20th WCP: Accommodating Pluralism: Liberal Neutrality and Compulsory Education

Various neutrality principles have long been associated with liberalism. Today I want to examine the general neutrality principle Rawls associates with his own liberalism.(1) I want to begin by getting clear on just what that principle is. Then I want to test it in the context of compulsory education.

Let us begin by noting that any basic social structure faithful to liberal principles of political justice will inevitably prove nonneutral in its effects on many comprehensive doctrines and ways of life. This will be true for politically unreasonable doctrines and ways of life (militantly theocratic doctrines, or ways of life centered on violating the basic rights of others). But it may also prove true for comprehensive doctrines and ways of life more or less unopposed to most liberal political values (perhaps the doctrines or ways of life of certain traditional or anti-modern religious sects).

Liberalism, Rawls tells us, cannot and should not promise neutrality of effects. But this should not count against it. Every basic social structure faithful to some conception of justice, liberal or nonliberal, will prove nonneutral in its effects on some comprehensive doctrines or ways of life.

Here one might think Rawls has missed the point. For what is problematic about his liberalism, it might be argued, is that it will prove non-neutral in its effects on doctrines and ways of life permissible on its own account of political justice. But Rawls has not missed the point. Rawls’s liberalism does not rest on a commitment to the value of, nor does it require, a social world maximally diverse with respect to comprehensive doctrines or ways of life willing more or less to accept liberal principles of political justice. Of course, Rawls’s liberalism would be in serious trouble were it to lead to a social world only weakly diverse. But so long as Rawls’s liberalism permits a healthy degree of diversity, to claim that its non-neutral effect on some comprehensive doctrine or way of life is unfair is to presuppose rather than establish the correctness of some competing conception of justice.

Liberalism cannot and should not promise neutrality of effects, but it can and should promise what Rawls calls neutrality of aim. This means that within a liberal political order citizens ought not purposefully arrange their basic institutions or adopt laws to favor or encourage, or disfavor or discourage, particular comprehensive doctrines or ways of life not radically inconsistent with liberal political justice.(2) Provided citizens do not intentionally arrange their basic social structure to achieve non-neutrality, the fact that a liberal regime may be expected to have non-neutral effects must simply be accepted with regret. This is the general neutrality principle of Rawls’s liberalism.

Now, to be faithful to liberal principles of political justice the basic social structure of any modern democracy will have to include a system of compulsory education. However this system is organized, its fundamental aims will, or should, be twofold.

Its first aim will be to provide every citizen with the education and training to which she is entitled as a matter of liberal distributive justice. We need not specify exactly the kind and degree of education and training required. We need note only that insofar as citizens are entitled to a system of equal basic liberties, to fair value for basic political liberties, and to fair equality of opportunity, they are each entitled to literacy, numeracy, various deliberative and critical skills, and a not insignificant degree of cultural, historical, philosophical, political, and scientific knowledge.

Its second aim will be to promote the capacities, virtues, and affections central to the ideal of liberal democratic citizenship. Again, we need not specify exactly what this requires; indeed, absent general consensus over a sharply defined citizenship ideal we may expect public disagreement over both the relevant capacities, virtues, and affections, and how best to promote them. Nevertheless, any ideal of liberal democratic citizenship plausible within the context of Rawls’s liberalism will demand a number of deliberative and critical capacities, political virtues, and civic affections not required by weaker ideals of the citizen as subject. Furthermore, to promote the ideal of liberal democratic citizenship effectively over successive generations, a liberal democracy will have to assign its system of compulsory education an active role in cultivating citizenship. We need note no more than this.(3)

Any basic social structure with a system of compulsory education adequate to the two foregoing aims will prove non-neutral in its effects on some reasonable comprehensive doctrines and permissible ways of life. Indeed, it may have in certain cases the following rather extreme effect: the elimination within a generation or two of such a comprehensive doctrine or way of life previously affirmed or pursued by some citizens for many generations.

This is the sort of case presented by the Old Order Amish in Wisconsin v. Yoder.(4) The Old Order Amish live quiet and withdrawn lives of faith within liberal political societies. They do not seek to employ the coercive power of the state to force their views or way of life on others. They generally do not seek or exercise political power. They are generally self-sufficient as a community and law-abiding as individuals. Their way of life, while antiquated and in many ways internally illiberal, strikes many non-Amish citizens as a valuable way of life, even if one they would never choose for themselves. Members of Old Order Amish communities are permitted to and sometimes do leave for other religious communities or for a life within the secular space of town or city life.

In the Yoder case, members of an Old Order Amish community objected to Wisconsin’s system of compulsory education which required their children to attend a state-certified school, public or private, into the high school years. They argued that forcing their children to attend a state- certified high school endangered (through exposure to various aspects of the modern world) their children’s salvation, their highest good, and threatened the very survival of their community and way of life. Several experts testified that immersing Old Order Amish children in the curriculum and culture of state-certified high schools would likely destroy the community and its way of life within a generation or two.

