Equality of What? and Intergenerational Justice

Equality of What? and Intergenerational Justice Kasper Lippert-Rasmussen Aarhus University, Denmark ABSTRACT. Luck egalitarian accounts of distributive justice presuppose a metric in terms of which people s relative positions are determined. This article addresses the issue of the metric in an intergenerational setting focusing on two problems: expensive tastes and the slavery of the talented. First, on the assumption that later generations will have more expensive preferences than earlier generations with goods that were luxuries for past generations becoming necessities, the availability of which generates little satisfaction for later generations welfare egalitarianism recommends redistribution of resources in favour of later generations. Indeed, welfare egalitarianism allows that one generation cultivates cheap preferences among future generations thereby allowing itself to deplete natural resources. I argue that neither of these implications defeats welfare egalitarianism as an account of intergenerational justice. Second, suppose, plausibly, that work by one generation positively affects the resources at the disposal of all later generations and assume that we apply Dworkin s idea of an auction of resources to an intergenerational setting. On these assumptions, there is a sense in which, ceteris paribus, earlier generations command more resources to wit, more in terms of the internal resources that valuable labour power constitutes than later generations. Accordingly, resource-egalitarianism implies that the first generation is required to be maximally productive. This implication represents a particularly damaging version of the so-called problem of the slavery of the talented, which has exercised resource egalitarians in an intragenerational setting. Overall, this paper indirectly supports welfare egalitarianism by deflecting possible criticisms of it, and by resisting resourcist proposals as to how one should accommodate the problem of the slavery of the first generation. KEYWORDS. Dworkin, equality, expensive tastes, intergenerational justice, luck egalitarianism, welfarism ETHICAL PERSPECTIVES 19, no. 3(2012): 501-526. 2012 by Centre for Ethics, KU Leuven. All rights reserved. doi: 10.2143/EP.19.3.2172301

ETHICAL PERSPECTIVES SEPTEMBER 2012 I. INTRODUCTION Patterned theories of distributive justice need, at some point, to incorporate an account of the currency, or metric, in terms of which people s positions are determined for the purpose of assessing the justice of the relevant distribution. Luck egalitarians have responded to this basic requirement with considerable care: Arneson s equality of opportunity for welfare (1989), Dworkin s equality of resources (1981a; 1981b; 2000), and Cohen s equality of access to advantage (1989) testify to this. However, their focus has been on intragenerational justice. On intergenerational justice the just distribution of goods across generations they have been less vocal. 1 Presumably this reflects the view that the shift from intragenerational to intergenerational raises no new issues. This view appears to be correct in the sense that both types of justice do not oblige us to introduce contrasting metrics: their metrics need not diverge. 2 But this does not mean that specific consideration of the intergenerational case is unnecessary. For one thing, such consideration might highlight consequences of the metric obscured or absent in the more familiar intragenerational context, and this may lead to modification of the metric itself. Second, an account of the metric of justice that is appropriately defended in the intragenerational setting may turn out to have surprising implications in the intergenerational setting. We might, on reflection, judge that we should accept these implications, even if they are at odds with our initial convictions about intergenerational justice. My goal here is explore two main problems the currency of justice is known to raise. I want to see how these problems play out in an intergenerational setting, since I believe that this focus offers a new angle on the metric of equality issue. The first main problem relates to expensive tastes and is dealt with in Sections II and III. Several theorists addressing the intragenerational issue have argued that an individual who is worse off through having expensive preferences is not worse off in terms of the relevant currency of egalitarian justice. Assuming later generations 502

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? tend to have more expensive preferences than earlier generations with goods that were luxuries for past generations becoming necessities, the availability of which generates little satisfaction for later generations welfarist egalitarianism would seem to recommend redistribution of resources in favour of later generations. Some might take this as an objection to the view that welfare is the currency of justice (II). Moreover, welfarist egalitarians seem to be comfortable with the notion that we might cultivate cheap preferences among future generations, thereby allowing the present generation to deplete natural resources, and some regard this as a problem with welfarist egalitarianism (III). Like the first objection, this also concerns expensive preferences, but here the focus is on one party s actively and deliberately cultivating cheap preferences in another party. The second main problem concerns the issue of the slavery of the talented (Dworkin 1981b: 90), to which Dworkinians have paid much attention in the more familiar intragenerational context. The idea is that insofar as internal resources, e.g. productive talents, are included in the measure of resources in terms of which people should be equal, one hour of leisure for a talented person will count as a much greater consumption of resources than one hour of leisure for an untalented person. Now suppose, plausibly, that work by one generation positively affects the resources available for all later generations; and suppose that all members of the first generation are untalented relative to the members of later generations in the sense that had one of the later generations been the first, the volume of resources conferred on later generations through the efforts of the first generation would have been much greater than they in fact are. On these assumptions, it seems that an intergenerational problem of the slavery of the first generation arises. I explain how this happens and discuss various resourcist responses to the problem in sections IV to VI. I argue that we make best sense of the problem from a welfarist perspective. Hence, while the article offers no direct defence of the view that welfare is the currency of intergenerational justice, it offers indirect 503

