Taking the "Public" Out of Our Schools: The Political, Constitutional and Civic Implications of Private School Vouchers1
"School choice" is the most contentious issue in American public education today. Traditionally, children attending public elementary and secondary schools have been required to attend a particular school in their district, regardless of the educational quality or character of that school. But in recent years, there has been an expansion of alternatives in education. The advent of magnet schools, charter schools, and school voucher programs-some of which allow children to attend private and parochial schools at the state's expense-has begun to provide parents the option to select the type of school their children will attend. This transition, from a static system of predetermined enrollment to a dynamic, consumer-oriented system of choice between competing schools, has the potential to change the very meaning of "public" education. It is a paradigm shift in the way that we conceptualize public schools, and it has aroused the passions (and brought out the pocketbooks) of various groups that rarely concern themselves with education reform.
What is it about school choice that arouses such controversy? For starters, the stakes in this paradigm shift are extremely high. Almost 90 percent of America's fifty-three million school-aged children attend primary or secondary schools funded by the government.2 And although only a quarter of American voters currently have school-aged children, everyone is connected in some way to the public school system: taxpayers finance it, employers hire its graduates, and more importantly, its effectiveness is widely understood to be a key measure of social and economic justice. Whatever else might be said about the American system of public education, all citizens seem to agree that such a fundamental change in its structure is likely to impact society well beyond the schoolyard in important and largely unforeseen ways.
If the warring factions agree about the stakes involved in the school choice debate, they seem to disagree about everything else. Proponents of school choice argue that free-market competition will force all schools to improve, since parents can "vote with their feet" and move their children to high-performing schools. They often claim that school choice initiatives will raise test scores, foster innovative teaching methods, mitigate segregation problems (which are a result of neighborhood schools being tied to segregated neighborhoods) and above all, serve the interests of justice by giving poor families more educational choices for their children. Opponents, however, argue against a market-based system on the grounds that it does not adequately value education, and that the market economy has consistently failed to serve the interests of lower-income Americans. Furthermore, they fear that school choice initiatives will decrease public accountability and oversight, aggravate segregation problems (by allowing or even encouraging "balkanized" schools with focused ethnic, religious or ideological curricula to compete against traditional "common schools"), undermine other public school reform efforts by shifting scarce resources away from public schools, and concentrate underperforming and problem students in inferior schools.
Despite the controversy, a variety of school choice initiatives are moving ahead with the support of federal, state, and local governments. Public magnet schools, first established in the 1970s, now number over four thousand, offering students across the nation specialized curricula in the arts, sciences, and other fields.3 Charter schools are a more recent phenomenon, and have proven to be quite popular. The nation's first charter school opened in 1992 in St. Paul, Minnesota, and today there are over 1,600 such schools in thirty-seven states and the District of Columbia.4 Presidents Clinton and (George W.) Bush have both offered federal grants totaling hundreds of millions of dollars to open new magnet and charter school programs nationwide. Existing voucher programs are also expanding. Ten years after Milwaukee initiated the nation's first school voucher program (for 300 inner-city children), almost 64,000 children in thirty-one states receive vouchers to attend the private or parochial schools of their (parents') choice.5 And it is clear that the demand for school vouchers far exceeds their supply: in 1999, for example, a private foundation offering 40,000 vouchers nationwide received 1.25 million applications.6 National media coverage of school choice debates in the federal, state, and local governments continues to fuel the demand, with no end in sight.
If the overall school choice movement is gathering momentum, however, the expansion of school voucher programs has stalled in the past year. Statistics regarding the explosive growth of school choice (as those noted above) and the attendant media buzz over political battles often serve to obscure the critical distinction between public and private financing schemes for vouchers. Over 80 percent of the current school vouchers are financed by philanthropic individuals or foundations, and the few publicly-funded vouchers that exist are concentrated almost exclusively in two cities, Milwaukee and Cleveland.7 None of the nation's largest cities have implemented voucher programs, and while at least twenty-five state legislatures are currently considering voucher plans, only Florida has passed statewide voucher legislation. Furthermore, well-financed ballot initiatives to establish statewide school voucher programs in California and Michigan were voted down in the 2000 election by a 2-1 margin.
School vouchers are indeed controversial for the many reasons noted above, but legislators and private citizens often vote against them for an entirely different (and decisive) reason: they are concerned that by using tax dollars to educate children in religious schools, publicly-- funded school voucher programs would violate the constitutional proscription against the establishment of religion. Privately-funded voucher programs are not affected by the Establishment Clause since they do not involve government expenditures. But most supporters assert that if vouchers are to have any real impact on the American educational system, they must be widespread-and that requires public financing. As several widely-watched cases on this issue wind their way through federal courts, supporters and opponents alike agree that the constitutional question must be resolved by the United States Supreme Court before vouchers can be widely implemented.
This essay is, in part, my answer to the constitutional question about school vouchers. I shall argue that a carefully crafted voucher program can indeed survive constitutional scrutiny, and that recent First Amendment jurisprudence suggests it will do so if put before the present Supreme Court. Given that legal sanction is not itself sufficient warrant for undertaking such a massive shift in public policy, I then ask what-if anything-might justify the implementation of school voucher programs. In so doing, the paper's final section considers the aims of education in a liberal democratic society, suggesting that private school vouchers ought to be distributed only by a public school system in extremis, and only then as a temporary measure. In order to make these constitutional and political arguments, however, we must first have a keen understanding of school vouchers themselves. What, exactly, are they designed to accomplish? Who supports them, and why? It is these questions to which we now turn.
THE PRINCIPLES AND PRINCIPALS BEHIND SCHOOL VOUCHERS
In the lexicon of educational reform, "school choice" refers to any initiative designed to allow students to attend a school other than their traditional local public school. The character of the choice program, however, varies widely according to the type of schools students can elect to attend. The most common type of school choice program allows students to choose from among any public school within their district, or (less frequently) public schools in a neighboring district. Most cities and states now supplement students' choice of schools by offering other, nontraditional public schools as well, including magnet and/or charter schools. Magnet schools place special emphasis on academic achievement or on a particular field such as science; they are designed to attract students from across the school district, and they are owned and operated by the government. Charter schools (sometimes called "community schools") offer curricula or educational themes that are tailored to community needs, usually with the guidance of teachers, parents, or foundations. Like traditional public schools, charter schools maintain open enrollment and receive public funding, but they largely operate independently of the traditional public school bureaucracy. Plans like these, which allow students to attend a variety of public schools, are usually referred to as "public school choice" programs, and they enjoy fairly broad support among the American public.8
"Private school choice" programs expand the range of educational options even further by offering students (through their parents) tuition subsidies in the form of "vouchers" that can be applied to private and/or parochial schools.9 Here the central distinction among varieties lies in the source of the funding for the vouchers: privately funded vouchers are more common and less controversial, while publicly funded vouchers are relatively rare and extremely controversial. Both types of vouchers give students the opportunity to attend private or parochial schools at little or no cost to their parents; in this sense they are functionally equivalent for the parents. Privately funded vouchers are essentially tuition scholarships, of the sort that civic groups, churches, and philanthropic organizations have been awarding needy students for many decades. Indeed, when private funds are used to pay the parochial school tuition of poor students, the effort is often lauded as philanthropy at its best. What makes the recent surge of privately-funded vouchers controversial-to the extent that they are controversial-is the intent behind the money itself. Whatever else privately funded voucher programs are intended to achieve, they are also explicitly intended to spur public school reform by offering incentives for students to attend private schools.10 The two types of voucher programs (namely, privately-funded and publicly-funded) also fall into completely separate legal categories, since private individuals and institutions can spend their money largely as they see fit, while governmental expenditures are constrained by the First Amendment's religion clauses. Let us forgo the constitutional question for the moment, however, and turn first to the principles-and principal supporters-behind school voucher proposals.
It is often noted that the politics of school voucher proposals has made strange bedfellows, each of whom supports some variation of voucher programs for quite different reasons. Specific voucher proposals vary widely enough that groups often favor one means of implementing vouchers but oppose others. For example, some plans would offer vouchers only to low-income students, while others apply to all students in a particular school, district, or state; some plans would allow voucher students at parochial schools to "opt out" of worship activities, while others would not; some would prohibit students from using vouchers at parochial schools entirely, while others encourage it. Given the multiplicity of both interest groups and voucher proposals, a complete analysis of the principals and principles behind the debate is impossible here. But at the risk of oversimplifying a thoroughly complex political phenomenon, we can group voucher supporters into three primary categories, as advocates for: (1) the free market, (2) the urban poor, and (3) religious education.11
At the front of the line of voucher supporters are those who believe free market economics can improve the quality of education in America. They argue that public schools, which educate 90 percent of American schoolchildren, have monopoly control over the educational system-"This is a monopoly that Bill Gates could envy," writes one observer--and that most parents have no real choice about where their children attend school, since private school tuition is often prohibitively expensive.12 According to free market advocates, vouchers will foster competition-and, ipso facto, improvement-in the public schools by giving parents the option of moving their children out of "under-performing" or "failing" public schools, thus placing those schools at risk of losing substantial numbers of students. That free-market supporters generally favor private school choice (i.e., vouchers) over public school choice (e.g., charter and magnet schools) often indicates their adherence to a widespread belief that the private sector can accomplish most tasks more efficiently and successfully than government.13
Why, then, must the government be involved at all? Why not simply motivate the private sector to sponsor enough vouchers to spur reform, or build enough private schools to lower the tuition costs? The answer lies in the sheer number of American schoolchildren. Despite the rapid growth of privately funded voucher programs, most observers agree that (barring an extraordinary shift in philanthropic priorities) the private sector cannot provide enough vouchers to engineer a truly competitive market in education. The largest provider of school vouchers today is the Children's Scholarship Fund, a $200 million foundation established by billionaires Ted Forstmann and John Walton that provides 40,000 vouchers annually. Despite this large sum of money, the CSF cannot claim to have dramatically altered the competitiveness of a public school system that educates 53 million children.
