Notre Dame Law School
Natural Law Forum
Distributive Justice: Some Addenda
Follow this and additional works at: http://scholarship.law.nd.edu/nd_naturallaw_forum
Part of the Law Commons
Rawls, John, "Distributive Justice: Some Addenda" (1968). Natural Law Forum. Paper 138.
This Article is brought to you for free and open access by NDLScholarship. It has been accepted for inclusion in Natural Law Forum by an authorized
administrator of NDLScholarship. For more information, please contact [email protected]
I wish to elaborate further the conception of distributive
sketched elsewhere.' This conception derives from
justice that I
the ideal of social justice implicit in the two principles proposed in the essay
"Justice as Fairness."' 2 These discussions need to be supplemented in at least
two ways. For one thing, the two parts of the second principle are ambiguous:
in each part a crucial phrase admits of two interpretations. The two principles read as follows: first, each person is to have an equal right to the most
extensive basic liberty compatible with a similar liberty for others; and second,
social and economic inequalities are to be arranged so that they are both (a)
reasonably expected to be to everyone's advantage, and (b) attached to positions and offices equally open to all.3 The two ambiguous phrases in the second
principle are "everyone's advantage" in part (a) and "equally open to all"
in part (b). I shall consider these ambiguities and the several interpretations of the two principles to which they give rise. I shall leave aside any
difficulties with the first principle and assume that its sense is the same throughout and dear enough for our purposes.
In "Distributive Justice" I discussed for the most part the interpretation
of the two principles which I believe the most defensible; and I tried to indicate how it would apply in practice. I did not compare it with the other
interpretations, except in passing; nor did I take up the various reasons for
preferring it. These reasons I now want to set out. In particular, I shall suggest that the difference principle (the meaning of which is explained below)
enables one to give a reasonable interpretation of the concept of fraternity
and of the Kantian idea that persons are always to be treated as ends and
* Given as The NATuRAL LAW FoRuv Lecture at Notre Dame on November 1, 1968.
I am grateful to the members of the editorial board for their many helpful comments and
criticisms. I have tried to take these into account in revising the lecture for publication.
I See the essay entitled Distributive Justice, in PHILOSOPHY, POLITICS, AND SOCIETY,
third series, 58 (ed. P. Laslett and W. G. Runciman, 1967).
2 67 PHILOSOPHICAL REVIEW 164 (April, 1958). See also Constitutional Liberty and
the Concept of Justice, 6 NoMos (ed. C. J. Friedrich and J. W. Chapman, 1963).
3 The formulation of these principles has been changed slightly, but not, I believe, their
NATURAL LAW FORUM
never as means only. Further, this principle takes account of the principle
of redress and appropriately regulates the influence of the distribution of natural assets and social contingencies on distributive shares. And finally, the difference principle seems to meet better than other traditional views a reasonable requirement of reciprocity. In considering these matters there is inevitably some overlap with the previous essays; but to keep this within bounds,
I summarize in the next section the main points to be kept in mind.
needs to be said about the contractarian doctrine from which
I believe the two principles can be derived. This aspect of "Justice as Fairness" can be left aside for the time being. 4 It suffices to remark that I think
of these principles as those that would be chosen, at least in preference to
other traditional alternatives, by rational persons in an original position of
equality as their common conception of justice. That is, presented with a
short list of alternatives containing, let us suppose, the two principles of
justice, the principle of utility (in its several variants), and a principle of the
common good, the two principles would be preferred. One of the essential
features of this initial contractual situation is that no one knows his social
position or his place in the distribution of natural assets (his intelligence,
strength, and so on). The original contract is made behind a veil of ignorance.
Another essential feature of the situation is that while the contracting parties
assume that they have no interests in one another's interests, or but a limited
one, they nevertheless have certain interests and responsibilities that require
them to protect their liberties. For example, they recognize an obligation to
secure the rights of their descendants and to insure the fulfillment of their
religious duties or cultural interests. The contracting parties are family heads
and not individuals. 5 The contention is that, given a full description of the
initial choice situation, the two principles of justice are the best alternative
on the list. But I do not wish to elaborate these matters here. Instead, I
want to compare the several interpretations of the two principles and to
consider some of the merits of the difference principle.
Now for several matters assumed on the basis of the earlier discussion in
"Distributive Justice." First, the primary subject of the two principles is the
basic structure of society, that is, the main institutions of the social system
4 This question is a difficult one, and I have nowhere given an adequate discussion of it.
Some intuitive arguments are sketched in Justice as Fairness, op. cit. supra note 2 at 171-74,
Constitutional Liberty, op. cit. supra note 2 at 107-23, and Distributive Justice, op. cit.
supra note 1 at 61n.
5 See Constitutional Liberty, op. cit. supra note 2 at 114-18.
and how they are combined. This structure includes the political constitution
and the principal economic and social arrangements which together define
a person's liberties and rights, and affect his life-prospects as a whole. Since
men are born into the social system at different positions, they have different
expectations and life-chances determined, in part, by the scheme of political
liberties and personal civil rights and by the economic and social opportunities open to those in these positions. In this way the basic structure of society
favors certain individuals over others, and these are the fundamental inequalities. It is to the social system as specifying inequalities of this kind, presumably inevitable in any society, that the two principles apply. These principles regulate the basic framework of the social order as a whole. They are
not meant for more limited associations or for the small-scale situations of
everyday life. How far they hold for these cases has yet to be considered.
But it is clear that so-called lifeboat situations, and other similar examples,
pose a different sort of problem. In situations of this kind we are confronted
with a severe conflict of principles aggravated by an extensive knowledge of
the particular circumstances. Often there seems to be no satisfactory solution.
