1. A Society of Individuals

What is a society of individuals? And why should we seek—or continue to seek—to realize it? My aim is to sketch some of the main features of such a society in a way that will make its attractions clear. I begin with a further question: who are the individuals who are to comprise this society? And the core of my answer is that they are, or at least have the frequently realized capacity to be, extended rational deliberators. Each considers what to do. Each is able to place, or comes to be able to place, his consideration of what to do in the context of his life. And his consideration is normative, not predictive. Each considers what he has reason to do, what makes sense for him to do.

I want to insist, as I have claimed elsewhere, that deliberative activity is primarily first-personal. Each member of society conceives himself as an independent agent, determining, through his own deliberations, what makes sense for him to do. To deliberate rationally, he need not see his deliberations as making any claim to universality. Although the considerations that weigh with him as reasons for acting in one way rather than another may not be self-directed, they are self-grounded. In considering what makes sense for him to do, he begins from his own situation. How he relates to other persons, over and above their impact on his situation, is not decided by characterizing him as a rational deliberator. He will be concerned with what others think makes sense for them to do insofar as this affects what they do and so how they impact on his situation. But he need not be concerned with the correctness of others’ beliefs about what they have reason to do, and he need not suppose that any conclusions about their reasons follow from his judgements of what he has reason to do.

To avoid a recurrent misunderstanding, I should emphasize the distinction drawn in the preceding paragraph between self-directed and self-grounded considerations. An agent whose concerns are self-directed may be said to be motivated by self-interest, by interests in the self. An agent whose concerns are self-grounded is motivated by interests of the self. But the object of these latter interests is left unspecified; they may be interests in family or friends, in humankind or in ecosystems, in the preservation of trolley cars and steam locomotives or in the prevention of tropical diseases. Fulfillment, rather than advantage or satisfaction, seems best to characterize an individual’s condition when her concerns, conceived as unrestricted in their objects, are realized. Each individual determines her own concerns; in a society of individuals, there is, and can be, no canonical list of concerns which all persons are expected to accept. Diversity of interests or concerns among individuals is assumed and indeed welcomed. A rational person exercises deliberative autonomy, in determining her own concerns which she then seeks to fulfill. But this is not the full story to be told about deliberation. A society of individuals, like any political and social order, makes deliberative demands on its members.

2. A Society of Individuals

A political order affects its members’ deliberations in two principal ways. First, it is constituted by a set of institutions which, like any set of institutions, introduce new possible actions, characterizable only in institutional terms. Second, it demands that the members recognize certain considerations as reasons, perhaps decisive reasons, for acting in one way rather than another. It expects compliance with its laws. A member may ask for the justification of these effects. Why should she willingly embrace a society’s institutional possibilities and demands within her own deliberations? In posing this question she affirms her normative priority to society, a priority implicit in my characterization of a rational deliberator. Why does it make sense for her, given her situation, to accede to society’s demands? In the absence of a satisfactory answer, she may see no reason not to subvert the institutions and ignore the demands, insofar as so doing seems good to her. This of course does not mean that others are constrained to acquiesce in her conduct, which represents in effect a rejection of society. What she may have reason to try to do, others need not have reason to let her do.

How shall we answer this demand for justification? What answer will give suitable recognition to the normative priority of each individual to society, and yet show that each should subordinate the unrestrained exercise of her deliberative autonomy in the light of the society’s demands? If the question initially seems unanswerable, if the juxtaposition of autonomy and subordination seems incoherent, this is only because we are not reflecting with sufficient care on the implications of autonomy for the interaction of rational deliberators. As an extended deliberator, each raises her sights from the most immediate practical questions; some come to ask, at least implicitly, how they shall live. Broadened and deepened, deliberative enquiry leads to a conception of the good life, but in keeping with the first-personal character of the quest, a conception decided and endorsed by each for herself. Thus when rational deliberators interact, each has her own distinct ends, and so her own distinct valuation of the possible outcomes of interaction, a valuation which need not correspond with those of her fellows. Now when persons with distinct valuations interact, if each seeks to realize her most favoured outcome, given her expectations of the actions of her fellows, then, as is well-known, the resultant outcome may be worse, in terms of each person’s valuation, than if all had acted differently. This familiar basic truth of interaction, holding even when each correctly anticipates the actions of her fellows and chooses her best response to their actions, and exemplified most simply in the Prisoner’s Dilemma, is the key to reconciling autonomy with subordination.

