LIBERTY, EQUALITY, AND PROPERTY - POLITICAL THEORY
1 INTRODUCTION
Political philosophers working in the analytic tradition have now spent nearly three decades debating the idea that distributive institutions should ensure that we share fairly in each other’s fortunes and misfortunes. Like various of their other recent debates, this one was launched in 1971 with the publication of John Rawls’s masterpiece, A Theory of Justice (Rawls 1971). This chapter examines how under-acknowledged assumptions about property rights, akin to those more frequently associated with Rawls’s foremost libertarian critic, Robert Nozick, have had an important influence on the debate. My aim is to show that Nozick’s challenge to egalitarians has played an important role in Ronald Dworkin’s alternative statement of liberal egalitarianism, and thus in indirectly shaping later non-Rawlsian egalitarianisms. Before examining those later views, I shall begin with some very brief remarks about Rawls’s initial formulation of the luck-sharing project.
2 RAWLS
Very few, if any, relatively aZuent individuals can credibly claim that those less well-oV could have enjoyed their standard of living had they been willing to make the same choices. Instead, as Rawls often emphasizes, it is more plausible to claim that the material inequalities present in modern societies arise from factors beyond the control of those less well-oV, most obviously their lesser luck in the social and natural lotteries that determine family and class background and genetic endowment. Rawls’s response to this socio- logical commonplace is ‘‘to look for a conception of justice that prevents the use of accidents of natural endowment and the contingencies of social circumstances as counters in a quest for political and economic advantage’’ (Rawls 1999a, 14, and chs. 11, 13, and 63). Proposing principles to share the eVects of luck, he focuses initially on inequalities in occupational opportun- ity, and appeals to the unease many feel toward inherited variations in career prospects (Rawls 1999a, 63; Marshall, Roberts, and Swift 1997).
Rawls then argues that because it is incoherent to limit our concern only to these involuntary inequalities, we should adopt a similarly critical attitude to inequalities in income and wealth that arise from diVerences in natural as well as social luck (Rawls 1999a, 64). Unlike more radical egalitarians (Temkin 1999), Rawls sees no reason to waste benefits by ‘‘levelling down.’’ So, he does not conclude that justice condemns all involuntary occupational and finan- cial inequalities. Instead his famous ‘‘diVerence principle’’ and principle of equality of opportunity require distributive institutions to arrange inequal- ities in income and wealth work to everyone’s benefit, with priority given in distributive conflicts to those who are less advantaged, and to ensure posi- tions are allocated through a fair competition.
3 NOZICK
As explained by Richard Arneson in this volume, there have been various critical reactions to democratic equality. Some endorse Rawls’s concern to share the eVects of luck in a fair way, but argue that his pursuit of that objective is insuYciently thoroughgoing. According to Susan Okin, for ex- ample, Rawls fails to recognize the extent to which his principles condemn injustice arising from gender-based inequalities in the distribution of labor within the family (Okin 1989; and Rawls 2001, 162–8), whilst G. A. Cohen has argued that Rawls’s defense of incentive-generating inequalities depends on an arbitrary restriction in the scope of his diVerence principle (Cohen 1997). One of the most widely discussed early critiques of democratic equality was far more hostile. In Anarchy, State and Utopia , Robert Nozick denied that justice requires any attempt to mitigate the diVerential eVects of the social and natural lottery (Nozick 1974, 213–31). The main positive strategy Nozick employed to support this conclusion involved claiming that an adequate account of distributive justice will be an historical entitlement theory.
In making this claim, Nozick meant not just that any such account must treat the past as relevant when assessing distributions. He also, and more controversially, assumed that individuals possess entitlements, or extensive private property rights (Waldron 1988, ch. 2) over their bodies and labor, the products of their labor, and non-produced, or natural, resources. These entitlements encompass not only claim rights against certain forms of inter- ference by others in an owner’s property, but also extensive powers to dispose of property via waiver, donation, bequest, and market exchange. Emphasizing their stringency, Nozick suggested that entitlements have near absolute im- portance, and implied that they can be defeated, if at all, only in exceptional circumstances; for example, when ‘‘catastrophic moral horror’’ would other- wise be unavoidable (Nozick 1974, 32).
