Res Publica

, 15:397

Rawls on Liberty and Domination


    • Department of PhilosophyFlorida State University

DOI: 10.1007/s11158-009-9102-6

Cite this article as:
Victoria Costa, M. Res Publica (2009) 15: 397. doi:10.1007/s11158-009-9102-6


One of the central elements of John Rawls’ argument in support of his two principles of justice is the intuitive normative ideal of citizens as free and equal. But taken in isolation, the claim that citizens are to be treated as free and equal is extremely indeterminate, and has virtually no clear implications for policy. In order to remedy this, the two principles of justice, together with the stipulation that citizens have basic interests in developing their moral capacities and pursuing their conceptions of the good life, are meant to provide a more precise interpretation of what is involved in treating citizens as free and equal. Rawls’ critics, however, have argued that satisfying the two principles of justice is not the most appropriate or plausible way to respect the status of citizens as free and equal. In relation to this debate, the present paper has two aims. The first is to examine Rawls’ account of the type of freedom that a just society must guarantee equally to its citizens. I will argue that those who think of Rawls as a theorist of freedom as non-interference are mistaken, because his notion of liberty resembles in important respects the republican notion of freedom as non-domination. Second, I will consider the extent to which Rawls’ principles of justice successfully protect the freedom as non-domination of all citizens so as to effectively treat them as free and equal.




One of the central elements of John Rawls’ argument in support of his two principles of justice is the intuitive normative ideal of citizens as free and equal (Rawls 1996, pp. 29–35; Rawls 2001, pp. 18–24). Rawls takes this ideal to play an important role in explaining and justifying the design of the original position. He also holds that the ideal is implicit in the public culture of contemporary democratic societies, and that it is present in the design and functioning of their political and legal institutions. For this reason he claims that the ideal can be regarded as widely shared. Perhaps this is true. Still, taken in isolation, the claim that citizens are to be treated as free and equal is indeterminate, and has virtually no clear implications for policy. In order to remedy this, the two principles of justice, together with the stipulation that citizens have basic interests in developing their moral capacities and pursuing their conceptions of the good life, are meant to provide a more precise and useful interpretation of what is involved in treating citizens as free and equal. The first principle of justice requires that society secure a scheme of equal basic rights and liberties for all its citizens; the second principle asserts that social and economic inequalities are permissible only when the following two conditions are met: (1) that there is fair equality of opportunity to access desirable social positions, and (2) that any such inequalities work to the benefit of everyone, particularly the least advantaged members of society (Rawls 1996, pp. 5–6). However, it is by no means obvious that satisfying Rawls’ two principles of justice is the most appropriate or plausible way to respect the status of citizens as free and equal—or so argue some of Rawls’ critics. In relation to this debate, the present paper has two aims. The first is to examine Rawls’ account of the type of freedom that a just society must guarantee equally to its citizens. I will argue that those who think of Rawls as a theorist of freedom as non-interference are mistaken, because his notion of liberty resembles, in important respects, the republican notion of freedom as non-domination. Second, I will consider the extent to which Rawls’ principles of justice successfully protect the freedom as non-domination of all citizens so as to treat them effectively as free and equal.

The Basic Liberties and their Worth

Although Rawls sometimes refers to liberty in general terms, it is clear from his discussion of the notion of liberty that he has in mind a set of specific liberties that should be granted to the citizens of a just society.1 In fact, he specifies the basic rights and liberties of citizens by means of a short list. The list is quite heterogeneous and consists of the following elements (Rawls 1996, p. 291)2:
  1. 1.

    Liberty of conscience and freedom of thought: Liberty of conscience includes the freedom to practice any particular religion, or no religion at all. It also includes the freedom to pursue a set of values grounded in philosophical, moral, or metaphysical views. Freedom of thought extends to the expression of belief regarding a wide variety of subjects, including, for example, art, science and literature.

  2. 2.

    Freedom of association: This is the liberty to associate with whomever one chooses and to form different sorts of groups with the goal of promoting a variety of ends.

  3. 3.

    Equal political liberties: These include the right to vote and to hold office, freedom of political speech, freedom of assembly, freedom to criticize the government, and to form and join political parties.

  4. 4.

    Rights and liberties that protect the integrity and freedom of the person: These include prohibitions against coercion, physical harm and enslavement, freedom to choose one’s occupation, and a right to hold a certain amount of personal property, among other things.

  5. 5.

    Rights and liberties covered by the rule of law: These include things such as freedom from arbitrary arrest and the seizure of one’s property, the right to a fair trial, the right against self-incrimination, and other due process rights. Rawls claims that the rule of law is a precondition for the enjoyment of a number of liberties, because it sets the boundaries between what is legally permissible to do and what is not, and allows citizens to form reasonable expectations about the behavior of others.