And, in fact, we should be prepared to recognize that most moral questions,
counting simply abstract possibilities, may have no reasonable answer. In
any case, I believe that it will turn out to be easier to settle the question of
social justice, understood as the justice of the basic structure, than to resolve
hard cases in everyday life. By ascending to more general matters and leaving aside particular information and the knowledge of certain contingencies,
the problem of justice becomes more accessible.
A second point is that the principles of justice presuppose that the main
question concerns the distribution of rights and duties, how these are to be
assigned in the fundamental arrangements of society. The formulation of the
principles assumes that we can usefully divide the social structure into two
more or less distinct parts, the first principle applying to one, the second to
the other. We are to distinguish between those aspects of the social system
which define and secure the equal liberties and opportunities of citizenship
and those aspects which establish or permit social and economic inequalities.
The fundamental freedoms, that is, the main political rights, liberty of conscience and freedom of thought, the basic civil rights, and the like -these
are all required to be equal by the first principle. The second principle applies
to the distribution of income and wealth, and to the structure of political and
economic institutions insofar as these involve differences in organizational
authority. While the distribution of wealth and authority is not required to
be equal, it must be to everyone's advantage, that is, each relevant representative man in the social system must reasonably expect to gain from it when
NATURAL LAW FORUM
he views the arrangement as a going concern. Further, offices and positions
are to be open to all. Questions of interpretation I leave aside for the moment.
What is to be emphasized here is that we are to try to design a system of
rights and duties such that, assuming the scheme to be properly carried out,
the resulting distribution of income to particular individuals and the allocation of particular items are just (or at least not unjust) whatever it turns out
to be. The system includes an element of pure procedural justice.
A closely related consideration is this. The problem of distributive justice
is not that of sharing out a given batch of things to particular individuals
with known tastes and preferences. We should not take as the typical case
instances of allocative justice, for example, situations in which, say, a wealthy
person is deciding how to divide his estate given a knowledge of the desires
and needs of the various individuals who are the possible objects of his beneficence. In this case a knowledge about particular persons is relevant to the
decision; and, moreover, their desires and needs are taken as fixed, whereas
in determining the justice of the social system as a whole, we want to consider
only general facts, that is, the general features of politics, economics, and
psychology, and so on, as these are expressed by the laws established by the
corresponding sciences. One is attempting to set up a workable and just
system, and since this system will itself affect the wants and preferences that
persons come to have, their wants and preferences cannot be taken as given.
Instead, one must choose between social systems in part according to the
desires and needs which they generate and encourage. To judge these desires
and needs, certain standards are necessary, and it is here that the conception
of justice has a role along with other moral principles.
Finally, two preliminary points about the phrase "everyone's advantage."
By "everyone" I understand all relevant representative men for judging the
social system. These I assume to be the representative (equal) citizens and
the representative individuals corresponding to the various ranges of income
and wealth. Thus the least favored man is the representative man in the
lowest income class that it is deemed feasible to distinguish. This group is
presumably more or less identical with the class of unskilled workers, those
with the least education and skills. In judging the social system one is to
view it from the standpoint of these representative persons. We are not required, say, in appraising the criminal law to look at it from the point of
view of the representative convicted felon. The other point is the following:
We need some way of defining the expectations of representative men, since
expectations determine what is to their advantage. Now I shall assume that
expectations are specified by the expected pattern of primary goods, that is,
things which rational persons may be presumed to want whatever else they
want. Men have an interest in having these goods however various their more
particular ends. For example, among the primary goods are liberty and
opportunity, income and wealth, health and educated intelligence. Perhaps
the most important primary good is self-respect, a confident conviction of the
sense of one's own value, a firm assurance that what one does and plans to
do is worth doing. Later on, the good of self-respect plays a part in showing
the merits of the two principles; but for simplicity the discussion is often in
terms of the other primary goods. Thus I shall say that when a representative
person can reasonably expect a larger index of primary goods, say a preferred pattern of liberty and opportunity, income and wealth, he is better
off. Inequalities are to everyone's advantage if they increase each representative person's expectations understood in this sense.
MucH, then, for a brief summary of some of the main points of "Distributive Justice." We are now in a position to take up in sequence three interpretations of the second principle of justice which specifies the conditions under
which social and economic inequalities are justified. To make things more
manageable, we may provisionally think of these inequalities as those of
income and wealth, but eventually a more comprehensive account is necessary.
Recall that the first principle of equal liberty is assumed to have the same
sense throughout, and therefore the several interpretations arise from the
various meanings of the two parts of the second principle.
Now both parts of the second principle have at least two natural senses;
and since these senses are independent of one another, there are four possible
interpretations of the two principles. These are indicated in this diagram:
(b) (b advantage"
(Principle of Mutual
NATURAL LAW FORUM
I shall examine in sequence these three interpretations: the system of natural
liberty, liberal equality, and democratic equality. In some respects this sequence represents the more intuitive development, but the second sequence
via the interpretation of natural aristocracy is not without interest, and I
shall comment upon it having set out the first sequence. As a problem of
interpreting the conception of justice as fairness, the question is which of
these four interpretations is to be preferred. I have suggested in "Distributive
Justice" that one should adopt that of democratic equality. By considering
these possibilities in order I wish to bring out some of the reasons for this
The first interpretation (in either sequence) I shall refer to as the system
of natural liberty. In this rendering the first part of the second principle is
understood as the principle of efficiency (Pareto optimality) adjusted so as
to apply to social institutions, or, in our case, to the basic structure of society;
and (b) the second part is understood as an open social system in which, to
use the traditional phrase, careers are open to talents. It is also assumed in
this, and also in the other interpretations, that the first principle is satisfied
and that the economy is roughly a free market system, although the means
of production may or may not be privately owned. The system of natural
liberty assumes, then, that a basic structure satisfying the principle of efficiency
and in which positions are open to those able and willing to compete for
them will lead to a just distribution. Assigning rights and duties in this way
is thought to give a scheme which allocates wealth and income, authority
and responsibility, in a fair way whatever this allocation turns out to be. The
doctrine includes an important element of pure procedural justice which is
carried over to the other interpretations.