In addition to these general assumptions, Nozick advanced some more specific claims about how best to flesh out an entitlement theory. Thus, he claimed that persons possess rights of full self-ownership that rule out not only involuntary slavery, but also any involuntary redistributive taxes on income from labor. Explaining how individuals could unilaterally acquire ownership rights in previously non-owned natural resources, he relied on a modified Lockean proviso, stating that appropriators of a previously unowned resource need only ensure that others are made no worse oV than they otherwise would have been had that resource remained unowned. Since non-ownership is ineYcient, Nozick argued his proviso would be easily satisfied, and so only infrequently require political action to redistribute wealth or limit powers of transfer.
Given Nozick’s entitlement theory contained no measures requiring a system of property rights to protect individuals from diVerential luck, it was relatively straightforward for him to conclude that it was possible for a system of property rights to be just without mitigating the diVerential eVects of the social and natural lottery, or eliminating destitution. In addition, he argued that justice prohibits many redistributive public policies. Such pol- icies, he objected, implicitly assume that resources are available for distribu- tion (Nozick 1974, 149–50), but that assumption is unsound since individuals are self-owners, and will have acquired unequal claims to material resources due by exercising their powers to appropriate and transfer property. Any policies required by the diVerence principle, Nozick concluded, could be justifiable, at best, as remedial measures to rectify previous violations of individuals’ entitlements, given ignorance about what would have happened in the absence of injustice (Nozick 1974, 231).
4 ECONOMIC LIBERTY
Because of humanity’s record of genocide, slavery, and violent expropriation, Nozick’s failure to defend any particular principles of rectification meant that his view’s positive implications for public policy were severely indeterminate. Even if they had been less indeterminate, however, it is unlikely that Nozick’s own version of entitlement theory would have won more converts. For although Nozick’s work on distributive justice is often memorably ingenious, and did much to renew political theorists’ interest in property (Becker 1977; Christman 1991; Reeve 1986; Ryan 1984; Waldron 1988; Sreenivasan 1995), it also generated many persuasive critiques (Cohen 1995; Nagel 1975; Scanlon 1975; Ryan 1977; WolV 1991).
One reaction to Nozick’s counter-intuitive conclusions about even modest redistribution is to object to their dogmatic reliance on under-argued as- sumptions about the types of right a theory of distributive justice should allocate. The charge is well-illustrated by Nozick’s infamous attempt to show ‘‘How Liberty Upsets Patterns’’ (Nozick 1974, 160–4; Hume 1998, 91).
The argument begins by inviting proponents of alternatives to the entitle- ment approach to imagine that their favored distribution obtains; for ex- ample, with everyone enjoying an equal share, or with shares distributed in proportion to some personal attribute, such as deservingness. Now suppose that Wilt Chamberlain is willing to display his prowess at basketball only if he receives 25 cents from each of his spectators, and all 1 million spectators are willing to make such an exchange. Granted these assumptions, Nozick first suggests we should accept that if the relevant exchanges voluntarily occur then the subsequent distribution, D2, is no less just then the initial distribu- tion, D1. To do so, however, we have to abandon our commitment to egalitarian or patterned principles since Wilt’s privileged position in D2 violates those principles.
Nozick then goes on to suggest his example also shows that the distributions favored by egalitarian or patterned principles can be maintained only by restricting the ability of individuals to dispose of their holding as they see fit, thereby implying that this renders them objectionable. As many critics have noted, the force of Nozick’s example, and the plausi- bility of his positive assessment of D2 and resistance to restricting transfers, depends not merely on the assumption that D1 confers on individuals the power to relinquish an extra quarter to see Wilt play. The example also implicitly assumes that individuals have the power to bestow on Wilt a highly unequal or disproportionate reward for his services, which Wilt himself may then use in ways that have even further disruptive repercussions. Given their eVects, however, it is far from obvious why any of us, not to mention egalitarians and pattern theorists, should share Nozick’s assumption about the extent of individuals’ powers.