Thus, Yoder presents a rather hard case for Rawls’s general neutrality principle. There was neither an allegation nor any evidence that the citizens of Wisconsin purposefully aimed through their system of compulsory education to disfavor or discourage the doctrinal beliefs or way of life of the Old Order Amish. While there were non-neutral effects, there were no non-neutral aims. And so, on Rawls’s view, the non-neutral effects must simply be accepted with regret.

But this will run against the intuitions and sympathies of many. Surely, we might think, when the basic social structure of a liberal democracy threatens to overrun within a generation or two the doctrinal unity and way of life of a particular social group unopposed to liberal political values and of no immediate threat to social stability, citizens ought to consider positively accommodating the members of that group. If that means permitting group members to withdraw their children from a system of compulsory education before it has run its course, thereby protecting them from some of its effects, then so be it. The question is whether these intuitions and sympathies may gain any foothold within Rawls’s liberalism.

Rawls’s liberalism underwrites no general right of parents exclusively to direct the education of their children. To be sure, it leaves room for a conditional right of parents to direct much, but not all, of a child’s early education. This right, however, rests not on any natural authority of parents over children, but rather on the reasonable assumption that giving parents a significant degree of control over the early education of their children may prove an effective and important part of a complex set of institutional arrangements, necessarily dominated to some degree by a system of compulsory education, aimed at distributing education and training so as to satisfy two basic aims: liberal distributive justice and the social reproduction of good citizens. Within Rawls’s liberalism, then, the idea of parental rights cannot by itself support parental claims to withdraw children from a system of compulsory education intended and well-designed to satisfy these aims. This is especially true in the context of Yoder where the issue was not parental rights during the very early years of childhood, when they are strongest, but rather during the high school years.

None of this is meant to suggest that Rawls’s liberalism cannot recognize that many parents reasonably assign great value to directing the education of their children and to raising them into a particular set of doctrinal beliefs or way of life. The point is just that a parent must undertake to secure these goods, like any other good, within the institutional framework of a just basic social structure. If it proves impossible to realize these goods fully within such an institutional context, then she must be prepared to subordinate certain desires or to revise her conception of the good.

Rawls’s liberalism does, of course, affirm a basic right to religious liberty. The religious liberty of Amish parents may be thought, then, to provide the needed foothold for accommodationist intuitions and sympathies in this case. Indeed, as we shall see, the Court in Yoder rooted its accommodationist result in just this soil. Yet, setting aside the Court’s reasoning for the moment, we may note that Yoder presents a peculiar conflict between the religious liberty of Amish parents and their children. Amish parents seek to exercise their religious liberty in a manner which may be thought to deny their children social resources essential to their enjoyment of a meaningful opportunity to exercise their own religious liberty as adults. Surely such a conflict must be resolved, within Rawls’s liberalism, in favor of the prospective religious liberty of Amish children.

William Galston has argued that the necessary foothold for the accommodationist intuitions and sympathies to which this case gives rise in many may be found in a liberal commitment to preserving a maximally diverse social world within the constraints of a (thinly) liberal political order.(5) But as I have already suggested Rawls’s political liberalism does not share this commitment, neither at the ground level nor at the level of normative institutional implications.(6)

Will Kymlicka has suggested one possible ground articulable within the terms of Rawls’s political liberalism which might provide accommodationist intuitions and sympathies some foothold in the case at hand.(7) Kymlicka argues that the basic moral powers of citizens at the heart of liberalism presuppose for their development and exercise a horizon of some relatively stable socio-cultural identifications. A quick and radical destruction or decentering of those identifications, therefore, poses a serious threat to what liberalism aims most basically to protect and cultivate. A liberal society ought, then, to avoid institutional arrangements which might radically disrupt the socio-cultural identifications of particular classes of citizens, assuming they pose no immediate threat to political justice or social stability.

With this as a possible philosophical foothold for accommodationist impulses in the case at hand, I want to turn now to the Court’s decision in Yoder, for examining it with the foregoing as backdrop may prove illuminating.

Early in its decision, the Court affirmed the compelling interest, indeed the responsibility, of a liberal democratic society to distribute education and training in a manner calculated to satisfy liberal distributive justice and to cultivate the capacities, virtues, and affections of citizenship. The Court noted that parents may not interpose their own desire to preserve a way of life, however virtuous, as a barrier to reasonable, democratically authorized state regulation of compulsory education. The Court noted also, however, that a system of compulsory education must not violate citizens’ basic liberties, including freedom of religion. Having identified a conflict between the religious liberty of Amish parents and the body politic’s compelling interests in compulsory education, the Court undertook to determine whether compulsory education into the high school years was the means least restrictive of the religious liberty of Amish parents through which the body politic might satisfactorily advance its compelling interests.