I shall not discuss the principle of efficiency in any detail. It suffices to
recall that as used in economics the Pareto principle says that group welfare
is at an optimum when it is impossible by any economic change, say a redistribution of goods, to make any one man better off without at the same
time making at least one other man worse off.6 In order to apply this principle to institutions I assume, as already remarked, that it is possible to associate with each relevant social position an expectation which depends upon
the assignment of rights and duties in the basic structure. Then the Pareto
principle for institutions holds that the pattern of expectations (and the basic
structure underlying it) is optimal if and only if it is impossible to change the
rules, to redefine the scheme of rights and duties, so as to raise the expectations of any representative man without at the same time lowering the expec6 This
principle was introduced by Pareto in his MANUEL D'ACONOMIE POMITIQU
and for several decades now has been the basic principle of welfare economics.
tations of some other representative man. In comparing different arrangements of the social system, we can say that one is better than another if in
one arrangement all expectations are at least as high as in the other, and
The principle of efficiency, as it is well known, does not by itself determine
a unique distribution. Even when we apply it to the basic structure, there
are presumably many arrangements that satisfy it. Thus the question arises
as to how the system of natural liberty chooses among the efficient distributions. Now let's suppose that we know from economic theory that under the
standard assumptions defining a competitive free market system, income and
wealth will be distributed in an efficient way and that the particular optimal
distribution which results in any period of time is determined by the initial
distribution of assets, that is, by the initial distribution of income and wealth,
and of natural talents and abilities. With each initial distribution, a different efficient outcome is arrived at. Thus if we are to accept the outcome
as just, and not as merely efficient, we must accept the basis upon which over
time the initial distribution of assets is determined.
Now in the system of natural liberty the initial distribution is regulated
by the arrangements implicit in the conception of careers open to talents as
earlier defined. These arrangements presuppose a background of equal liberty
(as defined by the first principle) and a free market economy. They require
a formal equality of opportunity in that all have at least the same legal rights
of access to all advantaged social positions. But since there is no effort to
preserve an equality, or a similarity, of social conditions, except insofar as
this is necessary to preserve the required background institutions, the initial
distribution of assets for any period of time is strongly influenced by natural
and social contingencies. The existing distribution of income and wealth,
say, is the cumulative effect of prior distributions of natural assets, that is,
natural talents and abilities, as these have been affected and their use favored
or disfavored over time by social circumstances and such chance contingencies
as accident and good fortune, and so on. The injustice of the system of
natural liberty is that it permits distributive shares to be improperly influenced
by these factors.
The liberal interpretation, as I shall refer to it, modifies the system of
natural liberty by adding to the requirement of careers open to talents the requirement of the principle of fair equality of opportunity. The thought here is
that positions should not only be open to all in a formal sense, but that all
should have a fair chance to attain them. Offhand it is not clear precisely
what the desired equality entails, but assuming that the analogy with games
is relevant, we might say that those with similar abilities and skills should have
NATURAL LAW FORUM
similar life chances. More specifically, assuming that there is a distribution
of natural assets, those at the same level of talent and ability and who have
the same willingness to use them, should have the same prospects of success
regardless of their initial place in the social system, that is, irrespective of the
class into which they were born. In all sectors of society there should be
roughly equal prospects of culture and achievement for everyone similarly
endowed and motivated. The expectations of those with the same abilities
and aspirations should not be affected by their social class.
The liberal conception of the two principles seeks, then, to mitigate the
influence of social contingencies on distributive shares. To accomplish this
end it is necessary to impose certain basic structural conditions on the social
system. One has to set free market arrangements within a framework of
political and legal institutions which regulates the overall trends of economic
activity and which preserves the social conditions necessary for fair equality
of opportunity.8 The elements of this framework are familiar enough, though
it may be worthwhile to recall the importance of preventing excessive accumulations of property and wealth and of maintaining equal educational opportunities for all. Chances of education should not depend upon one's class
position; and the school system, whether public or private, should be framed
to even out, or at least not to strengthen, class barriers.
While the liberal conception is dearly preferable to the system of natural
liberty, it is still defective. For one thing, even if it works to. perfection in
eliminating the influence of social contingencies, it still allows the resulting
distribution of wealth and income to be singled out by the natural distribution
of abilities and talents. Within the limits allowed by the background arrangements, distributive shares are decided by the outcome of the natural lottery;
and this outcome is also arbitrary from a moral point of view. There is no
more reason to permit the distribution of wealth and income to be settled by
the distribution of natural assets than by historical accident and social fortune.