ETHICAL PERSPECTIVES SEPTEMBER 2012 support for this view by deflecting possible criticisms thereof, and by resisting resourcist proposals as to how one should accommodate the problem of the slavery of the first generation. 3 One preliminary point: I write about redistribution across generations as if there were some way of taking resources from one generation, wherever it is located in time, and giving these to another generation, wherever it happens to be located in time. Obviously, direct backward redistribution across generations is impossible, and this, among other factors, means that the principles of regulation of intergenerational equality must differ from the fundamental principles governing that equality. My concern in the present contribution is with the latter only. Accordingly, to insist that redistribution in favour of past generations is impossible is simply no objection to the views I advance. The said impossibility does not prevent us from asking whether, if past generations had been as well off as the present generation, the world would have been more just the sort of question we ask if we are concerned with fundamental principles of justice, not principles of regulation (Cohen 2008, 5; 19-21). 4 Also, even though redistribution in favour of past generations is impossible, redistribution from present to future generations remains possible. II. THE EXPENSIVE TASTES OF FUTURE GENERATIONS One important challenge to a welfarist metric appeals to the idea of expensive tastes (Page 2007, 455). In his defence of primary goods as the currency of justice, John Rawls asks us to Imagine two persons, one satisfied with a diet of milk, bread and beans, while the other is distraught without expensive wines and exotic dishes. In short one has expensive tastes, the other does not (1982, 168). If welfare is the relevant measure of egalitarian justice, the second person merits more resources than the first, since that (unlike equal shares) will ensure equality of welfare. However, Rawls denies that this is just: As moral persons citizens have some part in forming and cultivating their final ends and preferences. It is not 504

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? by itself an objection to the use of primary goods that it does not accommodate those with expensive tastes (1982, 168). 5 The expensive tastes problem seems pressing in the intergenerational setting, particularly if we focus on what is now the developed world. Here, generations ago, most people would have been satisfied with milk, bread and beans. But many members of the present generation require at least occasional expensive wines and exotic dishes to get to the same level of satisfaction, and who knows what exorbitant tastes future generations will develop. 6 More generally, by saying that members of one generation, G 2, have more expensive preferences than members of another, G 1, I mean that the former would reach a lower level of satisfaction than that actually enjoyed by members of G 1 were they only to have the goods available to the latter, whereas members of G 1 would have a higher level of welfare than actually enjoyed by members of G 2 were they only to have the resources available to G 2. Before I proceed, I should say a little more about expensive tastes. Rawls s example may mislead, or even give expensive tastes a bad name, because it does not specify the grounds of the relevant satisfaction and dissatisfaction. Suppose my dissatisfaction derives from an anti-egalitarian preference to have greater satisfaction or consume more expensive goods than others. This might be an objectionable preference from an egalitarian point of view, but it does not follow that I have expensive preferences. 7 For a person to have expensive preferences, in the sense explained and shown to be theoretically fruitful by Dworkin and Cohen, is for this person to need more income, or, to put the point more generally, more resources, than others do simply to achieve the same level of [some form of] welfare as those with less expensive preferences (Dworkin 2000, 48), be that form of welfare satisfaction of preference, or selfdevelopment, or good experience, or whatever other form of welfare is brought into view (Cohen 2004, 5-6). People who are very good at converting resources into welfare may have an anti-egalitarian preference that they have more than others and still have cheap preferences overall in 505

ETHICAL PERSPECTIVES SEPTEMBER 2012 this sense e.g. because they are very good at converting resources into welfare and because they already have more welfare than others. 8 So construed, the concept of an expensive preference is essentially comparative. It describes a relationship between the preferences of two or more persons. Accordingly, I might have preferences that are expensive relative to one person and cheap relative to another. The baseline governing whether someone has expensive preferences and the expensiveness of satisfying other people s preferences may not converge. The baseline might be expensive relative to what it might have been had this person acted differently or been subjected to different conditions. Alternatively, preferences can turn out to be expensive when the difference between the quantity of resources required to achieve a certain level of satisfaction and the minimum quantity of resources that would have been required to achieve the same level of satisfaction had the agent acted differently is higher than it is in the case of other individuals. 9 Members of the present generation may have tastes that are expensive (relative to those of previous generations) according to Dworkin s definition, but not expensive (relative to those of previous generations) according to the definition just proposed. Although members of the present generation use more resources to reach a certain level of satisfaction than members of past generations, they could not have acted differently in a way that ensured that they were capable of reaching this level of satisfaction with fewer resources, whereas members of past generations could possibly have done so. Thus far I have discussed what it means for preferences to be expensive and, more specifically, for members of one generation to have more expensive preferences than that of another. I now turn to the moral significance of the fact that someone s preferences are expensive. Unlike Rawls and Dworkin, Arneson and Cohen do not take the expensive taste objection to show that the concept of welfare needs to be removed wholesale from the metric of justice. Rather, it shows that [w]e should replace equality of welfare by equality of opportunity for welfare (Cohen 506

