1. Introduction: Overlapping Consensus

In 1971 John Rawls published A Theory of Justice; in time, he appears to have become dissatisfied with the shape his theory had originally taken. Rawls traces the roots of this dissatisfaction to a paper he received from Samuel Scheffler in 1977 (published in 1979). Scheffler's paper argued that the account of stability developed in A Theory of Justice was in conflict with Rawls's “The Independence of Moral Theory,” section III. Rawls says of Scheffler's paper that he remembered it as “the moment … when I started thinking whether and how far the view of Theory needed to be recast” (PL, xxxii–xxxiii).

The problem with A Theory of Justice, Rawls says, is that he had assumed that the two principles of justice as fairness (the principle of equal basic rights and liberties and the principle of fair equality of opportunity paired with mutual benefit in outcomes and, in the ideal case, the greatest benefit of the least well-off group) would become part of an overarching moral theory in any well-ordered society in which these principles were the public principles of justice. Such a society would be stable because everybody in it would continue to hold to the two principles in the light of this overarching moral theory, which contained those principles as an integral part. But such uniform acceptance, Rawls now says, is implausible. (For Rawls's own account of the problem, see PL, xv–xviii; also “Justice as Fairness: Political not Metaphysical,” 414 n33.)¹

In Rawls's writings in the 1980s (especially in “Justice as Fairness: Political not Metaphysical,” “The Idea of an Overlapping Consensus” (IOC), and “The Domain of the Political and Overlapping Consensus”), he argues that there is going to be, in a continuing free and open society, an irreducible pluralism of reasonable comprehensive moral and religious and philosophical doctrines. This pluralism of principles is a permanent and ineradicable fact, now and for the foreseeable future. In his writings since 1985, Rawls seems especially concerned with the problem of assuring political stability in a pluralist or multicultural social environment. Rawls gives this preoccupation its most complete elaboration in Political Liberalism.

Rawls claims that his theory is specifically a political theory of justice, which is itself not a general or comprehensive critical moral theory. Rather, the most significant feature of Rawls's current, revised theory is that he takes the public political culture of a contemporary democratic society to be the deep background of the entire theory. For the leading ideas out of which the political conception of justice is to be constructed and by reference to which it is to be justified are said by Rawls to be implicit in that culture (PL, 13, 15, 175, 223).

Here political justification sets out from four “model conceptions” or “fundamental ideas” (PL, Lecture I). First is the idea of the person or citizen as free and equal and as having two distinctive capacities or moral powers and two corresponding “higher-order interests” in the realization of these capacities (PL, 74, 75). Thus, each person has, over that person's entire life, (i) an interest in being able to have, formulate, revise, promulgate, live according to, and advance one's particular determinate conception of the good, and (ii) an interest in exercising one's “sense of justice” and being motivated by it, providing others do so as well. To amplify point (ii), each person has an interest in living with fellow citizens on terms of mutual respect under a unified and stable scheme of basic political and economic institutions that has been organized by a shared set of principles of justice (PL, Lectures II and VIII). Next is the idea of fair cooperation by free and equal citizens for mutual (or, better, reciprocal) benefit. Third is the idea of the well-ordered society and its basic institutional structure, and last is the idea of a linking or mediating conception which lays out the standards for discussion and decision-making to which fellow citizens could be expected to adhere in reaching a decision respecting the governing principles of political justice (principles for the basic structure of their well-ordered society, in which they could expect to live their entire lives). This fourth idea is linked with what Rawls called “the original position” in his earlier book (TJ, ch. III). Herein would be included such ideas as sharply limited information (the so-called veil of ignorance), publicity, and unanimity. The function of this mediating conception is to help unify the other fundamental ideas into a single coherent whole from which one could then reason to certain principles and institutional arrangements.

Establishing terms for social cooperation for mutual benefit – principles for a fair distribution of certain primary goods (including such things as liberties, opportunities, social and economic positions, income and wealth) – continues to be the main object of Rawls's conception of justice. In the new account, though, the principles that emerge as preferred (from among a small set of historically available candidate principles) are the principles that are best supported by the background democratic ideas, from within the nexus formed by the four “model conceptions.” The preferred principles are the principles most appropriate to these fundamental ideas, under the assumption that there is going to be, in a continuing free and open society, an irreducible and incommensurable pluralism of reasonable comprehensive moral and religious and philosophic doctrines.

Rawls thinks that those best-supported principles will be his own two principles of justice, understood now as political principles. (See PL, 5–6 for Rawls's current version of these principles.) Or, to be precise, he thinks the preferred set will actually be a “family” of principles, among which are included the two he emphasizes. (See PL, xxxvii, xlv, xlvi, l–li, 7, 164, 439; ”The Idea of Public Reason Revisited” (IPRR), 581, 582, 584, 592, 594, 607, 611.)

This “family” is constituted by the set of “generic” liberal principles. Generic liberalism, as Rawls conceives it, has three main features: (1) certain familiar rights, liberties, opportunities are to be singled out and specified and maintained; (2) a certain priority is to be given to these rights, etc. over against “the claims of the general good and of perfectionist values”; (3) measures to help citizens make effective use of these rights (etc.), by having an adequate base of income and wealth, are to be set in place. (See PL, 6; also xlvi, 156–157, 375; IPRR, 581–582; LP, 14, 49.) These generic liberal principles are well designed to specify an acceptable distribution of primary goods in the context of existing democratic “fundamental ideas.”