We may express this truth more precisely. Let us say that an outcome of interaction is in equilibrium if and only if no person could achieve greater fulfillment by acting differently, given the actions of the others. An equilibrium outcome will then result from interaction in which each seeks her fulfillment in terms of her own valuations, correctly expects what her fellows will do, and acts correctly on the basis of her expectations. And let us say that an outcome is optimal if and only if no person could achieve greater fulfillment, again in terms of her own valuations, without some other person being less fulfilled. It is easy to see that persons can always interact optimally. It is not easy to see, but can be shown subject to some reasonable assumptions, that persons can always interact to yield an equilibrium. Then the basic truth of interaction to which I referred in the preceding paragraph is that, in many (though of course not all) situations, the set of possible outcomes of interaction that are both in equilibrium and optimal is empty. We may call this the ‘interaction problem.’

I have long argued that morals and politics are best understood as embracing practices and institutions whose roles are to enable persons to achieve optimal outcomes that each finds superior to mutually disadvantageous equilibria. This is a normative claim—a claim about the rationally justifiable forms that can be taken by morals and politics. If it were generally accepted, we should live in a very different and, I believe, very much better world than we do. But in this paper I shall not urge that part of this claim which concerns morality—indeed, I shall shortly embrace the view that individual autonomy must be expected to yield a diversity of moral outlooks, and that this is one of the core features of a society of individuals. I shall restrict my claim to politics, and insist that a political order justifiable to rational deliberators must be responsive to the interaction problem. The justification is contractarian.

Actual persons are of course born into political societies. But suppose, as a thought experiment, that persons were in a position to agree, ex ante, on the terms of their political association. Suppose we think of these persons as characterized each by a natural endowment—a set largely of potentialities for social realization—but without any of the particular advantages and disadvantages that social and political life has in fact brought them. Suppose that these persons were aware of the interaction problem, and were to recognize that, if they fail to agree on terms of association, each will be thrown back on trying to respond to his expectations of the others’ actions, with a consequent failure to achieve optimal levels of fulfillment. It would then make sense for these persons to seek and reach agreement on terms that promise mutually more fulfilling interaction, and that offer each a share of this enhanced fulfillment that fairly reflects her endowment and her potential contribution to their mutual fulfillment. Although these terms would require each to subordinate the unrestrained exercise of her deliberative autonomy, each would rationally exercise that autonomy in agreeing with her fellows to accept this subordination.

To this point I have argued in the hypothetical mode—if persons were to occupy a certain position, they would agree to political institutions and practices that afforded each fair mutual fulfillment in their interactions. Political contractarianism takes this argument, and, passing to the actual mode, makes it ‘the contractarian test.’ ^{Footnote 1} Real persons—you and I—have good reason to accept those political arrangements that we would or could have agreed to from the ex ante perspective of the test. Terms of social interaction may be justified to each member of society insofar as these terms provide fair mutual fulfillment to each, and terms are fair insofar as they would or could have been endorsed by persons aware of their natural endowments and social potentialities. Note that this justification makes no appeal to shared values, or to any allegedly objective standard of values. Note also that it rejects the idea of a veil of ignorance that would deprive a person of knowledge of her own capacities and interests. It proceeds from each person’s conception of her own concerns—a conception that, as I have noted, need not be self-directed but is self-grounded—and depends on the structure of interaction among persons, each of whom deliberates about what to do starting from those concerns.

Exercising their capacities as rational deliberators, persons develop individual conceptions of the good life. Contractarian justification establishes a single standard of the right, embodied in the political institutions and practices of society, as compatible with, and indeed expressive of, the differing individual good lives of the members of society and their deliberative autonomy. We may call this standard of the right ‘the social contract.’

3. Contractarian Justification

But why would persons want such agreement? Why should each person not do as seems best to her, given her expectations about what others will do? The answer emerges straightforwardly from the following considerations. Assume each person has a conception of her good which guides her deliberations, and in terms of which she evaluates possible states of the world. Suppose, as is surely plausible, that there is some difference in the orderings of states of the world that result from different persons’ evaluations. (Note that these different orderings reflect the different standards of evaluation the persons apply—each her own conception of the good—and not disagreement about the application of a common standard.) Now suppose that each person seeks to do what would realize the most fulfilling state of the world, according to her own evaluation, and given her expectations of what others will do, and suppose, as an ideal case, that each is successful. Then the actual state of the world realized through the interaction of different persons will satisfy an equilibrium condition: no one could have acted to bring about a more fulfilling state of the world, according to her own evaluation, and given the actual actions of the others.

