Michael Sandel
Michael Sandel; drawing by David Levine


The political system of the United States manages to contain, under conditions of peace if not civility, a remarkable range of moral, ideological, and religious conflicts. The conflicts are not so severe as those that led to the Civil War, but they are greater than those that divide most European countries—where public opinion occupies a narrower political range, and religion is not an important element. Because of its size and regional differences, and the historical shadow of slavery and the Civil War, the United States is radically divided over issues of war, taxes, welfare, race, religion, abortion, and sex.

These conflicts are not just about the best means to pursue generally accepted ends. They are about ultimate values. Yet they do not threaten the stability and legitimacy of the system. Except for a small lunatic fringe, citizens of the United States are prepared to accept the results of the political and legal process even when those results contravene some of their most fundamental convictions. Americans vilify one another as bigoted religious fanatics or morally depraved atheists, racist reactionaries or crypto-totalitarian socialists, but they know they will not be put up against the wall if their party loses an election. That Americans can share a common political system with people whose views they despise, and try to fight out their disagreements legally through the pursuit of power under that system, shows that the cohesion of American society is stronger than its divisions.

This cohesion is possible only because of a general commitment to the principles of limited government embodied in the Constitution. Individuals and groups can be confident that they will be protected by the rule of law from the arbitrary exercise of governmental power, and that the way they conduct their lives will be largely exempt from collective control based on majority preferences. The precise definition of the limits on governmental power is controversial, but no one doubts that they exist.

Their importance has been brought home to us again by the current administration’s contempt for the rule of law and its attempts to evade the limits on executive power, under the color of a war on terror. The worst of these abuses, like torture and indefinite detention, have been mostly inflicted on foreigners, but surveillance increasingly threatens domestic rights of privacy. I am optimistic enough to believe that our society’s attachment to constitutional limits will prevent the domestic abuses from going very far; but in the treatment of non-Americans, there is reason to expect the worst, precisely because of their weak or nonexistent legal protection.

Still, Americans continue to be embroiled in virulent conflicts, largely between conservatives and progressives. They disagree about what, if anything, the state should do about economic, racial, and sexual inequality; about the separation of church and state; about sexual and reproductive freedom; and about what limits, if any, to put on the freedom of expression. Conservatives are more interested in enforcing moral standards on the community and protecting private property, and less interested in protecting personal liberty and reducing inequality; with progressives, it is the reverse.

Within the progressive camp there is a further question about how some of these conflicts should be pursued. Should the argument be about “first-order principles”—fundamental beliefs about religion, abortion, and homosexuality, for example—or should it be about “second-order principles” concerning what kinds of first-order principles may be used to justify the exercise of political and legal power? The liberal strand of progressive thought, shaped by such thinkers as John Rawls,1 holds, for example, that to defend a woman’s right to terminate her pregnancy it is not necessary to prove that the Catholic position that the fetus is a person from the moment of conception is false. It is sufficient to show that under our system of rights, the first-order principles embodied in Catholic doctrine cannot legitimately be used to constrain private choice. This argument could be accepted as a political principle of limited government even by those who hold that abortion is always morally impermissible. The same issue arises about gay rights. In the liberal view, their defense need not depend on the argument that there is nothing wrong with homosexuality; it can be based instead on the narrower political principle that private sexual conduct should not fall under government control.

But another school of thought, which can be described as progressive but not liberal, holds that those who oppose conservative positions on such issues as abortion and homosexuality should engage directly with conservatives on the first-order moral and religious questions. Proponents of this view argue that in defending rights to abortion and to sexual freedom it is a political and philosophical mistake to rely on limits to the legitimate scope and grounds of collective control over the individual. Instead, they maintain, defenders of these rights should argue frankly that conservative religious views on sexual morality and abortion are false. Since that is what most liberals believe anyway, the claim not to be relying on it looks phony. Progressives should not devote their energies to defending individual rights against majority opinion. They should concentrate instead on changing majority opinion.