The “political conception of justice,” as Rawls calls it, is not limited to such principles alone. It also includes certain of the institutional arrangements that are required to put the principles into effect in a given society. These institutions – political, economic, social – are the sort of thing Rawls had in mind when he referred to the basic structure of a society. But he tends throughout to emphasize the basic political institutions.

Rawls's account of the political conception proceeds in two main stages (see PL, 64–65, 140–141, 385–388). The first stage is the one we have focused on up to now. The main project here is to settle on that principle or set of principles for distributing primary goods which is optimally appropriate, given the fundamental democratic ideas from which we started. This first line of justification (justification from democratic ideas in a democratic context) is said by Rawls to be “freestanding,” in the sense that it draws only on these background democratic ideas (PL, xlii, 10, 12, 25 n27, 40, 140, 374–378). Such justification is independent (PL, 144), and does not draw, in an essential way, on the ideas or values of any comprehensive moral or religious doctrine.

What Rawls calls overlapping consensus is a second stage in which the antecedently established “freestanding” justification is endorsed from the respective points of view of a variety of comprehensive ethical doctrines (such as Kant's moral theory or Mill's utilitarianism) and religious doctrines (such as contemporary Catholic Christianity). On this view, the political conception is a “module … that fits into and can be supported by various reasonable comprehensive doctrines that endure in the society regulated by it” (PL, 12; also 145, 387). But it need not be presented by reference to such support initially; rather it is established completely independently of direct consideration of any and all such doctrines (hence Rawls's description of it as “freestanding”).

At this second stage, we contemplate the support provided the political conception from within the confines of a variety of comprehensive views. In some of these cases such endorsement will follow a pattern of derivation; in others it will be more like endorsement based on an instantiation model – here the political conception counts as a feasible real-world exemplification of the comprehensive view in question. And, for yet others, it will be endorsement only in a weak sense: the political conception is said merely to be compatible with – that is, not incompatible with – the comprehensive doctrine in question (PL, xix, 11, 140, 169–171, 242 n31; see also 158–164). In any event, where several different comprehensive doctrines, each for its own reasons, can converge on and support a single political conception (for example, the generic liberal conception outlined in PL, 6) in one of these ways, we say that there is an overlapping consensus among these comprehensive doctrines, all centering on this common focal point.

2. Constitutional Consensus

Rawls is concerned that the idea of overlapping consensus might seem utopian. He tries to suggest that it is not by locating the overlapping consensus within an undergirding substructure which he calls constitutional consensus.

Constitutional consensus is one of the important new ideas in Political Liberalism, and one of the least discussed (for instance, there is no mention of it in Brian Barry's critical review essay, 1995, or in Weithman 2010 – see 311 – and no index entry for it in the Cambridge Companion to Rawls, Freeman 2003a). Constitutional consensus is first introduced in Rawls's argument (in PL, Lecture IV, §6) where he attempts to deal with the charge of utopianism by showing how it is possible to move away from a mere political modus vivendi to a quite different sort of social order.

2.1 Main Themes of Constitutional Consensus

In a mere modus vivendi, certain principles and practices are accepted as a way for people to live together without constant fighting and disruption. Rawls's own example of accepting religious toleration (in a time of deep intolerance) is one important instance. Both Catholics and Protestants in the sixteenth and to some extent the seventeenth century “held that it was the duty of the ruler to uphold the true religion and to repress the spread of heresy and false doctrine; [here] the acceptance of the principle of toleration would indeed be a mere modus vivendi, because if either faith becomes dominant, the principle of toleration would no longer be followed” (PL, 148).

The agreement that exists in such a case is not wide (it covers only a fairly narrow range of rights – for example, the right of religious toleration – and institutions). It is not deep (in that the reasons offered for the desirability of these accepted arrangements do not go beyond the idea of establishing a modus vivendi). And it lacks a distinctive political focus: fellow co-inhabitants have no shared conception of a public political life – no animating reasons, widely accepted, that would take them beyond the status quo, beyond the modus vivendi itself; they have accepted that status quo simply out of fear of a worse alternative.

There is nothing in a mere modus vivendi that goes beyond the horizon set by the co-inhabitants' own morality or their own religion or their own philosophical view of the world (though such reasons are often coupled with considerations of prudence and of strategy). No constituency has been created for a shared public political conception and there is no principled commitment to it (PL, 392).

A constitutional consensus comes about, then, as the agreed-upon area of rights and practices widens; it comes about as the ground under that area deepens (as convincing political reasons for having such arrangements, reasons that go beyond the mere utility of a modus vivendi, gain acceptance and are taken on board). And it comes about as a conception of public principles of justice, with greater focus and definition, gains widespread support. These moves away from a mere modus vivendi allow a space to be created for citizenship. This new dimension creates the availability of a new role for political co-inhabitants, that of fellow citizens. (Rawls's main discussion of constitutional consensus is found in PL, Lecture IV, §§3, 5–7.)