Thus successful interaction, among persons each of whom does what seems best to her, yields an equilibrium state of the world. Now let us ask: does such a state satisfy an optimality condition? Is there no alternative, which could have been realized had some or all of the persons acted differently, and which would have been more fulfilling, in terms of some persons’ evaluations, and worse, in terms of none? And here it is easy to show that an equilibrium state of the world need not be optimal. Furthermore, it is easy to show, by examples of which the Prisoner’s Dilemma is the simplest, that in many contexts, no equilibrium state of the world is optimal. That is, in many contexts, for any state of the world in which each succeeds in bringing about what is most fulfilling according to her own evaluation given the actions of the others, there is a realizable alternative in which some do better according to their own evaluations and none do worse. It follows that, if persons were in a position to agree on their terms of interaction, each would want to agree on terms that would yield states of the world mutually more fulfilling than those which would result if each simply sought to do what would be best according to her own evaluation, given her expectation of what the others would do. Each could only expect to gain from agreement on such terms of interaction, in comparison with no agreement. So the contractarian claims that terms of social interaction may be justified to each member of society insofar as these terms provide fair mutual fulfillment to each, arguing that it would have been rational for each to seek ex ante agreement with her fellows about such terms, and to accept agreement if it yielded terms offering each fair expected benefit in comparison with no agreement. Now note that this justification makes no appeal to shared values, or to any allegedly objective standard of values. It proceeds from each person’s conception of what is good—a conception that, as I have noted, need not be self-directed but is self-grounded—and depends on the structure of interaction among persons, each of whom deliberates about what to do starting from her conception of the good.

But there is a qualification needed in the argument I have just given. For suppose that, even though there are potential mutual gains from agreement, some of the persons would expect their lives to be better in the absence of some of the others. That is, suppose interaction itself is seen as costly from the standpoint of some persons. From their point of view, their fulfillment is not enhanced by the presence of at least some of their fellows. Imagine for example a world composed largely of peaceful farming families, with a small number of predators who raid the crops, rape the women, and engage in a variety of other offensive assaults on the peaceful and productive farmers. A pact in which the farmers cater to the desires of the predators in return for being left in peace might afford mutual benefit—and indeed, might do so fairly in relation to the original situation. But the farmers’ only reason for adhering to such a pact would be fear of the consequences of violation. They will see no reason not to violate the pact and subvert its arrangements were they to think they could succeed. They would prefer not to have to interact with the predators.

Successful contractarian justification depends on interaction being welcome. The valuations of the members of society must be such that each judges adherence to the social contract to be more fulfilling, in comparison both with interaction in its absence, and with no interaction. This is not to suppose any ideal correlation among the concerns of different persons. But it is to suppose that each person’s concerns can be furthered by mutual interaction, so that each can expect a more fulfilling life because of the presence of the others, given terms of interaction that would or could gain ex ante agreement. So far I have argued that the rationale for the contractarian argument depends on two factors. The first is the interaction problem, which is structural, depending not on the content of the valuations of the several parties to interaction but only on the different ways in which they order the possible outcomes. The second, which is weakly substantive, is that the valuations, or conceptions of the good, held by the interacting parties are compatible with each being welcomed by the others as such a party. There is a third factor, to which I now turn.

4. The Fact of Reasonable Pluralism

I suggested earlier that in a society of individuals deliberative autonomy must be expected to yield a diversity of moral outlooks. This is not because morality, like conceptions of the good life, arises from first-personal deliberation with no claim to interpersonal validity. Moral, as well as religious, views are put forward objectively, as universally binding in the case of morality and as universally acceptable in the case of religion or its absence. But I take it to be a well-established fact that there are several families of reasonable moral and religious views, with no prospect of resolving the disagreements among them. This is what John Rawls has felicitously termed the “fact of reasonable pluralism.” ^{Footnote 2} Moral and religious views are reasonable when they are held reflectively and compatibly with what we reasonably believe about our world and ourselves. These grounds are underdetermining, leaving room for deep disagreement, as between Kantians and utilitarians, theists and atheists. But these are disagreements we must live with, and so we need to consider how the fact of reasonable pluralism may be accommodated in a society of individuals. It is the third factor in the contractarian argument.

The key is that views which claim to be reasonable may of course deny the correctness, but may not deny the reasonableness, of their competitors. Reasonable holders of reasonable views must recognize that the grounds which may be offered in their behalf do not amount to proof. In the usual philosophic sense of sufficiency—p is sufficient for q if and only if one may not reasonably accept p and deny q—the grounds for reasonable moral and religious doctrines are not sufficient. In another sense of sufficiency—p is sufficient for q if and only if one may reasonably accept q because one accepts p—then of course the grounds are sufficient. This recognition of insufficiency in the first sense and sufficiency in the second establishes the mutual reasonableness of reasonable doctrines. It also establishes, as a constraint on the terms of the social contract, that it endorse no moral or religious doctrine and reject no reasonable doctrine. The contract in this way affirms the reasonableness of reasonable doctrines, while remaining agnostic about their truth. Contractarianism itself may of course be embraced as a comprehensive moral doctrine. But its political role is restricted to providing a defence for and a characterization of the social contract, and so justifying the idea of a society of individuals. We should not think that this role depends on accepting a comprehensive contractarian outlook. Rather, we should think of a society of individuals as defensible from the perspective of any reasonable person, given her recognition of the fact of reasonable pluralism. If reasonable persons holding different conceptions of the good life and at least partially incompatible moral and religious views are to interact in ways that recognize one another’s reasonableness, then the terms of their interaction must be viewed by each as ones to which she would rationally have agreed, as fair and beneficial, and must include a commitment to a form of democracy. Recall my earlier remark that, in demanding justification for social rules and practices, each person affirms her normative priority to society. We should now understand this as the idea that each reasonable person affirms the normative priority of her own moral and religious view, which embraces her conception of the good, in making her own judgment of the claims of society, while recognizing that other reasonable persons affirm the normative priority of their differing views in making their judgment. But the views they accept need not themselves defend the normative priority of the individual to society. Contractarian justification does not presuppose liberal individualism, although it is of course congruent with that outlook, but instead arises from recognition of the fact of reasonable pluralism. The idea of a society of individuals, a society in which political justification is addressed to each person and proceeds in terms of her concerns, reflects the basic constraint that reasonable pluralism imposes.