Rawls seems to have produced his list of the basic rights and liberties of citizens primarily by examining the historical development of actual democracies over the past few centuries. Certainly the list conspicuously includes the kinds of liberties that were singled out as central during the political struggles that led to their constitutional protection in a number of countries. But Rawls also uses a further criterion to refine the set of liberties that count as basic. This criterion derives from his own account of citizens as free and equal, with interests in developing and exercising their moral capacities (the capacity for a sense of justice and the capacity for a conception of the good), as well as in pursuing their conceptions of the good (Rawls 1996, p. 292–3). Because he holds this account, certain rights and liberties do not count as basic for Rawls: those that do not seem to him to be essential to the informed and effective exercise of moral capacities—for example, the right to own the means of production or the right to bear arms.

Although there has been some controversy concerning whether Rawls holds a negative or a positive conception of freedom, he himself explicitly tries to avoid labeling his account of liberty in one way or the other.3 Following Gerald MacCallum (1967) he claims that we can explain any liberty in terms of three elements: what counts as an agent, what counts as a restriction on freedom, and what activities agents are free to engage in or refrain from. In Rawls’ own words:

The general description of a liberty, then, has the following form: this or that person (or persons) is free (not free) from this or that constraint (or set of constraints) to do (or not to do) so and so. Associations as well as natural persons may be free or not free, and constraints may range from duties and prohibitions defined by law to the coercive influences arising from public opinion and social pressure. For the most part, I shall discuss liberty in connection with constitutional and legal restrictions. In these cases liberty is a certain structure of institutions, a certain system of public rules defining rights and duties. Set in this background, persons are at liberty to do something when their doing it or not doing it is protected from interference by other persons (Rawls 1999a, p. 177).

A careful reading of Rawls’ list of the basic rights and liberties shows that he takes the class of relevant basic agents to be limited to individual persons and that the liberties that associations may have to act as collective agents turn out to be derivative, and to be dependent on the liberties of individuals.4 As the above paragraph makes explicit, Rawls takes the relevant constraints on liberty to be those defined by law. That is, he excludes as possible infringements on freedom—in the case of the basic liberties—not only the effects of the informal sanctions of public opinion, but also natural obstacles and internal psychological obstacles to action. Finally, the list of basic liberties establishes a range of protections for what are considered the central interests of citizens. That is, the list provides a set of criteria that indicate which types of actions or activities it is most important that the laws protect.

A number of neo-republican critics of Rawls—including Fabien Spitz, Philip Pettit and Quentin Skinner—read Rawls’ claims as an endorsement of a negative conception of freedom; they take Rawls to understand freedom in terms of non-interference.5 They then go onto argue that this conception of freedom makes the absence of interference an ideal situation of liberty, so that ideal liberty can be enjoyed even (perhaps ideally) in the absence of other people (Spitz 1993, p. 332; Pettit 1999, p. 50). But while this latter claim might be true of a very simplistic conception of freedom as non-interference, it certainly does not apply to Rawls’ developed account of citizens’ basic rights and liberties. All of Rawls’ basic liberties are liberties of citizens, and are granted by political and legal institutions. Some of them—such as freedom of conscience, freedom of association, and political liberties—do not even make sense in the absence of other people. Furthermore, Rawls never says that citizens enjoy these liberties whenever they do not suffer from actual interference. Rather, he says that in order to enjoy them, there must be an adequate system of these liberties, so that citizens are effectively protected against interference (from the actions of individuals, groups, or the state) by political and legal institutions.

A further complication for efforts to put any well-established label on Rawls’ conception of liberty is that his theory of justice as fairness recognizes the need for economic resources as a necessary condition on citizens’ being able to take advantage of their liberties in pursuing their distinct conceptions of the good life. On Rawls’ view a just society is a society that respects and supports its citizens’ development of different conceptions of the good life, and that enables its citizens to pursue these conceptions as they themselves conceive them. This kind of activity is appropriately understood as an exercise of positive freedom—not, however, in the sense of involving a developed conception of personal autonomy. Rather, it amounts to granting citizens the authority and means to decide about how to live their lives.6

It is true that Rawls does not think that a lack of economic resources—or a lack of knowledge or ability—undermine the claim that citizens have the basic liberties. However, he does claim that such deprivations can affect the “worth” of these liberties. The worth of a liberty is a measure that depends both on the ability that a citizen has to make use of that liberty, and on the degree to which the exercise of that liberty is useful or essential to the realization of that citizen’s plan of life. Since a lack of money, knowledge, or ability can affect the usefulness of a liberty in the effective pursuit of a plan of life (which the theory is concerned to protect), the worth of a liberty will therefore always depend on that citizen’s values, inclinations, talents, material resources and social position. Each citizen will value his or her basic liberties in different ways: some will be very involved in political affairs and give more importance to their political liberties; others may work as journalists, poets, or artists and will make more extensive use of their freedom of expression; members of a third group may devote a significant part of their time and resources to religious observance, and so on. Because the second principle of justice allows certain inequalities in the allotment of desirable social positions and of economic resources, the exercise of basic liberties will turn out to be much easier for some citizens than for others. This might initially seem unfair, but minimal reflection shows that the goal of equalizing the worth of liberty is as difficult as that of equalizing individual wellbeing. As Rex Martin puts it,