Furthermore the principle of fair opportunity can only be imperfectly carried
out, at least as long as-we accept the institution of the family, as I have been
assuming. The extent to which natural capacities develop and reach fruition
is affected by all kinds of social conditions and class attitudes. Even the
willingness to )rnake an effort, to try, and so to be deserving in the ordinary
sense is itself dependent on fortunate family and social circumstances. It is
impossible in practice to secure equal chances of achievement and culture
for those similarly endowed, and therefore we need to adopt a principle
7 This definition follows a suggestion of Sidgwick's; see his THE METHODS OF ETHICS
285n. (7th ed., 1907). See also R. H. TAWNzY, FQUALITY ch. III, see. (ii) (4th ed., 1952).
s See Distributive justice, op. cit. supra note I at 69-72.
which recognizes this fact and which also mitigates the arbitrariness of the
natural lottery itself. That the liberal conception fails to do this is its chief
weakness from the standpoint of interpreting the two principles of justice
clearly is to find an interpretation of the two principles under
which distributive shares are not improperly influenced by the arbitrary contingencies of social fortune and the lottery of natural assets. The liberal conception is a step in the right direction but it does not go far enough. In
"Distributive Justice" I suggested that the first part of the second principle
be rendered in the following way. Assuming the framework of institutions required by fair equality of opportunity to obtain, the higher expectations of those better situated in the basic structure are just if and only if they
work as part of~ scheme which improves the expectations of the least advantaged members of society. The intuitive idea here is that the social order
should not establish and secure the more attractive prospects of those better
off unless doing so is to the advantage of those less fortunate. The basic structure is just throughout when the higher expectations of each more favored
group further the well-being of the class that is least favored, that is, when
a decrease in any of these advantages would make the least fortunate even
worse off than they are. The basic structure is perfectly just when the lifechances of the least favored are as high as they can be consistent with the
requirements of the other principles.
To illustrate, suppose that the least advantaged are represented by the
typical unskilled worker. The proposed principle requires that inequalities
in wealth and income- for example, the greater distributive shares earned
by entrepreneurs and professional persons-must be justified to the members of this group. Differences in wealth and income are just only if they
are to the advantage of the representative man who is worse off. Presumably,
given a background of equality of opportunity, and so on, the greater expectations of entrepreneurs and professional persons have the effect in the long
run of raising the life-prospects of the laboring class. They provide the necessary incentives so that the economy is more efficient, industrial advance pro-
ceeds more rapidly, and so on. All of this is familiar. I do not wish to say
that it is true in any particular case. But it may serve to illustrate the sort
of thing that must be argued if the suggested principle is t be satisfied. All
differences in wealth and income, all social and economic inequalities, should
work for the good of the least favored. For this reason I call it the difference
NATURAL LAW FORUM
principle. As we shall see, it is also a reciprocity principle expressing a natural
condition of mutual advantage.
Some of the finer points in interpreting this principle I have considered
elsewhere. 9 For our discussion here it suffices to mention two things. First,
the difference principle satisfies the principle of efficiency. For when it is
fully met, it is impossible to make any representative man better off without
making some other such man worse off, namely, the least favored representative man. Thus it defines a conception of distributive justice that is compatible
with Pareto optimality. Just as the liberal conception added to but did not
contradict the requirements of careers open to talents, so the democratic conception adds to the liberal conception. On this interpretation the two principles express a consistent and natural development from the system of natural
liberty. Secondly, it must be kept in mind that the difference principle applies
in the long run as this extends over generations. It is the expectation of the
least advantaged from one generation to the next that is to be improved consistent, as always, with the constraints of equal liberty and fair equality of
The case for the democratic interpretation rests on the fact that when
it is satisfied distributive shares are not improperly influenced either by social
contingencies or by the lottery in natural assets. So understood, the two principles define a fair way of dealing with accidental social circumstances and
the arbitrary distributions of nature; and in this respect they are superior to
I should now comment on the second sequence via the conception of
natural aristocracy. On this view no attempt is made to regulate the effects
of social contingencies beyond what is required by formal equality of opportunity, but the advantages of persons with greater natural endowments are to
be limited to those that further the good of the poorer sectors of society. The
aristocratic ideal is applied to a social system at least formally bpen, and the
better situation of those favored by it can be justified only by radiating benefits
and on the condition that were less given to those above, less would be attained
by those beneath them. 10 In this way the idea of noblesse oblige is carried
over to the conception of natural aristocracy. Now it should be noted that
both the liberal view and that of natural aristocracy are unstable. For once
o Id. at 66-68.
This formulation is derived from Santayana's description of aristocracy in the last part
of ch. IV, REASON AND SOCIETY (1906). I am indebted to the members of the editorial
board of the NATURAL LAW FORUM for calling my attention to the conception of natural
aristocracy as a possible interpretation of the two principles. And I am grateful to Professor
Robert Rodes for pointing out to me that a feudal system such as existed in the Middle
Ages might also try to fulfill the difference principle. A more complete discussion would have
to consider these further cases.
we are troubled by the influence of either social contingencies or natural
chance on the determination of distributive shares, we are bound, on reflection, to be equally concerned by the influence of the other. So however we
move away from the system of natural liberty, we cannot stop short of the
democratic conception. Indeed, as soon as we try to formulate an interpretation of the two principles of justice which treats everyone equally as moral
persons, and which does not weight men's share in the benefits of social cooperation according to happenstance and accident, and in particular by their
good social fortune or their luck in the natural lottery, it is clear that the'
democratic rendering is the only possible choice among the four alternatives.
I now wish to support this conclusion by presenting further reasons in support
of the difference principle.
sight the difference principle seems unworkable if not eccentric. A
necessary step in removing this impression is to keep in mind that it applies
to the basic structure of society and to representative groups defined by it.