Michael Sandel, a professor of government at Harvard University, is a prominent contributor to this debate. He is a progressive who is opposed to contemporary liberalism. He believes that liberal appeals to individual rights and to the broad values of fairness and equality make a poor case for the progressive cause, both as a matter of strategy and as a matter of principle. The country and the Democratic Party would be better off, he thinks, if progressives made more of an effort to inspire the majority to adopt their vision of the common good and make it the democratic ground for public policy and law. In the introduction to his new book, Public Philosophy: Essays on Morality in Politics, Sandel writes:

Fairness isn’t everything. It does not answer the hunger for a public life of larger meaning, because it does not connect the project of self-government with people’s desire to participate in a common good greater than themselves.

He goes on to argue:

Liberals often worry that inviting moral and religious argument into the public square runs the risk of intolerance and coercion. The essays in this volume respond to that worry by showing that substantive moral discourse is not at odds with progressive public purposes, and that a pluralist society need not shrink from engaging the moral and religious convictions its citizens bring to public life.

In place of liberalism, Sandel endorses a “republican” tradition of self-government that he identifies with the earlier history of the United States. In contrast to the individualism that he claims is at the heart of liberal theory, republicanism, as he understands it, gives primacy to the communal aim of shared self-government, and what he calls “soulcraft,” the cultivation of virtue in the citizenry by the design of political, social, and economic institutions. It gives personal morality a larger part in political life than liberalism does, and in this respect tries to meet the moralism of the right on its own ground:

The problems in the theory of procedural liberalism show up in the practice it inspires. A politics that brackets [i.e., that excludes from discussion] morality and religion too completely soon generates its own disenchantment. Where political discourse lacks moral resonance, the yearning for a public life of larger meaning finds undesirable expression. The Christian Coalition and similar groups seek to clothe the naked public square with narrow, intolerant moralisms. Fundamentalists rush in where liberals fear to tread.

Anyone concerned over the political success of conservatism in recent years must be interested in this critical analysis.

Public Philosophyis a collection of previously published essays and opinion pieces from the past twenty years, many of them only a few pages long but some quite substantial. Sandel’s general approach is familiar from his previous books, Liberalism and the Limits of Justice (1982) and Democracy’s Discontent (1996), but he usefully summarizes his ideas here and applies them to a range of current issues.


Sandel’s views on substantive issues of social welfare and personal liberty are not very different from those of most liberals. He supports affirmative action, gay rights, abortion rights, and stem cell research, opposes state lotteries and advertising in the classroom, and is doubtful about assisted suicide. In economic policy, he is a sentimentalist who thinks small is beautiful:

Where the liberal regards the expansion of individual rights and entitlements as unqualified moral and political progress, the communitarian [a name for Sandel’s position which he will eventually reject; see below] is troubled by the tendency of liberal programs to displace politics from smaller forms of association to more comprehensive ones. Where libertarian liberals defend the private economy and egalitarian liberals defend the welfare state, communitarians worry about the concentration of power in both the corporate economy and the bureaucratic state, and the erosion of those intermediate forms of community that have at times sustained a more vital public life.

Such nostalgic rhetoric may suggest that Sandel is uneasy about Social Security and Medicare, and would oppose a single payer health system as an extension of the faceless bureaucratic state. But so far as I know, he does not draw such conclusions.2

His theoretical differences with liberalism are more significant, and he has used this disagreement to define himself as a thinker. Unfortunately his understanding of liberal political theory is defective, and his description of the principles and arguments of those he wants to criticize is persistently inaccurate—for example, his claim that a liberal couldn’t have opposed slavery before the Civil War because it was too controversial. Caricaturing the opposition can be a polemical strategy, but in Sandel’s case I believe it is due primarily to philosophical weakness exacerbated by the difficulty of understanding a view that he thinks is wrong. This leads him to interpret it in a way that makes it incomprehensible how anyone else could believe it. To evaluate Sandel’s disagreement with the principles of contemporary liberalism, we have first to disentangle it from his faulty account of what those principles are.