Moreover, Nozick himself makes claims elsewhere that suggest he should not take for granted the content of individuals’ ownership rights. Thus, he disingenuously concedes that his Lockean proviso requires ‘‘a more complex principle of justice in transfer’’ (Nozick 1974, 179) that limits owners’ powers to dispose of their holdings, and the same presumably holds for the terms of a bequest that limit future owners’ powers. More generally, Nozick’s remarks about individuals partitioning their self-ownership rights in order to sell some of the elements (Nozick 1974, 282) indicate he accepts that ownership involves a complex bundle of rights, capable of disaggregation (Ryan 1977). Despite such acceptance, however, Nozick does little more to justify his assumption about individuals’ powers in D1 than ask the question, ‘‘If . . . people were entitled to dispose of the resources to which they were entitled (under D1), didn’t this include their being entitled to give it to, orexchange itwith, Wilt Chamberlain?’’ (Nozick 1974, 161). The fact that Nozick does so little to pre-empt a negative answer tends to support Thomas Nagel’s charge that Anarchy, State and Utopia is an example of ‘‘Libertarianism Without Foundations’’ (Nagel 1975).
5 NOZICK’s CHALLENGE
Despite its counter-intuitive and incomplete character, it would be a mistake merely to dismiss Nozick’s work on distributive justice. Self-styled left-liber- tarians have followed Hillel Steiner’s lead in arguing that measures designed to redress inequalities in fortune can be justified from within an entitlement theory, and are consistent with some version of full self-ownership (Otsuka 2003; Steiner 1994; Vallentyne 1998; Vallentyne and Steiner 2000). Whether such arguments are persuasive in their own right is debatable (Fried 2004, 2005; Risse 2004; Vallentyne, Steiner, and Otsuka 2005), but they merit attention and serve, at the very least, as plausible ad hominem arguments against libertarian critiques of the luck-sharing project.
There is an additional reason why Nozick’s work should remain important even for those who reject full self-ownership, or any presumption of full ownership right over impersonal resources, namely that it poses an important challenge for any philosophical account of distributive justice. Having argued that ‘‘the particular framework of property and contract rights which Nozick proposes does not constitute an adequate account of the claims of economic liberty’’ (Scanlon 1975, 25), T. M. Scanlon states the challenge well when making the following remarks about Anarchy, State and Utopia:
It is a virtue of the book that it forces us to consider economic institutions not merely as mechanisms for the distribution of goods but also, like political institutions, as placing restrictions and demands on us that raise questions of obligation. When things are seen in this way it becomes apparent that questions of economic liberty must be considered, along with political and civil liberty and fair distribution, as conditions for the legitimacy of social institutions. I hope that this will have an impact on contemporary moral and political philosophy, where economic rights and liberties have generally been neglected in favor of political and civil liberties and rights of other sorts.
Even if we reject Nozick’s implausibly permissive conclusions about inequal- ity and destitution, and his assumptions about income tax and extensive powers of transfer, Scanlon’s observation reminds us that Nozick shows we still have to decide what forms of control over their holdings individuals can reasonably demand of social institutions.
One simple way to make this challenge vivid is to imagine a scenario in which a group of equally capable individuals has to distribute fertile land and other natural resources amongst its members. Any adequate solution to the distributive problem facing the group will need to explain not only how to distribute those assets but also what to distribute, or the content of the ownership rights that a just distribution should confer on individuals. Sup- pose that we favor some egalitarian explanation in the first case and for the latter favor resourcist over welfarist explanations.2 We still need to answer various additional questions about what individuals owe one another when allocating rights to control and benefit from their resources. Individuals could enjoy more or less extensive rights to make various unilateral decisions about the transformation, consumption, or productive use of resources, and the generation of externalities. They might also possess quite diVerent rights to exclude others from the benefits generated by their decisions, and, as we have seen, to produce transfers in the distribution of rights over resources.
