Egalitarianism, broadly conceived, has received a lot of attention during the last thirty years. However, not much consensus has evolved. Instead, the discussion has been divided into a number of sub-issues, which are treated largely independently of each other. This is clearly reflected in the present anthology with papers from a conference held in Copenhagen in 2004. The fact that the editors have inserted a few cross references cannot conceal that the twelve papers hardly refer to the same issues.
The editors have written a very informative introduction, which presents some of the main lines of the recent debate on egalitarianism. Some of the papers relate only marginally to the development described in the introduction. However, this is much preferable to the more usual kind of introduction that attempts to construct a superficial connection between the papers in an anthology. The wide range of subjects subsumed under the notion of egalitarianism is also illuminated by the fact that there are many other issues not dealt with in the introduction; for instance, the discussion on equality of income in economics, measures of inequality and their relation to welfare functions, equality under conditions of risk/uncertainty, the use of lotteries and John Broome' theory of fairness, just to mention a few.
I shall attempt to give an impression of each of the papers while focusing on the details that have been of most interest to me. The anthology is divided into four sub-sections: Foundations for Equality (2 papers), The Nature of Equality (6 papers), Equality and Other Values (2 papers) and Applications (2 papers). The first paper by Thomas Christiano presents a long and winding argument for the claim that egalitarianism is the fundamental principle of distributive justice. This principle states that “individual persons are due equal shares of some fundamental substantial good” (which later turns out to be well-being).
Christiano derives this claim from a set of what he calls uncontentious premises: the generic principle of justice (that one ought to treat relevantly like cases alike and relevantly unlike cases unlike), the idea that humans have equal moral status, the principle of well-being (that it is better that individuals have more well-being than less) and the claim that there are no relevant differences between human beings to determine that one person should have more well-being than others. However, as far as I can see, some of these premises are heavily loaded with content. For instance, the generic principle is understood to exclude both utilitarianism and prioritarianism. Also, equal moral status is interpreted quite strongly. I find it hard to accept these premises as uncontentious. Consequently, I find the derivation a bit futile – what does the fact that one can “derive” an egalitarian position from quite strong premises tell us?
There is an interesting feature of Christiano's position, though. He considers the principle of well-being an integral part of egalitarianism (i.e. his position is a version of what is often called moderate egalitarianism). This feature allows him to avoid the levelling-down objection. We might say that, for every unequal state, there is some equal distribution that is more just. But it does not follow that any equal state is more just, not even in respect of equality. This is closely related to a point I make in Jensen (Reference Jensen2003): if moderate egalitarianism is defined by a primitive all things considered relation between states (and this is how I understand Christiano's suggestion), there is no separable respect of equality as the leveling-down objection presupposes. True, one can always from the all things considered relation derive a relation concerned with the aspect of equality. But this is not the right representation of the kind of egalitarianism that is concerned with benefiting people where benefits are weighted such that, for a given total, an equal distribution of them is better than an unequal. Similarly, the principle of well-being is an integral part of Christiano's egalitarianism. As he says (and I agree), what would be the point of equality if this were not so?
There is something bold about Ingmar Persson's short and concise defence of extreme egalitarianism. Extreme egalitarianism is the following claim: A state is just if and only if everyone (capable of being well or badly off) is equally well off in it. This claim is derived from the formal principle that a state is just if and only if everyone is equally well off in it unless there is something to make it just that some are better (or worse) off than others and the claim that there is nothing to make it just that some are better (or worse) off than others. The extreme character of this position is not that that it makes egalitarianism the only or dominant moral consideration. Justice should be balanced against beneficence. A Pareto superior inequality might be morally justified, even though the inequality is unjust, because the gain in welfare outweighs the injustice. Also, what is extreme about it is not that it includes conscious animals. Although he is very explicit in saying “everyone capable of being well or badly off”, Persson does not discuss the implications of treating animals justly. (However, as we shall see later, Peter Valentine, in his contribution, argues that it would have deeply problematic consequences.) The extreme character of the position is simply the negative claim that nothing can make it just that some are better off than others. Thus, Persson argues against desert and rights as candidates for just-making conditions. In particular, he argues that responsibility for the desert basis cannot serve as just-making condition, because there is a regress: one is not responsible for the facts that make one responsible in the first place. Apart from libertarians, Persson's main opponents appear to be luck-egalitarians who claim that people may justly become worse off through choices for which they are responsible themselves. However, Persson himself accepts that it is morally permissible to bring about an inequality through an autonomous choice. So the disagreement appears to be about whether such an outcome should be called “just” or, as Persson prefers, “neither just nor unjust”. In a way, I find Persson's approach more lucid. However, when he claims that the formal principle is true on logical grounds, I cannot follow him.
