, 8, 101, 102 Maranto, Robert, 62 Market: as normative theory, 3, 7; as positive theory, 3, 7; as religion, 1; black box and, 134–135; commonwealth values and, 48, 55, 72, 103; democratic values and, 7, 8, 9, 10, 11, 140, 142; human capital theory and, 24–25; ideology and education, 7, 9; models of education, 10–12; school goals and, 77–80; social stratification and, 11; theory and education, 4–7; values of, 2–3, 13, 102, 127, 145–146, 147. See also Civic capital, Commonwealth, Education, Public education Marschall, Melissa, 135, 140 Mayer, Susan, 21, 22, 128 McArthur, C.C., 111 McCabe, Barbara Coyle, 8, 13, 15, 82, 128 McClosky, Herbert, 106, 116 McNeal, Ralph, 111, 114 McPartland, James, 60 Meier, Deborah, 145 Meier, Kenneth, 1, 4, 5, 29, 33, 41, 55, 57, 66, 67, 78, 83, 112, 135, 159, 160 Merelman, Richard, 9 Miller, Paul, 21 Miller, Ross, 105, 113 Miller, Trudi, 61 Miner, Barbara, 138, 139 Mintrom, Michael, 6, 136, 140, 141, 148 Moe, Terry, 1, 2, 4, 5, 9, 11, 20, 22, 47, 51, 54, 56, 62, 76, 77, 81, 82, 83, 85, 89, 90, 91, 98, 128, 136, 140, 153n Molnar, Alex, 30, 54, 139 Mosher, Ralph, 140 Mueller, Jerry, 135
Muha, Michael, 110, 114, 117, 118 Muir, Ed, xi, 3 Murphy, Joseph, 9 Murphy, John, 1 Murray, Charles, 59, 151n Myrdal, Gunnar, xi National Center for Education Statistics, 30, 86, 91, 114, 163, 167 National Commission on Excellence in Education, 19, 26 Necochea, Juan, 54, 57 Nifong, Christina, 139 Niskanen, William, 5 Noblit, George, 10 Noyes, Chad, xi, 11, 47, 55, 62, 141 Nuthall, Graham, 85, 89, 90 Olson, Lyn, 33, 54 Olson, Mancur, 4 O’Neill, June, 28 Osborne, David, 77, 146 Ostrom, Vincent, 7 Pangle, Lorraine, 8, 48, 77 Pangle, Thomas, 8, 48, 77 Paris, David, 16, 19, 25, 26, 128 Parrish, T.B., 57 Parry, Taryn, 55 Paulsen, Ronnelle, 109, 111, 114, 119 Peevely, G.L., 54 Peterson, Paul, xi, 4, 11, 47, 55, 62, 102, 112, 128, 141 Phillips, Meredith, 29, 59, 151n Plank, David, 146 Policy sciences, xi–xii Posner, Daniel, 104, 105 Pounder, Diana, 80 Powers, Jeanne, 1, 2, 3, 11 Prather, James, 52 Public choice: assumptions of, 4, 81; education and, 5–7, 47. See also Market
Index Public education: as religion, 1; civic values and, 8, 102–103; democracy and, 2–3, 9, 12, 101–103; democratic control and, 10, 140; dissatisfaction with, 132; equality of educational opportunity and, 48–40, 57–60; failures of, 129–132; goals of, 10, 77–78, 128, 129, 146; inequity in, 13; political role of, 9; similarities with private education, 133; values and, 2. See also Commonwealth, Education, Market Putnam, Robert, 104, 105, 106, 113, 122, 154n Q equity measure, 66, 153n, 160, 162 Rademacker, J. Scott, 33, 35, 37, 152n Rahm, Wendy, 106 Ravitch, Diane, 1, 9, 55, 57 Ray, J.R., 54 Rebell, Michael, 1, 32, 54, 79, 83, 90, 102 Reyes, Pedro, 80 Rice, Tom, 105, 154n, 169 Riley, John, 27 Ritzen, Jozef, 11, 55, 83 Rizzuto, R., 54 Reich, Robert, 19 Rose, Lowell, 132, 154n Rosenstone, Steven, 106, 111 Rothman, Robert, 6 Rothstein, Richard, 26, 129 Rubison, Richard, 108, 114 Salamon, Lester, 24, 25 SAT. See Test scores Sayrs, Lois, 157, 158 Schiller, Jeffrey, 1 Schneider, Mark, 5, 11, 112, 128, 132, 140, 141, 148 Schools. See Education, Public education, Market, Commonwealth
195
Schultz, Theodore, 20, 24 Schwartzmantel, John, 3 Sclar, Elliott, 136 Seltzer, C.C., 111 Sharkansky, Ira, 34, 37, 43, 158, 159 Singell, L.D., 47 Simon, Herbert, xi, 85, 147 Simon, Julian, 53 Slavin, Robert, 54 Smith, Adam, 4, 20, 135 Smith, Kevin, 1, 4, 5, 29, 33, 35, 37, 41, 55, 57, 67, 78, 83, 89, 102, 135, 152n, 159 Social capital: defined, 104–105; public education and, 106–107, 141. See also Civic capital Solon, Gary, 22 Spencer, Bruce, 53 Sprehe, J. Timothy, 60 Statistically Weighted Analytic Technique (SWAT), 41, 68, 155n, 159–160 Stimson, James, 157 Sullivan, John, 106 Sumberg, Alexander, 105, 154n, 169 Swanson, Austin, 16, 22 Tanner, Daniel, xi, 3, 31, 52, 62, 134, 146 Taylor, Lori, 33, 52 Teske, Paul, 62, 133 Tesseract. See Education Alternatives Inc. Test scores, 23, 26, 27, 35, 47, 52, 53, 54, 58, 65, 62, 67, 68, 129–131, 148, 150, 152n, 153n, 158, 159, 160, 161, 162 Texas Assessment of Academic Skills (TAAS). See Test scores Texas Education Agency, 62, 72, 160, 161 Texas: charter schools in, 62–63, 153n; school governance in, 62, 153n; test scores and equality of educational opportunity in, 62–73
196 Texas Excellence in Education Project, 66, 153n Theiss-Morse, Elizabeth, 146 Thurow, Lester, 21 Tiebout, Charles, 4 Tullock, Gordon, 4, 7 Tyack, David, xii, 9, 12, 19, 62, 84, 85, 90, 128, 131, 146, 147, 148 Unnever, James, 53, 61 U.S. Department of Education, 150 Vandenberghe, Vincent, 5, 11 Van Dunk, Emily, 158 Verba, Sidney, 103, 105, 111, 119, 122, 124 Verstegen, Deborah, 6, 21, 23, 24, 52, 53, 62 Vinovskis, Maris, 8, 102 Vouchers, 11, 51, 55–57, 83, 98, 141, 148, 149, 155n Wachtel, P,. 54 Walsh, Mark, 138
Index Weber, Max, 28, 106 Weiss, Andrew, 27 Welch, Finis, 26 Wells, Amy Stuart, 29, 110, 125, 142, 143 Whittle, Christopher, 138, 149 Wiarda, Howard, 28 Wiley, David, 53 Williams, Rhonda, 28 Wilson, C. Dwayne, 54 Wilson, James, 85, 89 Witte, John, 3, 11, 13, 47, 51, 55, 132, 141, 148 Wolfinger, Raymond, 106, 111 Wright, R.G., 144 Young, Robert, 145 Youth Civic Involvement Survey, 114, 118, 167 Zinser, Jana, 33 Zweigenhaft, Richard, 111