Suppose, for example, we need to decide between ownership rights that enable individuals to use their holdings only as a means of consumption from rights that enable individuals also to use their holdings as a means of production and exchange. Moreover, if we allow productive use, and indi- viduals have equal productive talents, consider the choice between rights that distribute the proceeds of individuals’ decisions in diVerent ways; for example, rights that entitle the producer to retain their entire product versus rights that permit others to take an equal share of the product. It will not suffice to resolve these additional disputes merely by appealing to comparative convictions that it is unfair for individuals to possess more ownership rights than any other due to factors beyond their control. In addition, we need to decide the contours of the rights to be allocated to individuals.
One way to address these disputes appeals to strategic arguments about which regime of ownership rights provide incentives that optimally shape individual decisions. Rawls appears to adopt this approach when claiming that ‘‘the principles of justice are compatible with quite diVerent types of regime,’’ and suggesting that the choice between liberal market socialism and property-owning democracy is contingent on ‘‘the traditions, institutions, and social forces of each country, and its particular historical circumstances’’ (Rawls 1999a, 242, 249). Relying only on strategic arguments, however, is not entirely satisfactory. To require equally capable individuals to share every- thing they produce, for example, seems objectionable because of the limited control it gives individuals over their holdings, and that objection persists even if the requirement does not lead to ineYciency. Nozick’s impact, I suggest, was to challenge egalitarians to explore these issues in ways that assumed there were non-instrumental reasons why individuals could demand substantial decision-making powers over material resources.
6 DWORKIN
Ronald Dworkin was the most influential philosopher to take up Nozick’s challenge. In papers published over more than two decades (Dworkin 1981a, 1981b, 1987, 2002, 2004) and his book Sovereign Virtue (2000), Dworkin defended a theory of economic justice designed to ensure individuals share in each other’s fate whilst also enjoying a range of economic liberties. Dworkin’s description of his theory—equality of resources—is complex, but begins with a simplified illustration.
Suppose a group of shipwrecked survivors have to divide a desert island’s resources equally amongst themselves. When the survivors do so, Dworkin suggests that they should attempt to satisfy an appropriate version of what economists term the ‘‘envy test’’ (Dworkin 2000, 67).3 Thus, they should ensure that nobody prefers anyone else’s resources, and that each individual plays an equal role in determining the character of the resource bundles available for distribution. Dworkin then argues that a market is the best device to eliminate envy in this way. More specifically, he describes an auction in which everyone has the same bidding power, and an auctioneer continu- ously divides lots until the market clears and no bidder wishes to repeat the process. Dworkin then asks whether a market procedure remains appropriate once production, investment, and trade complicate the island’s economy, and diVerences in luck as well as ambition shape the islanders’ prospects.
Here Dworkin draws an important distinction between an individual’s luck in the decisions she makes, and in the conditions she finds herself in regardless of her decisions. As he explains these two types of luck, ‘‘Option luck is a matter of how deliberate and calculated gambles turn out— whether someone gains or loses through accepting an isolated risk he or she should have anticipated and might have declined. Brute luck is a matter of how risks fall out that are not in that sense deliberate gambles’’ (Dworkin 2000, 73). Where everyone has the same brute luck, and there is no variation in productive talent and other natural abilities, Dworkin argues that equality of resources entitles individuals to make productive use of their resources, and keep the proceeds. Dworkin also argues that individuals are entitled to use their resources in ways that expose them to diVering degrees of option luck. Thus, he concludes that if some islanders choose to gamble with their endowment, and have good option luck, then there is no reason to object to their having more resources than similarly situated islanders who declined to gamble, or who chose to gamble and had worse option luck.
