Choice for Me But Not for Thee

Passage of the “Charitable Choice” provisions of welfare reform in 1996, and the more recent publicity surrounding President Bush’s “Faith Based Initiative” have focused renewed public attention on the delicate relationship between church and state, and between politics and religion.

Choice for Me But Not for Thee?
                                                              Welfare Vouchers, School Vouchers and the Constitution


Passage of the “Charitable Choice” provisions of welfare reform in 1996, and the more recent publicity surrounding President Bush’s “Faith Based Initiative” have focused renewed public attention on the delicate relationship between church and state, and between politics and religion.
Critics of Charitable Choice and similar efforts seem not to realize that the government already pays religious organizations to house, clothe and counsel the poor, and has done so since the beginning of state-run social welfare programs. Religious providers range from 501(c) 3 affiliates of denominational entities, like Catholic Charities and Lutheran Social Services, to “pervasively sectarian” organizations like the Salvation Army, to individual congregations.  Indeed, proponents of “Charitable Choice” legislation and President Bush’s “Faith-Based Initiative” describe those policies as nothing more than an attempt to “level the playing field,” to ensure that government officials do not inappropriately require religious contractors to diminish or eliminate religious components of their services. Whether that claim is accurate or disingenuous, the fact that public tax dollars are being used to support sectarian providers of social services—and have been so used for decades—is beyond debate. Generally, government funds follow the individuals entitled to the benefits involved. For example, Medicaid patients may choose a nursing home or hospital, which then receives payment from the government. While these and similar benefits are not generally referred to as vouchers, they are functionally indistinguishable from vouchers. Direct contracts and other collaborations between  government units and pervasively sectarian organizations, including individual congregations, while less common, are also not unusual.
On the other hand, at least to date, the Court has refused to allow public money to flow directly to religious elementary and secondary schools, although other sorts of assistance have been permitted, and attempts to evade that restriction through voucher programs have generally been unsuccessful.
Public debate over Charitable Choice has highlighted these inconsistencies, and some constitutional commentators have predicted that they will be resolved by eventual approval of vouchers, and disapproval of more direct forms of funding, for both religious education and religious social services. While such a result would resolve the doctrinal tension, this article considers whether a principled case can be made for continuing the different constitutional treatment of vouchers in the areas of education and welfare. Recognizing the incoherence of current case law, is it nevertheless possible to construct an argument that there are differences of constitutional magnitude between the use of a voucher or other portable government benefit to purchase religiously based medical care or job counseling, and issuance of a voucher enabling parents to enroll their children in a religious elementary or secondary school?



Government financial support for religious social services has been a longstanding feature of public welfare programs.  In a 1969 study of findings from a 1965 survey of 406 sectarian agencies in 21 states, Bernard J. Coughlin reported that 70 percent of them were involved in some type of purchase of service contract with the government.  A 1982 study by F. Ellen Netting, focusing on government funding of Protestant social service agencies in one Midwestern city, found that some agencies received between 60 and 80 percent of their support from the government, and that approximately half of their combined budgets were government-financed. In 1994, government funding accounted for 65 percent of the nearly two billion dollar annual budget of Catholic Charities USA, and 75 percent of the revenues of the Jewish Board of Family and Children’s Services.
This allocation of public tax dollars to religious social services providers has gone virtually unchallenged. There are two major Supreme Court precedents: Bradford v. Roberts, an 1899 case permitting the flow of public dollars to religious hospitals, and Bowen v. Kendrick, decided in 1988. Bowen involved an Establishment Clause challenge to the Adolescent Family Life Act. The Family Life Act provided funding to a variety of local organizations, including religious organizations, to support counseling of teenagers about premarital sex and teenage pregnancy. The Court in Bowen acknowledged that counseling services might be delivered by sectarian groups in a manner that violated the Establishment Clause, but declined to find the Act facially unconstitutional merely because that danger existed.  According to the majority, a successful challenge would have to rest upon the particulars of a specific program; the mere inclusion of religious contractors in the program was held not to constitute a per se Establishment Clause violation.