To accept contractarian justification is to recognize the terms of the social contract as giving one reasons for acting. The members of society do not, then, deliberate each in terms of their own conceptions of good, and their comprehensive moral and religious views, but rather respond directly to their society’s laws and practices. To be sure, these must accommodate each person’s conception of the good, and recognize each person’s moral and religious outlook, so far as the fact of reasonable pluralism allows. But what this justification establishes is that the laws and practices are conducive to fair mutual fulfillment. And this guides the deliberations of the citizens in their political relations. Again, to avoid misunderstanding, I want to emphasize this last point. The members of a society of individuals are not to be viewed as engaged in an ongoing process of political bargaining in which each seeks simply to advance his own interests. Each person is of course free in his private life to pursue whatever goals bring him fulfillment within the framework set by the laws and practices of the society. But in his public life he is expected to aim at fair mutual fulfillment, whether he is actively participating in legislative or judicial decision-making, or whether he is merely, as a citizen, relating his actions to the laws and social customs.

Why he is motivated by the idea of fair mutual fulfillment or, if you prefer, by the idea of justice is not a matter which contractarianism as a political doctrine seeks to settle. Rather, it assumes that any reasonable moral view will provide the requisite motivation, given, as always, the expectation of fair mutual fulfillment and recognition of the fact of reasonable pluralism. A comprehensive contractarian moral doctrine may, as I argue elsewhere, motivate concern with fair mutual fulfillment by appealing to the desire each person has to be accepted by his fellows as a welcome member of society, given the beneficial opportunities that membership brings. ^{Footnote 3} But a defence of this particular grounding for the motive of justice is neither part of nor necessary to my present argument.

5. Principal Terms of the Social Contract

I want now to consider some of the principal terms of the social contract. These emerge quite directly from the idea that rational agreement assumes the possibility of mutual benefit, of each person achieving greater fulfillment than she could in the absence of agreement or the absence of interaction with the others, and then seeks to realize that mutual benefit in a way that each may accept as fair. Mutual benefit is the fundamental tie that binds the members of a society of individuals to each other; it is the only one that is both socially universal, transcending particular affections and sympathies, and independent of any particular moral and religious view. But it constrains the citizens to interact in ways that are mutually beneficial, and since the benefits which society provides result from their actions, it constrains each to assume a productive role in relation to his fellows. A citizen who had the capacity to play such a role but failed or refused to do so, and yet has shared in the benefits social interaction afforded, would be a net cost to his fellows, and so could not be welcomed by them in a partnership for mutual benefit.

But if each ‘normal person’ (described more fully below) has the obligation to perform a productive role, and so the obligation to shape her conception of the good so that it provides an appropriate place for productive action, society equally has an obligation to afford each person the training and the access necessary to give her an opportunity to choose among possible productive roles and to select one suited to her talents and interests. A society that, say, systematically blocks women from receiving types of education available to men, or denies blacks effective access to professions open to whites, will obviously fail to meet this obligation. Thus willingness to engage in a productive role and equal effective opportunity to choose such a role are correlative obligations that are demanded by a social contract that offers each the expectation of enhanced fulfillment. Equal effective opportunity is thus a right vested by the social contract in each citizen. We may think of this as including an equal right to productive resources and to jobs.