This realization has led some free market advocates to propose a universal, publicly financed voucher program that would be of such magnitude as to truly spur competition. Economist and Nobel laureate Milton Friedman is generally cited as the first to propose such a voucher system in the-United States. In 1955, he made the case for a "de-nationalized" educational system that would remove government involvement from almost every aspect of primary and secondary education-except its financing.14 He contended (and continues to do so15) that a universal voucher program would create a competitive marketplace in which low-performing schools would close (for want of students) and every school would be driven to improve (for fear of losing students and the voucher money that accompanies them).
Friedman's basic argument retains its appeal today among a large segment of voucher supporters, though they no longer set the terms of the political debate (as we shall see). Perhaps the most influential adaptation of the free market argument for vouchers came in 1990, with the publication of Politics, Markets and America's Schools, by John Chubb, senior fellow at the Brookings Institution (and chief education officer of Edison Schools), and Terry Moe, professor of political science at Stanford University. Chubb and Moe argued that the government's massive education reform efforts in the 197-0s and 1980s failed to solve the problem, precisely because government itself was the problem. Despite the best intentions of everyone involved, the nation's democratic institutions of school governance (e.g. state and local school boards and bureaucracies) served primarily to stifle administrative efficiency and student achievement alike in a thick haze of bureaucratic regulations. The solution, they argued, was to replace the current system of democratic control of schools with an indirect system of control by market forces and parental choice.16 Chubb and Moe are credited with bringing school vouchers (and school choice in general) into the mainstream of public policy circles during the 1990s.17 Many conservative politicians as well as some economists and education policymakers have warmly received their free market arguments. But a majority of voters and legislators, it seems, are not yet convinced that efficiency should be the primary concern of the public school system. "Every time open-ended [i.e., universal] market-based vouchers have been proposed in the states in recent years, they've been defeated, either in the legislatures or, as in major initiative campaigns in Colorado and California, at the ballot box."18 Indeed, even limited voucher programs in Cleveland, Milwaukee, and Florida were established only after the terms of the political debate had shifted from "efficiency" and "market economics" to "equality" and "justice." This shift in focus was brought about by the second group of voucher supporters noted above, namely the advocates for the urban poor.
Whenever politicians, academics, or policymakers pronounce a "crisis in American education," the epicenter of the crisis is usually located in "the inner city," often characterized by desperate poverty, crumbling schools, apathetic or unqualified teachers, and underachieving or even dangerous students. Indeed, by most measures (including standardized test scores, graduation rates, teacher qualifications, classroom size, and schoolyard crime), urban schools fare much worse than their rural and suburban counterparts, and have thus received the most attention from education reformers. Poverty and poor education (like wealth and good education), are highly correlated and are mutually reinforcing conditions in our society. Improvement in one area will likely yield improvement in the other, so it is no surprise that advocates for the urban poor (including, of course, the urban poor themselves) are intensely concerned with the condition of urban education. Exactly how to improve urban education, however, has been a divisive question for decades.
Contemporary media reports often highlight the "conservative" lineage of school voucher proposals, without mentioning the politically liberal scholars and policymakers who supported voucher plans over thirty years ago. In the late 1960s and early 1970s, Christopher Jencks, Theodore Sizer, and other self-proclaimed liberals argued that a limited system of publicly funded vouchers could improve the educational opportunities of the urban poor.19 For a time, educational vouchers were even discussed as a possible addition to Lyndon Johnson's "war on poverty," alongside another voucher program-food stamps.20
But if the free marketeers and the liberal reformers agreed that educational vouchers might be a good idea, their reasons (and thus their specific policy proposals) differed significantly. Whereas the former trusted the free market to improve education for all children, including those in the inner city, the latter were convinced that the unfettered market had done little but harm the urban poor. Why, they asked, would the private marketplace build hundreds of new schools in urban areas shunned by so many other businesses? Critics of Friedman-style voucher proposals also attacked the assumption that education was an "industry" that could be "deregulated" in the same way as banks or airlines. Common schools, they further argued, provide public goods like social cohesion, racial and ethnic integration, and civic education that "the market" consistently undervalued. In short, the free market proposals were characterized as cold-hearted toward the poor and fundamentally misguided. These critiques resonated with enough voters and legislators that school voucher proposals largely dropped from public view until Chubb and Moe's controversial book revived the debate in the early 1990s.
This renewed debate over school voucher proposals was taken up in the midst of a broad national discussion about school choice. As studies of urban schools continued to report bad news, innovative programs like magnet and charter schools caught the public's imagination, as did untested alternatives like vouchers. Having learned a political lesson decades earlier, free market advocates (or at least their legislative representatives) largely conceded that voucher plans should be targeted at disadvantaged and underserved populations, in the name of equity as well as efficiency.21 Despite firm opposition from the leadership of groups traditionally supportive of liberal politics-like the National Association for the Advancement of Colored People, the Urban League, and the National Education Association, all of which argue that vouchers undermine public schools by diverting scarce resources-an uneasy alliance has developed in recent years between voucher supporters on the left and the right. Many in the African-American community believe publicly funded vouchers for low-income students could serve as a "lifeboat" to lift them out of the stormy urban public schools and into calmer, safer private schools. These claims-that vouchers could "save" children otherwise doomed to poverty and violence, and that justice demands that such inequities be resolved without delay-carry moral weight and broad social appeal that "market efficiency" simply does not. As a result, the coalition of voucher supporters is slowly but consistently expanding.22
The third primary group of voucher supporters consists of those who believe that a good education is-in part, at least-a religious education, and that public schools have become increasingly hostile environments for religious expression and religious values. They point to a string of Supreme Court rulings since the 1960s that have outlawed mandatory- prayer, Bible-reading, and other religious devotions in public schools, and argue that parents should have the option to send their children to schools that respect their religious beliefs, even if that means public financing for parochial school tuition.23 This group of voucher supporters is often characterized as "religious conservatives," and indeed it includes self-described "conservative" organizations such as the Union of Orthodox Jewish Congregations of America, the National Association of Evangelicals, the Christian Coalition, and Focus on the Family. But school voucher plans are also supported in principle-though not always in practice-by the United States Catholic Conference (USCC), which defies easy characterization as "conservative" or "liberal." As William F. Davis, assistant secretary for Catholic Schools and Public Policy at the USCC, recently noted,
Not every parental choice proposal and every voucher program is a good one. School-choice programs need to give priority to those families with low and middle incomes. They need to respect the integrity and identity of private and religious schools, especially their mission, policies, and practices. These programs need to recognize the existing applicable civil rights laws, the need for accountability and adequate education about the available educational options so that parents can make appropriate and informed decisions. As the saying goes, "the devil is in the details," and Catholic educators reserve the right to review each federal, state, and local proposal to determine if it is acceptable and appropriate for our schools.24
While Protestant evangelical organizations sometimes differ with the USCC on the virtues of particular voucher proposals, they agree that any such program must allow students to apply the vouchers to both parochial and non-sectarian schools alike. Excluding parochial schools from voucher programs, they say, would represent an unconstitutional restriction on the free exercise of religion. Many opponents of voucher programs take exactly the opposite view, that publicly financed vouchers for students to attend religious schools constitute an establishment of religion expressly forbidden by the First Amendment. The constitutional battles over this issue continue to rage in state and federal courts and legislatures, with limited victories (and defeats) to date for both sides. If the unusual coalitions of voucher supporters and opponents can agree on one thing in this debate, it is that the United States Supreme Court must ultimately resolve the question-and the sooner the better. The next section, therefore, takes up the relevant constitutional questions through a careful examination of First Amendment jurisprudence.