Once Dworkin’s simplifying assumption about the absence of diVerential brute luck is relaxed, his conclusions about the fairness of option luck inequalities play a crucial role in equality of resources. To understand why, suppose that some individuals are sighted whilst others become blind (Dwor- kin 2000, 76), or that Adrian has a higher income than his similarly motivated counterpart Claude simply because he is more naturally gifted (Dworkin 2000, 83). To deal with such inequalities in fortune, Dworkin appeals to the idea of a hypothetical insurance market , where purchasers make decisions based on their own attitude to risk, but—unlike in actual insurance
2 It is notable that the economist Hal Varian (1975) highlighted the possibility of replying to Nozick by appealing to the idea of an envy-free competitive equilibrium some time before Dworkin presented equality of resources (Dworkin 1981b ) markets—are fairly situated because equally endowed and aware only of the overall distribution of brute luck rather than their personal fortunes. He then argues that the victims of brute luck are entitled to the level of compensation delivered by the average package of cover purchased in such a market. Such compensation is, furthermore, to be financed by a system of general taxation, which includes a progressive income tax.
To summarize, then, Dworkin’s core claim is that a distribution of re- sources is fair only if the individuals involved, given their convictions and ambitions, might have produced such a distribution through a specific market process. The process involves those individuals exercising certain rights to produce and trade, using resources they have acquired in an equal auction, and to pool risks in a manner mimicking a fair insurance market.
Thus summarized, the extent to which equality of resources continues the luck-sharing project initiated by Rawls, and opposed by Nozick, should be apparent. Various features in Dworkin’s argument serve to redress inequal- ities in fortune, most obviously individuals’ equal endowment in the initial auction and subjection to the same veil of ignorance in the later insurance market. At the same time, however, the argument accommodates convictions about economic liberty far closer to Nozick’s than Rawls’s view. For Dworkin it is a matter of principle that individuals are entitled to become private owners in the means of production rather than merely in what Rawls terms ‘‘personal property’’ (Rawls 2001, 114). Dworkin’s view, therefore, accepts far more readily Nozick’s suggestion that socialism objectionably restricts the liberty of potential entrepreneurs who prefer to transform their personal possessions into means of production, and employ others to labor with them (Nozick 1974, 162). Indeed, equality of resources might even provide more robust support for private ownership than the views of either Rawls or Nozick, given that both display only a contingent commitment to capitalism, which depends respectively on instrumental arguments or historical assump- tions about how earlier owners exercised their powers of bequest.
It is also noteworthy that assumptions about economic liberty play a fundamental role in Dworkin’s explanation of the way in which justice requires individuals to share in each other’s fortunes. As mentioned, equality of resources compensates individuals for misfortune in a way that depends on how actual individuals, given their values and attitude to risk, would have chosen to exercise certain rights to purchase protection against misfortune. Because it assumes the existence of such rights, and so seeks to secure ‘‘endowment-insensitivity’’ in an ‘‘ambition-sensitive’’ manner, Dworkin’s approach diVers significantly from that favored by Rawls (Williams 2004, 131–3). On the former view, what qualifies as misfortune, as well as the appropriate form and level of redress, is dependent on individuals’ diverse preferences, amended only to correct for standard cognitive and informa- tional errors. Rawls’s view, in contrast, does not involve any attempt to mimic hypothetical market behavior. Indeed, the thought that the type of protection against misfortune secured by democratic equality is optional seems alien to his enterprise.
So far, I have constructed a narrative in which Dworkin’s account of equality of resources figures as an attempt to combine elements from the apparently deeply opposed views of Rawls and Nozick. My aim has been largely diagnostic, although admittedly I have taken for granted the implaus- ibility of Nozick’s own version of historical entitlement theory. I did not suggest, however, that the brief history so far described is one of progress, in which Dworkin ingeniously combines the best elements from two opposed views. Nor did I suggest the history is one of decline, whereby Dworkin corrupts the luck-sharing project with an excessive zeal for rights in private property. My remaining remarks explore the latter possibility, whilst also introducing some of the other egalitarian proposals Dworkin’s view has generated.