Funding for religiously affiliated schools, on the other hand, has generated a significant jurisprudence, and one not notable for consistency; indeed, it has been aptly characterized as a “tangle of cases.”  The Supreme Court has upheld the allocation of federal funds to Catholic schools offering education to Native Americans and has ruled that state governments cannot require that students attend public schools. Public subsidies for textbooks and transportation have been upheld, as has state support of remedial instruction provided by public school teachers at parochial schools. States cannot constitutionally supplement the salaries of teachers at religious schools, so as to bring those salaries in line with those of public school teachers, but may provide sign language interpreters for deaf students attending Catholic high schools, or support to blind students studying for the ministry at Christian colleges. State laws allowing tax deductions for educational expenses (including tuition, textbooks, school supplies and transportation) for parents whose children attend parochial primary or secondary schools have been held not to violate the Establishment Clause. Religious colleges have been allowed to use state construction grants to build facilities devoted to pursuit of secular subjects, and the GI Bill has financed students’ educations at public, private and religious institutions of higher education.

A major cause of this tortured jurisprudence is judicial recognition that, whatever else the Establishment Clause may mean, it absolutely forbids government funding of religion. The mere fact that tax dollars are paid to a religious organization, however, is not equivalent to funding religion; government may constitutionally purchase services from sectarian sources, or enter into other partnerships that involve the transfer of tax dollars to such entities so long as the funds do not support religious activities. The relative lack of concern over government support for religious social services is at least partly because the secular nature of those services is readily apparent.  Hospitals and nursing homes are providing medical care; day care facilities are supervising children; job placement counselors, drug treatment facilities and the like have secular counterparts engaged in providing similar, if not identical, programs. While economists remind us that dollars are fungible, so that support for one activity frees up funds that can then be used elsewhere, it is relatively simple to calculate the costs of nursing services or child care, and reasonable to argue that if payment is only sufficient to cover those costs, it subsidizes only the secular activity.
In the school context, programs that have passed constitutional muster have generally been those involving an identifiably secular benefit available to all citizens—immunization, speech and hearing testing, transportation—where exclusion of children attending religious schools was deemed to burden the free exercise rights of parents opting for religious education. When public aid has been disallowed, the Court has pointed to the “pervasively sectarian” nature of the religious school, and the supposed impossibility of ensuring that only secular programs benefit from the expenditure of public funds. If public money does flow to a pervasively religious institution, the Courts have held that any benefit accruing to that institution must be incidental to a secular purpose. In Agostini v. Felton, the Court (quoting from Witters v. Washington Department of Services for the Blind) explained that in order to be constitutionally permissible, any public money earmarked for secular purposes that ultimately goes to pervasively religious institutions must do so “only as a result of the genuinely independent and private choices of individuals.”


If public dollars have been allocated to support a secular program, and if there is a “genuinely independent and private” choice of service provider, then there is no constitutionally persuasive reason to prevent an elderly person seeking nursing home care from spending her benefits in a nursing home run by her religious denomination, or to prevent a recipient of the GI Bill from using his benefits to study pharmacy or English literature at a Catholic university. As the Court struggles for neutrality in its application of the religion clauses, as it searches for a formulation that neither burdens nor benefits religious practice and belief, the exercise of intervening independent choice sufficient to insulate government from a charge of endorsement would seem to be the fairest way to achieve evenhandedness.  Is it time, then, to abandon the Court’s historic reluctance to allow public money to flow to religious primary and secondary schools?  Is that reluctance simply a lingering vestige of anti-Catholic bias, as many scholars assert? Or are there constitutionally compelling reasons for treating vouchers for primary and secondary schools as a separate and distinct case?
I would suggest that there are three such reasons, all of which are direct consequences of the legal status of, and the state’s duty to, young children: the best interests of the child, whose lack of autonomy means that the choice of school will be exercised by parental or other proxies; the state’s vested interest in the education of children, over and above its acknowledged interest in their general welfare; and the inherent conflict between the measures required to protect those state interests, on the one hand, and the free exercise and equal protection rights of private religious schools on the other.