The term “liberalism” applies to a wide range of political positions, from the libertarianism of economic laissez-faire to the democratic egalitarianism of the welfare state. In its European usage the term suggests the former rather than the latter; in American usage it is the reverse. But all liberal theories have this in common: they hold that the sovereign power of the state over the individual is bounded by a requirement that individuals remain inviolable in certain respects, and that they must be treated equally. The state is a human creation, and it is subject to moral constraints that limit the subordination of the individual to the collective will and the collective interest. Those constraints have to be embodied in political institutions. They include not only the familiar freedoms of religion, expression, association, and privacy, but also equality of political status, equality before the law, and, in the welfare state version, equality of opportunity and fairness in the social and economic structure of the society.

Liberal constraints on the exercise of collective power are therefore both negative and positive. The purely negative liberalism of Friedrich Hayek and Robert Nozick is unusual; most American liberals favor not just the protection of individual rights but a form of distributive justice that combats poverty and large inequalities perpetuated by inheritance and class. It is this form of egalitarian liberalism, particularly as represented by its leading theorist, John Rawls, that Sandel has spent his career opposing.

Sandel’s point of attack is Rawls’s central claim that individual rights and principles of social justice should take precedence over the broad advancement of human welfare according to some standard of what constitutes the good life. Rawls’s argument makes precise the familiar moral intuition that the end does not always justify the means, that there are principles of right—principles that govern how individuals should be treated by the state—that may not be violated on grounds of expediency. As Sandel writes:

For Rawls, as for Kant, the right is prior to the good in two respects, and it is important to distinguish them. First, the right is prior to the good in the sense that certain individual rights “trump,” or outweigh, considerations of the common good. Second, the right is prior to the good in that the principles of justice that specify our rights do not depend for their justification on any particular conception of the good life.

The distinction between these two kinds of priority is important because some liberals accept the priority of rights in practice (the first kind) but not in justification (the second kind). John Stuart Mill, for example, was a great defender of individual rights to freedom of expression and to other forms of liberty as a bar to the tyranny of the majority. But in contrast to Rawls, Mill justified those rights on the ground that protecting them strictly was the best way to serve the cause of general human happiness. As a utilitarian, Mill believed that the only way to justify a moral or political principle is to show that it will promote good lives, which to him meant happy lives.

While acknowledging the achievements of utilitarianism, Rawls argues that this derivation of rights and justice from a particular conception of the general welfare is morally mistaken. His main criticism of utilitarianism is that while the promotion of good overall outcomes is important, there is another type of moral requirement that underlies rights and social justice. This is the requirement that a society should treat its members as equals, and it explains directly why there are limits on the degree to which individuals can be subordinated to the collective interest, the general welfare, or the preferences of the majority. Equal treatment means protecting the equal rights of all members of a society, even if they belong to an unpopular minority, and refusing to allow any members to be excluded from social or economic opportunities or to fall below some decent minimum standard of living. Furthermore, since “the right,” so understood, has a different moral foundation from the promotion of good overall outcomes, its principles can be identified without settling some of the major disagreements about the ultimate ends of life that divide the citizens of a typically pluralistic modern society.

Utilitarians have not been persuaded, and neither has Sandel. Sandel is not a utilitarian, since he believes in goods other than the sum of happiness, such as communal solidarity, strong family ties, and the search for higher meaning in our lives. But he agrees with utilitarians, in opposition to Rawls, that individual rights and principles of distributive justice are subordinate to the collective good, and have to be justified by reference to it: “Principles of justice depend for their justification on the moral worth or intrinsic good of the ends they serve.”

The protection of religious or sexual freedom, or freedom of expression, depends, he believes, on whether those freedoms serve valuable ends. So he says protection should be extended to demonstrators against racism and segregation, but not to Nazi demonstrators. On that principle, those who regard homosexuality as sinful should be opposed to allowing a parade for gay rights. Sandel acknowledges that “on any theory of rights, certain general rules and doctrines are desirable to spare judges the need to recur to first principles in every case that comes before them.” In other words, he might accept a fairly strict rule protecting political speech because it would be too time-consuming to decide in every case whether it was on balance beneficial or harmful. But he seems to think that if there are limits on censorship, they have no more fundamental justification than that.