7 THE AGENCY OBJECTION
Anti-egalitarians sometimes argue that certain plausible assumptions about responsible agency, individual liberty, and personal liability provide grounds to reject egalitarian distributive principles. As an example, consider the following argument, which I shall term the agency objection:
We are responsible agents, capable of acting freely in a sense that renders our conduct subject to moral appraisal as blameworthy or commendable. As such, provided we respect the entitlements of others, we should be at liberty to make our own decisions about how best to advance our aims. Moreover, under appropriate conditions, it is unfair to make us liable to bear certain costs arising from others’ decisions, or to relinquish certain advantages gained through our own eVorts. Egalitarian principles, however, demand that outcomes remain within a certain range. Consequently, implementing those principles involves unjustifiably denying our decision-making powers, or forcing us to share the costs of others’ decisions. Egalitarian principles should, therefore, be rejected on the grounds that they unjustly limit liberty or unfairly spread liability.
My previous remarks should indicate why equality of resources suggests a relatively conciliatory response to this objection, which dispenses with the outcome-based conception of egalitarianism the objection targets, allocates various choices to individuals, and then holds them liable for their diVerent responses.
Since the original presentation of equality of resources in 1981, political philosophers have followed Dworkin’s lead in defending other examples of what I shall term post-libertarian egalitarianism. They include rival resourcist proposals, such as Eric Rakowski’s equality of fortune (Rakowski 1991) and Philippe Van Parijs’s real libertarianism (Van Parijs 1995), as well the left- libertarian views already noted. In addition, there are welfarist alternatives, such as Richard Arneson’s equality of opportunity for welfare (Arneson 1989), along with hybrid proposals like G. A. Cohen’s equality of access to advantage (Cohen 1989).
Several post-libertarians appear to have accepted that well-informed, vol- untary choice against a background of equal opportunity can render any unequal outcome just, even if some individuals fare extremely badly as a result. For instance, in the example involving the risk of blindness mentioned earlier, Dworkin writes that ‘‘if everyone had an equal risk of suVering some catastrophe that would leave him or her handicapped, and everyone knew roughly what the odds were and had ample opportunity to insure . . . then handicaps would pose no special problem for equality of resources’’ (Dwor- kin 2000, 77).
In the case of two individuals who face the same risk of blindness, and the same insurance options, but who make diVerent purchas- ing decisions, Dworkin then adds: ‘‘the bare idea of equality of resources, apart from any paternalistic additions . . . would not argue for redistribution from the person who had insured to the person who had not if, horribly, they were both blinded in the same accident’’. Finally, he draws the stark conclu- sion that ‘‘the situation cannot be diVerent if the person who decided not to insure is the only one to be blinded’’. Dworkin has more recently appealed not only to ‘‘the need to protect people from mistakes they are very likely to regret’’ but also to the need to correct for imperfections in insurance markets in order to show that, in practice, equality of resources would not condone extreme forms of voluntary inequality (Dworkin 2002, 114). His argument will not satisfy those who are convinced such inequal- ities are objectionable in principle, and even in the absence of likely regret or ineYciency.
Similarly, both Arneson and G. A. Cohen have suggested that the appro- priate genesis can render even an extremely unequal outcome just. Thus, Arneson writes that, ‘‘When persons enjoy equal opportunity for welfare . . . any actual inequality of welfare in the positions they reach is due to factors that lie within each individual’s control. Thus, any such inequality will be non-problematic from the standpoint of distributive equality’’ (Arneson 1989, 88). Similarly, on G. A. Cohen’s conception of egalitarian justice, ‘‘When deciding whether or not justice (as opposed to charity) requires redistribution, the egalitarian asks if someone with a disadvantage could have avoided it or could now overcome it. If he could have avoided it, he has no claim to compensation, from an egalitarian point of view’’ (Cohen 1989, 920).