Perhaps the most significant distinction between vouchers for social services and vouchers for education is the non-autonomous nature of the intended beneficiary. In the social service arena, an adult voucher recipient exercises independent choice. If the recipient is unable to do so, she may designate her own proxy by executing a power of attorney or similar document. If incompetence occurs prior to such designation, a court will oversee the choice of a guardian who must satisfy the court that s/he will act in the ward’s best interests. Children, however, are unable to exercise genuine, independent choice; in the absence of extraordinary circumstances (allegations of abuse or neglect, abandonment, etc.) they are also unable to change their parent, guardian or other proxy. Furthermore, since procreation is a constitutionally protected right, there are no minimum requirements—mental, physical, emotional, financial or otherwise—for being a parent, and no way of assuring that a given parent will act, or is even capable of acting, in the child’s best interests. As a consequence, the state owes a duty to ensure that its policies protect the interests of children—a duty that the courts have recognized in a wide variety of contexts.
The fact that the decision which theoretically insulates the state from responsibility is being made by an unchosen and unsupervised proxy rather than by the child who is the intended beneficiary is further complicated by the fact that the state has a significant, independent interest not only in the general welfare of children, but in the transmittal of core civic values. In the welfare arena, where the state’s interests are more limited, it nevertheless furthers those interests by requiring licensing of medical and child care facilities, by conditioning contracts upon compliance with nondiscrimination laws and other relevant regulations, and/or by requiring periodic audits and documentation of acceptable outcomes. If public funds flow to religious schools, the magnitude of the state interest will require much more stringent accountability measures; however, the free exercise rights of the schools and those who choose them are arguably inconsistent with the significant degree of regulation that would be required. 

Whose Choice?

As we have seen, the vast majority of social service funding follows an adult recipient to the nursing home, hospital, or other provider of that recipient’s choice. Establishment Clause decisions involving higher education have been based upon the belief that the older students involved are adults or, as the Court has framed it, “less impressionable,” less susceptible to indoctrination, and more able to exercise independent judgment than younger children. The ability of a voucher recipient to exercise a significant degree of autonomy has been an important, although often tacit, element of First Amendment analysis. Young children, however, are not autonomous. Indeed, American case law accords children a unique status, and reflects a corresponding ambivalence with respect to their rights, and government’s duties to them.
Historically, American laws concerning children have been informed by two very distinctive paradigms: a paternalistic approach that has focused on nurturance and protection; and an individual-rights approach that has emphasized the child’s age-appropriate right to exercise personal autonomy.  To the first approach we can attribute legislation making school attendance compulsory, establishing juvenile courts, and prohibiting child labor.  These  measures assume a large degree of dependence by children on adult society.  In the second category are court cases such as Brown v. Board of Education, which recognized children as rights-bearing individuals, or In re Gault, which entitled juveniles to court-appointed counsel, or Tinker v. Des Moines, which protected the right of minors to express political opinions in a non-disruptive manner in a public-school setting.  The legal system has struggled to determine when it is appropriate to categorize a particular individual as a child requiring nurturance, and when as an emerging adult entitled to individual autonomy and voice. This struggle has been made more difficult because the rights and duties of parents and (to a lesser but not inconsiderable extent) the interests of the state are also implicated.  As one commentator has noted, the courts have struggled to mediate the conflict between “democratic ideals of individual freedom and the sanctity of the family unit.” The result has been a sharp distinction between public and private responsibility for children’s welfare, with the public assuming responsibility when the private actors have defaulted.  The degree of power given to parents and family rests on a belief that children need authority rather than autonomy, that “rights” so understood are not in the child’s best interests, and that the family is better suited than the state to determine the child’s needs.