To assure equal value, we would have to create a homogeneous class of persons—alike, for example, in vocational interests—and place them in similar circumstances. Since Rawls assumes throughout that there will be significant differences among persons in the real social world, it is no part of his program to achieve equality in fact for the actual value of liberties to various persons (Martin 1985), p. 55.

However, Rawls claims that the difference principle ensures that the worth of the liberties of the worst off members of society will always be greater than it would be if economic resources were distributed more equally.7

In an article written in the early seventies, Norman Daniels criticized Rawls’ claim that citizens’ liberties could be equal while the worth of their liberties was significantly unequal (Daniels 1989). Because the second principle of justice does not set an absolute limit on the degree of permissible inequality, Daniels pointed out that there is a plausible reading of the theory of justice as fairness that legitimizes very significant inequalities of power, authority and economic resources among citizens. Daniels then argued that historical experience has shown that even if wealthy and poor citizens have the same formal political liberties, the unequal worth of these liberties will result in wealthier citizens having a systematic advantage in their exercise of political influence. Wealthier citizens are far more capable of influencing the selection of candidates by contributing to political campaigns, and they are more capable of influencing public opinion by controlling the media. And many other similar facts could be cited. As a result, individually unobjectionable instances of the exercise of political liberty—instances that the theory treats as permissible—will have a tendency, taken together over time, to produce structural results such that the interests of the wealthier and more powerful citizens are ultimately over-represented in democratic processes and the resulting laws and policies. This might seem to be a merely empirical argument, but Daniels correctly points out that the likelihood of this tendency challenges the stability of Rawls’ ideal model of a just society. It suggests that the theory is not realizable under favorable conditions, or that, even if it were realizable, that support for the principles of justice could not easily be sustained over time as inequalities became more and more pronounced (Daniels 1989, p. 255).

In response to Daniels’ criticism, Rawls conceded that allowing the worth of all the basic liberties to be unequal brought with it a number of unfortunate foreseeable consequences. He therefore modified the first principle of justice to include a requirement that guaranteed the equal worth of a certain subset of basic liberties. This subset is made up of the specifically political liberties. This new requirement, according to Rawls, can be met by restructuring the funding of elections, by subsidizing public debates, and by other measures designed to equalize the opportunities that citizens have to influence the political process.8 This modification of the first principle of justice may seem to be an ad hoc response to Daniels’ criticism. But it can also be seen as a principled response to the fact that the political liberties have a competitive aspect. That is, the relative quantities of resources that citizens have available, and which they can use to exercise their political liberties, will significantly affect their prospects of influencing the outcome of political processes (Brighouse 1997). It is worth noticing that Rawls seems to think that the other basic liberties are different from the political liberties in this respect. In particular, he thinks that for those citizens who are interested in exercising one of their basic non-political liberties, it is always better simply to secure the highest possible amount of economic resources. That is, in determining whether one economic arrangement is better for them than another with respect to the worth of a given non-political liberty, Rawls seems to think that there is no need to consider the relative economic positions of different people. For example, my interest in exercising my freedom of conscience is necessarily satisfied to a higher degree if I have a greater amount of economic resources (and leisure) available to devote to religious activities, than if I have fewer economic resources (and leisure); and this is true even if no one has more economic resources than me in the latter scenario, while this is not true in the former. Thus, Rawls’ idea that it is consistent with justice that citizens have equal liberty but unequal worth of liberty presupposes that the interests of citizens in exercising their liberties (excluding the political liberties) will be served in a fair way by the workings of the second principle of justice. This implies that the least advantaged citizens’ interest in making use of these basic liberties is better satisfied by the fair equality of opportunity principle taken together with the difference principle than by any alternative principles that directly aim at equalizing the economic resources of all citizens. Rawls therefore has more resources than some might think to defend the claim that he has captured the central sense of “freedom” at work in the ideal of citizens as free and equal.