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? 2010, 11; cf. Arneson 1991; 1989; italics mine). Hence, if the person with expensive tastes and the person with inexpensive tastes initially had equal opportunities for preference satisfaction, justice does not require the person who freely adopted more costly tastes to be brought up to the level of welfare now enjoyed by the person with cheaper tastes. Cohen thinks this response is truer to the underlying motivation of the expensive taste objection. 10 Applied to intergenerational justice, this view strikes me as plausible. It seems correct to insist that we should transfer resources to future generations with preferences that are more expensive than ours only if the expensiveness of those preferences is not the result of choices freely made by members of the relevant future generations. Rawls might disagree on the grounds that his position on non-compensation for expensive preferences is informed by his view of justice as a feature of relations of social cooperation, relations that do not obtain between past generations with cheap preferences and later generations with expensive preferences. According to his view, therefore, there is no injustice in either compensating or not compensating expensive preferences across generations. This Rawlsian reply is suspect, because it ties justice to the existence of relations of social cooperation. Also, I do think we can imagine situations in which justice requires that we take into account expensive preferences and do not simply let the frustration these generate (in the absence of a greater allocation of resources) fall on those who have such preferences. Consider the following case. G(eneration) 1 can choose between two policies both of which will give G 2 and G 3 the same quantity of resources. The first policy will cause G 3 to have expensive preferences; the second policy will cause neither generation to have expensive preferences. Assuming everything else is equal, I believe it would be unjust for G 1 to choose the first policy thereby causing inequality of welfare between future generations. If so, from a third person perspective at least, an expensive preference is not, pace Rawls, simply something for which the bearer of the preference bears full responsibility. 11 507

ETHICAL PERSPECTIVES SEPTEMBER 2012 In a later treatment of the topic, Cohen (2004, 7-8) revises the position espoused in his 1989 article. First, he stresses that what matters is not whether preference holders are responsible for having the preference, but whether they can reasonably be held responsible for the fact that their preference is expensive (cf. Arneson 1991). So if I cultivate a preference for transportation by private car rather than by train, which I could not possibly have foreseen would turn out to be expensive, because of soaring oil prices, according to Cohen I may be due compensation. Second, Cohen s 2004 position distinguishes between preferences that are informed by valuational judgments and those that are not. 12 An example of the former is a preference for opera, where this preference reflects the value judgment that a good life requires acquaintance with the fine arts. An example of the latter might be a strong preference for back massage that one has simply because it is pleasant without in any way judging massage or the pleasant experiences it gives to have any value. 13 Cohen believes that someone who has an expensive preference, which is informed by valuational judgment and could be unschooled, or which the holder deliberately chose to develop, may be entitled to compensation on account of his or her expensive preference: the subject s approving identification with their expensive tastes [ ] is something that I regard as a reason for offering it, since, where identification is present, it is, standardly, the agents very bad luck that a preference with which they strongly identify happens to be expensive, and to expect them to forgo or restrict satisfaction of that preference (because it is expensive) is, therefore, to ask them to accept an alienation from what is deep in them (2004, 7). 14 Let us apply Cohen s 2004 view to the issue of intergenerational justice. More specifically, we shall consider the suggestion that, according to this view, there is very little intergenerational injustice. Our preferences may be much more expensive than those of people who lived in medieval times. To reach the same level of satisfaction prior generations reached by having Sundays off, members of the present generation need the whole weekend off as well as several weeks of summer holidays, and the 508

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? preferences of future generations might well be more expensive than those of the present generation. According to Cohen s 1989 view, egalitarians would be unwilling to compensate people for such preferences if they could school themselves out of them. Our greater use of resources than previous generations might, therefore, involve an injustice it will do so if we are capable of jettisoning the expensive preferences and habituating to scarcity. Similarly, there might be no injustice if our present use of resources leads to the frustration of later generations expensive preferences, as long as these are preferences our descendants will be capable of schooling themselves out of. However, on the assumption that the expensive preferences are backed by valuational judgments say, unlike previous generations we believe leisure to be important to living a good life neither of these claims follows on Cohen s 2004 view. On that view, the present generation should transfer resources to allow future generations to satisfy their very expensive preferences, because these are informed by valuational judgments. This looks like a powerful objection to Cohen s 2004 position. The implication of that position that justice is compatible with massive intergenerational inequalities provided that these are suitably related to valuational judgments is surely highly questionable and, accordingly, Cohen s 1989 position delivers a more plausible view of intergenerational justice than his 2004 view. 15 III. CAUSING FUTURE GENERATIONS TO HAVE CHEAP PREFERENCES I now wish to consider a quite separate challenge to the welfarist perspective on intergenerational justice. Suppose one generation can determine whether the next generation has expensive or inexpensive tastes or, as I shall put it, suppose one generation can engineer the next generation s tastes, e.g. through the way in which they educate its members or through autonomy-circumventing techniques such as drugs or subliminal communication techniques. Imagine G(eneration) 1 faces the following choice: to preserve equality of welfare it can either abstain from using certain 509