Children have no political power and are not usually the focus of laws and decisions that affect them.  Child labor laws went nowhere until labor unions decided that such laws would be economically beneficial to their members; similarly, court opinions affecting children typically arise in the context of other conflicts: child custody disputes, religious liberty cases, the criminal justice system and conflicts around medical treatment and intervention.  Those disparate cases reflect America’s evolving (and uncertain) approach to the complex relationships among children, parents and the state.
The earliest cases defining children’s status focused on conflicts between parents and the government, and established parental prerogatives as constitutionally protected privacy rights.  In 1923, Meyer v. Nebraska overruled a state law prohibiting parents and schools from teaching children in the German language.  The case was decided largely on First Amendment freedom of expression grounds. Two years later, Pierce v. Society of Sisters established the right of parents to remove their children from the public school system; however, the decision also upheld the right of the state to mandate education through a particular age, or for a particular number of years.  The Court thus recognized both the right of parents to instruct their children in a manner consistent with their own values and beliefs, and the right of government to protect society’s vital interest in assuring an educated citizenry.  It would not be until Wisconsin v. Yoder in 1972 (and then only in a dissent by Justice Douglas) that the rights of the children involved would be explicitly considered.
Despite longstanding judicial deference to parental rights, those rights have often been constrained by public policies intended to protect children from exploitation. A case in point is Prince v. Massachusetts, in 1944, where the Court upheld the constitutionality of child labor laws and the imposition of penalties upon parents who violated them.  The case signaled that parental rights, while still paramount, could and would be limited if the state decided that the interests of children required such limitation.
In the late sixties and early seventies, courts continued to send mixed signals about the rights of young people and the interests of parents and the state: In re Gault, handed down in 1967, required the state to provide legal representation for children facing criminal proceedings, extending to juveniles a right previously enjoyed only by adults.  The next year, in Ginsberg v. New York, the Court upheld a state law forbidding sale of a “girlie” magazine to a sixteen year old.  In 1969, in Tinker v. Des Moines, the Supreme Court issued its now-famous (and generally ignored) edict that children “do not leave their constitutional rights at the schoolhouse gate,” and permitted the young plaintiffs to wear black arm-bands to class to express disapproval of the Vietnam War.  In 1971, in McKiever v. Pennsylvania, the Court declined to extend the right of trial by jury to young people. 
In 1972, Wisconsin v. Yoder held that Amish parents had First and Fourteenth Amendment rights to withdraw their children from formal schooling after completion of the eighth grade.  The case continues to be widely discussed and debated, because it addresses the persistent tension between the free exercise rights of parents and the interest of the state in fostering civic cohesion and citizenship education.  What it did not address, as Justice Douglas pointed out in dissent, was the possibility that Amish children might have rights distinct from those of their parents that the state (and the Court) might have a duty to consider.
If parents have often been accorded unreasonably wide latitude over childrearing decisions, the Court has just as steadfastly refused to limit the scope of state interference with children’s liberty interests.  In 1979, in Parham v. J.R., the Court declined to require  a hearing before a child could be institutionalized; and in 1984, in Schnall v. Martin, it permitted the “preventive” detention of juveniles said to “pose a risk” of committing crime.  The Court in Schall held that a minor’s liberty interest was subordinate to his “best interests” as the state might choose to define them. Hazelwood School District v. Kuhlmeier, a 1988 decision, upheld the right of public school officials to censor student newspapers.  Despite language in these and other cases suggesting a significant state interest in their welfare, however, government’s power over children has not always been accompanied by a corresponding duty; in DeShaney v. Winnebago County Department of Social Services, the Court declined to find county social workers liable for severe brain damage suffered by a child at the hands of his father, despite the fact that the agency had been aware of the situation for months, had removed the child from the home at one point, and had later returned him and allowed him to remain despite  documented evidence of continued abuse. 