But, as noted in "Distributive Justice,"'" the crucial point is that the difference principle can be regarded as an agreement to consider the distribution of natural assets as common property and to share in the benefits of this
distribution whatever it turns out to be. Those who have been favored by
nature, whoever they are, may gain from their good fortune only on terms
that improve the situation of those who have done less well. The naturally
advantaged are not to gain merely because they are more gifted, but rather
in order to cover the costs of training and cultivating their endowments and
for putting them to use in a way that helps the less fortunate as well. No
one is thought to deserve his greater natural capacity or to merit a more
favorable starting place in society. But it does not follow that these distinctions should be eliminated, or scaled down. There is another way to deal
with them. The basic structure can be arranged so that these contingencies
work to everyone's advantage; the system can be justified to all, and therefore
to the least favored representative man. One is led to the difference principle
if one wishes to frame the social system so that no person gains or loses from
his arbitrary place in the distribution of natural endowments or from his
initial position in society without going on receiving compensating benefits in
return. The reasons favoring the principle invoke this basic idea in various
I begin by noting that the difference principle gives some weight to the
See Distributive Justice, op. cit. supra note 1 at 68.
NATURAL LAW FORUM
considerations singled out by the principle of redress. This is the principle
that undeserved inequalities call for redress; and since inequalities of birth
and natural endowment are undeserved, these inequalities are to be somehow
compensated for. 12 Thus the principle of redress holds that in order to treat
all persons equally, to provide genuine equality of opportunity, society must
give more attention to those with fewer native assets and to those born into
the less favorable social positions. The idea is to redress the bias of contingencies in the direction of equality. In pursuit of this principle greater resources might be spent on the education of the less rather than the more intelligent, at least over a certain time of life, say, the earlier years of school.
The principle of redress has not, to my knowledge, been proposed as the
sole criticism of justice, as the single aim of the social order. It is plausible,
as most such principles are, as a prima facie principle, one that has to be
weighed in the balance with others. For example, we are to weigh it against
the principle to improve the average standard of life, or to advance the common good.13 But whatever other principles we hold, the claims of the principle of redress are to be taken into account. It is thought to represent one
of the elements in our conception of justice.
Now the difference principle is not, of course, the principle of redress.
It does not require society to move in the direction of an equality of natural
assets. We are not to try to even out handicaps as if all were expected to
compete on a fair basis in the same race. But the difference principle would
allocate resources in education, say, so as to improve the long-term expectation of the least favored. If this end is attained by giving more attention to
the better endowed, it is permissible; otherwise not. And in making this decision, the value of education should not be assessed only in terms of its productivity effects, that is, its realizing a person's capacity to acquire wealth. Equally
important, if not more so, is the role of education in enabling a person to enjoy
the culture of his society and to take part in its affairs, and in this way to
provide each man with a secure sense of his own worth.
Thus, although the difference principle is not the same as that of redress,
it does achieve some of the intent of the latter principle. It transform the
aims of the basic structure so that the total scheme of institutions no longer
emphasizes social efficiency and technocratic values. The ideal of equal
opportunity ceases to be the right to leave the less fortunate behind, and
hence the callous aspects of a meritocratic regime are avoided.
12 See H. Spiegelberg, A Defense of Human Equality, 53 PHILOSOPHICAL REVIEW 101,
113-23 (March, 1944); and D. D. Raphael, justice and Liberty, 51 PROCEEDINGS OF THE
ARISTOTELIAN SOCIETY 187ff. new series (1950-51).
Is See, for example, Spiegelberg, op. cit. supra note 12 at 120f.
advantage of the difference principle is that it provides an interpretation of the principle of fraternity. In comparison with liberty and equality, the idea of fraternity has had a lesser place in democratic theory. It is
thought to be less specifically a political concept, not in itself defining any
of the democratic rights but conveying instead the attitudes of mind without
which the values expressed by these rights would be lost sight of. 1 4 Or closely
related to this, fraternity is held to represent a certain equality of social esteem
manifest in various public conventions and in the absence of manners of
deference and servility.' 5 No doubt fraternity implies these things, as well
as a sense of civic friendship and solidarity, but so far it expresses no definite
requirement. The difference principle, however, corresponds to a natural
meaning of fraternity: namely, to the idea of not wanting to have greater
advantages unless this is for the benefit of others who are less well off. In
the ideal conception of the family, the principle of maximizing the sum of
advantages is rejected, and no one wishes to gain unless he does so in ways
that further the interests of others less well situated. But wanting to act on
the difference principle has precisely this consequence. Those better circumstanced are willing to have their greater advantages only under a scheme in
which this works out for the welfare of the less fortunate.
The idea of fraternity is sometimes thought to involve ties of sentiment
and feeling which it is unrealistic to expect between the members of society.
But if it is interpreted as incorporating the requirements of the difference
principle, it is not an impracticable criterion. It does seem that the institutions and policies which we most confidently think to be just satisfy its demands, at least in the sense that the inequalities permitted by them contribute
to the well-being of the less favored. On this interpretation the principle of
fraternity is a feasible standard to follow. We can now associate the ideas of
liberty, equality, and fraternity with the democratic interpretation of the two
principles of justice as follows: liberty corresponds to the first principle,
equality with equality of fair opportunity, and fraternity with the difference
principle. We have found a place for the notion of fraternity in the democratic interpretation of the two principles, and we see that it imposes a definite
requirement on the basic structure of society.