SCHOOL VOUCHERS AND THE FIRST AMENDMENT
The United States Constitution protects religious freedom in this country primarily through two pithy clauses in its First Amendment: "Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof. . . ." Michael Perry has aptly described these clauses as "anti-discrimination" provisions that, together, institutionalize the American conception of religious freedom by prohibiting the government from discriminating on the basis of religious belief or practice.25 On their face, these provisions apply only to the federal Congress, but over the course of the nineteenth and twentieth centuries, they have been interpreted by the Supreme Court to apply to the executive, judicial, and legislative branches of all levels of government.26 Today the extension and "incorporation" of the First Amendment's religion clauses is rarely challenged, and the basic principles they codify-that the government cannot legally establish a religion or prevent citizens from the free exercise of their religious beliefs-are firmly embedded in the American political and cultural consciousness. Though there is widespread agreement in the United States about the anti-discriminatory nature of the religion clauses, there are vast and resilient differences of opinion as to whether they are more than anti-discrimination provisions, and what, exactly, they enjoin the government to do (and not do). These differences are especially profound in relation to private school vouchers, in part because the Supreme Court has yet to offer definitive guidance on the subject. In this section, therefore, I aim to clarify the constitutional questions presented by private school vouchers by examining the First Amendment's religion clauses and the relevant Supreme Court decisions that have guided their interpretation. In short, I will argue that the First Amendment neither requires nor prohibits private school voucher programs (provided they meet several important conditions), and thus the wisdom of their implementation must rest upon other considerations, which we shall take up in the final section of this essay.
While the First Amendment's religion clauses are closely related and inextricably joined, they nevertheless are distinct conceptions. Indeed, they are in constant tension with one another, and an expansive interpretation of one clause invariably requires a restrained interpretation of the other.27 For its part, the Free Exercise Clause forbids the government to engage in prohibitory action that disfavors religious practice as such (i.e., because it is religious). This means that the state cannot prohibit actions (e.g., worship, religious education, public expression of religious belief, proselytizing, or even a way of life guided by religious beliefs) simply because legislators or voters believe the actions exemplify or are motivated by false theological doctrine, misguided notions of "the good life," or the like.28 This is not to say, of course, that any action undertaken for religious reasons is thereby exempt from regulation; the constitutional guarantee of free exercise is not absolute, and the safety, order, and health of the general public provides legitimate reasons to regulate all sorts of conduct. In an extreme case like human sacrifice, there is general agreement that the state's interest in maintaining public safety and the due process of law outweighs the harm accrued to the would-be murderer (who is thereby restricted from fully "practicing" his faith). The difficulty, rather, lies in striking a balance, in determining the exact points at which the state's various interests are sufficiently compelling to warrant restriction of a citizen's free exercise of religion, and when they must yield to those free exercise rights.
So how would an expansive interpretation of the Free Exercise Clause impinge upon the Establishment Clause? Two situations will perhaps serve to illustrate the point. First, a generous interpretation of free exercise rights can tend toward establishment when a single religious person or group is granted an exemption from a generally applicable law. Such was the case in Wisconsin v. Yoder, creating a situation in which the government appeared to favor one religion over others.29 Second, an expansive view of free exercise can create tension with the Establishment Clause if the former is construed as imposing a positive obligation on the state to provide space for the free exercise of religion. Here we may return to the issue of school vouchers for an example. A plausible case can be made that public education burdens the free exercise of religion when it compels students to act in ways their religion condemns (e.g., by wearing "immodest" attire for physical education classes) or when it inculcates values or lessons abhorrent to their faith (e.g. sex education or biological evolution). To be sure, parents have the right to direct their children's education by sending them to parochial schools,30 but many poor families are unable to exercise this right for purely financial reasons. "There are precedents," argues William Nord, "for claiming that government has a special obligation to provide the funds that enable poor people to act on fundamental rights."31 Indeed, the government must provide attorneys for indigent defendants in criminal trials, and it is commonly (if controversially) argued that the right to privacy established in Roe v. Wade (1973) cannot be said to extend to all citizens unless the government provides abortions for poor women who could not otherwise afford them. In like fashion, some argue, the government has the obligation to ensure that poor people can exercise their religious freedom by sending their children to parochial schools at the government's expense. The tension with the Establishment Clause is apparent in this case: government responsibility for the religious indoctrination of children is problematic at best, and tantamount to religious establishment at worst.
The argument that private school vouchers are required on free exercise grounds, however, falters on several points. The First Amendment's guarantee of "free exercise" of religion is not a guarantee of absolute freedom; as we have seen, the state has a variety of interests that, at times, may burden or restrict a citizen's religious practices or beliefs. The key questions are whether these burdens are warranted, and whether they are the least intrusive means of pursuing the state's interests.32 In the case of public education, the state's interest in maintaining an environment of neutrality toward religion and non-religion-- an environment intended to protect students from government-sponsored indoctrination-may very well create a burden for those who wish to promote, defend, or express their faith in the classroom. But as repugnant as this may be for some, this stance is neither unreasonable nor unacceptably intrusive, even if it seems anti-democratic at times.33 Furthermore, while it is true that the state provides attorneys to indigent defendants in order to uphold their constitutional right to equal protection under the law, providing legal representation cannot be equated with providing religious education. In the first place, criminal defendants have no alternative venue in which to exercise their rights to equal protection and due process. Second, unlike religious indoctrination through education, there are no constitutional proscriptions on the state's provision of legal representation. More generally, though, the existence of a right enumerated in the Bill of Rights does not require state funding for those who cannot afford to exercise that right in the manner they favor. The government has no more obligation to provide religious education in light of the First Amendment than it has an obligation to arm the nation in light of the Second Amendment. Indeed, in the school voucher debate, the Free Exercise Clause is something of a red herring; it neither proscribes nor prescribes the use of publicly funded vouchers for religious education. The thorny constitutional questions about school vouchers arise instead primarily from the Establishment Clause.
There is little risk in the foreseeable future that federal or state governments will officially and explicitly "establish" one sect or religion as the "national religion," but there are myriad subtle ways in which the state supports religious groups or practices, both directly and indirectly. Churches are exempted from income taxation; the government employs clergy in prisons, the military, and both chambers of Congress; and government funds are given to religiously affiliated hospitals, universities, primary and secondary schools, and social service organizations. The key to assessing whether a particular practice represents an unconstitutional establishment of religion is not whether a religious person or group receives some financial benefit from the government, but whether that benefit is given (or withheld) because the person or group is religious.34 To put it in Michael Perry's terms, the Establishment Clause forbids government from engaging in prohibitory action, whether direct or indirect, that favors one or more religions as such: "No matter how much some persons might prefer one or more religions, government may not take any action based on the view that the preferred religion or religions are, as religion, better along one or more dimension of value than one or more other religions or than no religion at all."35
Perry's characterization of the Establishment Clause as an "antidiscrimination" provision is appropriate, though the nature and extent of the prohibited "discrimination" is widely disputed. A "separationist" interpretation of the Establishment Clause usually draws upon Jeffersonian language to contend that a "wall of separation" ought to be constructed between the institutions of church and state, thereby committing the state to a thoroughly "secular" or non-religious stance. Separationists argue that state support or encouragement of religion amounts to discrimination against non-religion, or discrimination in favor of one or more religions, which are generally Judeo-Christian. Diametrically opposed to this view is the "accommodationist" interpretation, which contends that such strict 11 separation" amounts to discrimination against religion as such, in favor of secularism, and that in fact the Establishment Clause creates a positive obligation to accommodate religion as such, even if it discriminates against secularism. These competing interpretations serve to illustrate rather clearly the tension between the Free Exercise and Establishment Clauses, with accommodationists emphasizing the importance (and perhaps preeminence) of the former, and separationists emphasizing the importance (and perhaps preeminence) of the latter.