The Court found that compulsory education through Grade Eight was sufficient, in the case at hand, to secure the state’s compelling interests. Thus, after Grade Eight, the religious liberty of Amish parents out-weighed the state’s compelling interests. The Court reached an accommodationist result.

Yet, two features of the Court’s reasoning prove problematic. Both concern the Court’s understanding of the state’s compelling interests in a system of compulsory education.

With respect to compulsory education and the social reproduction of citizenship, the Court relied on an impoverished conception of the citizen as subject, emphasizing that the Amish were generally law-abiding and self-sufficient, and thus good, if non-participatory, citizens. Given this conception of citizenship, the Court concluded, not unreasonably, that compulsory education through Grade Eight was sufficient in this case to satisfy the state’s compelling interest in the social reproduction of citizenship. But Rawls’s liberalism entails a far richer, more demanding conception of citizenship.

With respect to compulsory education and liberal distributive justice, the Court reduced the kind and amount of education and training to which Amish children were entitled as a matter of political justice to that which was needed to keep them from becoming a burden on society. The State of Wisconsin argued, consistent with Rawls’s liberalism, that Amish children were entitled to the kind and amount of education (which they pegged at some high school education) needed to insure a meaningful opportunity to make use of basic rights and liberties, to underwrite in public life the fair value of basic political liberties, and to pursue careers or ways of life outside the Amish community. The Court took the State to be arguing only that Amish children must be given enough education and training to insure they don’t become a social burden. The Court then noted that few Amish children grow up to leave the Amish community, and that when they do, they rarely become a social burden (by joining the welfare rolls, etc.). The Court concluded that the State failed to show that in this case education through Grade Eight could not satisfy either of its compelling interests in educating Amish children.

It should be clear that if the Court had drawn on Rawls’s liberalism to reach its understanding of a liberal democracy’s compelling interests in a system of compulsory education, it may well have reached a different result. I do not see how those interests, understood within the context of Rawls’s liberalism, may be satisfactorily advanced without some education into the high school years.(8)

But this leaves us with an unresolved tension or conflict. On the one hand compulsory education into the high school years looks like an essential ingredient (at least in modern democracies) of any system of compulsory education adequate to the demands of Rawls’s liberalism. On the other hand, such a system may have in some cases extreme non-neutral effects which, if Kymlicka is right, Rawlsian liberals have good reason to try to avoid or soften. This tension or conflict is complicated by the fact that the Amish present something like a strongest possible case for accommodation. Unlike other communities of belief threatened by the non-neutral effects of a system of compulsory education (Hasidic Jewish communities, certain fundamentalist Islamic or Christian communities), the Amish neither seek nor exercise political power, nor do they heavily engage or make much use of basic social institutions like open economic markets. This somewhat unique fact of Amish withdrawal lies at the root of and strengthens the accommodationist impulse in the Yoder case.(9)

Seizing on this fact, Jeff Spinner has suggested that the idea of partial or merely formal citizenship may underwrite a principled accommodation in the Amish case without running the risk of over inclusion.(10) The Old Order Amish, Spinner notes, are highly atypical of groups likely to suffer from the nonneutral effects of a liberal regime. They do not seek or exercise political power, and do not participate in economic markets or civil associations. While they benefit from public services and infrastructure to some degree, they pay their taxes and do not consume much in the way of public goods. Given these facts, Spinner suggests, the compelling interests of a liberal society in educating Amish children do not extend to cultivating in them the full range of capacities, virtues, and affections central to the ideal of liberal democratic citizenship, but rather just to insuring Amish children do not wind up staying within Amish communities because they do not know of or are radically ill-prepared to pursue any other options. The satisfaction of this interest may require no more than compulsory education through Grade Eight.

Spinner’s thought seems to lie just below the surface of the Court’s reasoning in Yoder; the Court emphasized the uniqueness and importance of Amish withdrawal. If the idea of partial or merely formal citizenship may be squared with Rawls’s liberalism—perhaps an unlikely prospect given the univocality of Rawls’s conception of citizenship—then Spinner and Kymlicka may provide the foothold needed to support relaxing a system of compulsory education to accommodate the Old Order Amish.

But the Old Order Amish are unique, and the Court has proved unwilling to extend Yoder much beyond its facts. And so we are left with the task of deciding whether to accommodate or assimilate (with perhaps serious consequences for the assimilated) when faced with significant to extreme nonneutral effects on social groups not fully withdrawn from the body politic and committed to comprehensive doctrines or ways of life more or less compatible with liberal political values (certain Native American, Mormon, or Jewish communities, for example). Here we must proceed case by case, with close examination of the facts (which in the case of Native Americans include a long history of oppression) and the cause of the nonneutral effects. If the accommo- dationist impulse remains strong in many such cases, then it may prove necessary to revise Rawls’s general neutrality principle.