It is also possible to use the difference principle to give an explication of
the Kantian idea that men are always to be treated as ends and never as
14 See J. R. PENNOCK, LIBERAL DEMocnmcY: ITS MERITS AND PROSPrCTS 94f. (1950).
15 R. B. PERRY, PURITANISM AND DEMOCRACY ch. 19, sec. 8 (1944).
NATURAL LAW FORUM
means only.' 6 I shall not examine Kant's view here; instead, I shall freely
render this notion in the light of the two principles. Now this suggestion of
Kant's obviously needs interpretation. Certainly we cannot say that it comes
to no more than treating everyone by the same general principles. This interpretation makes it equivalent to formal justice, that similar cases (as defined
by the system of principles, whatever it is) are to be treated similarly; but
even a caste or slave society can meet this condition. In accordance with the
conception of justice as fairness we could say that treating persons always as
ends and never as means only signifies at the very least dealing with them as
required by those principles to which they would consent in an original position of equality. For in this situation men have equal representation as moral
persons who regard themselves as ends, and the principles they accept will
be rationally designed to protect the claims of their person. This conception
of justice, then, gives a sense to the Kantian idea. But this interpretation is
abstract, and the question arises as to whether this idea is expressed by any
substantive principle; or put another way, if the contracting parties wished
to express visibly in the structure of their institutions that they wanted always
to regard each other as ends and never as means only, what principle would
they acknowledge in the original position?
Now it seems that the two principles of justice on the democratic interpretation achieve this aim: for all have an equal liberty, and the difference principle insists that none gain from basic inequalities in the social system except
in ways that further the advantage of the less fortunate. The difference principle explicates a distinction between treating men as means only and treating
them as ends in themselves. To regard a man as an end in himself in the
basic design of society is to agree to forego those gains which do not contribute
to his expectations. By contrast, to regard a man as a means only is to be
prepared to impose upon him lower prospects of life for the sake of sufficiently
compensating higher expectations of others and so to promote a greater sum
of advantages. The principle of utility subordinates persons to the common
good, or to the end of attaining the greatest net balance of satisfaction, in a
way that the conception of justice as fairness excludes. For this principle
allows, at least in theory, that the greater gains of some, even when they are
better off, mayj offset the losses of others less fortunate.
The utilitarian may object to this that his principle also gives a meaning
to the Kantian idea, namely, the meaning provided by Bentham's formula
(attributed to him by J. S. Mill) "everybody to count for one, nobody for
16 See I. KANT, THE FOUNDATION OF THE METAPHYSICS OF MORALS 66f. (2nd ed.,
more than one."' 17 This implies, as Mill remarks, that one person's happiness
assumed to be equal in degree to another person's is to be counted exactly
as much. The weights in the additive function that represents the utility
principle are the same for all individuals, and it is natural to take them as
one. Therefore the utilitarian can say that his principle treats all persons
always as ends and never as means only, since the well-being of all is given the
same weight, and no one's happiness is counted for naught.
There is no reason to'deny that utilitarianism can give this interpretation.
Rather, the two principles of justice with the difference principle provide a
more plausible rendering, or at least a stronger one. For it seems that a
natural sense of treating others as means is to require them to have less, to
make still further sacrifices, for the purpose of raising the prospects of others
who are more favored. To ask a man to accept less for the sake of improving the welfare of others whose situation is already better than his is to treat
him as a means to their well-being. The principle of utility, one might say,
treats persons both as means and as ends. It treats them as ends by assigning
the same weight to the welfare of each; it treats them as means by allowing
the benefits to some to counterbalance the losses of others, especially when
those who lost out are the less advantaged. The difference principle provides
a stronger interpretation to the Kantian idea by ruling out even the tendency
to regard men as means to one another's welfare. It gives a meaning to a
more stringent variant of the notion, namely, always to treat persons solely
as ends and never in any way as means.' 8
A FURTHER consideration in support of the difference principle is that it
satisfies a reasonable standard of reciprocity. Indeed, it constitutes a principle
of mutual benefit, for, when it is met, each representative man can accept the
basic structure as designed to advance his interests. The social order can be
justified to everyone, and in particular to those who are least favored. By
contrast with the principle of utility, it is excluded that any one worse off
than another should be asked to accept less so that the more advantaged can
have more. This condition seems an essential part of the notion of reciprocity,
and the difference principle fulfills it whereas utilitarianism does not.
It is necessary, however, to discuss in more detail how the condition of
mutual benefit is satisfied. Consider any two representative men A and B,
and let B be the one who is worse off. Actually, since we are most interested
" See J. S. MILL, UTILITARIANISM ch. V, par. 36.
28 For this idea I am indebted to Anton-Hermann Chroust.
NATURAL LAW FORUM
in the comparison with the least favored man, let's assume that B is this
individual. Now clearly B can accept A's being better off since A's advantages have been gained in ways that improve B's prospects. If A were not
permitted to win his better position, B would be even worse off than he is.
The difficulty, if there is one, is to show that A has no grounds for complaint.
Perhaps he is required to have less than he might since from it there would
be no additional benefit to B. Now what can be said to the more favored
man? First of all, it is clear that everyone's well-being depends upon a
scheme of social cooperation without which no one could have a satisfactory
life. Second, such a scheme should draw forth the willing cooperation of
everyone taking part in it, and this can be asked for only if reasonable terms
are proposed. The difference principle, then, seems to be a fair basis on which
those better endowed, or more fortunate in their social circumstances, could
expect the willing cooperation of others when some workable scheme is a
necessary condition of every one's welfare. 19
There is a natural inclination to object that those better situated deserve
their greater advantages whether or not they are to the benefit of others. But
at this point it is necessary to be clear about the notion of desert. It is perfectly true that given a just system of cooperation as a scheme of public rules
and the expectations set up by it, those who with the prospect of improving
their condition have done what the system announces that it will reward are
entitled to their advantages. In this sense the more fortunate deserve their
better situation; their claims are legitimate expectations established by social
institutions, and the community is obligated to meet them. But this sense of
desert presupposes the existence of the cooperative scheme; it is irrelevant as
to how in the first place the scheme is to be designed, whether in accordance
with the difference principle or some other criterion.