Kasper Lippert-Rasmussen's paper is an analysis of Parfit's distinction between telic and deontic egalitarianism. According to this analysis, telic egalitarians are concerned with badness and deontic egalitarians are concerned with injustice, more precisely injustice that “essentially involves agency or representations”. Lippert-Rasmussen unpacks three foci for egalitarianism within Parfit's distinction, genesis-and-outcome, pure-genesis and pure-outcome, and he shows that each of them is open for both telic and deontic egalitarianism. From this, he draws the implications, contrary to Parfit, firstly that deontic egalitarianism is not necessarily narrower in scope than telic egalitarianism, and secondly, that deontic egalitarianism does not necessarily avoid the leveling-down objection.
This is all right as it goes. However, Lippert-Rasmussen leaves it unclear what the point of drawing the distinction is – as a matter of fact, he calls it “insignificant” in the title. My impression has always been that Parfit's intention was to base the distinction on the distinction between teleological and deontological positions, such that telic egalitarianism is a teleological consideration, and deontic egalitarianism is a deontological consideration. The problem with the way Parfit draws the distinction is that he overlooks the generally recognized fact that teleology can attach value to outcomes as well as their genesis. This is basically what gives rise to the “insignificance” Lippert-Rasmussen points at. However, the remedy, in my view, should be to distinguish deontology from teleology in a way that does not rely on the (insignificant) distinction between the value of acts and the value of outcomes. Nils Holtug presents a thorough defence of prioritarianism as compared with telic egalitarianism. Basically, he reworks in more detail Parfit's argument that telic egalitarianism is exposed to the leveling-down objection whereas prioritarianism is not. However, it is interesting that he, at one point, describes egalitarianism as the position that lets the value of a benefit depend on the welfare level of others. Given moderate egalitarianism (benefits are always valued positively), precisely this definition of egalitarianism is not exposed to the leveling-down objection (cf. the discussion above).
Holtug commits himself to a person-affecting morality, and as a consequence he accepts the challenge from the non-identity problem. His solution is a principle, which makes it person-affecting to come into existence. To me, this solution appears ad hoc. However, I should note that Holtug has given more elaborated arguments for the principle elsewhere. Holtug ends the densely argued paper by answering some further objections to his prioritarianism.
Dennis McKerlie contributes a paper on “Egalitarianism and the Difference Between Interpersonal and Intrapersonal Judgments”. This paper reconstructs the history of McKerlie's thoughts on equality and the time dimension. Contrary to the then dominant view (made explicit by Norman Daniels) that egalitarianism is concerned with life-time well-being, McKerlie was among the first philosophers to insist on the claim that there is a case to be made for equality between people at a time. McKerlie himself interprets this in the light of an analogy between the interpersonal and the intrapersonal case. On “the standard egalitarian view”, the separateness of persons prevents maximization of total well-being, whereas nothing prevents maximization of well-being over time within a single life. But according to McKerlie, the case for equality between lives at a time suggests an analogical case for equality (or priority, as McKerlie now prefers it) between lifetimes within a single life.