ETHICAL PERSPECTIVES SEPTEMBER 2012 natural resources, which will then be available to satisfy G 2 s preferences that will be identical to those of G 1, or it can somehow ensure that G 2 develops cheap preferences, in which case G 1 can use more resources and leave fewer resources to future generations. Does G 1 act unjustly if it fixes things so that G 2 has cheap preferences? If an injustice is involved, that will count against not only equality of welfare but also equality of opportunity for welfare, since the two generations have equal opportunities for welfare (Vrousalis forthcoming, 12). 16 Call this the preference engineering objection. In my view the preference engineering objection does little to undermine welfarist egalitarianism. First, I doubt that there is an egalitarian objection to this sort of engineering. To see that the case sketched above does not give rise to an egalitarian complaint, consider a scenario in which a generation develops cheaper preferences to achieve equal satisfaction for it and future generations with expensive tastes. This seems wholly compatible with equality. If so, why should it make a difference, from an egalitarian point of view, whether a generation develops cheap preferences or engineers a future generation to have cheap preferences? Second, I do not think that engineering cheap preferences as such is objectionable. Cheap preferences, perhaps especially their cultivation, are often disparaged. This happens partly because it is assumed that such preferences differ in kind from most of our preferences. But, as a matter of fact, it is hardly surprising that few of us have the desire to direct operas, to own the largest private collection of motorcycles, or to be the first tourist who travels three times into space. Most of our preferences are to varying degrees informed by the costs of their satisfaction. This is good. It saves us a lot of frustration. 17 And in some, but by no means all, cases this is unproblematic. The cases where making people holding cheap preferences rather than expensive ones is unproblematic are those in which the preferences are non-brute, i.e. reflect valuations. Our preferences are sometimes in conflict with our evaluations e.g. the preference for meat. Engineering 510

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? to ensure that future generations have no such brute, expensive preferences would not be objectionable. 18 In cases where the preferences are based on valuations there is nothing problematic about ensuring that people hold cheaper preferences provided they can reach those preferences through a sound deliberative route from their present evaluations. In fact, I am inclined to take a stronger view of the range of cases in which it is permissible to engineer future generations so that they hold cheaper preferences. In this instance, we may well be not just permitted, but even morally required to cause future generations to hold cheaper preferences. Indeed some respectable environmental aims e.g. changing people s preferences in relation to private versus public transportation or in relation to air travel and holidays involve precisely this. More specifically, if we can achieve equality across generations by ensuring that future generations will have cheaper preferences than we have, we might well be morally required to do so. Note that whether a preference is classified as cheap or expensive will depend on the baseline, and in the case of future generations there are several possible suggestions as to what the relevant baseline might be. If, for instance, the baseline is how costly the relevant generations preferences would have been had we not intervened, there is an obvious question about what, exactly, is problematic about our acting with the aim of dictating whether preferences of future generations will be costly. 19 I conclude that reservations about the engineering of preferences do not generate a strong objection to welfarist metrics. In fact, this particular line of attack has nothing specifically to do with welfarism versus resourcism. If we were to determine people s share of resources through Dworkin s auction, we could imagine future generations preferences being engineered in such a way that the members of those generations do not envy the resource bundles of previous generations, in which case no inequality of resources obtains across generations. If preference engineering is a problem for welfarism, it is also a problem for Dworkinian resourcism. 20 Note, finally, that the line of argument explored in this 511

ETHICAL PERSPECTIVES SEPTEMBER 2012 section only presupposes that some ways of engineering future generations to have cheap preferences e.g. non-autonomy-violating ways are morally permissible. Hence, it avoids commitment to the view that all ways of engineering preferences of future generations are permissible. IV. DWORKIN S RESOURCISM AND SLAVERY OF THE TALENTED I now set aside expensive preferences and turn to worries about the intergenerational variant of Dworkin s slavery of the talented: the slavery of the first generation. To understand this concern we need some background. Dworkin s favoured metric is resources. In his view, therefore, people should be equal in the quantity of resources that they command. But how do we compare different bundles of resources? According to Dworkin s view, we should apply the envy test (2000, 68-69), which says that we should distribute resources and see if anyone prefers a bundle of resources possessed by another. If this is the case, equality of resources does not obtain and we should redistribute until we reach an envy free distribution. 21 Which resources are relevant for the purpose of applying the envy test? Clearly, external resources such as land and water should be included, but what should we say about internal resources such as labour power (or organs)? The question is tricky because, on the one hand, labour power is clearly a resource, and people who command more of it, or more valuable kinds of it, than others command more resources. On the other hand, labour power is closely tied to a person s talents and personality in such a way that people possess very different amounts of it, and addressing this inequality appears problematic. It is at this point that Dworkin s concern about the slavery of the talented enters the scene. Dworkin imagines that external resources are distributed through an auction at which everyone who participates bids and where all have equal bidding power. The price of popular resources rises simply because they have many bidders. Suppose internal resources are included in the auction. Since people have very different skills and talents, talented people 512