Perhaps some will think that the person with greater natural endowments
deserves those assets and the superior character that made their development
possible. Because he is more worthy in this sense, he deserves the greater
advantages that he could achieve with them. This view, however, is surely
incorrect. It seems widely agreed that no one deserves his place in the distribution of native endowments, any more than one deserves one's initial
starting place ,'n society. Also a man's deserving the superior character that
enabled him to cultivate his abilities is equally problematic; for this character
depends upon fortunate family and social circumstances for which he can
claim no credit. The notion of desert seems not to apply to these cases. Thus
the more advantaged representative man cannot say, and being reasonable,
29 For this argument I am Indebted to Allan Gibbard.
will not say, that he deserves and therefore has a right to a scheme of cooperation in which he is permitted to acquire benefits in ways that do not
contribute to the welfare of others. There is no basis upon which he can
make this claim.
The difference principle should be acceptable, then, both to the more
advantaged and to the less advantaged man. The principle of mutual benefit
applies to each increment of gain for the more favored individual, a unit
increase, so to speak, that improves the situation of this individual being,
allowed provided that it contributes to the prospects of the least fortunate.
The principle of reciprocity applies each step of the way, the increments for
the better situated continuing until the mutual benefit ceases. It is evident
that, in general, the principle of utility does not satisfy the principle of reciprocity; there is no definite sense in which everyone necessarily benefits from
the inequalities that are authorized by the utilitarian conception. It seems
irrelevant to say that everyone is better off than he would be in a state of
nature, or if social cooperation were to break down altogether, or even that
all are better off than they were in comparison with some historical benchmark. We want to be able to say that as the social system now works, the
inequalities it allows contribute to the welfare of each.
Now the fact that the two principles of justice embody this reciprocity
principle is important for the stability of this conception. A conception of
justice is stable if, given the laws of human psychology and moral learning,
the institutions which satisfy it tend to generate their own support, at least
when this fact is publicly recognized. Stability means that just arrangements
bring about in those taking part in them the corresponding sense of justice,
that is, a desire to apply and to act upon the appropriate principles of justice.
Assuming as a basic psychological principle that we tend to cherish what
affirms our good and to reject what does us harm, all those living in a basic
structure satisfying the two principles of justice will have an attachment to
their institutions regardless of their position. This is the case since all representative men benefit from the scheme. In a utilitarian society, however, this
is not guaranteed; and therefore to the extent that this psycholqgical principle holds, the principle of utility is likely to be a less stable conception.
This consideration is reinforced by the fact that a social system acknowledging
the stronger variant of the Kantian idea, that is, the idea of always treating
persons solely as ends and never in any way as means, is bound to underwrite
a more secure foundation for men's sense of their own worth, a firm confidence that what they do and plan to do is worth doing. For our self-respect,
which mirrors our sense of our own worth, depends in part upon the respect
shown to us by others; no one can long possess an assurance of his own value
NATURAL LAW FORUM
in the face of the enduring contempt or even the indifference of others. But
it is precisely by publicly affirming the two principles of justice on the democratic interpretation that a society acts on the Kantian idea and makes visible
in its institutions the respect that its citizens have for one another as moral
persons. 2 0 Since a sense of our worth is perhaps the most important primary
good, these considerations constitute strong arguments for the democratic
interpretation, as well as for the two principles of justice in preference to
the principle of utility.
It is remarkable that Mill appears to agree with this conclusion. In
Utilitarianismhe notes that with the advance of civilization men come more
and more to recognize that society between human beings is manifestly
impossible on any other basis than that the interests of all are to be consulted.
The improvement of political institutions removes the opposition of interests
and the barriers and inequalities that encourage individuals and classes to
disregard one another's claims. The natural end to this development is a
state of the human mind in which each person has a feeling of unity with
others. This state of mind, if perfect, makes an individual "never think of,
or desire, any beneficial condition for himself, in the benefits of which they
are not included." One of the individual's natural wants is that "there
should be harmony between his feelings and aims and those of his fellow
creatures." He desires to know that "his real aim and theirs do not conflict;
that he is not opposing himself to what they really wish for, namely their
own good, but is, on the contrary, promoting it."21
Now the desire Mill
characterizes here is a desire to act upon the difference principle and not a
desire to act on the principle of utility. Certainly it is curious that Mill does
not notice this discrepancy; in any case, however, he seems intuitively to
recognize that a perfectly just society in which men's aims are harmoniously
reconciled would be one that followed the concept of reciprocity expressed
by the difference principle. A stable conception of justice which generates
its own support and elicits men's natural sentiments of unity and fellow
feeling is more likely to follow this ideal than the utilitarian standard.
of these arguments shows, I believe, that the democratic inter-
pretation of the two principles is a reasonable conception of justice. It turns
20 These remarks parallel a stronger sense of fraternity introduced by Perry in the section
following the one cited above; for example, when he says that the "full spirit of fraternity
acknowledges the just pride of others, and gives in advance that which other's self-respect
demands. It is the only possible relation between two self-respecting persons."