A comparison of the juvenile justice system with health care law provides stark examples of the legal system’s ambivalence about the rights and capacities of children and the corresponding duties of government.  Prosecutors are waiving more and younger juveniles into adult courts, insisting that they be held legally and morally responsible under adult criminal standards, while at the same time other courts are tightening parental notification laws based on the assumption that female adolescents are incapable of assuming responsibility for their own reproductive decisions.  There are states in which an 18-year-old can obtain an abortion without parental notice, but cannot legally miss school for a doctor’s appointment without a note.  Finally, there are a number of cases widely reported and heatedly debated cases where the state’s interest in the well-being of children has been held to justify overruling parents right to direct children’s medical treatment, even when the parental decision rested upon sincerely held religious beliefs.
What does this judicial ambivalence about the rights of parents, the interests of the state and the status of children suggest for the issue of school vouchers?  Nancy E. Walker, Catherine M. Brooks and Lawrence Wrightsman have put the issue succinctly:
“Who are the schools most obligated to serve—children, parents, or society? If parents disagree with something being taught, which is the greater right, the parents’ right to control their child’s education, the state’s need for informed citizens, or the child’s right to know?”
In “School Vouchers: Inviting the Public into the Religious Square,” James G. Dwyer suggests a child-centered answer to that question. As he notes:
“When addressing questions of constitutional interpretation or political theory presented by conflicts over any aspect of child rearing, one must first recognize that a proper analysis will necessarily differ in important respects from an analysis of situations involving only adults.  Because standard modes of constitutional analysis and theorizing envision that only competent adults are involved in disputed matters, and rely substantially on premises that are true of competent adults but not of children, the standard modes of analysis cannot simply be extended without modification to childrearing contexts. Whenever constitutional scholars or political theorists turn their attention to schooling or any other child welfare issue (e.g. parents’ religious objection to medical care or core family law issues like child abuse) they must restructure their analytical apparatus.”
Dwyer makes a convincing case that parents should be viewed as fiduciaries rather than rights holders. As he notes, “guardians for incompetent adults generally do not possess rights in connection with decision making for their wards, but rather are deemed to enjoy a privilege of exercising limited authority over the lives of their wards.” He concludes that the best interests of the children, as independently determined by the state, must take priority over the interests of the parents. To hold otherwise is to suggest that parents may use their children to fulfill their own expressive or religious needs—to treat children’s lives instrumentally. That, as he notes, would be both morally and constitutionally improper. While Dwyer proposes a new paradigm of children’s rights, his emphasis upon the duties rather than the prerogatives of parenthood has ample support in case law.
If courts are obliged to consider the best interests of children, independent of the desires of their parents, what is the effect on the constitutional analysis of education vouchers? Dwyer believes that a proper concern for the rights of children requires government to  fund and control all schools—public, private, secular and religious. Indeed, he argues that even in the absence of public funding, the state has an obligation to regulate all schools, in order to ensure that children do not attend schools where they are physically unsafe, where they receive an inadequate secular education, or where they are exposed to curricula or practices that subvert democratic and constitutional values.
While Dwyer’s argument has the appeal of intellectual consistency, acceptance of it would give the state a degree of power substantially at odds with our American libertarian tradition and jurisprudence. It would also give short shrift to the free exercise rights of families and religious schools. A better approach would be to acknowledge (and accommodate to the extent possible) the rights of parents and religious schools to freedom from unwarranted government interference, while also recognizing the state’s substantial interest in encouraging civic cohesion and democratic values, and explicitly protecting children’s right to be educated in a manner that will allow them to exercise meaningful autonomy as adult members of a democratic polity. 
Achievement of such a balance, however, requires a more searching evaluation of the provider than is necessitated by the choice of job training or medical facilities by an adult using her own welfare voucher, or even a parent using a voucher to secure day care or medical services for her child.  In the social service arena, the state’s interest is in assuring a minimum level of professional competence: the job trainer should have appropriate certifications, the medical facility and day care provider should be licensed.  Facilities should meet code requirements, and books and records should be open to audit in order to assure proper application of funds. In the school context, these requirements are also present, and in addition, the state’s interest in the civic adequacy of the education being provided must also be addressed. 

The State’s Interest


Institutions accepting education vouchers must withstand a higher level of scrutiny than those accepting welfare vouchers not only because use of the voucher is being controlled by someone other than the young and vulnerable beneficiary, but also because the state has a compelling, independent interest in the transmittal of common civic values.  It has been suggested that existing education jurisprudence cannot be understood without taking into account the distinction between the purely academic function of the schools and their “social” function, which is to teach children to be responsible and participatory members of a particular society, grounded upon certain distinctive principles and values.  This interest is not present in the welfare context, and it requires greater judicial deference than is required when reviewing free exercise claims in the context of welfare policy.  As Stephen Macedo has recently framed the issue,
“It may be that a greater reliance on educational choice and school competition makes sense from the point of view of educational policy. And it may be that we would do well to utilize the particular virtues and capacities of private groups and civil society institutions in the delivery of social services. The last thing we should do, however, is to simply ignore or assume away the civic ambitions that have been at the core of common schooling from the beginning.
It is one thing to disagree with particular means toward the end of creating a shared citizenship; it is quite another to altogether neglect the importance of that basic end.”
Macedo notes that public schools have been the venue where “the tension between diversity and the felt need to promote shared values has been played out most dramatically.”
That tension has been manifest even in the cases most widely thought to have elevated the rights of parents above the interests of the state. It is instructive to reread Meyer and Pierce; both are more nuanced decisions than is often appreciated. In Justice McReynold’s majority decision in Meyer, for example, there is explicit recognition of the legitimacy of the state’s interest in fostering a “homogeneous people with American ideals” and of the right of the state not only to compel school attendance, but to regulate the schools and to require instruction in English as the common language of citizenship.  The state fell afoul of the First Amendment only when it tried to forbid the teaching of other languages in furtherance of its otherwise legitimate citizenship goals. The opinion in Pierce also recognized—and approved—a significant state interest in requiring education; the law at issue, however, required all students to attend the public schools, and that limitation was held to violate both the property rights of those who operated private schools and the right of parents to direct the upbringing and education of their children free of unreasonable state restraints. 
The state’s interest also includes a concern for promoting toleration and inter-group amity; as the nation becomes more and more diverse, that interest becomes more compelling. Recent research suggesting that educational vouchers promote segregation by race, religion and ethnicity raises concerns that expansion of educational choice programs will further erode social capital and undermine community. To the extent vouchers are used in religious schools, the concern may be particularly acute; as Robert K. Vischer has recently written
“In the school voucher debate, one critical issue still unaddressed is the potential for vouchers to increase racial segregation. Specifically, the widespread implementation of vouchers could increase segregation by fostering the creation and expansion of church-affiliated elementary and secondary schools. Given that American churches are already segregated to a great extent—a 1998 study found that only 10 percent of American churches can be considered racially integrated—increased attendance at church-affiliated schools may further exacerbate school segregation.”
Vischer notes that 85% of existing private schools are religious, and that churches enjoy an immense organizational advantage over others seeking to enter the education market, making it likely that religious schools would continue to dominate the private school landscape in a privatized system.  John Witte, an educational researcher who has evaluated the voucher program in Milwaukee and supports it, has expressed similar concerns, noting that the Milwaukee experiment has led to more segregated schools.
            “Regardless of one’s beliefs or hopes, the overwhelming social factor of