Adherents of these positions in the contemporary school voucher debate can find ample precedent in constitutional case law to support their arguments for and against vouchers. On numerous occasions in the last hundred years, the Supreme Court has considered the government's proper relation to religious education, with decidedly mixed results for -all involved. Before 1971, the Court generally took an accommodationist stance toward the state's involvement with and regulation of religious schools, in the sense that it protected religious schools from excessive government interference and allowed direct aid to flow to such schools under certain circumstances.36 By applying the Establishment Clause to the states in Everson v. Board of Education (1947), the Court completed a procession of decisions that incorporated the entire First Amendment into the due process provisions of the Fourteenth Amendment.37 The Everson opinion is perhaps best known as the quintessential judicial statement of the separationist position. Writing for the Court, Justice Black summarized the position in memorable terms:
The "establishment of religion" clause of the First Amendment means at least this: Neither a state nor the Federal Government can set up a church. Neither can pass laws which aid one religion, aid all religions, or prefer one religion over another.... No tax in any amount, large or small, can be levied to support any religious activities or institutions, whatever they may be called, or whatever form they may adopt to teach or practice religion. Neither a state nor the Federal Government can, openly or secretly, participate in the affairs of any religious organizations or groups and vice versa. In the words of Jefferson, the clause against establishment of religion by law was intended to erect "a wall of separation between Church and State."38
This application of Jefferson's famous remark, which also appeared in Reynolds v. United States (1879), has been criticized from both the Supreme Court bench and the academy as lacking historical merit, but its influence on the American public's perception of establishment law has been profound nonetheless.39
Often lost in the controversy over Everson's separationist language, however, is the fact that the decision upheld a state program that reimbursed parents who sent their children to school on buses operated by the public transportation system. It also affirmed the ability of states to provide publicly funded transportation to religious and public schools alike. In affirming the New Jersey law in question, the Court established a crucial distinction between direct government aid to religious schools (which was prohibited) and indirect aid given to parents to use according to their own choice (which was allowed). Services like transportation and police protection that are separated from the religious function of parochial schools ought not be denied such schools, wrote Justice Black. Indeed, the First Amendment "requires the state to be neutral in its relations with groups of religious believers and non-believers; it does not require the state to be their adversary. State power is no more to be used so as to handicap religions, than it is to favor them."40 In 1968, the Court extended this logic to allow the State of New York to lend textbooks (on "secular subjects") to all students, ineluding those at religious schools.41
The Court's views on indirect aid to religious schools took a decisive turn in 1971 with the adoption of a new legal standard of review for establishment cases in Lemon v. Kurtzman. Using criteria drawn from several recent opinions, the Court articulated a three-pronged test for laws dealing with religious establishment. To be constitutional, a statute must have "a secular legislative purpose," its "primary effect" must neither advance nor inhibit religion, and it must not foster "an excessive government entanglement with religion." The Lenton test, as it came to be known, was used in the present case to strike down several state laws that supplemented the salaries of teachers in religious schools; the "cumulative effect" of such programs, the Court held, was an excessive entanglement of government and religion.42
Several additional cases in the next two years applied the Lemon test to further tighten restrictions on government aid to religious schools. " Among these, Committee for Public Education v. Nyquist (1973) stands out as particularly important, for (among other reasons) it was cited by a federal appeals court as controlling legal precedent when it struck down Cleveland's school voucher program in December 2000.44 In short, the Nyquist Court held that indirect distribution of government funds to private school parents did not, by itself, make the law constitutional, since the benefits accrued only to parents of children in nonpublic schools. The state must maintain, rather, "an attitude of `neutrality,' neither 'advancing' nor 'inhibiting' religion, and it cannot, by designing a program to promote the free exercise of religion, erode the limitations of the Establishment Clause."45
That statement perhaps best summarizes the goals of the Lemon test, which was designed to serve as a middle ground between the separationist and accommodationist positions.46 Locating this middle ground while enforcing "neutrality,"" however, proved to be more difficult than expected. By 1985, a series of increasingly tortuous applications of the Lemon test had led to a nearly incomprehensible legal environment for educators and lawyers alike. In Aguilar v. Felton (1985), an exasperated Justice Rehnquist condemned the Court's overzealous application of the "excessive entanglement" prong of the Lemon test, suggesting in dissent that "the Court takes advantage of the `Catch-22' paradox of its own creation, whereby aid must be supervised to ensure no entanglement but the supervision itself is held to cause an entanglement."47 Rehnquist's assessment of-if not his solution for-- the Court's rulings on governmental aid to religious schools was widely shared by legal scholars. John Witte has called Establishment Clause case law "a bizarre byzantine code"; Michael Perry suggests it is "an unholy mess"; and Leonard Levy marveled in 1986 that "the Court has managed to unite those who stand at polar opposites on the results that the Court reaches; a strict separationist and a zealous accommodationist are likely to agree that the Court would not recognize establishment of religion if it took life and bit the Justices."48
Although the Court's Establishment Clause rulings remain confused (and confusing), the last twenty years has brought a shift away from the strict separation of church and state demanded by vigorous application of the Lesion test. Instead, the justices seem to be seeking a balance between the Free Exercise and Establishment Clauses that maintains state "neutrality" toward religion and non-religion while expanding protection for individual (i.e., parental) decisions about religious education. The explicit prohibition (per Lemon and Nyquist) of direct aid that primarily benefits religious schools remains "good law," but various types of indirect aid have been upheld in a number of recent cases.49
It appears that the "modified Lemon test," first articulated in Agostini v. Felton (1997),50 will continue to guide the Court's current decisions on government aid to religious schools, but whether private school voucher programs in particular will clear the lowered constitutional hurdle remains an open question for the present. The Court has yet to rule explicitly on voucher programs51 and its most recent decision on aid to religious schools, Mitchell v. Helms (2000), revealed fractures among the justices on the question. The six-member majority of the Mitchell Court held that a federal program providing funds for educational materials and equipment used for "secular, neutral, and nonideological" programs in public and private elementary and secondary schools is "not a law respecting an establishment of religion" simply because many of the private schools receiving the aid are religiously affiliated.52 The law was challenged in Louisiana's Jefferson Parish, where 30 percent of such federal funds were allocated to private schools, most of which are Catholic.
Though the decision upheld a law providing federal aid only for educational materials and equipment, a plurality (led by Justice Thomas, joined by Rehnquist, Scalia and Kennedy) took the opportunity to write a broad opinion reaffirming Agostini's position that governmental aid to private schools is neutral to religion and non-religion so long as that aid "results from the genuinely independent and private choices of individual parents." When parents choose to send their children to private or religious schools-bearing aid "allotted on the basis of neutral, secular criteria that neither favor nor disfavor religion, and is made available to both religious and secular beneficiaries on a nondiscriminatory basis"--the religious indoctrination that occurs cannot reasonably be attributed to governmental action.53 In a concurring opinion, Justices O'Connor and Breyer voted to uphold the law under review, but declined to join Thomas et al. in their broader opinion; they instead drew a distinction between direct aid to schools for materials, and vouchers that are used for religious as well as secular studies. Justices Souter, Stevens, and Ginsburg dissented, arguing that the plurality opinion "espouses a new conception of neutrality. . .[that] breaks fundamentally with Establishment Clause principle, and with the methodology painstakingly worked out in support of it."54
On 20 February 2002, this divided Court heard oral arguments in Simmons-Harris v. Zelman, a case from the sixth federal circuit that struck down a nationally-watched voucher program in Cleveland on First Amendment grounds. The Cleveland Scholarship and Tutoring Program, initiated in 1995, allows roughly 3,800 students to use vouchers worth up to $2,250 to attend the public, private, or parochial school of their choice. No public schools outside the district are currently participating in the voucher program, however, and the 46 parochial schools that participate (out of 56 total participating nonpublic schools) enroll 96 percent of the program's voucher students.,55 Affirming the district court's ruling, the Sixth Circuit held in December 2000 that this concentration of voucher students in religious schools was an impermissible effect of a program that offered financial disincentives for public and private nonsectarian schools to take voucher students.56 "Unlike [the aid programs approved in] Mitchell, Agostini, Witters and Mueller," wrote the Court, "the [Cleveland] scholarship program is designed in a manner calculated to attract religious institutions and chooses the beneficiaries of aid by non-neutral criteria."57 The opinion also contained several lines of reasoning that may have prompted the Supreme Court to review the case. First, it took Justice O'Connor's concurring opinion in Mitchell as controlling precedent, on the grounds that the narrowest opinion must be followed in a plurality decision. It then used O'Connor's argument regarding "factual similarity" of cases to determine that Nyquist was the most applicable precedent. Finally, it did all this with a tip of the hat to the Supreme Court, quoting Agostini's reminder that that Court alone retains "the prerogative of overruling its own decisions."58
It is perhaps needlessly speculative to predict how such a divided Supreme Court will rule on Zelman v. Simmons-Harris.59 But regardless of this particular case's merits, the Rehnquist Court's recent decisions demonstrate an increasing willingness to allow the government to fund certain activities, including education, conducted by religious groups. Private school vouchers are distinct from other government aid to religious schools only as a matter of degree, so it is certainly plausible that they will, in some form, stand up to constitutional scrutiny.60
What, then, might a voucher program that passes constitutional muster look like? To begin with, it would be established in order to improve the education of all students, or to improve the educational opportunities for disadvantaged students;61 these, and possibly others, are what the Lemon test calls "legitimate, plausible secular purpose[s]." (Despite substantial revisions in recent years, the Lenum test remains the guiding framework for Establishment Clause jurisprudence.) If the program were intended to promote or hinder one or more religions (or non-religion), it would be struck down as unconstitutional, whether or not it actually did so in practice.
A constitutional voucher program would also take extraordinary precautions to avoid discrimination on the basis of religion or non-religion, in any of its criteria-for determining which schools may participate in the program, which students receive vouchers, or for any other decision. This means that any school (whether its teachings promote Muslim, Christian, Secular Humanist, nonsectarian, or even anti-- Muslim, anti-Christian, etc. perspectives) must be allowed to participate in the voucher program on an equal basis. Students must be given as wide a choice as possible of schools at which they may apply their voucher, since, for example, a Muslim student may have little practical use for a voucher redeemable only at Christian schools. That being said, it is impossible to say exactly what constitutes "enough" choice to pass constitutional muster. It hardly seems reasonable to strike down a voucher program simply because there are not any participating schools of a particular kind (Mormon, Lutheran, Montessori, etc.), yet it is manifestly clear that some choice must be available to students, and that the voucher program must accept new schools on an equal basis that open to fill the demand.