Equal opportunity to productive positions in society is a right afforded to each party to the social contract. It is an unusual right, however, insofar as it is an inalienable ‘use right.’ The most important implication of treating the right to productive opportunities as a use right is that it renders the right non-marketable. Suppose, to adapt an example from Philippe van Parijs, ^{Footnote 4} I want to spend my life surfing at Malibu. You on the other hand would like an additional share of productive resources and an extra job. Why shouldn’t I sell you my rights, thereby gaining the wherewithal to live as a surfer? Although I perform no productive role—no one is prepared to pay for my surfing—it would seem that my opportunity rights constitute an entitlement that I can use to gain access to the benefits of society. But this proposal is mistaken. If my rights to productive resources and jobs are marketable, then I am in a position to benefit from interaction with others, against a no-interaction baseline, in ways that are costly to them, measured against the same baseline. My rights would not enable me to live as a surfer were I to exercise them; they benefit me only if I can find a buyer for them. But, in my absence, what I have to sell—resources and a job—would be available to some other person or persons without cost. And so my presence would be unwelcome. Marketable rights give rise to ‘parasitism.’ Rights to resources and a job benefit others only if the holder exercises them; they are part of a mutually beneficial social framework only as rights of use. And so I conclude that the right to equal effective productive opportunities vested in each citizen by the social contract must be a right to hold a job and employ resources productively. A person who chooses to be unproductive has no claim on someone who uses the resources and holds the job that, had the first person chosen otherwise, would have been available to him.

Structural unemployment complicates the argument, however. Although I cannot examine the idea of a productive role in depth, we can see the complexity involved in determining what is genuinely productive by reflecting on the situation of those who are unemployed. One might think that the right to productive opportunities afforded to each under the social contract demands that society takes steps to keep the unemployment level very low. But suppose that the attempt to achieve a negligible rate of unemployment leads an economy to overheat in ways that lead over time to diminished rather than enhanced overall productivity. Under such conditions, ensuring directly productive employment for each results in diminished production for all. In these circumstances, unemployed persons are necessary to achieve maximum productivity, and so, while not directly productive, their being unemployed makes an indirect productive contribution. They contribute by relinquishing their rights to productive resources and a job—not by selling these rights to someone wanting to exercise them, since what is required is that they go unused.

The unemployed are then entitled to compensation, provided it does not make those employed worse off than they would have been under full employment. No one is disadvantaged by such compensation. The idea that unemployed persons must constitute a cost to their fellows, first in reducing the total social product and then, insofar as they are supported, in taking from what others produce, although it may be intuitively plausible, may reveal ignorance of the workings of complex economic systems. (I am assuming in this example that those entitled to unemployment compensation are willing to work and play more directly productive social roles but find themselves involuntarily unemployed under a social policy aimed at achieving an optimal level of unemployment; they are not in the position of the would-be surfer and they are not ‘free riders’ or parasites (persons who take benefits from the productive activities of others without contributing to their production or in excess of their contribution).) ^{Footnote 5} van Parijs draws different conclusions about the entitlement of the would-be surfer than do I: he defends the surfer’s right to a basic income. That is unjustifiably indulgent to the prodigal, and unfair to the productive.

van Parijs is not the only theorist willing to impose dues on productive members of society to support even those who are able but unwilling to work (the voluntarily unemployed). Some utilitarians, perhaps some traditional communists, and on one interpretation, Rawlsians, would impose what contractarians take to be unjustifiable demands on the able and willing. Likewise, we should reject such theories if they entail that each person has an obligation to maximize her contribution to social productivity or to be as productive as possible. Recognizing such a duty would severely limit persons’ choices among productive roles. In a society of individuals each need only choose to be productive—to add to the social stock so that she is a welcome partner in interaction. Persons are permitted to pursue careers that contribute less to the social good than alternatives they could pursue that would yield greater amounts of good. If someone prefers gardening to practicing medicine, say, then even if she would contribute much more to society by taking up medicine and would herself receive greater compensation for that work on account of its social value, she must be free to choose according to her preferences (and accept the reduced remuneration for her chosen line of work). Such a choice gives her fellows no cause for complaint. In a society of individuals, no one has the right to exploit the talents of other members, or to treat the talents of her fellows as part of a common pool of resources. To suppose otherwise would be to fail to recognize her normative priority to society.

6. Proportionality between Contributions and Returns

I have already mentioned the return from one’s social contribution, and at least implicitly suggested that return and contribution must be proportional. The social contract will require society to be structured so that, for each normal person, a choice of roles is available, each of which will offer a return sufficient to provide the holder of the role with resources sufficient for a fulfilling life. Each person has the opportunity to construct a conception of the good and form a life-plan based on it, compatible with the role she chooses and the share of the social product she receives in return for performing it. Each is also entitled to the education, in the broadest sense, of her capacities and sensibilities needed for her to find fulfillment in the pursuit of a life-plan compatible with her social role. Although persons are free to form a conception of the good and choose a life-plan constrained only by the obligation to contribute to social production, these conceptions and life-plans are not taken as inputs which society must seek to accommodate in the way in which it must accommodate reasonable moral and religious views. Not every conceivable life-plan satisfying the social contribution requirement need be feasible or effectively available. The institutions and practices of any society will limit the ways of life effectively socially available to a sub-set of all of the ways that can offer human fulfillment. It must suffice if each normal person may expect to find an occupation suited to her talents, and to develop interests that she may fulfill compatibly with her occupation. The normative priority of the individual to society does not require society to adapt to prior, fixed individual demands, but rather to afford a range of options among which the individual may expect to find one or more which she can happily adopt. What is distinctive about contractarian society is that the constraints on available life-plans are not set by any supposedly objective social value or values, but rather by (1) effective compossibility, (2) the obligation that each make a productive contribution, and (3) the need to respect the range of moral and religious views which reasonable persons may hold.