As for freedom of religion, Sandel holds that it depends on the assumption “that religious belief, as characteristically practiced in a particular society, produces ways of being and acting that are worthy of honor and appreciation—either because they are admirable in themselves or because they foster qualities of character that make good citizens.” So if someone believes that this is not true of most religions other than his own, or of atheism, he has no reason, according to Sandel’s theory of rights, to support their toleration. He may in fact have an overwhelming reason to suppress heresy, since it puts other members of the society in spiritual danger. Sandel would presumably regard such a person’s religious beliefs as mistaken, and he may even hold that the value of a religion doesn’t depend on its truth. But he could not fault the opponent of religious liberty for failing to conform to Sandel’s theory of rights or justice.


As these examples show, there is something paradoxical about liberalism, for it asks us on moral grounds to refrain from using the power of the state to enforce on others some of our most deeply held moral convictions about how people should conduct themselves—religiously, sexually, or expressively. This liberal restraint comes from our special moral relation to fellow members of our society—a collectivity that can coerce each of its members, but only if it claims to act in the name of all of them. Sandel’s response, that there are no legitimate rights which cannot be derived from the good, and therefore no limits, in principle, on the use of political power to pursue the good as the majority sees it, is much simpler and easier to understand. It belongs to the enduring tradition of teleological theories, according to which all moral principles are just means to an end, whether it be happiness, salvation, or human perfection.

This disagreement runs through the history of moral philosophy, and will continue to do so. But when Sandel attempts to argue for his position, the result is deeply confused: Sandel finds Rawls’s nonteleological liberalism so incomprehensible that he misinterprets it as a teleological theory with a very special conception of the good, based on a peculiar conception of the self, one that calls to mind Sartre’s existentialism. This is the notion of “the unencumbered self, a self understood as prior to and independent of its purposes and ends.”

The unencumbered self describes first of all the way we stand toward the things we have, or want, or seek. It means there is always a distinction between the values I have and the person I am. To identify any characteristics as my aims, ambitions, desires, and so on, is always to imply some subject “me” standing behind them, at a certain distance, and the shape of this “me” must be given prior to any of the aims or attributes I bear….It rules out the possibility of what we might call constitutive ends. No role or commitment could define me so completely that I could not understand myself without it. No project could be so essential that turning away from it would call into question the person I am….

Only if the self is prior to its ends can the right be prior to the good. Only if my identity is never tied to the aims and interests I may have at any moment can I think of myself as a free and independent agent, capable of choice.

In Sandel’s view, then, liberalism depends on the absurd assumption that no end is valuable unless it is freely chosen by a completely featureless self, without ties, obligations, values, or commitments. From this he draws the political consequence that the liberal state must be a neutral system of rights that refuses to choose among competing purposes and ends, leaving each individual free to pursue those ends on which his choice has conferred the only value it can have. Liberalism is therefore unable to make sense of the idea of the common good. Instead it favors “a procedural republic, concerned less with cultivating virtue than with enabling persons to choose their own values.”

To this caricature of Rawls and other liberals Sandel offers the following counterargument:

Certain moral and political obligations that we commonly recognize—such as obligations of solidarity, for example, or religious duties—may claim us for reasons unrelated to a choice. Such obligations are difficult to dismiss as merely confused, and yet difficult to account for if we understand ourselves as free and independent selves, unbound by moral ties we have not chosen.

I know of no liberal theorist who subscribes to the extreme freedom from moral ties that Sandel describes. It is preposterous to attribute such a view to Rawls, whose argument in A Theory of Justice is explicitly based on the importance of religious commitments and family ties, and who regards communal solidarity and concern for a just version of the common good as fundamental requirements of justice.3

For Rawls, the requirement of limited political neutrality among religions or comparably ambitious secular ideals such as hedonism is based on the overwhelming importance and self-defining character of commitments and values that different members of a society may not share. It is precisely because we care so deeply, in ways we cannot change, about very different things that it is so important to protect individual liberty and avoid the wholesale imposition of ultimate values. Evangelical Christians, atheistic libertines, and Buddhist monks do not have a common vision of the good life. Rawls sees the task of liberalism as that of upholding a form of solidarity and a conception of the common good that respects these differences. If we are going to treat as equals in a collective social enterprise those whose religious convictions we reject, we have to define the common good and the legitimate aims and applications of political control in a way that does not exclude them from the outset.