            the integration movement was white flight from our cities. It is difficult for me to

see how a [wider]market model of choice would do anything but accelerate the growing balkanization of our schools and country. Whether that result will be class or racially motivated, is in one sense irrelevant given the correlation between the two.” 
These concerns implicate core democratic principles, and must be weighed in any constitutional analysis of proposed education voucher programs. Indeed, the difficult balancing of the state’s interest in an educated and tolerant citizenry against the free exercise rights of parents has been the central theme of an entire series of cases involving religion and the public schools. The importance of the state’s interest has never been seriously disputed, and while it is rarely sufficient to totally “trump” the free exercise rights of parents, the magnitude of the state’s stake in forging a citizenry nevertheless gives government a duty to “tip the scales” toward civic cohesion. If the state cannot, consistent with the free exercise clause, require children to attend government-run schools, neither is it required to facilitate—or fund—their exit from those schools. A useful analogy exists in the creation science and school library cases. While parents retain the right to direct the religious upbringing of their children, recognition of that right does not require schools to eviscerate biology curricula, or purge  school libraries of books containing ideas contrary to parents’ religious beliefs. The balance in these instances is struck in favor of communal, rather than individual, norms. 
It is the civic function of the schools that makes them fundamentally different in kind from providers of social services, and that implicates constitutional interests in ways those services do not. One cannot understand the American experience without understanding the importance we place on individual liberty and personal autonomy; however, it is equally misleading to view the national history only through the lens of radical individualism.  Beginning with Aristotle, political theorists have described citizenship as first and foremost a sharing, a process of forming community around things held in common.  That sharing and commonality has been as important a part of the American character as has our individualism.   As Francis Kane has suggested, there are few public issues that do not presuppose a civic understanding of, and broad agreement with, a common purpose, a shared vision of the public good.   This constant tension between notions of a public or common good and our commitment to the rights of the individual is a truism of Constitutional law and political debate. As Kane notes, 
The balancing of these two poles, at once repelling the state’s unwarranted intrusion into the private lives of its citizens, and at the same time attracting those same citizens to the sweet joys and harsh sacrifices of community life, is what the American experiment is, in large measure, all about.
In the last several years, renewed academic and popular concern with the health and importance of civil society, and the communitarian “backlash” against the liberal democratic emphasis on individual rights, have focused attention on the importance Americans place on the public good and the depth of our yearning for community. Robert Putnam and others have sparked a renewed interest in the importance of social capital: those features of social organization, such as networks, norms, and trust that facilitate coordination and cooperation for mutual benefit. In the face of an increasingly multicultural America, Stephen Macedo cautions that

“Talk of diversity and difference too often proceeds without taking adequate account of the degree of moral convergence it takes to sustain a constitutional order that is liberal, democratic, and characterized by widespread bonds of civic friendship and cooperation.” 
Perhaps Will Kymlicka puts the issue most succinctly by asking the question: what is the basis of social unity and political stability in a liberal state that contains significant ethnic [and, we might add, political and cultural] cleavages?   In many ways, Americans continue to struggle with the implications of Aristotle’s original insight.  If communities are created and sustained by the things we have in common, by mutual engagements that build social capital, how do we transmit, support and reinforce those essentials without unduly infringing upon the liberties of individual citizens?  In large measure, we have depended upon our schools to accomplish that essential act of community building.
In 1996, the Twentieth Century Fund issued a report on school privatization, in which the authors noted:
Under [vouchers], the educational funding stream flows directly from the government to private individuals without the mediation of a public system. Thus, education ceases to be a collective public undertaking and becomes instead a private relationship between each family and its school.  Schooling ceases to be part of the public sphere; no longer a public service, it becomes a consumable item. (emphasis supplied)
If schools are more than providers of technical skills to consumers, if they do play an essential role in the creation of a public infrastructure and the formation of a workable polis, the state’s interest in that process is compelling.