There is a further constraint on life-plans, which is of the first importance. If children are to grow into adults with the real opportunity to make the choices among productive activity and life-plans that are to be available to them, then parents must not raise their offspring in ways that would nullify this opportunity. This is a real constraint on the parents’ life-plans, arising directly from the recognition that each member of society is a full party to the social contract and so entitled to the same benefits, as well of course as subject to the same obligations, that the contract extends to all other members of society. In thinking of agreement on the terms of interaction, parents must recognize that their children are equal participants with themselves. For in considering what persons would agree to, were they in a position to choose together their terms of interaction, we should think of each person, of whatever generation and of whatever actual abilities, as represented by an ideal agent, aware of the talents and interests and circumstances of the person he represents, and able to negotiate on equal terms with the others. This agent is unhindered by the possible lack of bargaining skills, or of awareness of self or others, or of any other factor that would in real negotiations stand in the way of the actual person obtaining agreement on terms of interaction that would fairly advance her prospects of fulfillment. And so this ideal agent must be unhindered by the actual effects of parental control, insofar as these could stand in the way of the person negotiating terms of fair mutual fulfillment.

7. Political Terms of the Social Contract

Let us now focus on some of the more narrowly political terms of the contract. And so let us imagine ourselves in the position of those choosing ex ante their terms of interaction. They choose to structure society in such a way that each normal person is expected to contribute to the social product, and may in return claim the resources needed to form and execute a fulfilling life-plan. How this choice is to be implemented is a task for institutional designers, not philosophers. But the need for an overseeing body, a government, that will take the decisions necessary to establish and maintain an appropriate social structure should not be in doubt, and so we may ask what form of government seems best suited to do this.

I propose to approach this question by considering why we should disagree with the views of the thinker whose general understanding of society is actually closest to that which I am sketching—Thomas Hobbes. I make this claim of affinity because Hobbes sees social and political structures as remedying the ills resulting from natural interaction, and, without having our current vocabulary at his disposal, in effect identifies these ills as the mutually disadvantageous, and indeed disastrous, outcome of actions each undertaken as the agent’s best means of promoting his own well-being, given what he expects his fellows to do. Natural interaction fails to yield an optimal level of individual fulfillment. There are of course important differences between Hobbes’ account and mine. Hobbes treats the concerns of each agent as self-directed and not merely self-grounded. He conceives each person’s good narrowly, focusing on “conservation” and “delectation.” ^{Footnote 6} Particularly relevant to our discussion of the form of government, he assumes self-interested motivation throughout, so that citizens do not willingly accept the constraints which they recognize as conducive to mutual benefit. But notwithstanding these differences, Hobbes’ argument for political society has the same contractarian structure as the one I have presented.

Hobbes argues that all persons will agree to authorize one person, or assembly, as in effect their universal agent, entitled to act in the name and with the power of each, to the end that they may live in peace and safety, and enjoy “all other contentments of life, which every man by lawful industry, without danger or hurt to the commonwealth, shall acquire to himself.” ^{Footnote 7} He supposes this authorization to be unrestricted, both in scope and time, and although he allows the possibility of authorizing an assembly, he strongly advocates authorization of a single person, who then has the power to name his successor, and so is in effect a hereditary monarch. For, Hobbes argues, the centrifugal forces created by the self-directed actions of the individual members of society are so great that they can be balanced only by the strongest centripetal force that human beings can devise, and this is the power that can be directed and exercised by a single person, who has the permanent and unconditional authorization of his fellows.

But, if the formal structure of our problem is akin to Hobbes,’ its content is significantly different. First, political contractarianism rejects the gloomier interpretation of Hobbes’ celebrated dictum that “Covenants, without the sword, are but words, and of no strength to secure a man at all,” ^{Footnote 8} insisting rather on the normal willingness of persons to comply with the terms of the social contract, seen as affording each a fair expectation of fulfillment, provided others are expected to be similarly willing. Following Gregory Kavka, we may distinguish offensive violations of the terms of the contract, in which persons unreasonably seek their fulfillment in ways that the contract rules out as being at the expense of the fair fulfillment of their fellows, from defensive violations, in which persons reasonably seek to protect their fulfillment from the effects of others’ violations. Political contractarianism supposes that persons are normally, though not universally, willing to abstain voluntarily from offensive violations. The role of the Hobbesian sword is then only to provide sufficient security against those who, without compulsion, would not abstain from offensive violations so that no occasion for defensive violations arises. And this is a limited role.