In order to challenge Rawls’ claim that an equality in the rights and liberties of citizens is consistent with a significant inequality in the resources citizens have, and which are required for the exercise of such liberties, one would have to show that Rawls’ theory fails in some important way to protect the status of citizens as free and equal. One way to do this would be to try to show that the non-political basic liberties have some comparative or relational aspect as well: that one’s position relative to others has an impact on the extent to which one can enjoy all the basic liberties characteristic of free citizens. Philip Pettit’s neo-republican theory of freedom as non-domination can be used as a basis for developing such a criticism, although Pettit himself has not criticized Rawls on precisely these grounds. In the following section I will briefly present the main elements of Pettit’s account of freedom as non-domination in order to determine whether this particular type of freedom can be adequately protected by Rawls’ principles of justice. I will argue that freedom as non-domination captures a sense of freedom that is significant enough to require some adjustments to the application of the theory of justice as fairness and that it has the kind of comparative aspect that Rawls acknowledged as problematic in the case of the political liberties.

Freedom as Non-Domination

Philip Pettit has argued that there is a conception of freedom as non-domination that is distinct from the negative notion of freedom as non-interference and the positive notion of freedom as self-mastery (or autonomy) that Isaiah Berlin famously distinguished (Berlin 2002). Pettit’s distinct understanding of freedom is present in the intuitive judgment that to be free, in contrast to being in the condition of a slave or a colonized nation, is to be not subject to the arbitrary will of another. It hardly needs argument that it makes perfect sense to say that the slave or the colonized nation lack freedom. But many traditional accounts of freedom as non-interference or freedom as self-determination or autonomy cannot satisfactorily explain this judgment. Consider: the slave may have a benevolent master who does not interfere in her life to any very significant degree. Nevertheless, she lacks freedom. Similarly, the colonized nation may be ruled by a benevolent despot who does not interfere in their lives even to the degree characteristic of a democratic government. Nevertheless, they are not a free nation. As regards their autonomy, slaves and colonized nations may be autonomous if autonomy is understood as some sort of capacity for self-determination that they have succeed in developing in spite of their situation—though they would not count as autonomous under other conceptions that require the relatively unimpeded exercise of capacities for self-determination.9 It is the relational situation of domination (which is consistent with a lack of interference and with an agent’s possession of developed capacities for self-determination) that better explains the lack of freedom of slaves and colonized nations. Thus, Pettit calls his distinctive conception of freedom “freedom as non-domination.” The central theoretical notion in Pettit’s account—domination—can be understood as follows:

An agent A dominates another agent B to the extent that A is in a position to interfere arbitrarily in some of the choices and actions of B.10

For Pettit, there is a wide variety of actions—including coercion, threats and manipulation—that count as interference, since these actions “intentionally worsen an individual’s choice situation.” But of course not all interference is arbitrary. Pettit uses the expression ‘arbitrary interference’ to mean ‘interference that is not subject to suitable controls.’ There are two quite different types of arbitrary interference, depending on whether the agent that interferes is (a) a private individual or a private collective agent, or (b) the state. When the interfering agent is a private agent, A’s interference in B’s affairs counts as arbitrary whenever A can practice it at will and with relative impunity, when it is not controlled in a way that ensures that it tracks the individual interests and opinions of B. In contrast, if the interfering agent is the state, A’s interference with B’s affairs counts as arbitrary when it is not controlled in a way that ensures that it tracks the common interests of citizens as a class.11 With this understanding of domination in hand, Pettit’s view is essentially that someone is free, in his sense, if she is not dominated.

In order to clarify the negative notion of freedom as non-domination, it may be useful to contrast it with the negative notion of freedom as non-interference.12 There is some obvious overlap in situations that proponents of freedom as non-domination and advocates of freedom as non-interference would describe as situations in which people lack freedom: these are situations in which people suffer from actual arbitrary (or uncontrolled) interference. But the notion of freedom as non-domination makes room for the possibility of suffering interference that is not arbitrary and does not compromise freedom, as in the case of interference of appropriately established laws. But even here advocates of freedom as non-interference may agree with Pettit that such laws do not amount to a genuine loss of freedom, provided that the laws are justified in an appropriate way. The most significant difference between the two conceptions of freedom appears when one considers cases in which someone is dominated (that is, when someone is vulnerable to arbitrary interference), but in which that person is not suffering any actual interference. In these cases, someone concerned with freedom as non-interference would be concerned with the degree to which it is likely that there will be actual interference, rather than with the structural relationship of domination itself. Thus, if employees are in a position to suffer arbitrary interference from their employers because of the risk of unemployment if they are fired, or if wives are in a position to suffer arbitrary interference from their husbands because they would be destitute if they divorced or because there are no well enforced laws against spousal battery, a consistent defender of freedom as non-interference would be more concerned if such interference was statistically likely to happen. In contrast, Pettit views the structural relationship of domination as bad in itself. On his view, such relationships involve a loss of freedom even when the probability of the dominated person’s suffering from actual interference is very low. A focus on the structural relationship itself allows Pettit to note that sometimes the absence of interference—or its low probability—can be explained by the activities of the dominated person, who may resort to strategies of seduction, avoidance or ingratiation to make interference less likely. Different advocates of freedom as non-interference may make different assessments of these kinds of situations. Some may say that when people are led to strategic planning and to the adoption of servile behavior of this sort, they are in situations that resemble those in which someone is explicitly threatened. Such advocates of freedom as non-interference will be able to claim, with Pettit, that these situations compromise freedom. But others might say that since there is no explicit threat, these are not situations in which freedom is compromised. For example, this is the position that Isaiah Berlin would take (Berlin 2002).