Our hypothetical program would also provide vouchers directly to parents, who would then redeem them at the school of their choice. Recent Supreme Court opinions have indicated that "indirect aid" may avoid judicial criticism as governmental involvement in "religious indoctrination" (which is of course forbidden) when this indoctrination occurs "only as a result of the genuinely independent and private choices of individuals."62 A persuasive analogy can be made between the indirect aid provided to religious schools via school vouchers, and indirect aid provided to religious universities via the G.I. Bill and Pell Grant programs, or the indirect aid provided to religious hospitals via Medicaid. In all these cases, money that was once in the hands of the government ends up, through private choices, being put to a religious use. The Rehnquist Court has often looked at such private choices as an aspect of the liberty of conscience, and will plausibly continue to do so when a voucher program is scrutinized.
Finally, a voucher program that passes constitutional muster will have to take care to avoid what the Lemon test calls "excessive entanglement" between church and state. The decision as to what constitutes excessive entanglement is, ultimately, arbitrary, though the Rehnquist Court has indicated its dissatisfaction with overly broad interpretations of the phrase.63 Some entanglement would clearly be necessary. Private schools that accept voucher students would certainly be required to meet basic health and safety regulations, but the state may require compliance with other policies as well, including open enrollment (to prevent discrimination in race, socioeconomic status, religion, etc.), limits on the number of voucher students per school (to prevent excessive reliance on government funds), reporting of educational statistics, and some basic curricular requirements.64 The state's entanglement with the church would be "excessive," however, if it required religious schools to adopt an "opt-out" policy wherein voucher students could avoid the religious "indoctrination" at mandatory worship services. Such a policy would be both practically impossible (because indoctrination occurs in the classroom as well as the chapel) and constitutionally unsound (since the indoctrination would actually be voluntary, in the sense that the parent or child chose to attend the school in the first place).
A voucher program with these characteristics likely could sustain constitutional scrutiny. The program would have to be conceived and implemented with extraordinary attention to the effects it has (or might have) on religious beliefs and institutions, but it can be accomplished within the boundaries of the law. The more important question, however, is whether it should be accomplished. By my understanding, the Establishment Clause does not prohibit school voucher programs that include religious schools-indeed, it would be unconstitutional to exclude them. Nor does the Free Exercise Clause require the state to institute school voucher programs to ensure religious freedom. There is, however, a vast territory between can and ought. The fact that a government program is or can be constitutional does not make it a good public policy (let alone a moral imperative), any more than that the program is unconstitutional makes it a bad public policy (or a moral hazard). As this essay's final section argues, whether we ought to establish school voucher programs depends on why we might want them, what we hope they will accomplish, and how they fit into the aims of education in a liberal democratic society.
SCHOOL VOUCHERS AND DEMOCRATIC EDUCATION
In examining the political and constitutional arguments for publicly financed private education, this essay has thus far largely avoided the prior (and often unspoken) question: why bother with public education at all? What value does society glean from a system of education that is "public" rather than "private"? Asking such questions reframes the debate as part of a larger inquiry into the nature of education in a democratic society, and helps to clarify what might be lost or gained when private school vouchers are implemented.
The proper aims of education in a political community are an endless source of debate for political and educational theorists; for, as Immanuel Kant noted, "Two human inventions . . . may be considered more difficult than any others-the art of government and the art of education; and people still contend as to their very meaning."65 Contention as to the "meaning" of education (including what and how we should learn, who should teach it, and why) has not precluded, however, widespread agreement in the post-Enlightenment West about the value of education as such-i.e., the intrinsic value of learning about ourselves and the world around (and beyond) us.66 In the contemporary American context, this intrinsic value of education is understood to correspond to a wide variety of extrinsic values as well. For example, numerous statistical analyses in the United States over the last fifty years have demonstrated that high levels of education are tightly correlated with (though not necessarily the cause of) increased levels of political and civic participation (including voting), higher-paying and more stable employment, and even good health.67 As an influential "blue ribbon" report entitled "A Nation at Risk" stated in 1983,
The people of the United States need to know that individuals in our society who do not possess the levels of skill, literacy, and training essential to this new era will be effectively disenfranchised, not simply from the material rewards that accompany competent performance, but also from the chance to participate fully in our national life. A high level of shared education is essential to a free, democratic society and to the fostering of a common culture, especially in a country that prides itself on pluralism and individual freedom.68
The socio-political political ramifications of a "shared education" have not been lost on political theorists through the ages, whether or not they advocate a "free, democratic society." Indeed, it is a virtual axiom of political theory that the education of children-future citizens, as it were-is central to the perpetuation and stability of a society or regime. Children must be inculcated with a society's values, initiated into its traditions, and educated about its institutions if the state is to persist in its existing form. Normative theories of civic education are thus closely tied to political theories, and both are invariably tied to some conception of human nature. Plato, for example, would have the children of his Republic taken from their parents at birth, segregated by natural ability, and inculcated with "the noble lie" of their metallic bloodline, in the service of creating a self-sustaining and unified polis wherein each member rightly performs his (sole) function in ordered harmony.69 Though Aristotle condemned Plato's drive toward absolute unity,70 he nevertheless agreed that the singular importance of "mold[ing] youth to suit the form of government under which he lives" required a system of public education. "Since the whole city has one end," he wrote in The Politics, "it is manifest that education should be one and the same for all, and that it should be public, and not private. ; the training in things which are of common interest should be the same for all."71 In such a scheme, the state inculcates in every child an understanding of the good life appropriate to him.
Opposed to political and educational theories that see public education as vital to a state's interest are those that grant exclusive educational authority to parents. The latter type of theory typically justifies parental control of education on the basis of either the perceived beneficial consequences for children, or the rights and liberties of parents.72 The consequentialist argument claims that parents ought to retain exclusive educational authority over their children because only they are naturally inclined to protect and nurture their children's best interests. Other theorists and theologians, drawing upon Thomas Aquinas, argue that this educational authority is a natural right of parenthood.73 Modern liberals who support parental authority in education often look to John Locke, who claimed that the right of every adult to be free from the arbitrary will of another extends to parents' educational decisions for their children. The state may not intrude on the educational realm, Locke wrote, since political power and paternal power are "perfectly distinct and separate" and given to different ends.74
Jean-Jacques Rousseau well understood Locke's distinction between paternal and political interests in educating a society's children. He saw an inherent contradiction between a family-based "natural education," which seeks to fulfill the child's true nature by training him to be independent yet cosmopolitan (i.e., identifying with humanity as a whole), and a state-based civic education, which trains children to become responsible citizens through interdependence and loyalty to a particular state. In Rousseau's famous phrase, "Forced to combat either nature or social institutions, you must choose between making a man and making a citizen, for you cannot do both at the same time." In the case of mile's education, at least, Rousseau chose to make the man before he made the citizen, lest the state corrupt the young man's natural virtue.75
We need not accept any of the aforementioned theorists' understandings of human nature to appreciate the power of their educational theories in our present context. "We the people" have a critical interest in the education of our society's children, and we must therefore ask whether the interests of our democratic society are sufficiently served when the state's role in education is minimized or completely excised. Amy Gutmann, I believe, has persuasively answered this question in the negative; she argues that education in a liberal democracy must balance the interests of the state, the parents, and the children by restricting the freedom of all three: "Neither parents nor a centralized state have a right to exclusive authority over the education of children. Because children are members of both families and states, the educational authority of parents and of polities has to be partial to be justified."76 Restrictions on parental authority are sometimes decried as anathema to Lockean-style liberal values, but it is nevertheless the case that "we the people" have deemed it necessary to enact laws that protect children from their parents at times.77 Furthermore, Locke himself did not envision the possibility of democratically governed schools. From the standpoint of a pluralistic, democratic society, it is not that parents cannot be trusted to educate their children in the values and interests they understand to be right, but rather that the values and interests of particular families are not necessarily the values and interests of the society as a whole. The democratic process of deliberation and disagreement over educational values and institutions not only serves to reconcile these divergent interests, it also exemplifies the democratic values themselves. How we deliberate about education, in other words, is itself an important part of democratic education, one that a fully privatized system neglects at society's peril. As Jeffrey Henig has written, "The real danger in the market-based proposals for choice is not that they might allow some students to attend privately-- run schools at public expense, but that they will erode the public forums in which decisions with societal consequences can democratically be resolved."78 Maintaining a strong system of public education, governed by democratic institutions, upholds this tradition of democratic deliberation.
If we agree that the American educational system must reflect democratic values in order to stabilize and perpetuate our political culture, the question becomes which democratic values our educational system ought to reflect. Liberal democracy is predicated on a balance between freedom and equality, and this balance is as difficult to strike in education as it is in other realms. We have already seen how the First Amendment's religion clauses exemplify the tension between freedom and equality when the free exercise of religion jeopardizes equal treatment of religions and non-religion (or vice versa). I have argued that while the Free Exercise Clause clearly protects the right of parents to educate their children in private schools, it does not create a requirement that the state finance private education, in whole or in part. So if the freedom at stake in the school voucher debate is the freedom to exercise one's religion (or non-religion) by attending a private school at public expense, what are the demands of equality in this context?