The second difference in the content of our problem arises from the transformation in the conditions of human existence which has taken place in the three centuries since Hobbes wrote. Briefly, we live, as Hobbes did not, in a world in which billions of human beings enjoy a reasonably full life span in material conditions which, if not affluent, permit a reasonably comfortable existence that includes valuable vocational opportunities and adequate leisure for the significant pursuit of various avocational goals. In Hobbes’ time, no society could organize itself in ways that promised a productive role and a fulfilling life for everyone. But this is well within the reach of the developed societies of North America, Europe, and increasingly the world as a whole (at least where war (civil and international) have not eroded the gains made possible by modernization). For the contemporary contractarian, then, governments provide the legislative, executive, and judicial mechanisms such societies require. Hobbes thought that, while government required the authorization of the citizens, it could not be answerable to them for its acts. But in the present setting, government can both draw support from the citizens in performing its facilitating role, and be answerable to them for the effectiveness of its performance. The contemporary social contract will then not contain the Hobbesian blank cheque, signed by the citizens, with the amount to be filled out by the government at its pleasure. Rather, it will afford only an authorization of powers limited in scope, and to persons only for a limited time. The government will be the universal agent of its citizens, but whereas in Hobbes’ argument the citizens, although principals, have no right to revoke the authorization they give or in any way question its exercise by the sovereign, in a contemporary contractarian society this will be among their primary rights. Treating the government as their agent, the citizens will expect to be able, individually and collectively, to challenge the government’s actions should it exceed its mandate to implement and maintain the terms of the social contract, but also able, collectively, to retain or dismiss the particular persons serving as their agent depending on how effectively these persons carry out this mandate.

This is the basis of contractarian democracy. First, since each person sees the government as her agent, each will insist on sharing equally with her fellows whatever rights of control over that agent there may be. And second, the rights of control that each person will insist on sharing are primarily the rights of appointment, dismissal, and challenge. Without these rights, citizens cannot expect to be secure in the other rights and freedoms which directly affect their opportunities for fulfillment. And so a government whose actions are not open to judicial challenge, and whose members are not appointed and dismissed by the citizens, lacks contractarian legitimacy. The right of challenge is exercised through a judicially-enforced constitution. The rights of appointment and dismissal are exercised through periodic elections in which the voters have an effective opportunity to dismiss the incumbents and put other persons in their place.

Governments, contractarians insist, exist to facilitate productive and fulfilling individual lives. It is therefore appropriate that those who govern be responsible and responsive to the individual citizens whose good lives they facilitate, and the most effective way to keep them responsible and responsive is to give the citizens the opportunity, at reasonably frequent intervals, to choose those persons who will in fact make legislative and executive decisions. Governments are not, the contractarian also insists, primarily mechanisms either for the brokerage of interests or for the struggle to determine which interests dominate. Equally they are not primarily devices for the determination and implementation of collective or community values. When class-based, or culture-based, or gender-based, or race-based, or identity-based, or faith-based, or ideology-based issues take over the political agenda, then the contractarian ideal of a society of individuals has failed to take hold. Then the only form of political and social life feasible may indeed be a mere modus vivendi, in which some compromise is reached among persons and groups of persons who see their values and concerns as deeply opposed.

In such circumstances citizens may judge it sufficiently important to be represented in government by those who share their collective identity or ideology that they accept the political fragmentation, the lack of direct voter control, and the emphasis on latent conflict and reluctant compromise that such representation invites. But to view the political order in this way is to see it as addressed to concerns which, from the contractarian perspective, arise only when it has failed in its real task. Social identities and ideologies take their appeal from either inability or failure to achieve optimal outcomes in the face of reasonable pluralism. A contractarian society readily accommodates three respects in which individuals differ. It accommodates individual conceptions of fulfillment and plans for a good life, by endeavouring to make it possible for each to realize her own fulfillment by and through contributing to the fulfillment of her fellows. It accommodates reasonable moral and religious views by providing a framework for interaction, which each can recognize as reasonable given the fact of reasonable pluralism. (I shall partly defend and partly weaken this claim in considering a possible objection presently.) And it accommodates political views, understood as views about the appropriate policies and personnel to realize the first two forms of accommodation, by establishing procedures which allow meaningful choice between political alternatives. Taken together, the accommodation of these differences among individuals is the object of what I call ‘political agreement.’

8. Providing for the Nonproductive?

Now that I have considered briefly some of the more narrowly political features that underlie the idea of contractarian democracy, I want to remark on the implications of the earlier references in my discussion to ‘normal persons.’ I have been assuming that society can be organized in such a way that persons with normal human physical and mental capacities may be contributing members, persons whose presence may be welcomed by their fellows because their activities increase the social product. But not all persons enjoy normal physical and mental powers. Some lose these powers through misfortune, and here we may expect them to be protected by disability insurance. But others will be congenitally handicapped so that a straightforward negotiation, recognizing their inabilities to contribute to the net social product, would give them no return. The parties to the contract would agree ex ante to insure to some degree against misfortune, since they know only its likelihood but not whether they themselves will be victims, but they would not agree to insure against congenital defects, since they know their own capacities and so whether or not they are handicapped.