The story told in Political Liberalism (respecting constitutional consensus) largely emphasizes a vigorous version of generic liberalism (as found in PL, Lecture VI, §5). It is, presumably, where democratic political culture is at today. (i) Here the consensus has become wide enough to embrace most of the well-known constitutional rights, liberties, and opportunities and to cover the main contemporary democratic political institutions – universal franchise, contested voting, and majority rule decision making. And it is wide enough to embrace such further basic structure institutions as an open and competitive market (but ultimately subject to political overview) and a system of free state-supported schooling. (ii) And the consensus has become deep enough to draw, as background, on a well-established, viable democratic political culture and to draw on the “model conceptions” latent there as the ultimate ground of justification of a “freestanding” political conception of justice. (iii) This political conception has gained increasing focus and detail. It embraces the main features of the “family” of liberal conceptions and accepts the idea, expressed there, that “these elements can be understood in different ways, so there are many variant liberalisms,” including Rawls's two principles of justice as fairness as one of the options (PL, 6, 164, 223; IPRR, 581–582).

Rawls takes the term “constitutional consensus” from Kurt Baier, but Rawls means by it something very different. For Baier, in the paradigm case, a constitutional consensus is procedural in character – a mere “agreement on the process of adjudication,” as he puts it – and would lack, essentially lack, agreement on a conception of justice; in particular, it would exhibit no convergence on such things as the principles of generic liberalism or Rawls's own two principles of justice. (See Baier 1989, 775, 776, 789.) For Rawls, a constitutional consensus is never simply an institutional matter, never simply an agreement on political procedures (PL, 149, 164), and it is open to the possibility of agreement on political principles of justice, certainly to agreement about some version of generic liberalism.

2.2 Temporal Development and Orderly Contestation

One important idea in constitutional consensus is that it involves change over time. Rawls presents the generic liberal constitutional consensus, just described, as itself based on successive and agreed-upon moves away from acceptance of a mere political modus vivendi (PL, xxxix, 159, 208).

At the stage currently reached (in a vigorous version of generic liberalism), the set of generic principles and the practices that apply them are in place because they constitute, on the balance of reasons, a sound interpretation and application of the “model conceptions” in the background democratic political culture. These principles and practices reflect the institutional history embedded in that particular culture and the profile of existing basic structure institutions that this history has currently given rise to.

This process of change can be presumed to continue over time. Such change could involve the fundamental ideas themselves, or the constitutional and other institutional essentials, or the liberal principles themselves and the “family” they constitute (PL, li; IPRR, 583). But there is a deeper background change that should be noted here. In the late seventeenth and in the eighteenth century, as the idea of constitutional government took hold, parliamentary or republican models competed with theories of royal absolutism and became in time the main options. The ideas of electoral democracy and representative government began to grow in prominence in the nineteenth century and became, in the twentieth century, the preferred view in liberal democratic societies.

We need to take account of this particular shift to egalitarian democracy as we consider Rawls's own ideas. In Political Liberalism, Rawls modified the statement, the actual formulation, of his first principle (the principle of equal basic rights and liberties) to give special place to the political participation liberties. He added there that “the equal political liberties, and only those liberties, are to be guaranteed their fair value” (PL, 5). By “fair value” Rawls means that steps are taken to make people substantively equal (or more nearly equal) in their exercise of these liberties, in voting and in campaigning (LP, 197–199; JF, 148–152). Among the steps designed to do this are such things as public funding of elections (JF, 131, 149).

I have, up to now, emphasized change as one of the main motifs in Rawls's theory of constitutional consensus. But there's one change he doesn't contemplate; he doesn't really consider a change that takes us beyond the family of liberal principles in the direction of one single agreed-upon liberal principle or set of principles. At one time, I thought that Rawls believed that, eventually, constitutional consensus would single out justice as fairness and its two principles as the overall preferred liberal theory (Martin 2001, 86–87). I now believe that this view was a mistake.

It isn't borne out by a careful reading of Rawls's post-1993 texts. But there's an even stronger reason for rejecting my earlier reading. If we look at the history of democratic liberal societies from its beginning point in the nineteenth century, we note that democratic liberal societies are all characterized by the fact of having two- or three-party political systems (or, in some cases, multiparty systems) that controlled contestation at the electoral level and at the parliamentary one. In each case the number of viable liberal theories probably exceeded the number of political parties, certainly in the two- or three-party systems. Political systems that had only one-party governments were characteristically autocratic and antidemocratic/antiliberal in character.

The rise of egalitarian democracy has changed and added to the public political culture of liberal societies. And, just as significantly, it has altered the shape of liberal theories, moving them away from the ideal of a single agreed-upon principle or closely integrated set of principles toward a family of rather variegated liberal principles – a family of principles having certain generic features in common and offering varying and even competing interpretations not merely of policy issues but also of the constitutional and institutional essentials in the various liberal polities and even of the fundamental ideas of a democratic public political culture as such (see IPRR, 585–586 n35).

The idea of a family of liberal theories accords nicely with fundamental democratic thinking and practice; the ideal of a single liberal principle, agreed upon by all, does not (see Dreben 2003, 322, 338). One might wonder, then, what is the place of justice as fairness and its two principles in constitutional consensus, as Rawls sees that consensus.

Rawls thinks that justice as fairness would continue to be one of the primary contending perspectives on political justice within the family of liberal conceptions. Indeed, it would be at the “center of the focal class” there because, in his view, it fits in well with the fundamental democratic ideas and is compatible with the developing pattern of democratic institutions. And Rawls, personally, regards justice as fairness as the “most reasonable” member of this family. (For the quoted phrases, see PL, 168, xlvi–xlvii; also 164, 167 and IPRR, 582 n27.)