At minimum, equality in the American political tradition means that every person is entitled to the same basic respect and dignity (since, as the Declaration of Independence claims, "all men are created equal"), as well as the equal protection of the laws (stipulated by the Fourteenth Amendment). This does not mean, nor should it mean, that every person is in all respects the same (in which case we would be identical, not equal), or that, in the context of education, absolute equality of outcomes is to be enforced across the nation.79 Real differences in persons and places demand different approaches to education, and even a modicum of respect for liberty demands recognition that equal outcomes are neither possible nor desirable in educational contexts. But we need not enforce absolute equality of outcomes in order to guarantee a minimum level of education for every student. Determining what constitutes a minimally "good" education is a matter for democratic deliberation, but it would surely include basic levels of literacy, numeracy, and civic education such that children will eventually be able to participate intelligently as adults in the political and economic processes that shape their society. They will also reap the extrinsic and intrinsic rewards that education carries with it in our society. We may call the guarantee of such minimal standards "equality of educational opportunity."
While equality of educational opportunity is hardly a controversial goal, it has nevertheless been an elusive one to achieve. Decades of scholarly research confirm what is apparent to even the casual observer: America's public schools educate some students (mostly suburban and white) very well, and others (mostly urban and black or Latino) quite poorly. This kind of inequality not only deprives millions of children of the intrinsic and extrinsic rewards of a good education, it also provides what is in effect a very bad civic education. Children learn about injustice and inequality of opportunity in a very concrete way when they attend a school with dilapidated facilities, few college-- preparatory classes, and little funding for extracurricular activities, while another public school just a few miles away enjoys beautiful facilities and an abundance of honors classes and extracurricular clubs and teams. One need not agree with the many reports and studies (most influential among them "A Nation at Risk") labeling the situation a "crisis" to recognize that such inequality has a profound influence on our shared political life.80
If, as I have argued, the public school system is an essential institution of civic education, in part because it serves as a forum for democratic deliberation, and equality of educational opportunity ought to be the primary concern of that system, where does this leave school voucher programs? First, we must reject in principle a universal, market-based voucher system (like those proposed by Friedman or Chubb and Moe), since the lack of democratic oversight that is its hallmark is also its greatest flaw. Even if we set aside the objection that such a system fails to reflect or serve the interests of a democratic polity, there is little reason to accept (and many reasons to doubt) Chubb and Moe's claim that universal school choice "is a panacea ... [that] has the capacity all by itself to bring about the kind of transformation that, for years, reformers have been seeking to engineer in myriad other ways."81 We ought to be wary of anyone who suggests that a single radical reform can easily and finally solve the nation's educational problems, regardless of how seductive the proposal or endemic the ills. Rather than searching for a single silver bullet, we should commit ourselves to a decentralized democratic process of educational governance by parents, educators and concerned citizens at the local level.82
This process may very well lead to the conclusion that some form of school choice ought to be among the many tools a school district has at its disposal to employ in certain situations. Public school choice has already been adopted by a great number of school districts (though the actual number of students attending charter and magnet schools remains relatively low), and in some places it offers many of the benefits of private school choice-e.g., fewer bureaucratic regulations that can hamper innovation, greater responsiveness to community needs, and more efficient budgeting-without jettisoning the democratic governance that justifies its public financing. Private school choice, however, ought to be considered a last resort, undertaken only following an admission that the public school district simply cannot provide a minimally good education to its charges, and that this situation is likely to persist in the foreseeable future. In this context, the extraordinary (and temporary-) measure of sending public schoolchildren to private schools may be an appropriate public policy, under certain conditions. First, the voucher program should focus its resources on the financially neediest students and families in the affected district, since they are least able to pay for a private education on their own; such means testing is presently a hallmark of the Cleveland and Milwaukee voucher programs. Second, participating private schools must be required to conform to state regulations for the health and safety of children, as well as minimal curricular requirements that would include the administration of statewide tests to measure student performance. Third, the voucher program must allow religious and nonreligious private schools to participate on an equal basis, and they in turn must admit all voucher students on an equal basis; any other arrangement would be unconstitutional. We nonetheless ought to remember that the problem vouchers are intended to solve is not the lack of religious schools to choose from, but rather the lack of good schools.
This leads to the final condition, that the voucher program be continually assessed to determine its impact on all parties: the children who receive them, the children who do not, the private schools that accept them and the public schools that lose students because of them. A voucher program ought be maintained only if it can improve the quality of public education (for those not receiving vouchers) while providing enhanced educational opportunities for the most disadvantaged students. This is in fact the central claim of many voucher advocates, and coincidentally the most legitimate reason to try a voucher program. It is certainly possible that private school vouchers can positively influence public schools, but scholars and policymakers are a long way from documenting any such results. Researchers from various interest groups, universities, and governmental agencies have conducted several studies of the Milwaukee and Cleveland voucher programs, but there is nothing even approaching a consensus interpretation of the educational effects across the system.83
The one thing the present data does make clear is that the extreme claims of voucher evangelists and Cassandras alike were misplaced; limited voucher programs have been neither panacea nor plague for these cities' public school systems. The jury, as they say, is still out on the empirical question, so we should pay close attention to the ongoing research in this area, recognizing that vouchers may very well be a psychological and financial distraction from other, more effective reforms. Perhaps, however, researchers will come to the conclusion that a particular voucher program has, for whatever reason, actually improved public education to the point that it once again meets the minimal educational standards every student needs and deserves. If we arrive at that point, citizens and policymakers alike can congratulate themselves for accomplishing an important goal, and then phase out the voucher program and enjoy our improved, democratically governed public schools.
1. Editor's Note: This essay was submitted before the U.S. Supreme Court upheld the Cleveland, Ohio school voucher program to which the author makes reference. The 5-4 decision, Zelman v. Simmons-Harris, was delivered on 27 June 2002.
2. National Center for Education Statistics, Digest of Education Statistics 2000 (Washington, D.C.: United States Department of Education), 12, Table 3.
3. Thomas Good and Jennifer Braden, The Great School Debate: Choice, Vouchers, and Charters (Mahwah, N.J.: Erlbaum, 2000), 97.
4. Office of Educational Research and Improvement, The State of Charter Schools 2000: Fourth-Year Report (Washington, D.C.: ti.S. Department of Education), 10-11.
5. Jodi Wilgoren, "Florida's Vouchers a Spur to 2 Schools Left Behind," New York Times, 14 March 2000.
6. The vouchers were offered by the Children's Scholarship Fund. Evan Thomas and Lynette Clemetson, "A New War Over Vouchers: Poor Parents Want Them, But Civil-Rights Leaders Are Split," Newsweek, 22 November 1999.
7. Wilgoren, 'Florida's Vouchers," A17.
8. One recent survey indicates that 75 percent of Americans support public school choice, and that this support climbs to 88 percent among inner-city parents. Terry M. Moe, Schools, Vouchers and the American Public (Washington, D.C.: Brookings Institution Press, 2001), 334. Support for particular school choice programs can be much lower, however, when evidence of mismanagement and poor oversight are reported. See Dennis Willard and Doug Oplinger, "Whose Choice?" Akron Beacon Journal, 12-15 December 1999.
9. Public school choice programs sometimes offer students "vouchers" to pay for education in public schools outside their district, but the term "vouchers" is more commonly used in reference to private and parochial schools. It should be noted that while supporters of private school vouchers now often call them "scholarships" instead of "vouchers" (in an attempt to avoid voters' negative perceptions about the latter term), this essay follows the more established usage of "voucher."
10. For an excellent review of the stated goals of a variety of recent private voucher initiatives, see Joseph Viteritti, Choosing Equality: School Choice, the Constitution, and Civil Society (Washington, D.C.: Brookings Institution Press, 1999), 92-98.
11. There are, of course, many people who support vouchers for none of the reasons noted here. For example, parents of schoolchildren in rural areas sometimes support vouchers as a means of avoiding the long commutes that often accompany consolidation of school districts. The three groups noted above, however, represent the most prominent supporters in the contemporary voucher debate.
12. Tim De Roche, "Why Vouchers?" Education Week, 6 June 2001.
13. See, for example, John Cassidy, "Schools Are Her Business," The New Yorker, 18-25 October 1999, 144-60.
14. Milton Friedman, "The Role of Government in Education," in Educational Vouchers: Concepts and Controversies, ed. George La Noue (New York: Teachers College Press, (1972), 8-18.
15. Friedman continues to write on the topic. See Milton Friedman. "Public Schools: Make Them Private," Education Economics 5, no. 3 (December 1997): 341-44. The Milton and Rose D. Friedman Foundation vigorously promote universal vouchers (and other school choice plans). Available online at http://www.friedmanfoundation.org.