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? can expect that their labour power essentially their time will cost much more than the labour power of the untalented. Hence, if I am a talented person who would like some leisure, I will have to bid at the auction for some of my own time; and because, alternatively, I could be very productive in the hours I would like to spend in leisure, others will offer high bids for those hours. Accordingly, I will have to use a lot of my bidding power to ensure that I obtain some leisure time. Similarly, if I am a talented person who wants to spend some of my time in a job I find fulfilling, but in which I am not particularly talented and which in any case results in products few wish to buy, I will need to use a lot of my bidding power to ensure that this is possible. Indeed, if we allow coalitions of buyers to pool their resources and buy the labour power of those with talent, I might end up owning none of my time, with the owners of my labour power dictating what I do. 22 This, Dworkin submits, is unfair. The result of the auction, so construed, would be [ ] that each would have to spend his life in close to the commercially most profitable manner that he could, or, at least if he is talented, suffer some very serious deprivation if he did not. For since Adrian, for example, is able to produce prodigious income from farming, others would be willing to bid a large amount to have the right to his labor and the vegetables thereof, and if he outbids them, but chooses to write indifferent poetry instead of farming full time, he will then have to spend a large part of his initial endowment on a right that will bring him little financial benefit. This is indeed the slavery of the talented (1981, 311-312; 2000, 90). 23 For those subscribing to welfarist theories of distributive justice, the slavery of the talented objection is of perverse interest because it seems to pose no problem at all for such theories. Welfarist egalitarians think that people should be equal in terms of welfare, whether understood in terms of (ideal) preference satisfaction or in terms of pleasant mental states. They do not require talented people to have less welfare than others as a result of spending their lives close to the commercially most profitable 513

ETHICAL PERSPECTIVES SEPTEMBER 2012 manner. They might see no injustice in talented people working more hours than untalented people if, and to the extent that, this is compatible with equality of welfare. Such compatibility might be preserved if, for example, the kind of work talented people do is typically more rewarding than the kind of work untalented people do. However, this hardly qualifies as slavery. Hence, insofar as the slavery of the talented is a problem for egalitarian theories of justice, it is a problem for resource-egalitarian theories only. As a resource-egalitarian with real concerns about the slavery of the talented, Dworkin obviously needs to address this problem. He does so by excluding internal resources, e.g. labour, from the auction and the envy test. However, before I sketch Dworkin s response to the problem of the slavery of the talented more fully in the next section, I want to consider how the problem arises in an intergenerational context. Like most theorists of justice, Dworkin focuses on intragenerational issues and, thus, does not address the problem in an intergenerational setting. So let us consider what happens when one applies the envy test and the idea of an auction to a simple intergenerational setting. Imagine a hundred generations. They do not overlap: in an admittedly far-fetched biological progression, the members of one generation die at the very moment those of the next are born. For some reason there will only be one hundred generations. Suppose, moreover, that labour expended by one generation improves the situation of all ensuing generations in that it boosts the productivity-increase of all later generations. Hence, the labour of the first generations makes a difference to the value of all later generations, and in particular it makes a tremendous accumulated difference to the value of the labour of the last few generations. The labour of the last generation makes no difference to any other generation. Suppose, finally, that members of one generation can direct members of any generation, including previous ones, to employ their labour power in whatever way they see fit provided they own it. 24 So, by analogy to the intragenerational case, if I own the labour power of a member of a previous generation I can employ it in way that benefits me the most. 25 514

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? In a case like this, the value of the labour of the first generation will be much higher than that of the later generations. Many members of later generations will want to bid for first-generational labour power, so members of the first generation will be obliged to use a lot of their bidding power simply to ensure that they have any leisure at all. Indeed, if coalitions of buyers from several generations are permitted, they might find that others completely outbid them in an auction for their time. If outbid, members of later generations will then employ the labour power of the first generation in a maximally productive way, thereby conferring maximum benefits on later generation. It is important to see that in this case the first generation is not talented in any ordinary sense of the word, but in the sense that employing its labour power results in much greater benefits for others (as a result of how this employment positively affects the productivity of all later generations) than the employment of, say, the labour power of the last generation (which has no effect on the productivity of any other generation). Its members could be less talented than those of later generations, and their labour has greater value than that of later generations simply because it is positioned at the start of the biological progression, where it is for convenience, let us say super-productive f. 26 It can be seen, then, that Dworkin s slavery of talented objection arises as a variant involving super-productivity f in the intergenerational setting. Accordingly, if that objection is effective against resource-egalitarian theories of intragenerational justice, obliging us to treat labour power and other resources differently, something similar would seem to follow in the case of resource-egalitarian theories of intergenerational justice. 27 V. WHY DWORKIN S RESPONSE TO THE SLAVERY OF THE TALENTED FAILS IN THE INTERGENERATIONAL CASE Having presented the challenge of the slavery of the talented objection in an intragenerational and intergenerational setting, I now describe 515