What we can reasonably expect is that the process of broadening, of deepening, and of gaining greater specificity and focus in the family of liberal principles will go on for the foreseeable future. Justice as fairness can be expected to continue to be among the viable principles in that family. And political debate will be framed by the interpretations that the principles (in this family of generic liberalism) take from the public political culture, with its ingredient fundamental ideas, and apply to the constitutional and institutional essentials and to debates and decisions about matters of policy, insofar as such essentials bear on these matters.

There is no teleology in Rawls's theory beyond what I've just described, no ultimate goal picked out by history or by philosophical reason (be that goal justice as fairness per se or any one of the other liberal principles, within generic liberalism, that contend with it). Rawls says:

It is inevitable and often desirable that citizens have different views as to the most appropriate political conception; for the public political culture is bound to contain different fundamental ideas that can be developed in different ways. An orderly contest between them over time is a reliable way to find which one, if any, is most reasonable.

(PL, 227)

A mere modus vivendi is not a political conception of justice; it lacks the dimension of depth altogether. But all the liberal political conceptions of justice have this dimension (sufficient to count as freestanding). The liberal conceptions have one other feature in common, besides the dimension of depth.

They all have a fairly high degree of consensual acceptance and support by their citizens. (1) The specific rights and practices the conceptions range over (dimension of breadth) are widely accepted. (2) The various articulated liberal conceptions of justice (within the family of generic liberalism) can be supported, to an appreciable degree, by sound arguments. Citizens may differ in that some think one conception is the best supported, and others others; but they agree in believing that a certain threshold of argumentative adequacy has been passed by them all – or at least this is the (rebuttable) presumption (PL, xxxvi, xlvi–xlviii, 223, 226–227, 241; IPRR, 578, 581–583). Here a pattern of support (or justification) by reference to the fundamental political ideas is in place for the various principles within that family; these principles allow cooperation on the basis of “political terms that everyone can accept” (PL, 163). (3) Finally, there is consensus over “the general structure of political authority” (PL, 393), that is, consensus over the legitimacy of a fairly detailed set of constitutional essentials, the features of which “all citizens may reasonably be expected to endorse” (PL, Lecture VI, §§5–6; see 217 for the quotation; see also xlviii, 137, 226, 241, 393, and IPRR, 578–579, 594). This consensual element, as identified in these three points, is a large part of what Rawls intends when he says a given liberal political society is well-ordered (see PL, Lecture I, §6).

It seems, then, if we find a sizable degree of citizen support at these three points, an important degree of stability is built into the public political conception of justice (of the sort formed by the family of liberal principles with certain generic features in common). A vigorous version of constitutional consensus (focused on the family of liberal political conceptions) would represent a high level of development – in the dimensions of breadth, depth, and focus – and would enjoy considerable consensual support. The degree of internal stability of any such family of liberal conceptions would, arguably, be considerable; such an order would be less likely to be overturned by divisive issues (see PL, 161–164, Lecture IV, §7). The political consensus here covers more important areas, for deeper and better reasons. And the public has come to recognize the “great values of the political virtues” – the virtues of tolerance, civic peace and order, civility, cooperation and trust – embodied in such a political consensus (PL, 157, 171).

A constitutional consensus robust enough to attract and to support overlapping consensus would itself go a long way toward solving the problem of stability on its own. Thus, the same measure – that is, constitutional consensus – which solves the utopianism problem for Rawls also appears to undercut and perhaps even remove much of the raison d'être he had offered, in the first place, for overlapping consensus.

3. Overlapping Consensus: Stability or Public Political Justification

This raises the question, much disputed in contemporary Rawls scholarship, of what the main point of overlapping consensus is: is it stability (as Rawls initially had suggested) or is it justification? Some – Charles Larmore (2003, 377), Thomas Scanlon (2003, 159–160), and Charles Beitz (2009, 76–77) – have put the emphasis on stability, to the exclusion of justification; others – Samuel Freeman (2003b, 36–37) – have made justification the main concern.

Paul Weithman's recent account falls somewhere in between. He sees stability (an idea which he develops in a very complex and nuanced way) as the principal object of overlapping consensus. But he also sees that drawing on comprehensive moral and religious doctrines, as found in an overlapping consensus, will provide a distinctive sort of background support for political stability. However, he does not see overlapping consensus as providing justification per se for the political conception; instead his discussion of any such support is always deflected back onto the issue of stability (Weithman 2010, ch. 10, §§6–8). Weithman's middle ground is closer to the stability pole than to the justification pole.

These two emphases (on stability and on justification) can be contrasted. Stability is basically a matter of fact (of likelihood). In Rawls's view a political system tends to be stable (i) when, drawing on its own resources, it is able to generate from one generation to the next widespread support for itself (adequate to satisfy the citizens' sense of justice), and (ii) when there is congruence over time, for most citizens, between a conception of their own good (as informed, in the typical case, by one comprehensive doctrine or another) and the operating principles of justice in their society. In a situation of longstanding and irreducible pluralism of reasonable comprehensive moral and religious and philosophical doctrines, overlapping consensus would contribute to stability so understood.

Justification is a normative issue. Overlapping consensus – in deploying several different comprehensive doctrines that converge, each for its own reasons, on a single political conception – plays a normative role (alongside freestanding justification) by providing a distinctive form (moral in character) of public justification for the principles of a liberal political order.