16. John E. Chubb and Terry M. Moe, Politics, Markets and America's Schools (Washington, D.C.: The Brookings Institution, 1990), 1-3, 35.
17. Viteritti, Choosing Equality, 86-87.
18. Peter Schrag, "The Voucher Seduction," The American Prospect, 23 November 1999. Available online at http://www.prospect.org/print/Vll/1/schrag-p.html.
19. See, for example, Christopher Jencks, "Giving Parents Money to Pay for Schooling: Education Vouchers," New Republic, 4 July 1970; Theodore Sizer, "The Case for a Free Market," Saturday Review, 11 January 1969; John E. Coons and Stephen D. Sugarman, "Family Choice in Education: A Model State System for Vouchers," California Law Review 59 (1971).
20. Schrag; available online at http://www.prospect.org/print/Vll/l/schrag-p.html.
21. Joseph Viteritti, "School Choice: Beyond the Numbers," Education Week, 23 February 2000.
22. The growing support for school vouchers is demonstrated in Moe's Schools, Vouchers and the American Public.
23. Specific court cases will be discussed in detail in the next section, though it should be noted that several recent federal rulings run counter to this perceived trend of general hostility toward religion in public schools. See, for example, Brown v. Gilmore, in which a federal appeals court upheld Virginia's mandatory moment of silence each school day (Brown v. Gilmore, No. CA-00-1044-A [4th Cir. July 24, 2001]), and the Good News Club, in which the Supreme Court ruled that religious and secular groups must have equal access to school buildings for after-school meetings (Good News Club v. Milford Central School, 121 S. Ct. 2093 [June 11, 2001]).
24. Rev. William Davis, "Remarks to the Faith and Reason Institute," 9 May 2000 (http:// www.frinstitute.org/ rrvoucher.html). See also United States Catholic Conference, "Principles for the Educational Reform in the United States" (Washington, D.C.: USCC, 1995). 25. Michael J. Perry, Religion in Politics: Constitutional and Moral Perspectives (Oxford: Oxford University Press, 1997), 13, 15.
26. The First Amendment, in part, forbids Congress from making a law "respecting an establishment" of religion, which at the time of ratification (1.791) meant it could neither establish a federal religion nor interfere with existing established religions of the several states. The latter restriction was rendered moot in 1833 (when Massachusetts became the last state to disestablish its church), and rendered impossible in 1947 when the Establishment Clause was incorporated against the states through the Due Process Clause of the
Fourteenth Amendment. Michael McConnell, "Accommodation of Religion: An Update and Response to the Critics," George Washington Law Review 60 (1992): 690, n. 19.
27. As Justice Lewis Powell noted in 1973, "[T]his Court repeatedly has recognized that tension inevitably exists between the Free Exercise and the Establishment Clauses ... and that it may often not be possible to promote the former without offending the latter. 11 See Committee for Public Education and Religious Liberty (CPERL) v. Nyquist, 413 U.S. 756 (1973) at 788.
28. There is an important distinction in constitutional jurisprudence between religious actions and religious beliefs. Religious beliefs (like all matters of conscience) fall completely outside the realm of governmental regulation, and religious speech (like all forms of speech) is given the highest level of protection-though not complete exemption-from regulation. Religious activity, however, has long been construed as a legitimate area of regulation, though the distinction between speech and action (and the application of the First Amendment to both) is often contested. One oft-cited example is the congressional ban on polyg
amy, which the Supreme Court upheld in the late nineteenth century in Reynolds v. United States (1879). The Reynolds Court held that belief in and teaching of polygamy was protected by the First Amendment, but the practice of polygamy was prohibited. See John Witte, Religion and the American Constitutional Experiment: Essential Rights and Liberties (Boulder, Colo.: Westview, 2000), 102-04.
29. Writing for the Yoder court, Chief Justice Burger anticipated this concern and argued for a balance between free exercise and non-establishment: "The Court must not ignore the danger that an exemption from a general obligation of citizenship on religious grounds may run afoul of the Establishment Clause, but that danger cannot be allowed to prevent any exemption no matter how vital it may be to the protection of values promoted by the right of free exercise." Wisconsin v. Yoder, 406 U.S. 205 (1972) at 220-21, quoted in Warren A. Nord, Religion and American Education: Rethinking a National Dilemma (Chapel Hill, N.C.: UNC Press, 1995), 405, n. 45.
30. This right was established in Pierce ro. Society of Sisters, 268 U.S. 510 (1925).
31. Nord, Religion and Arican Education, 433, 11. 74.
32. The specific criteria for making these judgments were set forth in the "Sherbert test," from Sherbet v. Verner (1963), which held that any law that substantially burdens the free exercise of religion must (1) serve a "compelling state interest," and (2) be narrowly tailored to achieve that interest with the least possible intrusion on free exercise rights. Following a struggle between Congress and the Supreme Court in the 1990s-in which the Sherbert test was renounced bv the Court in Employment Division v. Smith (1990), restored by the Congress via the Religious Freedom Restoration Act (RFRA) of 1993, and partially overturned by the Court in City of Boerne v. Flores (1997)-today the Sherbert test guides free exercise jurisprudence in federal courts, while Smith remains the controlling precedent for free exercise claims against state and local governments.
33. "Anti-democratic" in this context is used in the sense of being counter-majoritarian; the fact that 70 percent of American voters favor allowing daily spoken prayers in public school classrooms, for example, does not deprive the other 30 percent of their rights. See Mark Gillespie, "Most Americans Support Prayer in Public Schools," Gallop News Service, 9 July 1999; available online at http://vw.gal hip.com/poll/releases/pr990709.asp.
34. To the extent that the religious institutions listed above receive government benefits on an equal footing with non-religious institutions, the Court has generally deemed the programs by which they benefit constitutional.
35. Perry. Religion in Politics, 15.
36. Reuben Quick Bear v. Leupp 210 U.S. 50 (1908); Meyer v. Nebraska, 262 U.S. 390 (1923); Pierce v. Society of the Sisters of the Holy Names of Jesus and Mary, 268 U.S. 510 (1925); Farrington v. Tokushige, 273 U.S. 284, 298 (1927); Cochran v. Louisiana State Board of Education, 281 U.S. 370 (1930). Each of these cases, it should be noted, were handed down before the Supreme Court had formally applied the First Amendment's religion clauses to state and local governments.
37. Gitlow v. New York (1925) began the march toward "selective incorporation" of the First Amendment clauses by applying the free speech provision to state/local laws. A number of cases over the next twenty years commented on the process of selective incorporation (see especially Palko v. Connecticut ) and incorporated the remaining provisions, including free press (Near v. Minnesota ), free assembly (De)onge v. Oregon ), free exercise of religion (Cantwell v. Connecticut ) and establishment of religion (Everson v. Board of Education ).
38. Everson t;. Board of Education of the Township of Ewing, 330 U.S. 1 (1947) at 15-16. 39. Then-Associate Justice Rehnquist's dissenting opinion in Wallace v. Jaffree (1985) offered a lengthY historical treatise on the Founding Fathers' views on the Establishment Clause, concluding "there is simply no historical foundation for the proposition that the Framers intended to build the `wall of separation' that was constitutionalized in Everson" (472 U.S. 38 at 106). George Marsden recently bemoaned the citation in Everson as the elevation of a personal remark "to virtual constitutional status." George Marsden, The Outrageous Idea of Christian Scholarship (New York: Oxford University Press, 1997), quoted in David Lyle Jeffrey, "Civic Religion and the First Amendment," in One Nation Under God? Religion and American Culture, eds. Marjorie Garber and Rebecca Walko-itz (New York: Routledge, 1999), 22.
40. Everson, 330 U.S. 1 (1947) at 18
41. Board of Education v. Allen, 392 U.S. 236 (1968). 42. Lemon v. Kurtzman, 403 U.S. 602 (1971) at 612-13.
43. Lemon v. Kurtzman [111, 411 U.S. 192 (1973); Norwood v. Harrison, 413 U.S. 455 (1973); Committee for Public Education and Religious Liberty (CPERL) v. Nyquist, 413 U.S. 756 (1973); Sloan v. Lemon, 413 U.S. 825 (1973).
44. Simmons-Harris v. Zen , 2000 FED App. 0411P (6th Cir.). 45. CPERL v. Nyquist 413 Li.S. 756 (1973) at 788-89.
46 Witte. Religion and the American Constitutional Experiment, 153.
47. Aguilar v Felton, 473 U.S. 402 (1985) at 420-21, Justice Rehnquist, dissenting.
48. Witte, Religion and the American Constitutional Experiment, 172; Perry, Religion in Politics, 23; Leonard W. Levy, The Establishment Clause: Religion and the First Amendment (New York: MacMillan, 1986), 163; quoted in Witte, Religion and the American Constitutional Experiment, 182.