Here contractarians may appeal to other grounds for providing some measure of protection for those unable, through no unwillingness of their own, to be net social contributors. First, there are the particular ties of affection with family members, who are unwilling to see needs of loved ones unmet. From their point of view, one of the benefits made possible by society is a greater degree of assurance that this will not happen. Second, there is natural human compassion. In circumstances in which persons view themselves as sharing fairly in the greater fulfillment which social cooperation makes possible, we may expect compassion to motivate them to want to share this fulfillment with those whose handicaps prevent them from contributing to it. But note that these arguments call on human characteristics which in themselves do not enter into the contractarian account. And so we may ask if there is deeper ground for providing for the congenitally handicapped.

To consider this question, we need first to introduce the distinction between comprehensive and political contractarianism. If, as the comprehensive theory holds, morality is altogether a matter of agreement, then only those who are parties to the agreement have a fundamental claim to share its benefits. That they extend consideration to others, if they do, must depend on contingencies such as those just mentioned. But political contractarianism in itself lacks any basis for determining the scope of moral entitlements over and above the claims established in the agreement itself. Although those who are both contributors to and recipients of the fulfillment which social cooperation makes possible are entitled to share in the social product in proportion to their contribution, their shares need not exhaust that product. The reasonable moral and religious views of the persons represented in the agreement must be taken into account in deciding whether persons who, although willing to play productive roles are unable to make a net contribution to the social product, nevertheless are entitled to share in it.

In a contractarian society, the moral and spiritual autonomy of the citizens is fully recognized. Each is entitled to practice his own morality and religion, insofar as this is compatible with the similar practice of his fellows. But there are circumstances in which society must maintain a common moral practice despite reasonable differences among the citizens. And prominent among these circumstances are those in which there is disagreement about the persons to whom consideration is owed. The productive members of society may want to provide for their congenitally handicapped fellows. But the issue is not one of sympathy. Are the congenitally handicapped owed support by those more fortunate, in circumstances in which support is feasible and compatible with reasonable fulfillment for all? Insofar as many reasonable moral views would answer with a clear affirmative, the social contract must find some means of accommodating them.

Here the resort to a hypothetical insurance market may be helpful. Although the political contract, unlike its Rawlsian counterpart, is not formulated behind a veil of ignorance, we may surely ask, were persons to treat congenital handicaps as ex ante risks, to what protection might they reasonably agree? The relevance of this question is set by the demand, stemming from views that all can recognize as reasonable even if some do not consider them sound, that provision be made for the support of those willing but unable to be net contributors to the social product. But it is posed in such a way that it can be answered without appealing to any particular moral view. The answer cannot of course be determined with any precision. Since many congenital handicaps would affect a person’s sense of her identity, the exercise of imagination required to give meaningful consideration to the question is not easy or straightforward. And, of course, knowledge of the nature and cost of measures that would enable persons with different congenital handicaps to achieve reasonable fulfillment is not sufficiently available. Nevertheless I think that by representing the question as involving a trade-off between fulfillment under adverse conditions and fulfillment in normal conditions, we should be able to generate plausible proposals, based not on unexamined sentiment but on a realistic consideration of costs and benefits.

9. Contractarian Democracy

Does contractarian democracy offer a way of achieving political agreement, given the fact of reasonable pluralism? Not fully, given unresolvable disagreements about moral status. Nevertheless, contractarian procedures offer a way of containing disagreement, in ways that advocates of opposed positions may recognize as a lesser evil than open conflict. Whether over time they will come to view each other as reasonable, even if wrong, is more than I can predict. But I want to conclude on a more positive note. For, if pluralism is the cause of problems, it is also, and more importantly, the expression of true autonomy. We live in a world in which people cannot only afford to differ, but can take pleasure from the multiplicity of their individual outlooks and concerns. The local societies of the past may have differed greatly one from another, but they lacked the internal richness of the developing universal society of which cosmopolitans dream and whose features we can at least sometimes seem to glimpse. Because our society can accommodate a multiplicity of concerns, it can provide individuals with the opportunity to choose their own fulfillment, provided only that they contribute to the conditions making fulfillment for all possible. Differences need not then be disagreements. Our political theorists have heard the inevitable discordant notes, and have written about Democracy’s Discontent, or Democracy and Disagreement. ^{Footnote 9} But in witnessing the emergence of a society of individuals, we might be led to another phrase, Democracy’s Fulfillment.