Whatever the precise relation between freedom as non-interference and freedom as non-domination, it should be clear that concern with freedom as non-domination leads to a distinctive focus on inequalities of power as something that compromises freedom, at least when such power is not checked. If one accepts that the idea of freedom as non-domination captures an important sense in which people can be free, one important practical question is what kinds of policies can best secure or promote this kind of freedom. Pettit claims that the best means for reducing the capacity for arbitrary interference by private parties is by the rule of law, supplemented by policies of redistribution of resources that increase the bargaining power of vulnerable individuals. For example, well enforced criminal laws serve to protect individuals from physical harm and threats of physical harm, divorce laws that establish rights to alimony and adequate levels of child support increase the bargaining power of women in traditional marriages, and labor laws that mandate holidays and benefits protect employees from entering into exploitative labor contracts. With regard to assurances against arbitrary interference by the state, Pettit’s main focus is on the design of democratic political institutions. Among other beneficial effects, well designed democratic institutions can prevent the accumulation of power in a few hands by means of mechanisms of checks and balances. They can also check the performance of public representatives by means of regular elections. And they can make it easier for citizens to contest public decisions in a variety of ways, in order to make the law more responsive to their interests.

Justice as Fairness and the Prevention of Domination

With the above presentation of freedom as non-domination in hand, we are now in a position to determine whether Rawls’ principles of justice are capable of protecting citizens from domination both by private parties and by the state. If it is true that freedom as non-domination captures an important sense of freedom—one that ought to be included in the ideal of citizens as free and equal—a society that fails to protect its citizens from serious forms of domination could not plausibly count as just. I will argue that Rawls’ understanding of the basic liberties as legal protections from interference in central areas of one’s life comes very close to Pettit’s understanding of freedom as non-domination. In my view, Rawls’ conception of the basic liberties as protections of key areas of one’s life is completely different from the caricature of freedom as non-interference that sometimes appears in the literature as the “liberal conception of freedom”. This simplistic account of freedom as non-interference sees the law as a threat to freedom and takes ‘greater freedom’ to be virtually synonymous with ‘more options for action’. One important feature of the caricature of freedom as non-interference is that it does not differentiate, in a qualitative way, the kinds of activities that individuals can engage in or, therefore, the kinds of interference that are more serious. It counts any trivial restrictions on action as reductions of freedom. For example, on this view a detour on my normal route to work counts as a reduction of my freedom because I cannot go in my normal way. On Rawls’ understanding, this does not count as a reduction of basic liberty—not even as a trivial one—because driving on one street rather than another is not a central activity of citizens. Rawls’ list of basic rights and liberties identifies the kinds of activities of citizens that should be protected.

Now, it is true that Rawls does not generally consider the reduction of citizens’ freedom posed by relationships of domination.13 One reason for this is that his main concern is to justify an ideal theory of justice that assumes that citizens comply with the principles of justice and that there are no historical obstacles to the just functioning of social institutions. Nevertheless, it is worth asking whether Rawls’ theory manages, non-accidentally, to protect the freedom as non-domination of citizens. If justice as fairness is to secure the status of citizens as free and equal, it should protect them from the possibility of arbitrary interference by the state or private agents. Our question can be split into two distinct sub-questions, reflecting the dual nature of Pettit’s notion of domination:
  1. 1.

    To what degree are the choices of individual citizens protected from arbitrary interference by the state?

  2. 2.

    To what degree are the choices of individual citizens protected from arbitrary interference by private parties?

As regards the sort of freedom at issue in (1)—freedom as non-domination from the state—it should be fairly uncontroversial that the theory of justice as fairness adequately protects citizens via the basic rights and liberties. These are rights and liberties which are guaranteed by the political institutions; they are inalienable and the state cannot legitimately violate them. The only justification for allowing certain restrictions of the basic rights and liberties is that doing so is required to secure an adequate system of equal liberties for all citizens. Rawls thinks that a society cannot be just unless it is a democracy, the central social and political institutions of which embody the ideal of citizens as free and equal.14 Granting all citizens the right to vote, to express their political views, to participate in a variety of political associations, or to engage in civil actions such as lawsuits serves to ensure that citizens are not dominated by the state.15 Although Rawls does not enter very deeply into issues of institutional design, his discussion of the political liberties shows that he is aware of the need for an adequate design of democratic political institutions (Rawls 1996, pp. 324–63). Moreover, the theory of justice as fairness guarantees the fair value of political liberties. This guarantee is meant to prevent the richest from controlling the political process, with the result that their interests are overrepresented in state policy. Rawls acknowledges that when economic and social inequalities are too large, there is a risk that one part of society will dominate the rest by controlling the machinery of the state. He puts the point in the following way, explicitly using an intuitive notion of domination:

When those kinds of inequalities are large, they tend to support political inequality. As Mill said, the bases of political power are (educated) intelligence, property, and the power of combination, by which he meant the ability to cooperate in pursuing one’s political interests. This power allows a few, in virtue of their control over the machinery of the state, to enact a system of law and property that ensures their dominant position in the economy as a whole. Insofar as this domination is experienced as a bad thing, as making many people’s lives less good than they might otherwise be, we are again concerned with the effects of economic and social inequality (Rawls 2001, p. 131).

In my view Rawls’ list of citizens’ basic rights and liberties can be used to provide a plausible interpretation of Pettit’s more indeterminate notion of the common interests of citizens—a notion that is essential to Pettit’s account of state domination (Costa 2007). Moreover, the basic rights and liberties will be protected by law in a Rawlsian just society. Thus, in order to assess whether the state or particular laws dominate citizens, we can focus on the question of whether the state respects and upholds citizens’ basic rights and liberties. It is, however, not enough that these basic rights and liberties are simply listed in the constitution. Rather, they must also be adequately protected by means of effective laws and adequate institutional design. This is not very different from Pettit’s proposal for dealing with the problem of domination by the state. But Pettit avoids providing a list of basic rights and liberties of citizens, appealing instead to the idea of that state interference should be checked in a way that is forced to track the common interests of citizens, whatever those interests turn out to be. When Pettit does consider the strategy of talking about the basic liberties of citizens instead of using ‘liberty’ as a mass noun, he decides that it is best to leave the basic liberties unspecified. In place of an explicit list such as Rawls’, he offers a number of conditions that a list of liberties must meet if it is to count as a list of basic liberties. For example, the liberties should be capable of being equally enjoyed, they should be important in the life of people, and they should be as extensive as possible (Pettit 2008). Despite this and other differences from Rawls, I take Pettit’s recent work on the basic liberties as an acknowledgement of the merit of Rawls’ proposal that we should select some particular liberties as the most significant.

Let us now turn to the sort of freedom at issue in (2) above: freedom as non-domination from private parties. At first sight, it might not be entirely clear whether the theory of justice as fairness manages to protect this type of freedom. The criminal law certainly protects individuals and groups from certain kinds of private interference, such as physical violence or threats of physical harm. But the theory of justice as fairness does not seem to protect individuals or groups from the full range of forms of private domination. This is because, outside of the political realm in which all citizens are considered equal, a Rawlsian just society might include highly unequal relationships of power in civil society and in the domestic sphere. For example, there could be relations of domination between husbands and wives (depending on what the law says as regards marriage and divorce), or between employers and employees (depending on the content of labor laws and the nature of unemployment benefits). But we should be cautious in claiming that Rawls’s view allows such domination, because whether or not this is true depends crucially on how the prescriptions of the second principle of justice—both the fair equality of opportunity principle and the difference principle—are interpreted. I will argue that there is some textual support for an interpretation of the second principle of justice on which it functions to eliminate domination by private parties. But first let us examine the principle of fair equality of opportunity and the difference principle and some plausible readings that would leave the problem of private domination unresolved.

The principle of fair equality of opportunity requires that everyone should have ‘the same legal rights of access to all advantaged social positions’ (Rawls 1999a, p. 62). This means that there should be no legislation that enjoins different treatment to different groups, such as legal discrimination against women or an official system of apartheid. Moreover, the principle requires that individuals’ life chances should not be determined by the social position in which they are born. Rawls claims that ‘those who are at the same level of talent and ability, and have the same willingness to use them, should have the same prospects of success regardless of their initial place in the social system’ (Rawls 1999a, p. 63). If this principle is interpreted in a way that allows employers to practice informal forms of discrimination, excluding, say, women or racial minorities from jobs due to a prejudice that they are not committed workers, then the principle would not prevent the domination of members of these groups. Of course, one might be able to argue that the principle requires non-discrimination laws and perhaps certain quotas to overcome patterns of historical prejudice.16 Andrew Mason has suggested another problem with standard interpretations of the principle of fair equality of opportunity. Suppose that we assume that qualifications and motivation for jobs include the ability to work extra hours at short notice, or to work sixty-hour weeks. If so, then those with childcare commitments, who were unable (or unwilling?) to work under those conditions, would not count as ‘equally talented and motivated’. As a result, there would be no transgression of equality of opportunity if their life-chances were worse than others’ (Mason 2006, p. 72–73). This interpretation would deny that women’s limited access to many desirable social positions is a sign of their being dominated in society. Rather, it would construe their situation as a result of their life choices. Later I will argue that this is the wrong interpretation of Rawls.