The willingness of post-libertarians to accept that extreme inequality can be just when cleanly generated has recently provoked a backlash against their view, which Elizabeth Anderson has dubbed ‘‘luck egalitarianism.’’. Although I persist in thinking the aim of sharing in each other’s fortunes should be central to egalitarianism, I agree with critics of luck egalitarianism in regarding the post-libertarian attitude to voluntary inequality as implausibly permissive. Adopting that attitude is too high a price to pay to show that egalitarianism can withstand the agency objection. Nevertheless, as we shall now see, problems remain in deciding which element of the post-libertarian view to reject.
8 THE TRILEMMA
One natural way to avoid the excesses of the post-libertarian view is to adopt a mixed conception of justice concerned with absolute as well as relative deprivation, which I shall term sufficientarian egalitarianism.6 On this view, individuals have weighty claims against suVering certain forms of absolute deprivation that cannot be relinquished through voluntary decisions, no matter how favorable the background conditions. Thus, only some inequal- ities can be justified by appealing to personal responsibility. Egalitarians who adopt the mixed conception can endorse Marc Fleurbaey’s view of the reckless motorcyclist who brings disaster upon himself, namely that ‘‘how- ever criminal and stupid his behavior may have been, there is a limit to the kind and amount of suVering he should endure’’ (Fleurbaey 1995, 41). Simi- larly, they can share Anderson’s conviction that ‘‘justice does not permit the abandonment of anyone, even the imprudent,’’ and that ‘‘starting gate theories, or any other principles that allow law-abiding citizens to lose access to adequate levels of . . . [certain] goods, are unacceptable’’.
The suYcientarian view provides a less conciliatory response to the agency objection than post-libertarianism, and challenges the objection’s fundamen- tal moral assumptions rather than its portrayal of egalitarianism. More specifically, the view rejects the objection’s refusal to limit liberty or to extend liability in order to save some from absolute deprivation. It is important to note, however, that there are at least two possible variants of this response. Those variants diVer depending on whether they recommend limiting liberty rather than extending liability.7 Consequently, they also diVer in how they treat externalities , or the unintended effect on others of individuals’ decisions. One variant argues that limiting our liberty to relinquish certain claims against others is justified when necessary to avoid absolute deprivation without requiring some to bear the costs of others’ decisions. I shall refer to this view as internalizing sufficientarianism since it aims to protect individuals from bearing costs arising from others’ decisions. The other variant claims that it is defensible to force some individuals to bear the costs of others’ decisions if doing so is necessary to avoid absolute deprivation without restricting individual liberty. Since this response countenances cost displace- ment, I shall refer to it as externalizing sufficientarianism.
Suppose that compromising suYciency, limiting liberty, and extending liability are all unpalatable options to some degree, but that the agency objection shows that egalitarians sometimes must select at least one of these options. One way to sum up my remarks is to conclude that egalitarians appear to face a trilemma that invites at least three diVerent responses. The first post-libertarian response exhibits a (more or less explicit) willingness to sacrifice suYciency. In sharp contrast, both suYcientarian variants refuse such a sacrifice. They diVer, however, in the compromises they recommend in order to safeguard suYciency. Thus, internalizers limit liberty and curtail liability, whilst externalizers uphold liberty and extend liability.
To illustrate the trilemma, and the diVerences between the post-libertarian, internalizing, and externalizing responses, consider some individuals who voluntarily decide to engage in some potentially harmful activity against a background of equal risk and opportunity.8 If some urgently need medical care as a result, post-libertarians are most likely to favor funding it only from private health insurance, and countenance denying it to those who exercised an entitlement not to insure. Internalizers and externalizers both reject such a denial, and supply care even to those who would decide not to insure. Internalizers, however, will prefer special taxes on the activity, compulsory insurance, or even outright prohibition, whilst externalizers might fund medical care through general taxation. Note also that these are pure views. It is clearly possible to devise impure views, which compromise more than just one of the three objectives.