ETHICAL PERSPECTIVES SEPTEMBER 2012 Dworkin s response to the problem in an intragenerational setting and then explain why this response fails in an intergenerational setting and, drawing on an objection by G. A. Cohen, explain why it does so in a way that also casts doubt on whether Dworkin s response works in the intragenerational setting. As I indicated above, Dworkin seeks to avoid the problem of the slavery of the talented by excluding labour from the auction and the envy test. At the auction he settles upon, one cannot bid for the labour power of others, and when one is determining whether some envy the resource bundles of others, one should disregard the labour power they command. By making this move, Dworkin s theory avoids the problem, as he diagnoses it, but it is also clear that he runs into another. Some people might simply be endowed with high-value labour power, and Dworkin is committed to the view that justice should be endowmentinsensitive. He thus needs some mechanism to address the inequality between those who are endowed with high-value labour power and those who are not. For that purpose, he proposes an insurance scheme in which people insure themselves against deficiencies in talents. In principle, such a scheme might lead to the same problematic results as Dworkin presumes will follow if we allow people to bid for each others labour power at the auction. However, in practice it will not, since people would be unwilling to insure themselves against not being among the most talented. Supposing they were willing to do this, they would have to pay a very large premium if they turned out to be among the very most talented, in which case they would, in effect, be talented slaves. 28 In the light of this response to the slavery of the talented, one might suppose that one can make a similar move in an account of intergenerational justice by subjecting the labour of the talented to an insurance scheme rather than the envy test and an auction. Unfortunately, it is not immediately apparent that Dworkin s insurance scheme can deal adequately with the intergenerational case. For whereas people in Dworkin s scheme would insure against not being sufficiently talented in the standard sense being untalented in this sense implies having fewer 516

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? marketable skills and, thus, being deprived of benefits one would otherwise have had they would not insure against not being super-productive f in the way the first generation is relative to later generations. The benefits of being super-productive f accrue to others, not to those who possess super-productivity f and, accordingly, not being super-productive f does not leave one worse off than one would otherwise have been and, thus, is not a situation one would want to insure against. Indeed, assuming people are ignorant of the generation to which they belong and wish to insure against belonging to some particular generation, they will insure against belonging to the first and poorest generation. At this point, some might object against the idea of insuring against being a member of the first generation that there is no way in which intergenerational compensation can be awarded in this case; but I do not think this is true. The compensation or at least, something relevantly like (self-) compensation might take the form of the first generation using a disproportionately large quantity of natural resources, leaving much less to ensuing generations than they would have done had they merely used their equal share. 29 In any case, and as already noted, I discuss the justice of distribution irrespective of the possibility of redistribution. This discussion of Dworkin s insurance device in an intergenerational setting reveals something important: that the insurance device is not essentially tied to the elimination of inequalities of resources. One insures to guarantee some kind of compensation in disadvantageous situations. In the familiar intragenerational setting, Dworkin assumes that being untalented is a disadvantage and that people gain from being talented; but in the intergenerational setting not being super-productive f (because one belongs to the last generation) is not a disadvantage the accumulative benefits that, ex hypothesis, result from exercising this talent accrue to members of later generations. However, if the quantity of resources one commands can diverge from the extent of the advantages that one takes into account when making up one s mind about the purchasing of insurance, equality of resources does not guarantee equality of the relevant kind of advantage. 30 517

ETHICAL PERSPECTIVES SEPTEMBER 2012 By way of further support for my claim that the insurance device is not a device for ensuring equality of resources, because whether one takes out an insurance against having a particular set of productive talents is not only determined by the level of resources one commands in virtue of these talents, consider Cohen and Dworkin s disagreement regarding the difference between welfarist and resourcist conceptions of egalitarian justice in connection with a rather different sort of case. To defeat Dworkin s purely resourcist view, Cohen imagines a person who is very good at controlling the movement of his limbs and very adept at moving them. So in an important dimension he is more resourceful than the rest of us. The problem is that he experiences pain whenever he moves his limbs. In Cohen s view, this person is worse off than the rest of us in a way that is relevant from the perspective of egalitarian justice. Since he does not have fewer resources than the rest of us he is very capable of moving his body and the experience of pain is not a resource, but merely something that affects his welfare, this shows, Cohen argues, that welfare is likewise relevant from the perspective of egalitarian justice. 31 The bearing of this dispute on Dworkinian auctions whether in an intragenerational or intergenerational setting can be readily appreciated if we imagine a situation in which talented people do not need to concentrate on what they are doing in order to exercise their talents. Suppose that there are great differences in people s talents and hence productivity, but that exercising one s talents in productive labour is a bit like driving on a deserted road in that it does not require one s attention: thus, for example, a talented brain surgeon can compose and even write down poems while performing surgery. Whether or not they attend to what they are doing, non-talented people cannot perform surgery, but the talented person can do it inattentively without loss of quality or speed in his or her work. The point here is that being at work does not prevent one from doing the things one likes or prioritizes. Hence, there is no reason for talented people, who put in many hours a day, to envy untalented people, who do not. Of course, talented people do not count as less talented for 518