The alternative would be to allow the majority to use state power to promote the good of everyone as defined by their religion. But if I were part of such a majority, I would then be treating the minority in a way that I could not accept as politically legitimate if I were subjected to it myself—if, that is, my own unconditional religious commitments put me in the minority instead. My sense of justice and equal respect should therefore inhibit the blanket enforcement of my own views in ways that are not essential to the public pursuit of the common good. Taxes, military spending, and environmental policies have to be decided collectively. Religious observance doesn’t—though it took a long time to realize this.


Sandel’s misunderstanding becomes particularly clear in his comments on abortion. Liberals propose to “bracket,” or set aside, the question whether abortion is morally wrong, and to defend the legal right to abortion on the ground that women’s liberty in a personal matter of this kind may not be overruled simply because of the religious convictions of the majority. Sandel’s reply is that we can “bracket” the moral question only if we first determine that the Catholic position is false. He argues:

The more confident we are that fetuses are, in the relevant moral sense, different from babies, the more confident we can be in affirming a political conception of justice that sets aside the controversy about the moral status of fetuses.

This is not a counterargument but a mere begging of the question: to use as a premise the falsity of the Catholic position on abortion is not to “bracket” the question but to answer it, so it cannot be a condition for setting it aside. Sandel has again interpreted the priority of right as being intelligible only if it serves the good.

Sandel’s obtuseness reaches its peak in the longest essay in Public Philosophy, a review of Rawls’s second book, Political Liberalism. There Sandel ridicules Rawls’s aim of trying to set aside conflicts among conceptions of the good life in determining principles of justice by claiming that such a principle of neutrality would have favored Stephen Douglas’s position in the debates with Lincoln over slavery. Because the free and slave states were so divided on the matter, Douglas advocated leaving the choice concerning the authorization of slavery to the individual states and territories, and avoiding a national decision. Sandel sees this as an example of the sort of liberal neutrality with respect to deeply contested moral issues that Rawls favors.

But the state is supposed to be neutral not about all contested moral questions, but only about those that do not have to be decided politically. Slavery, unlike religion or private sexual relations, was a public institution, part of the legal system of property. It carried profound implications for political representation and equality (according to the Three-Fifths Compromise of 1787, each slave counted as three fifths of a person in both the distribution of taxes and the apportionment of seats in Congress). Slavery was part of the basic structure of American society, and had been a central issue of justice since the Revolution.

Insofar as Rawls favors state neutrality and limits to the enforcement of the majority’s values, he does so only in regard to those questions concerning values that can be left to private choice, and that do not have to be answered collectively in order to reach important political decisions. This boundary is itself contested, but slavery was never regarded, either by its defenders or by its opponents, merely as a question of private, personal morality. The idea makes no sense, and Sandel’s invocation of it shows how deeply he misunderstands the liberal position.4

He points out, as if it were an objection to liberalism, that moral disagreements about justice and rights are just as deep as disagreements about religion and sexual morality. But liberals have always known this. They claim only that there are some disagreements about the good life and ultimate values that we don’t have to settle in order to decide collectively how we will pursue justice and the common good. This leaves plenty of disagreements that we do have to battle over, and that demand all our resources of solidarity and trust to settle peacefully.


Notwithstanding his many misinterpretations of Rawls, Sandel’s disagreement with liberalism is real and important. He denies that the right is prior to the good, and he opposes even the limited state neutrality about non-political values that liberals favor.

In the flood of response that followed the publication of Rawls’s A Theory of Justice in 1971, Sandel was grouped with Michael Walzer and Alasdair MacIntyre as a defender of “communitarianism,” as opposed to individual rights, and their argument was called the communitarian critique of liberalism. Sandel doesn’t like the term, and in this book he explains why it doesn’t accurately describe his position:

The term “communitarian” is misleading…insofar as it implies that rights should rest on the values or preferences that prevail in any given community at any given time. Few, if any, of those who have challenged the priority of the right are communitarians in this sense. The question is not whether rights should be respected, but whether rights can be identified and justified in a way that does not presuppose any particular conception of the good.