                        Accountability and the Rights of Religious Schools

In the social service arena, institutions are licensed if they meet certain recognized professional standards, and individuals may spend state “vouchers” only with providers who are licensed or otherwise approved by the state. Government’s refusal to license a nursing home that depended on prayer and faith-healing in lieu of medical care would not be likely to lose a free exercise challenge; the expenditure of public dollars requires accountability, and professional medical standards are required to insure a minimal level of care. In the private and religious school context, however, accountability issues become more problematic.
Even the most ardent proponents of vouchers for social services and education concede that Establishment Clause principles constrain the use of public funds. Whenever tax dollars have supported programs at religious institutions, the Court has stressed the need for accountability measures to ensure that the public funds are being spent only for secular purposes. Often, regulatory oversight is simply assumed; for example, in Mitchell v. Helms, a case widely considered to be a breach in the “wall of separation,” the plurality held that
“[I]f the government, seeking to further some legitimate secular purpose, offers aid on the same terms, without regard to religion, to all who adequately further that purpose…then it is fair to say that any aid going to a religious recipient only has the effect of furthering that purpose.” (emphasis supplied)
Obviously, any assessment of “adequacy” will require oversight. Concurring opinions in Mitchell stressed the importance of assuring that government support not be “diverted” to support religion. As Dwyer has noted, the constitutional requirement that aid actually be used to further the state’s secular purpose

“…generates substantial regulatory requirements, including imposition of comprehensive and robust educational standards on all recipient schools. This is necessary to ensure that the aid is being used to provide secular instruction rather than religious instruction under the guise of state-mandated courses. The available empirical evidence suggests that there are many schools in this country today—particularly Fundamentalist Christian schools—that ostensibly teach state-mandated courses but so infuse them with religious content and so distort the secular content to conform to religious beliefs that the courses in effect become religious activities. The state cannot constitutionally pay for that sort of practice, even if the schools themselves characterize it as secular instruction.”
Dwyer concedes that the amount of regulation necessary would “substantially transform” private education, and would have the effect of discriminating against religious schools that “do not value, and perhaps even discourage, critical thinking and mastery of scientific methods, or that do not want their students to learn mainstream views on certain topics.” Nevertheless, he argues that such a result is required by the special duty that government owes to children.
While Dwyer’s approach might seem radical, it is not substantially different from views expressed by some state courts or by Stephen Monsma and Stephen Macedo, both of whom have endorsed the proposition that government ought not—and need not—include   “antidemocratic” providers in voucher programs. While a full analysis of that position lies outside the scope of this article, it raises important and troubling civil liberties issues.  If nonprofit providers—secular or religious—accept government dollars, may the government make those dollars contingent upon adherence to “approved” practices? If the answer to that question is yes, as the Court clearly held in Rust v. Sullivan, the only option available to religious organizations that hold and act on “undemocratic” beliefs (and are unwilling to change those beliefs) is to refuse to participate in voucher programs. But if submission to significant regulation and compliance with government-approved professional and educational standards and practices is the price of participation, what are the implications for religious freedom and equal protection, not to mention free speech?  
In the course of public debate over Charitable Choice, a number of ministers have expressed their concerns over the potential for state intrusion with a pithy phrase: “With the government’s shekels come the government’s shackles.” Even under an accomodationist reading of the First Amendment’s religion clauses,  “shackles” ensuring accountability by schools accepting vouchers will be constitutionally required. If the state supplies vouchers to fund the education of children who cannot exercise independent choice—children whose welfare and education implicate compelling state interests— substantial regulatory oversight will be required to insure that those interests are being protected.


However incoherent and Byzantine it may be, the Court’s current jurisprudence reflects an underlying recognition that the education of young children is a qualitatively different matter than other government enterprises. Education policy implicates complex issues and compelling state interests that are not present in the social welfare context. There is a   distinction of constitutional magnitude to be drawn between vouchers for social services and vouchers for the education of young children, and the Court should explicitly draw it.