As regards the difference principle, it is quite indeterminate in its policy implications, both in ideal and in non-ideal circumstances. It is, after all, extremely difficult to establish what set of rules for the functioning of economic institutions would make the life expectations of the worst off representative individuals as good as possible.17 It would be controversial even among social scientists and economists what set of taxes, market regulations and other programs would work to increase the lifetime expectations of the least advantaged so that they could be fully participating members of society on an equal standing with the more advantaged. But the difference principle requires precisely this: that the life prospects—in terms of access to income and wealth and other social primary goods—of the least advantaged working members of society be maximized, understanding ‘least advantaged’ to refer to those with the fewest marketable skills. But indeterminacy is just one problem. On some plausible interpretations of the difference principle, that principle is not violated if women do not do paid work, or do not have an independent source of income, and are consequently in a position to be dominated by their husbands. It may also be possible for employers to have such a high degree of power over employees that it amounts to a form of domination, on the plausible assumption that it is not always easy for employees to leave their jobs and find another source of income when they are not satisfied with their present working conditions. If one puts emphasis only on maximizing access to economic resources, the fact that employees suffer from domination may be overlooked when one assesses alternative methods of structuring the workplace.

All the forgoing might suggest that Rawls has little to offer by way of reducing private domination, at least by reference to the second principle. However, Rawls’ remarks on a property-owning democracy belie this suggestion (Rawls 2001, pp. 135–148). The idea of a property-owning democracy involves the widespread distribution of land and capital, laws limiting intergenerational transmission of property and wealth, and governmental policies that promote equality of opportunity in education. Richard Krouse and Michael McPherson argue that fair equality of opportunity in education would equalize the acquisition of human capital and would increase the number of people with marketable skills, leading to a reduction in the rent generated by possession of such skills. The result would be the equalization of the difference in earnings between the more favored and the least favored income groups (Krouse and McPherson 1988, pp. 91–92). Because the system of a property-owning democracy prevents excessive concentration of economic power, partly by supporting intergenerational redistribution of property, it provides an obstacle to the formation of interpersonal relationships of domination that are based on socio-economic inequality. Moreover, widespread access to education and training will mean that more citizens are capable of supporting themselves, and this will impede the formation of interpersonal relationships that lack viable exit options.

At this point I would like to make some brief remarks regarding the problem of the private domination of women, since it matters for our purposes that the second principle of justice not allow the systematic domination of women as a group. This is, unfortunately, a topic that cannot be adequately discussed in a short space, but Rawls’ responses to feminist criticisms point in the right direction. In the following passage, Rawls clearly indicates that the principles of justice should operate to prevent the systematic domination of women.

Since wives are equally citizens with their husbands, they have all the same basic rights and liberties and fair opportunities as their husbands; and this, together with the correct application of the other principles of justice, should suffice to secure their equality and independence” (Rawls 2001, p. 164, my emphasis).

Rawls acknowledges that women have traditionally borne a disproportionately large share of childcare responsibilities, and that they have been seriously disadvantaged by divorce laws. And he also sees that the childcare work done by women should be financially rewarded, and that divorce law should entitle women who have left the labor market in order to care for children to an equal share of the increased value of a family’s assets during the period the marriage lasted (Rawls 2001, p. 167). I would add that fair equality of opportunity for women also requires that there be policies that redistribute childcare responsibilities, as well as laws requiring the workplace to accommodate workers with young children.

The possibility of private domination must be taken into account when one tries to apply Rawls’ theory to non-ideal contexts. In the real world, people sometimes resort to different forms of coercion, including threats and manipulation, in order to control the behavior of others. A thoughtful application of the second principle of justice should recognize the importance of dispersing power and protecting those who are vulnerable to serious forms of interference by others. It might seem that in order for a policy successfully to eliminate or minimize domination it must transform the structural relationships between individuals, so that it is not possible for any person to have unchecked power over another. Examples of such policies include, but are not limited to, high quality publicly funded education, minimum wage laws, mandatory paid holidays and retirement savings, spousal support in case of divorce, and workplace accommodations and childcare support for workers with children.

Concluding Remarks

I hope to have shown that there is in Rawls’ account of the basic liberties more than a germ of a notion of freedom as non-domination: a notion of freedom as security against interference by others. I am aware that this is not exactly the same as Pettit’s account of freedom as non-domination. Some differences are to be expected, since Rawls does not tend to think of the liberties in relational terms, and therefore does not pay sustained attention to the possibility of interpersonal relationships of domination and servitude. Despite some differences in the underlying conception of freedom, the theory of justice as fairness gives, I have argued, adequate protection to citizens from state domination, and contains a number of valuable tools for dealing with the problem of private domination. How extensive the protection from private domination is crucially depends on how we interpret and apply the second principle of justice. The fact that Rawls’ main concern is to provide a model of an ideally just society with basic institutions that work according to the principles of justice and with citizens who are motivated to comply with the requirements of justice obscures the need to think about how to deal with interpersonal domination. This task is left to the application of the theory to non-ideal contexts. One way to guide such application is by appeal to Rawls’ fundamental intuition, that citizens are free and equal. This allows one to argue that the second principle requires policies geared to the prevention of domination by private parties, by showing that domination is a serious harm and that citizens who suffer it are not treated as free and equal.