Granted, one can distinguish these two emphases (on stability and on justification); but there's no need to separate them. They can have complementary and mutually supporting roles in Rawls's account. The second of these emphases has received less attention. I want in this section to develop a view which stresses the justificatory character of overlapping consensus.

Rawls thinks that an overlapping consensus presupposes and arises in concert with constitutional consensus. In the case envisioned, there is already an independent, widespread, and long-lived support by citizens for the public political conception – that is, for a family of liberal principles. (See PL, 164, 168.) An overlapping consensus would arise, then, where the great bulk of citizens could also affirm, upon reflection and given experience, that the fundamental ideas and the governing principles and institutional essentials of the public political conception were compatible (or could be made so), in each of their respective cases, with the various comprehensive moral and religious and philosophical doctrines that they individually held. (See PL, 140, 160, 187–188, 210, 386.)

The general run of citizens in that society don't regard the perspectives they individually have as incompatible, in general or in principle, with the overall public political conception there. If a sizable number of the citizens actually are as reflective as ordinary people can be expected to be, then it follows, for the variety of diverse perspectives that happen to be held by these citizens en masse, that these diverse views constitute (or can be regarded as constituting) an overlapping consensus on a family of liberal principles and on a given set of institutional essentials. Rawls describes this as a “concordant fit between the political conception and the comprehensive views together with the public recognition of the great values of the political virtues” (PL, 171; see also 158).

Here it is not so much that various comprehensive doctrines (understood as “isms”) converge on a single public political conception of justice; rather, it is that a considerable bulk of citizens, coming from diverse perspectives, do. In the end it is the citizens themselves – “a substantial majority of … politically active citizens,” as Rawls puts it – who decide (PL, 38). This is consistent with his claim that here there are no “experts,” no philosopher kings (PL, xxvi–xxvii, 383, 426–427; also Scheffler 1994, 11–12).

An overlapping consensus, like the public political conception on which it focuses, is directed to the basic framework, to the institutional essentials themselves; it does not require or imply an agreement on all matters of policy. Indeed, it is compatible with disagreement, even considerable disagreement, on such matters – on detailed pieces of legislation, say (see IPRR, 604–607).

Such an overlapping consensus would, arguably, be stronger and more enduring were it to attach itself to a suitably broad, deep, focused public political conception, for example, to a family of liberal political conceptions (with justice as fairness among its members). An overlapping consensus of this sort would, of course, occur gradually; it would take time to gel. (See PL, 160 n25.) None of this shows, of course that an overlapping consensus will occur: nothing is guaranteed. It shows merely that such a consensus plausibly could occur, in the way Rawls envisioned. (See PL, xlv–xlvi.) Its occurring in that way is not utopian.

What would overlapping consensus add here? What overlapping consensus provides is not political stability per se, but rather “stability for the right reasons” (PL, xxxvii, xli, 388 n21, 390, 391, 394; IPRR, 589).

A political conception, simply on its own, is always a consensus within and from public political reasons; in the case at hand, the reasons appropriate to a liberal democratic society. As such it lacks a certain dimension: it lacks moral credentials of the sort afforded by a comprehensive critical moral theory. Several different critical moral and religious doctrines can be drawn on, at a given time, in a liberal society. Each of them is controversial; no one of them is accepted by everybody. Insofar as we are concerned, then, with anything like a full public justification of the political conception, using accredited critical moral doctrines, we must accept that the doctrines there admissible as premisses in that form of justification, doctrines such as the utilitarian general happiness principle and Catholic natural law, are clearly not acceptable to all reasonable citizens. The only form a full public justification could take in a morally and religiously pluralistic society and still have authority outside a narrow circle of partisan sentiment would be as an overlapping consensus – with the public political conception as focal – of these various doctrines.

Citizens in general could not participate as adherents in each of these lines of moral or religious justification. What they could do, as adherents of one such line (or even of none), is to note and register the fact of overlapping consensus. This fact is a public fact and general acknowledgment – common knowledge – of this fact is the form that overlapping consensus would take insofar as it was itself a matter of full public justification of the political conception. (For discussion of what Rawls calls a “full justification” and of “public justification of the political conception by political society,” as both of them aspects of overlapping consensus, see PL, 385–394, also 67.)

If it could be established as a matter of public fact (based on settled judgments of compatibility by the great bulk of citizens) that various of the main present-day comprehensive moral doctrines and religious faiths could converge – each for its own reasons (PL, 134) – on one and the same public political conception of justice, as a common focal point, then that particular conception would be fully and publicly accredited by the standards of these various comprehensive doctrines. And this means that each citizen can affirm that the political conception is a reasonable one; it can be regarded as mutually intelligible among them and as justifiable, under one or another of the constituent doctrines in the overlapping consensus, on normative grounds that all (taken severally) can endorse. In an overlapping consensus, the political conception is “affirmed on moral grounds” (PL, 147; see also ”Justice as Fairness: Political not Metaphysical,” 410–411; IOC, 422, 432; “The Priority of Right and Ideas of the Good,” 470–471; PL, Lecture IV, §3, and xli, 126, 150–151, 168–169, 208, 211 n42).