49. See, for example, Mueller v. Allen 463 U.S. 388 (1983); Witters v. Washington Department of Social Services, 474 U.S. 481 (1986); Zobrest v. Catalina Foothills School District, 509 U.S. 1 (1993); and Agostini v. Felton, 521 U.S. 203 (1997).
50, In Agostini v. Felton (1.997), the Court modified the Lemon test by emphasizing "three primary criteria the Court currently uses to evaluate whether government aid has the effect of advancing religion: It [must] not result in governmental indoctrination, define its recipients bv reference to religion, or create an excessive entanglement. Nor can [a government] program reasonably be viewed as an endorsement of religion." Agostini v. Felton, 521 U.S. 203 (1997) at 230-31, reversing Aguilar v. Felton, 473 U.S. 402 (1985).
51. See footnote 1.
52. Mitchell v. Helms, 530 U.S. 793 (2000)
543 Ibid., Justices Souter, Stevens and Ginsburg, dissenting.
55. Zelman v. Simmons Harris, 2000 FED App. 0411P (6th Cir.)
56. The financial disincentives, the court held, stem from the fact that the vouchers (worth a maximum of $2,250) are worth less than the per-pupil cost of the neighboring public schools ($7,097), as well as the per-pupil payment that local charter schools received (between $4,400 and 86,000). There is thus a financial incentive for private schools to "convert" to charter schools, thereby diminishing the available supply of nonsectarian schools. This process of conversion was documented in an influential twelve-part investigative newspaper report, "Whose Choice?" in the Akron Beacon Journal (see 11. 8).
57. Zelman v. Simmons-Harris, 2000 FED App. 0411P (6th Cir.).
58. Agostini v. Felton, 521 U.S. 203 (1997) at 237.
59. See footnote 1.
60. Opponents of school vouchers have argued that the move from state assistance for educational materials or facilities at religious schools to state assistance for religious school tuition represents a qualitative shift, since the latter aid is "divertible" to religious use. But the Court has rejected claims in Zobrest, Witters, and Mitchell that divertibility necessarily renders aid unconstitutional.
61. The Milwaukee program limits vouchers to families with incomes below 175 percent of the federal poverty level; in Cleveland, priority is given to those with incomes below 200 percent of the poverty rate.
62. Agostini v. Felton, 521 U. S. 203 (1997) at 226 (quoting Witters v. Wash. Dept. of Services for Blind, 474 U.S. 481 (1986) at 487).
63. See especially Agostini v. Felton, 521 U.S. 203 (1997).
64. Curricular requirements for religious schools are bound to be a matter of some contention, but this entanglement between church and state is not necessarily excessive. I believe the analogy between schools and hospitals is largely (though not entirely) apt here as well: despite the state's interest in a minimally common educational curriculum for students, it is not clear that school voucher programs create more of an entanglement with the content of services provided by religious schools than with the content of services provided by, say, religious hospitals. In the latter case. federal regulations constrain the content of medical services through the use of drug formularies, treatment guidelines, etc.
65. Immanuel Kant, Kant on Education (Ueber Padagogik), trans. Annette Churton (Boston, Mass.: D.C. Heath & Co., 1900), 12, quoted in Amy Gutmann, Democratic Education (Princeton, N.J.: Princeton University Press, 1987), 3.
66. Even vehement critics of post-Enlightenment rationalism like Stanley Hauer\,as do not challenge the importance or difficulty of education per se, but rather its presumed abstraction from communities of discourse and meaning. straction from communities of discourse and meaning.
67. National Center for Education Statistics, The Condition of Education 2001 (Washington, D.C.: UT.S. Department of Education, July 2001), 30-32. For a condensed bibliography of the research literature on the subject, see Viteritti, Choosing, 267-68, n. 1.
68. National Commission on Excellence in Education, "A Nation At Risk: The Imperative for Educational Reform" (Washington, D.C.: U.S. Department of Education, April 1983), 7. 69. The Republic of Plato, 2nd ed., trans. Allan Bloom (New York: Basic Books, 1991), 9394 (414d-415c), and 101 (423d).
70. "Unity there should be, both of the family and of the state, but in some respects only. . . The state is a plurality, which should be united and made into a community by education." Aristotle, The Politics and Constitution of Athens, ed. Stephen Everson, trans. Jonathan Barnes (New York: Cambridge University Press, 1996), 37 (Politics 11.5, 1263b3038).
71. Ibid., 195 (Politics VIII.1, 1337a12-27).
-- NJ x s.,'
73. According to Aquinas, "A child is by nature part of its father ... so, according to the natural law, a son, before coming to the use of reason, is under his father's care. Hence it would be contrary to natural justice, if a child, before coming to the use of reason, were to be taken away from its parents' custody, or anything done to it against its parents' wish" (Sout Theologica, 11-II q.10 a.1). Further, "Nature intends not only the begetting of offspring, but also its education and development until it reach the perfect state of man as man, and that is the state of virtue" (Summa Theologica, Supplementum 111, q.1 a.1). These passages were cited by Pius Xf in his encyclical "On Christian Education," which went on to declare: "The family therefore holds directly from the Creator the mission and hence the right to educate the offspring, a right inalienable because inseparably joined to the strict obligation, a right anterior to any right whatever of civil society and of the State, and therefore inviolable on the part of any power on earth" (Pius XI, "On Christian Education (Rappresentanti In Terra)," 31 December 1929, 32-33).
'74. John Locke, "The Second Treatise of Government," in Two Treatises of Government, 2nd ed., ed. Peter Laslett (London: Cambridge University Press, 1988), 314 (72), and ch. 6. See also John Locke, The Educational Writings, ed. J.L. Axtell (London: Cambridge Universit, Press, 1968).
75. This decision perhaps says more about Rousseau's (lack of) faith in the French monary in 1762 than his belief in the importance of citizenship, since he prescribed a thoroughly
civic (and public) system of education in his Considerations on the Government of Poland (1772). Preceding quotation from Jean-jacques Rousseau, Emile, or On Education, trans. Barbara Foxley (London: Everyman, 1993), Book I, 22.
76. Gutmann, Democratic Education, 30.
77. In the case of medical treatment, the state's interest in children's health is seen to override, at times, even the First Amendment right to the free exercise of religion. See Jehovah's Witnesses v. King County Hospital, 390 U.S. 598, where blood transfusions may be given to a child even when the parents are opposed on religious grounds.
78. Jeffrey Henig, Rethinking School Choice: Limits of the Market Metaphor (Princeton, N.J.: Princeton University Press, 1994), 200. Or as Amy Gutmann has similarly argued, "The problem with voucher plans is not that they leave too much room for parental choice,
but that they leave too little room for democratic deliberation." See Gutmann, Democratic Education, 70.
79. By "outcomes," I mean here the measurable results of schooling such as literacy, numeracy, etc.
80. See "The Political Meaning of `Crisis,"' ch. 2 of Henig's Rethinking School Choice. 81. Emphasis in original. Chubb and Moe, Politics, Markets, and America's Schools, 217. 82. For a discussion of decentralized, democratic educational governance as a "third way" between decentralized market control and centralized state control, see Amy Gutmann, "Why Should Schools Care About Civic Education?" in Rediscovering the Democratic Purposes of Education, eds. Lorraine McDonnell, P. Michael Timpane, and Roger Benjamin (Lawrence, Kans.: University Press of Kansas, 2000), 73-90.
83. John F. Witte, a professor of public policy and former evaluator of the Milwaukee school choice program, cautions that "a daunting number of unanswered empirical research questions remain, the answers to which will have a major bearing on the more general policy and institutional issues." See his The Market Approach to Education (Princeton, N.J.: Princeton University Press, 2000), 24. For a list of reports and studies heralding positive results in Cleveland and Milwaukee, see the special "school choice" section of the Center for Education Reform's web site (www.edreform.com/school_choice/research.htm); for a list of reports claiming insignificant or negative effects of vouchers, see the National Education Association's web site ( A .nea.org/issues/vouchers/resources.html). For one example of the heated nature of the debate over voucher effectiveness, see Kate Zernike, "New Doubt Is Cast on Study That Backs Voucher Efforts," New York Times, 15 September 2000.
* ERIK C. OWENS (B.A., Duke University; M.T.S., Harvard University Divinity School) is a doctoral candidate in religious ethics at The University of Chicago Divinity School and is a research associate at the Pew Forum on Religion and Public Life. He is co-editor (with John D. Carlson) of The Sacred and the Sovereign: Rethinking Religion and International Politics and is co-editor (with Eric Elshtain and John D. Carlson) of A Call for Reckoning: Religion and The Death Penalty. His articles will appear in two forthcoming volumes, The Encyclopedia of Religious Freedom (2003) and The Encyclopedia of American Religion and Politics (2003) Special interests include the role of religion in international relations, political theory, and civic education. He is grateful for many comments and helpful suggestions on this essay from Franklin Gamwell, Jean Bethke Elshtain, and the Public Theology Working Group at the University of Chicago Divinity School's Martin Marty Center where an edited version of this essay was presented on 7' March 2002.…