H. L. A. Hart has pointed out that there was a tension in A Theory of Justice between some passages that refer to liberty in general terms and some that refer to the basic liberties of citizens. Rawls acknowledged this criticism and revised his theory accordingly, focusing only on the distribution of basic liberties. See Hart (1989); Rawls (1996, pp. 289–371).


This explanation of the activities covered by each basic liberty borrows heavily from Freeman (2007, pp. 44–59). For an assessment of the list see Martin (1985, pp. 45–61); Nickel (1993–1994).


For discussions on Rawls’ account of freedom and its compatibility with republican theories see Spitz (1993); Swift (1993); Pettit (1999, p. 50); Laden (2001); Larmore (2008, pp. 187–189).


The only place where Rawls talks about the freedom of collective agents is in his theory of international justice, where there are principles that require respect for the freedom of “peoples”. In this case, freedom amounts to self-government. See Rawls (1999b).


Spitz (1993); Pettit (1999); Skinner (1983). However, Pettit (2008) is in many respects closer to Rawls’ approach to the problem of liberty.


In his earlier work, Rawls makes use of a broadly Kantian notion of autonomy to justify the selection of the principles of justice. See Rawls (1999c). However, after the “political turn” he avoids appealing to this notion of moral autonomy because it is a comprehensive ideal that not all reasonable citizens share. But he continues to endorse the value of political autonomy understood as “the legal independence and assured integrity of citizens and their sharing equally with others in the exercise of political power.” See Rawls (1999b, p. 146).


For further considerations against the feasibility of equalizing the worth of the basic liberties see Krouse and McPherson (1988, p. 85).


Because the principle of equal liberty has priority over the second principle, Daniels thinks that if institutional design did not suffice to protect equal opportunities for political influence, the new version of the theory would support limiting economic inequalities to prevent them from generating a highly unequal distribution of political power. See Daniels (Daniels 1989), p. xxiv. Rawls’ endorsement of the economic system of a property-owning democracy seems to acknowledge that the equalization of political power among citizens requires a dispersion of property: that is, that it requires a reduction of economic inequality.


One problem with appealing to autonomy in a political theory of justice is that there are too many accounts of what counts as being or acting autonomously. Some of these accounts are quite demanding and would label many people as lacking autonomy. Depending on the account of autonomy, domination may either impede its development, or seriously restrict its exercise, or block its exercise completely.


This is not a direct quotation, but seems to be the clearest statement of his present view. See Pettit (1999, pp. 52–58); Pettit (2005, pp. 92–94).


In my view, Pettit’s notion of arbitrary interference by private parties is clearer than his notion of arbitrary interference by the state. One reason for this is that we have a fairly clear pre-theoretical sense of what the central interests of individuals are. But the notion of arbitrary interference by the state is more problematic because of difficulties in giving a precise sense to what the common interests of citizens are, particularly when it comes to deciding policies about which there is reasonable disagreement. For discussions of the problem of determining what counts as state domination see McMahon (2005); Pettit (2006); Richardson (2006); Costa (2007).


The following two paragraphs are based largely on my Costa (2009).


But see Rawls (2001, pp. 130–132).


For a detailed examination of the democratic elements of Rawls’ theory see Cohen (2003).


The requirement that laws dealing with constitutional essentials and matters of basic justice be justified in terms of public reasons is another element in the theory that provides some assurance against state domination of citizens. But it is worth noting that this requirement only applies to constitutional essentials and matters of basic justice. Moreover, the requirement that citizens use public reasons in political debates is a moral duty of citizens, according to Rawls. As such, it cannot be enforced. And for this reason it cannot alter structural relationships of domination. See Rawls (1999b, pp. 136).


It is a controversial question, what kinds of policies of affirmative action the fair equality of opportunity principle would support in non-ideal contexts. See Taylor (2009).


For excellent discussions of the implications of the difference principle see Van Parijs (2003) and Schaller (1998).



I would like to thank Joshua Gert and two anonymous reviewers for helpful comments on an earlier version of this paper. I am also grateful for funding from a grant issued by the Argentine Agency for the Promotion of Scientific Research and supervised by Maria Julia Bertomeu.

Copyright information

© Springer Science+Business Media B.V. 2009