Thus, even in the face of a continuing and very likely ineradicable pluralism, we would have achieved stability (as provided by a public political conception embedded in a constitutional consensus) and for the right reasons (as provided by an overlapping consensus, and not a mere compromise, among the various relevant critical moral and religious doctrines). A mere compromise would be a tenuous solution, continually subject to renegotiation as the balance of powers and interests shifts (PL, 148, 161). By avoiding such constant renegotiation, or the continuing threat of defection, overlapping consensus reinforces the existing stability of a suitably broad, deep, and focused public political conception; and this is the contribution it makes to the stability of such a regime.

Rawls had assumed in A Theory of Justice that, since his preferred principles of justice came out on top in the contest with utilitarianism and with perfectionist values such as Platonic aristocracy or Nietzschean elitism, these principles would in effect be endorsed by everybody, and for the same reasons. Thus these principles would become the moral theory or part of the moral theory of any well-ordered society whose principles of justice were constructed in the original position behind the veil of ignorance, subject to the constraints of publicity and unanimity. Such a society would be stable, in short, because everybody in it would adopt the two principles as among their moral principles. In fact, there would likely be (given the argumentation of A Theory of Justice) a high level of consensus, amounting to virtual unanimity, in favor of one comprehensive moral doctrine in particular; and that one moral doctrine would be a version of contractualism. Ultimately, then, it is this almost universal convergence upon a single justifying moral theory that underwrites Rawls's account of stability in A Theory of Justice.

Rawls makes the point here, that his “premiss” in A Theory of Justice was that there would be an almost universal convergence on a single comprehensive moral doctrine, quite explicit in his later writings. (See PL, xv–xviii, xl, 388 n21; IPRR, 614; see “Commonweal Interview with John Rawls,” 617, for the point about contractualism.) Such uniform acceptance, Rawls now says, is inconsistent with the idea that a pluralism of reasonable comprehensive moral and religious doctrines is here to stay, at least in any modern political society committed to free and open discussion. (See PL, xv–xviii, xxiv, xl, 4, 36–37, 129, 144.)

It was this problem, of moral justification under conditions of pluralism, which A Theory of Justice had conspicuously failed to solve. And overlapping consensus, one of Rawls's new ideas in Political Liberalism, was called upon to address this self-professed defect.

In Political Liberalism the problem of political stability, in a world of moral pluralism, was tackled first and on its own, using political devices: constitutional consensus and the attendant idea of institutional essentials that all citizens “may reasonably be expected to endorse.” The solution to the next problem, that of critical moral justification, was first tailored to confront the fact of a presumably permanent and unresolvable pluralism of moral and religious doctrines and was then brought to bear, in the idea of overlapping consensus, on a “freestanding” political conception of justice and on a preexisting political solution to the problem of internal stability. The job of overlapping consensus is to provide an independent critical moral grounding for each, for the public political conception itself and for the inbuilt stability afforded by that conception. Overlapping consensus is directed at the issue of stability for the right reasons. The argument I have sketched here suggests that a pluralistic society, given freestanding justification and overlapping consensus, could be well-ordered and stable.

4. Utilitarianism and Overlapping Consensus

There remains an important concern with overlapping consensus. Can it actually draw on most, if not all, of the major comprehensive moral and religious doctrines that are prominent in the societies that call themselves liberal? In particular, the question is whether traditional utilitarianism could be a contributor to and supporter of an overlapping consensus that does not assign priority to the general good or to the greatest happiness.

Contrary to what he says elsewhere, in Political Liberalism Rawls suggests that utilitarianism could support not only equal “rights, liberties, and opportunities” but also their priority over “claims of the general good,” as he puts the matter in his discussion of generic liberalism (PL, 6). At one time, I thought Rawls's latter claim, concerning priority, was simply wrong (see Martin 1994, 757–761). The problem was that utilitarianism, where we stick with the prevailing interpretation of it, cannot allow for moral or constitutional rights that have a built-in, standing priority over considerations of general (or common) good, understood as maximized aggregate welfare.

In an effort to deal with this problem, significant attempts have been made within utilitarianism (under the name of “rule” utilitarianism) to address and perhaps resolve it. The theorists of rule utilitarianism assert that direct appeals to the greatest welfare are self-defeating, all things considered, and that putting standing constraints on the basic utilitarian principle – such as a system of moral rules or a coherent set of civil or constitutional rights justifiable by the utilitarian standard – in fact produces the greater well-being.

Rule utilitarians do not, however, assert that moral rules should never be overridden nor individual rights ever broached. Rather, on their view, where rules conflict or rights do, some sort of appeal to the greatest happiness is in order. Here is where the notion of indirect utilitarianism – as developed by John Gray (1996), David Lyons (1994, chs 2–5), and others – comes crucially into play. Its advocates argue that the happiness principle should not directly determine what is to be done even here. Rather, the principle operates only indirectly in all such cases. Here the general welfare principle is used merely to help contribute, in a continuing way, to an ongoing cumulative ranking of “the opposing obligations” so as to achieve a refined and resultant clear ranking of those obligations, for use on particular occasions of conflict. (See Lyons 1994, 61.) In indirect utilitarianism there is never a direct and determining appeal to the general happiness principle as to what to do, what act to perform or what rule to follow, on a particular and given occasion.