McKerlie points out a very strong pragmatic reason for the claim that egalitarians should be concerned about equality at a time, namely that if the concern were for equality between lifetime well-being, we would not know whether there is reason to redistribute between people before their life has ended. This seems to me absurd, and perhaps this consequence has not received sufficient attention. I am not yet convinced by McKerlie's analogy, though. He relies on intuitions mostly, but I think that we need a firmer grip on the relation between well-being at a time and lifetime well-being in order to be able to establish an analogy like this.
Bertil Tungodden and Peter Valentyne present a formal analysis of Rawls' difference principle. Most writers interpret the difference principle as leximin, implying that we should give absolute priority to the worst-off individual (and in case of indifference, the second worst off and so on). But Rawls himself had the worst-off group in mind. It is the average level of this group that should receive absolute priority; only, he did not define the worst-off group precisely. Tungodden and Valentyne investigate under which conditions the difference principle thus conceived can diverge significantly from maximin. Very roughly, they show that it can, if the least advantaged group is defined by an absolute cut-off level. If the definition satisfies the condition that the best-off individual is not a member of the least advantaged group (which rules out absolute cut-offs), then (given some further conditions) the difference principle cannot diverge significantly from maximin. But, interestingly, if this condition is weakened to the requirement that the welfare of any of the non-worst-off individuals can be increased so that they are not members of the least advantaged group, it can diverge.
Linda Barclay's paper presents the feminist attention to the relations between people and their relevance for egalitarianism. However, the paper is mostly negative in focus, involving a critique of Elizabeth Anderson's notion of “democratic equality” and her claim that what should be equalized are indeed the relations between people, rather than well-being, resources or capabilities. This seems all right, since Anderson appears to base her critique of luck egalitarianism and her own “democratic egalitarianism” on somewhat crude claims. However, Barclay accepts the feminist idea that egalitarianism should pay more attention to the patterns of socialization, social relations and the like. I share this intuition, but the proper way to cash it out remains to be seen.
Peter Vallentyne writes on equality between humans and animals. Using “mice” and “men” as placeholders, he argues that if morality requires us to promote equality among individuals with moral standing, and if mice have moral standing, then since mice have much lower well-being than humans, the problematic conclusion follows that morality requires “a massive shift of resources away from most humans (. . .) to most mice”. Vallentyne then examines different attempts to avoid this conclusion. Vallentyne dubs his egalitarian premise “moderate”. But in my view, it is quite strong. It is defined as the requirement that “we significantly promote equality of fortune among individuals with moral standing”. The question is how to weigh up the concern for equality of well-being against the concern for the size of total well-being. Vallentyne accepts that the egalitarian premise “perhaps” is subject to “certain relatively weak constraints” (which does not sound moderate to me). And it turns out that “a relatively weak constraint” is a constraint that “does not typically rule out a high proportion of the Pareto optimal distributions of well-being”. This is to avoid levelling-down.
Vallentyne admits that a weak constraint on equality promotion is crucial for generating the problematic conclusion. He refers to utilitarianism as a contrast case. On utilitarianism, the conclusion would not follow. However, it seems to me that he overlooks the middle area (which I would call “moderate”), where the concern for equality is weighed up against the concern for maximizing the total of well-being. Suppose there was to be a massive shift of resources from men to mice. Then human production would decrease dramatically and consequently, in the future, there would be far less resources available for the generation of well-being. I think that a less massive shift of resources with only marginal effects on future production is much preferable to this problematic conclusion. In my view, the concern for the total of well-being has great weight in this case. But it could still be the case – contrary to the demands of utilitarianism – that we should accept a certain amount of equalizing at the cost of reducing the total slightly. But since this outcome would be Pareto incomparable with the problematic conclusion, Vallentyne's egalitarianism rules it out. Hence, as I see it, the problematic conclusion only follows because the constraint on equality promotion is too weak, or in other words, because Vallentyne's egalitarianism is quite extreme.
Andrew Williams' paper attempts to use Scanlon's account of choice and responsibility to illuminate their role in contemporary egalitarianism. Though sympathetic to Scanlon's contractualist approach, Williams in his interpretation identifies some aspects in need of clarification.