Suppose, like the critics of luck egalitarianism, we reject the post-libertarian willingness to sacrifice suYciency. Even so, we might be unsure about the relative merits of the internalizing and externalizing suYcientarian responses. Reflection on actual critiques of post-libertarian egalitarianism is unlikely to eliminate our doubts. For those critiques normally focus on the post-libertarian view’s most counter-intuitive implications rather than its animating assumptions about liberty and liability. Dwelling on that view’s most apparent defect rather than its underlying appeal, they do not clearly diVerentiate the two ways of securing suYciency. Nor do they evaluate the relative importance of individuals possessing powers to make their own choices rather than enjoying immunities from bearing the costs of others’ choices.
Anderson, for example, does remark that luck egalitarians ‘‘have been most responsive to criticisms of equality based on ideals of desert, responsi- bility and markets’’ (Anderson 1999, 291), but provides little diagnosis of the appeal of their view. Her claim about desert is unsupported, and there seems no reason to believe desert plays a larger role in defending luck egalitarianism than its egalitarian predecessors, including justice as fairness. Admittedly, her claims about responsibility and markets are more relevant, particularly in relation to resourcist forms of luck egalitarianism, but Anderson omits to note that those considerations are salient to luck egalitar- ians because they endorse a particularly expansive conception of economic liberty that empowers individuals to jeopardize their own access to minimum levels of certain essential goods. Moreover, she fails to recognize that there is a price to be paid for maintaining individuals’ access to those goods since doing so requires a more restrictive conception of economic liberty or a less restrictive conception of our liability to bear the costs of others’ exercising their liberty.
4 CONCLUSION
According to Will Kymlicka’s influential history of the recent debate over egalitarian justice, later egalitarians have extended Rawls’s argument that welfare inequalities are a matter of personal responsibility because individuals can avoid relative frustration by exercising an ability to revise their ends (Kymlicka 2002, ch. 2; Rawls 1999b, 369–70). There are good reasons to doubt Kymlicka’s interpretive suggestion, which have been pressed by Samuel Sche- Zer (ScheZer 2003a ). There are also good reasons to doubt that much would be lost by abandoning Rawls’s argument, given the availability of persuasive objections to welfare egalitarian principles that eschew appeal to contested assumptions about what is avoidable. For example, anti-welfarists might fall back on Rawls’s own worries about the informational demands of welfarist principles, or argue those principles are objectionable because they either pander to individuals with voluntarily-acquired expensive tastes or penalize those with involuntarily-acquired inexpensive tastes (Dworkin 2000, 48–59; Williams 2002, 379–80).
Pursuing G. A. Cohen’s suggestion that ‘‘Dworkin has, in eVect, performed for egalitarianism the considerable service of incorporating within it the most powerful idea in the arsenal of the anti-egalitarian right: the idea of choice and responsibility’’ (Cohen 1989, 933), I have sketched an alternative history that takes seriously the presence of non-Rawlsian elements in more recent statements of egalitarianism. Although I am less sanguine than Cohen about the extent to which Dworkin’s achievement is a service to egalitarians, I have suggested that Cohen is correct to recognize the role played in equality of resources by conceptions of economic liberty and liability more often asso- ciated with critics of equality, such as Nozick. I have also argued that those conceptions render post-libertarian egalitarianism too hospitable to cleanly generated inequality. To avoid that problem, egalitarians now need to exam- ine more carefully the relative merits of liberty-restricting and liability- spreading means to ensure that inequality remains within acceptable limits. It is only by scrutinizing the assumptions about liberty and property that Dworkin injected into contemporary egalitarianism that they can hope to do so.