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? this reason, or as people who do not command more internal resources than untalented people. Hence, the ground for envy in Dworkin s scenario cannot be that untalented people have more resources than talented ones if labour is put up for bidding at the auction. 32 I have proposed that the envy test registers welfare as well. VI. SLAVERY OF THE FIRST GENERATION: AN OBJECTION So far I have proceeded on the assumption that resource-egalitarian accounts of justice need to avoid productivity-sensitive enslavement. But this is not something that can simply be assumed. Thus, in connection with the intragenerational case some observers have argued that Dworkin fails to give convincing egalitarian reasons for thinking that talents and labour should not be auctioned. Miriam Cohen Christofidis pursues this line of argument. In her view, the situation of the talented in intragenerational auction is nothing like slavery in the literal sense (2004, 32). Admittedly, talented people might face a severely restricted choice. They may have to devote most of their purchasing power at the auction to ensuring that they control at least some part of their time. But it is not clear that they end up worse off than others in a sense that ought to trouble egalitarians. That would follow only if all that people cared about, in relation to their jobs, was the number of hours they worked the fewer, the better. However, people care about other things too, including how fulfilling their working lives are (Christofidis 2004, 36); and since, typically, the jobs of talented people are more fulfilling than those of untalented people, it may well be that hardworking talented people are not envied by much less hardworking untalented people. Can this line of argument be pressed into service to assuage worries about the slavery of the super-productive f? I do not think so. In real life there is every reason to believe that the further back we go in history the less fulfilling jobs were. Nor would it matter if this contingent, historical claim were incorrect. We can simply stipulate that in the case of 519

ETHICAL PERSPECTIVES SEPTEMBER 2012 super-productivity f the jobs of the first generation are less fulfilling than later jobs without exception. The ideal of equality of resource should give the right answer in this hypothetical case as well. It does not. Christofidisstyle doubts about the slavery of the super-productive f cannot be marshalled at this point. Let me summarize the discussion of the two last sections. We have been concerned with resource-egalitarian accounts of distributive justice. Many believe that, in the intragenerational setting, these accounts have a problem avoiding the enslavement of the talented. If that is so, then, in the intergenerational setting specifically, in the far-fetched case involving one hundred generations we have described there is a similar problem with the slavery of the first generation, or as we have put it, the super-productive f. While the problem may be softened in the case of talented people, because they tend to have more fulfilling jobs, there is no analogous way of softening it in relation to the super-productive f. Moreover, it is not clear how Dworkin s insurance scheme can be applied in an intergenerational setting. In fact, insurance seems to be problematic in both intragenerational and intergenerational settings. I conclude that the slavery of the first generation, or the super-productive f, constitutes an important challenge to Dworkin s resourcist position. VII. CONCLUSION I have argued that two objections to welfarist accounts of the metric of (intergenerational) justice, one from expensive tastes and the other from the engineering of preferences, are unsuccessful. I have also presented reasons for thinking that Dworkin s resourcist metric is incapable of handling a variant of the slavery of the talented what I have called the slavery of the first generation, or of super-productivity f convincingly. 33 520