Sandel rejects majoritarianism or the authority of community values:

The mere fact that certain practices are sanctioned by the traditions of a particular community is not enough to make them just. To make justice the creature of convention is to deprive it of its critical character, even if allowance is made for competing interpretations of what the relevant tradition requires.

Instead, Sandel thinks justice and rights depend on what is actually good, and what rules or institutions serve those ends; he is not a relativist. So there is for him no substitute for moral argument and moral reasoning about what is truly valuable in determining the character of a just society.

He applies this principle most effectively in the essay “Moral Argument and Liberal Toleration: Abortion and Homosexuality.” There he distinguishes between two styles of argument,

the “naive” and the “sophisticated.” The naive view holds that the justice of laws depends on the moral worth of the conduct they prohibit or protect. The sophisticated view holds that the justice of such laws depends not on a substantive moral judgment about the conduct at stake, but instead on a more general theory about the respective claims of majority rule and individual rights, of democracy on the one hand, and liberty on the other.

He prefers the naive view and it is hard not to sympathize with him. He believes he can show that abortion is not wrong and that homosexuality is just as good as heterosexuality, and he is willing to stake the legal issue on those claims. As I have said, there is something paradoxical about the sophisticated, liberal alternative. Why should anyone, except for strategic reasons, want to defend the legal right to abortion and toleration for homosexuals in a way that someone who regards both practices as sinful can accept?5

The answer depends on a certain understanding of the complexity of moral theory. This is a deep issue: Do all moral standards derive from a single principle, or are there different basic principles for different kinds of entities? Rawls, for example, believed that the moral standards for social and political institutions were not derived from the standards for personal life, or from a common principle that yielded them both. Rather, he thought that justice, which is the special virtue of social institutions like the state, depended on the distinctive moral character of the state itself, as an immensely powerful form of collective agency.

As citizens, we are subject to the will of the majority, coercively enforced. The extent of that control, and the grounds on which it is exercised, are what is at issue. Rawls believed that constitutional limitations and requirements should reflect the democratic ideal that each member ought to be able to regard the system as acting on his authority, even when he may disagree with particular decisions or policies. He believed that this required not only political equality and civil liberties, but also strong measures to combat racial, sexual, and socioeconomic inequality. This is the fairness that Sandel derides.

Sandel’s alternative, the untrammeled pursuit of the good, means that the state is not governed by special moral standards because of its collective power. He thinks we should join together to decide on the true ends of life, and then use the power of the state to create virtue and give everyone’s life a meaning as part of something larger than themselves. This will lead to liberal toleration if we accept Mill’s conception of human good. Or if we accept what seems to be Sandel’s ideal, it will lead to an unmaterialistic culture of closely knit communities and strong family ties. But it will lead to theocracy, fascism, or communism for those who accept alternative conceptions of the human good.

The question, finally, is whether there is a special moral requirement that applies to political institutions and that has enough force to inhibit the coercive enactment of more comprehensive ideals. Sandel is convinced that we cannot justifiably subordinate our deepest personal convictions about the ultimate good for humanity to a system of respect for individual rights. Rawls and other Kantian liberals think that respect for our fellow citizens provides the moral resource needed to justify the protection of rights, and that it defines the restricted political terrain on which we ought to argue about our common institutions.

Liberalism may be a minority conviction in the world at large. To most people values are values, and political power should be used to implement them: What else is it for? But Sandel’s ideal republic of comprehensive virtue would abandon a form of civic respect that has been of inestimable value, and threaten one of the indispensable grounds of political stability in our free, stormy, magnificently diverse nation. To use a phrase of Jürgen Habermas, constitutional patriotism should be enough to satisfy what Sandel calls our “hunger for a public life of larger meaning.” A hunger that demands more from the state will lead us where history has shown we should not want to go.

This Issue

May 25, 2006