On the indirect utilitarian account, it is possible to have policies for action (to have both moral rules and rights) that are justifiable by the standard of general happiness and at the same time to shield these policies from direct confrontation with (and possible overthrow by) the happiness principle on individual occasions. And it is possible to do so while still allowing these policies to remain sensitive to what produces the greater or more general benefit – a sensitivity that is registered in the differential weights assigned the various rights and policies, an assignment that occurs gradually (over time and with experience) and cumulatively. Utilitarianism casts a wide net. Jeremy Bentham famously claimed the “greatest happiness of the greatest number” to be the master principle of utilitarianism. This well-known formula can be and has been read in two distinct ways: as enjoining the greatest aggregate amount of happiness (of net well-being) or, alternatively, as setting as the goal the general happiness (the happiness of the greatest number of persons).

We need to refine our view of utilitarianism. With Bentham's formula as background, we can distinguish and develop two issues: (a) the utilitarian standard of right, and (b) how that standard figures in deliberation and in actual attempts to realize or approximate the standard, in general and on particular occasions. Two of the utilitarian theories we have considered follow the “greatest happiness” part of Bentham's formula.

The classical or orthodox strand singles out that action whose total consequences will have the greatest aggregate benefit; that particular act token is the right-making standard. And agents determine what in particular is to be done on given occasions by appealing directly to this standard, making careful estimates of consequences, in the factual setting they currently occupy.

The other main strand (traditional rule utilitarianism) identifies, as the right-making standard, an ideal rule or set of rules specifying types of action that, if generally accepted and generally conformed to, would have a higher expectable benefit than would alternatives (such as having no rules at all or having simpler rules that excluded explicit and well-grounded exceptions). Acting according to these ideal rules would approximate, as closely as is humanly possible, the overall goal of achieving the greatest happiness. Here agents are concerned with installing and abiding by that set of rules which optimally conforms to the right-making standard. On given occasions, rules within that set might conflict; here agents would attempt to determine which of the constituent rules would probably give the best result if consistently followed in similar cases of conflict.

The third type of theory we looked at, so-called indirect utilitarianism, doesn't emphasize the greatest aggregate amount of happiness but, rather, (following John Stuart Mill) adheres to a more distributive view – to the improvement of the general happiness. For indirect utilitarians the right-making standard is conduct (act types) which has a pronounced tendency toward the goal of achieving the happiness, at an acceptable level, of the greatest number of persons, ideally, of each and all.

Indirect utilitarians don't begin by trying directly to ascertain what the general happiness principle might require and then reasoning directly from that to precepts, as would ideal-rules utilitarians; instead, indirect utilitarians first make a decision about exceptions or conflicts by reasoning from existing tried-and-true precepts which embody right-tending conduct (taking account, in so doing, of reasons for the success of these precepts, making judgments of fairness and appropriateness, factoring in maxims and paradigm cases, and so forth). For indirect utilitarians an appeal to the principle of utility doesn't involve going outside the rules so much as it involves building upon them.

The version of indirect utilitarianism I'm stressing, as developed by Gray and Lyons, is not merely a decision procedure. It has a distinctive right-making standard (founded on the notion of the general happiness) and is thereby quite different from either orthodox utilitarianism or from ideal-rules utilitarianism. Of course, there are some who describe themselves (or are described) as indirect utilitarians and who adopt a hybrid approach. They combine an act-token standard of maximizing general happiness with indirect utilitarianism simply as a strategy, a decision procedure, in which one uses a rule or precept to decide how to act on all given occasions, regarding it as a more reliable procedure to follow than consistently going directly to the principle of general happiness to so decide (for example Sumner 1987, 180–198).

Indeed, we'll find that for each of the main utilitarian theories examined here, there are existing or imaginable alternative or variant (often hybrid) versions, sometimes more than one. Bentham's master principle, the “greatest happiness of the greatest number,” does not reduce to any one of these theories; it can embrace them all. As internal debates within utilitarianism make clear, there are multiple perspectives for assessment within utilitarianism, all of them presumably legitimate by the lights of that theory. (See Lyons 2000.)

In an overlapping consensus involving various comprehensive moral and religious doctrines that exist in the Western world today, orthodox utilitarianism is prominently mentioned by Rawls as one of the doctrines that would support and endorse the family of liberal principles and the basic structure institutions involved with that family. (See PL, 169–170.)

But it is doubtful, in my view, that the utilitarian principle of greatest happiness, as an aggregative and maximizing one, could join any such consensus. That principle could not support the sort of radically distributive program we associate with rights – could not support the assignment of constitutionally guaranteed benefits and protections, to each and every individual person in advance, so to speak, and across the board. I don't believe then, contrary to Rawls's claim, that the utilitarian theories of Bentham and Sidgwick could accept the idea that the rights and liberties appropriate to the family of generic liberal principles could have a standing priority over policies favoring the general good (understood as a form of aggregative and maximizing well-being). The aggregative/maximizing version of philosophical utilitarianism is simply incompatible with the notion of basic rights and their priority developed by Rawls, and others. (See also Scheffler 2003, 451–453.)

But indirect utilitarianism (if all its arguments and presumptions are allowed) seemingly establishes that utilitarianism is compatible with human rights or basic constitutional rights and with the priority of such rights – at least in the case of those rights that are themselves justifiable in accordance with the general happiness principle. Certainly, it preserves the priority of such rights over against considerations of general good or corporate good that mark but a marginal increase in aggregate welfare. In any event, the point of indirect utilitarianism is not to maximize aggregate welfare but, rather, to improve the general happiness – the happiness, at an acceptable level, of the greatest number of persons, ideally, of each and all. For these reasons indirect utilitarians can plausibly be counted as joining into an overlapping consensus that has generic liberalism and its institutions as a focus.