KASPER LIPPERT-RASMUSSEN EQUALITY OF WHAT? WORKS CITED Arneson, Richard. 1989. Equality and Equal Opportunity for Welfare. Philosophical Studies 56: 77-93. Arneson, Richard. 1991. Liberalism, Distributive Subjectivism, and Equal Opportunity for Welfare. Philosophy & Public Affairs 19/2: 158-194. Arneson, Richard. 2000. Welfare Should Be the Currency of Justice. Canadian Journal of Philosophy 30/4: 497-524. Cohen, G. A. 1989. On the Currency of Egalitarian Justice. Ethics 99: 906-944 (reprinted in Cohen 2010, 3-43). Cohen, G. A. 2004. Expensive Tastes Ride Again. In Dworkin and His Critics. Edited by Justine Burley, 3-29. Oxford: Blackwell. Cohen, G. A. 2008. Rescuing Justice and Equality. Cambridge, MA: Harvard University Press. Cohen, G. A. 2010. On the Currency of Egalitarian Justice, and Other Essays in Political Philosophy. Princeton: Princeton University Press. Cohen Christofidis, Miriam. 2004. Talent, Slavery, and Envy. In Dworkin and His Critics. Edited by Justine Burley, 30-44. Oxford: Blackwell. Dworkin, Ronald. 1981a. What is Equality? Part I: Equality of Welfare. Philosophy & Public Affairs 10: 185-264 (reprinted in Dworkin 2000, 11-64). Dworkin, Ronald. 1981b. What is Equality? Part I: Equality of Resources. Philosophy & Public Affairs 10: 283-345 (reprinted in Dworkin 2000, 65-119). Dworkin, Ronald. 2000. Sovereign Virtue. Cambridge, MA: Harvard University Press. Dworkin, Ronald. 2004. Replies. In Dworkin and His Critics. Edited by Justine Burley, 339-395. Oxford: Blackwell. Elster, Jon. 1983. Sour Grapes: Studies in the Subversion of Rationality. Cambridge: Cambridge University Press. Hansen, Rasmus Sommer and Søren Flinch Midtgaard. 2011. Sinking Cohen s Flagship or Why People with Expensive Tastes Should Not Be Compensated. Journal of Applied Philosophy 28: 341-354. Keller, Simon. 2002. Expensive Tastes and Distributive Justice. Social Theory and Practice 28: 529-552. Knight, Carl. 2009. Egalitarian Justice and Valuational Judgment. Journal of Moral Philosophy 6: 482-498. McMahan, Jeff. 2002. The Ethics of Killing: Problems at the Margins of Life. Oxford: Oxford University Press. Page, Edward. 2007. Intergenerational Justice of What: Welfare, Resources or Capabilities? Environmental Politics 16: 453-469. Rawls, John. 1982. Social Unity and Primary Goods. In Utilitarianism and Beyond. Edited by Bernard Williams and Amartya Sen, 159-186. Cambridge: Cambridge University Press. Scanlon, Thomas M. 1975. Preference and Urgency. Journal of Philosophy 72: 655-669. 521

ETHICAL PERSPECTIVES SEPTEMBER 2012 Schaller, Walter E. 1997. Expensive Preferences and the Priority of Right: A Critique of Welfare-Egalitarianism. Journal of Political Philosophy 5: 254273. Sumner, L. Wayne. 1996. Happiness, Welfare, and Ethics. Oxford: Clarendon. Vrousalis, Nicholas. Meeting Posterity s Needs: Flourishing, Sustainability, and the Metrics Question. Forthcoming. NOTES 1. An exception is Page (2007). 2. If they did, might not the metric of life-time equality similarly differ from the metric of time-slice equality? And might not the relevant metric for comparing G n with G n+1 differ from the relevant metric for comparing G n+2 with G n+3? 3. Ultimately, I am not certain that welfare is the metric of justice. The intrinsic preferences people have for states of affairs other than those that involve their self-interest may count as well from the point of view of egalitarian justice. In a world in which people have only other-regarding preferences for species of non-human animals, it might be unjust if some people s preferences in favour of penguins welfare are satisfied, while other people s preferences in favour of pigs welfare are unsatisfied. I do not address this issue elsewhere in this paper. 4. I assume that an unjust distribution can arise even if everyone acts in the best way possible to promote justice. Hence, I depart from a Dworkinian statist view, according to which justice is primarily concerned with whether the state treats its citizens with equal concern and respect. I believe this view is too narrow anyway (e.g. it does not allow us to ask questions about international distributive justice). 5. Dworkin (2000, 48-59) presses a similar objection to equality of welfare. 6. While my argument does not rest on an empirical claim to the effect that later generations tend to have more expensive preferences than earlier ones, the well-known mechanism of adaptive preferences (Elster 1983, 109-140) supports this claim, given economic growth. 7. Even if the preference represents an objectionable stance, it remains unclear that egalitarians will not want to accommodate it. Similarly, a preference that one has less satisfaction than others is an anti-egalitarian preference. It would, ironically, qualify as an expensive preference if it is expensive to (reliably) satisfy, say, because comparisons of levels of satisfaction are very costly. 8. A hyper-egalitarian preference is one that is informed by the desire to have no greater satisfaction, or not to consume more expensive goods, than others. 9. According to Cohen s view of the normative significance of expensive tastes, this concept of expensive tastes might be normatively more relevant than the concept Dworkin defines. 10. Dworkin thinks otherwise. For him identification rather than choice is of the essence (cf. Scanlon [1975] on pilgrimage and compensation). 11. Normally, expensive preferences are discussed in relation to cases in which it is the bearer of the preference who can cultivate it or refrain from so doing. 12. Cohen does not say what kind of valuational judgment he has in mind: prudential value judgments or just any kind of value judgment, e.g. a judgment to the effect that although the 522