21 See MILL, op. cit. supra note 17 at ch. III, pars. 10-11.
out that despite its initial implausibility the difference principle expresses a
number of aspects of the democratic ideal. I should now like to comment
on the contrast between the conception of justice as fairness and the utilitarian doctrine. The essence of the contrast, as I have suggested before, is
that justice as fairness takes the plurality of persons as fundamental whereas
utilitarianism does not.2 2 The latter tries to maximize some one thing, it being
indifferent in which way it is shared among persons except insofar as it affects this one thing itself. Utilitarianism extends the principle of rational
choice for one person to the group, to the plurality of persons. For one person
may properly count his advantages now as compensating for his own losses
earlier or subsequently, but the principles of justice rule out the analogous
kind of reasoning between persons. The principles of practical reasoning for
an entity capable of decision are to be decided upon by that entity. Thus the
principle of choice for a rational individual is his to make; but, similarly, the
principles of social choice must be adopted by the association itself, that is,
by the individuals that constitute it. The fundamental idea of the doctrine
of the social contract, of which justice as fairness is simply an elaboration, is
that this decision should be made in an initial contractual situation in which
fair representation is given to each individual as a moral person. The principles that would be agreed to in this original position (the possible alternatives
being presented by a list of traditional principles) single out the most reasonable conception of justice.
Now the development of the theory of justice brings out that the idea of
the initial contractual situation involves many elements and can be defined
in various ways. This situation is the analogue, in the contract theory, of the
point of view of the impartial sympathetic spectator in utilitarianism; and
these constructions, though distinct, become all the more easy to assimilate to
one another when the original position is conceived, as it must be, as a purely
hypothetical occurrence. Thus it is possible to formulate conceptions resembling the original position of the contract doctrine which do not lead to
the two principles of justice but to the principle of utility instead. In this way
a weak variant of the contract construction might be used as a roundabout
and perhaps more sophisticated justification of the utility principle. As I
cannot examine these matters thoroughly here, I shall only remark that in
each case the variant of the original position loses sight of the core of the contract view, namely, the idea that since the essence of society is a plurality of
distinct individuals, this plurality of persons must be conceived as choosing
together before one another their common conception of justice. Thus if we
22 See Constitutional Liberty, op. cit. supra note 2 at 124f; and Distributive Justice, op.
cit. supra note 1 at 58-60. See also D. GAUTHIER, PRACTICAL REASONING 126 (1963).
NATURAL LAW FORUM
suitably combine the notions of impersonality and lack of information concerning social position with that of an individualistic risk-taker, then Harsanyi
has shown that one is led to the principle of utility. 2 3 The traditional utilitarian doctrine since Hume has been based on the fact that an ideally impartial and sympathetic spectator who identifies with all the interests in conflict and who thinks of each interest as if it were his own likewise judges in
accordance with the utilitarian criterion. 24 And it is a striking fact that even
writers who seek to incorporate certain Kantian elements into their ethical
theory present an account which turns out to be utilitarian. A clear cxample
of this is Nelson, and more recently Hare. 25 Of course, for a utilitarian this
outcome is confirmation of his doctrine; but for the contract view, which is
the traditional alternative to utilitarianism, such a conclusion would be a
disaster. The ascent to the original contract would have come to nothing.
We know, then, the notions of generality of principle, universality of application, and limited information as to natural and social status are not
enough by themselves to characterize the original position of the contract
theory. They are necessary but not sufficient. The initial situation must be
one of group choice; it must not be that of one person, whether this person
is a rational risk-taker or an impartial sympathetic spectator. We are to conceive of the contracting parties as a group of family heads, say, who have
responsibilities to their descendants and to certain religious or cultural interests. It is also important that they are choosing principles to apply to the
basic structure; and since their choice is an agreement to which they are bound
to one another in perpetuity, the principles acknowledged must be ones that
they can be reasonably sure that they will be able to honor and abide by. In
this and other ways, I believe that the initial contractual situation can be
characterized so that the two principles of justice on the democratic interpretation would be chosen. This characterization is preferable to the parallel constructions leading to the principle of utility on the grounds that it is closer
to the relevant conditions of human life and more adequately expresses the
reasonable restrictions on the choice of first principles.
The bearing of these remarks, which I cannot adequately explain here, is
that the previous arguments for the difference principle, while they may stand
23 See J. C. Harsanyi, Cardinal Utility in Welfare Economics and in the Theory of RiskTaking, 61 JOURNAL OF POLITICAL ECONOMY 434 (October, 1953); and Cardinal Welfare,
Individualistic Ethics, and Interpersonal Comparisons of Utility, 63 JOURNAL OF POLITICAL
ECONOMY 309 (August, 1955).
24 See DAvD HUME, A TREATISE OF HUMAN NATURE bk. III, pt. III, sec. 1.
25 See L. NELSON, SYSTEM OF ETHICS 110-16 (trans. N. Guterman, 1956); R. M. HARE,
FREEDOM AND REASON ch. 7 (1963). The attraction for Hare of the utilitarian conception
of the impartial sympathetic spectator who conflates all persons into one by identifying with
all interests as if they were his own is particularly evident on p. 123.
on their own, can also be regarded as presented by the contracting parties to
one another in the original position where they acquire additional force. The
discussion relating the difference principle to that of reciprocity can be imagined as taking place not between two persons one of whom knows that he is
better off and another who knows that he is rather worse off, but between any
two persons who know only that it is possible that they hold these positions,
and who wish to reach an agreement that will be most acceptable to them no
matter how things turn out. The notion of an agreement is a central feature,
even though the similarity of the parties' situations often allows one to consider principles from the standpoint of a single individual. Since the original
position gives an equal place to all individuals as equal moral persons, the
principles that would be adopted are fair, and because they apply to the basic
structure of society, they are the principles of social justice. By ascending to
this initial contractual situation, we have reached a fair status quo, one that
relates persons on equal terns irrespective of the outcome of the natural
lottery or of social circumstances. The principles of justice can be viewed,
then, as an understanding between moral persons not to exploit for one's own
advantage the contingencies of their world, but to regulate the accidental
distributions of nature and social chance in ways that are mutually beneficial
for all. Interpreted in this way, we can, I think, construe the original position
so that it leads to the two principles of justice.