I have argued elsewhere (Martin 2008 and 2011) that Millian utilitarianism has several essential points in common with indirect utilitarianism. Rawls thought that Mill's view (as interpreted by John Gray) exemplified a form of indirect utilitarianism (IOC, 433–434, esp. n20). In his view, Mill's theory, like contemporary indirect utilitarianism, was compatible with a constitutional regime that included both rights and liberties and their priority over the general good (understood aggregatively).

Establishing this point is important. If some versions of utilitarianism, including some historic or classical utilitarians, can be included in such an overlapping consensus, then that fact would, I think, make Rawls's idea of such a consensus plausible, both philosophically and practically.

This chapter draws on my work published in “Conceptions of Rights in Recent Anglo-American Philosophy,” IVR Encyclopedia of Jurisprudence, Legal Theory, and Philosophy of Law, 2007, at http://ivr-enc.info/index.php?title=Conceptions_of_Rights_in_Recent_Anglo-american_Philosophy (accessed May 2013); and “Rawls,” in David Boucher and Paul Kelly (eds), Political Thinkers: From Socrates to the Present (554–574), 2nd edn (Oxford: Oxford University Press, 2009).

1 Page numbers for Rawls's works included in Collected Papers refer to that collection.

Works by Rawls, with Abbreviations

Collected Papers (CP), ed. Samuel Freeman. Cambridge, MA: Harvard University Press, 1999.

“Commonweal Interview with John Rawls” (1998), in Collected Papers (616–622).

“The Domain of the Political and Overlapping Consensus” (1989), in Collected Papers (473–496).

“The Idea of an Overlapping Consensus” (IOC) (1987), in Collected Papers (421–448).

“The Idea of Public Reason Revisited” (IPRR) (1997), in Collected Papers (573–615).

“The Independence of Moral Theory” (1975), in Collected Papers (286–302).

“Justice as Fairness: Political not Metaphysical” (1985), in Collected Papers (388–414).

Justice as Fairness: A Restatement (JF), ed. Erin Kelly. Cambridge, MA: Harvard University Press, 2001.

The Law of Peoples, with “The Idea of Public Reason Revisited” (LP). Cambridge, MA: Harvard University Press, 1999.

Political Liberalism (PL), expanded edn. New York: Columbia University Press, 2005.

“The Priority of Right and Ideas of the Good” (1988), in Collected Papers (449–472).

A Theory of Justice (TJ), rev. edn. Cambridge, MA: Harvard University Press, 1999.

Other References

Baier, Kurt (1989) “Justice and the Aims of Political Philosophy.” Ethics 99: 771–790.

Barry, Brian (1995) “John Rawls and the Search for Stability.” Ethics 105: 874–915.

Beitz, Charles (2009) The Idea of Human Rights. Oxford: Oxford University Press.

Dreben, Burton (2003) “On Rawls and Political Liberalism.” In Samuel Freeman (ed.), The Cambridge Companion to Rawls. Cambridge: Cambridge University Press.

Freeman, Samuel (ed.) (2003a) The Cambridge Companion to Rawls. Cambridge: Cambridge University Press.

Freeman, Samuel (2003b) “Introduction: John Rawls – An Overview.” In Samuel Freeman (ed.), The Cambridge Companion to Rawls. Cambridge: Cambridge University Press.

Gray, John (1996) Mill on Liberty: A Defence. 2nd edn. London: Routledge.

Larmore, Charles (2003) “Public Reason.” In Samuel Freeman (ed.), The Cambridge Companion to Rawls. Cambridge: Cambridge University Press.

Lyons, David (1994) Rights, Welfare, and Mill's Moral Theory. New York: Oxford University Press.

Lyons, David (2000) “The Moral Opacity of Utilitarianism.” In Brad Hooker, Elinor Mason, and Dale E. Miller (eds), Morality, Rules, and Consequences. Edinburgh: Edinburgh University Press.

Martin, Rex (1994) “Rawls's New Theory of Justice.” Chicago-Kent Law Review 69: 737–761.

Martin, Rex (2001) “Rawls on Constitutional Consensus and the Problem of Stability.” In David Rasmussen (ed.), Social and Political Philosophy, vol. 11 of Proceedings of the Twentieth World Congress of Philosophy (81–95). Bowling Green, OH: Philosophy Documentation Center.

Martin, Rex (2008) “Two Concepts of Rule Utilitarianism.” Journal of Moral Philosophy 5: 227–255.

Martin, Rex (2011) “Mill's Rule Utilitarianism in Context.” In Ben Eggleston, Dale Miller, and David Weinstein (eds), John Stuart Mill and the Art of Life. New York: Oxford University Press.

Scanlon, Thomas (2003) “Rawls on Justification.” In Samuel Freeman (ed.), The Cambridge Companion to Rawls. Cambridge: Cambridge University Press.

Scheffler, Samuel (1979) “Moral Independence and the Original Position.” Philosophical Studies 35: 397–403.

Scheffler, Samuel (1994) “The Appeal of Political Liberalism.” Ethics 105: 4–22.

Scheffler, Samuel (2003) “Rawls and Utilitarianism.” In Samuel Freeman (ed.), The Cambridge Companion to Rawls. Cambridge: Cambridge University Press.

Sumner, L.W. (1987) The Moral Foundation of Rights. Oxford: Clarendon Press.