Richard Arneson defends the idea that desert has a role to play alongside priority to the worse off constituent in our fundamental principles of justice. He takes the side that the desert base must be one for which the individual is responsible. He meets Persson's challenge that there is nothing for which an individual is ultimately responsible by pointing at the conscientious striving to realize what is right. Persson would object that an individual is not responsible for possession of the motivation to do so, but Arneson answers that desert is based on doing the best one can, given whatever the initial circumstances are. Persson does not appear to have considered this Kant-flavoured approach.
Arneson thus suggests that people have priority, the more deserving they are, and the worse off they are. At one point he says that “either desert or priority might determine right conduct when they conflict”, but elsewhere he appears to suggest that both considerations should be combined without specifying how. At any rate, the desert part seems to be utterly difficult to turn into practical policy. To say that it is a regulative ideal, and that we in some cases can find proxies to guide our practice, is hardly an answer to this worry.
The papers I have found most interesting are the two final papers on “applications”. Jonathan Wolff asks a question raised by the Titanic: Why do we find it intolerable that there were lifeboats for first class only, whereas steerage was expected to go down with the ship? Wolff sees this question as part of the more general problem about how and when to regulate markets involving variations in safety standards. The traditional answer is that we have reason to regulate safety in cases of asymmetry of knowledge, externalities and monopoly. The trouble is that none of these reasons apply to the Titanic case. However, Wolff argues, safety on the Titanic might be considered a “high-fidelity slow-release” good, which means, roughly, that it is consumed some time after it is paid, and that its quality is under control by the supplier. The point is that it is hard to create fluid markets for these goods, and therefore the market for them does not function well. It is a very stimulating but rather short paper. I believe there is much more to be said on this issue and the more general problem about equality under conditions of risk and/or uncertainty.
Susan Hurley, who sadly died from cancer recently, has examined what different egalitarian positions say about health. Does health have a special status among the goods that are relevant for justice? An unhealthy or disabled person is often believed to be less efficient in generating welfare out of resources than a healthy person, and this is the motivation behind the egalitarian aim to equalize the welfare between people even though it results in a lower total of welfare. However, to consider the unhealthy or disabled less efficient in generating welfare may be a prejudice; these people often learn to adapt and perhaps become equally effective as the healthy. Then if we want to maximize the total of welfare, efficiency does not favour the healthy; but the flipside of coin is that there no longer are egalitarian reasons for compensating the unhealthy or disabled or even remove their disability, if possible, and this appears counter-intuitive. I do not recall having seen this consequence pointed out so clearly before.
Hurley takes this to provide a case against welfare egalitarianism. Because welfare egalitarianism does not distinguish between poor health and expensive tastes but simply considers both a cause of inefficiency, there might be reasons to turn to resourcism which considers health an (internal) resource, and hence subject to equalizing, whereas expensive tastes are considered a matter of personal choice, not a resource. (This step is a bit too hasty. Hurley assumes a preference-based account of welfare. The counter-intuitive consequence could instead provide a case for an objective list account of welfare. But then on the other hand, on an objective list account, the difference between welfare egalitarianism and resource egalitarianism need not be big.)
According to resourcism, health is just one resource among others. For Daniels, who understands health from a Rawlsian perspective as a condition of equal opportunity, health takes priority over the distribution of other goods. Luck egalitarians, on the other hand, apply the concept of equal opportunity to all goods: egalitarianism should only compensate for bad brute luck, but not for the bad outcomes of autonomous choices (including choices that compromise health status). All in all, considering the different positions on what to equalize, Hurley finds no decisive reasons for regarding health as special. However, considering the difference between the priority view and egalitarianism, health may be different from other goods in that inequalities in health do not necessarily make everyone better off than equality in health. Moreover, it appears particularly repugnant to level down people in health for the sake of equality. So from this perspective, two important reasons for giving health special status emerge.
The papers are generally well written, but my judgment is that only few of them have the potential to become classics. Still, the anthology might serve as a useful introduction to important discussions within the diverse range of issues related to distributive justice.
