Egalitarianism and Distribution to the Disabled
Egalitarianism and Distribution to the Disabled
Abstract and Keywords
Welfare egalitarianism distributes too many resources to some who are disabled, while resource egalitarianism distributes too few resources to some who are disabled. Both of these egalitarian theories are flawed because they are insensitive to relative benefit and do not take into account the extent to which people can benefit from resources. This chapter explores egalitarian approaches to disability and distribution. After drawing a number of distinctions between various egalitarian theories, it considers the problems faced by these theories and compares them with those associated with utilitarianism. It discusses moderate and extreme forms of egalitarianism as well as marginal egalitarianism and looks at four limits to redistribution under welfare egalitarianism: satiation, counterproductivity, equalization of welfare, and sympathy.
There are many kinds of utilitarianism, but there are even more kinds of egalitarianism. The distinction I most stress in this book is that between resource egalitarianism, which would distribute too few resources to some who are disabled, and welfare egalitarianism, which would distribute too many resources to some who are disabled. The common fault of both these egalitarian theories is that they do not consider the extent to which people can benefit from resources. Because resource egalitarianism is insensitive to relative benefit, it distributes too few resources to disabled people who could benefit greatly from additional resources. Because welfare egalitarianism is insensitive to relative benefit, it distributes too many resources to those severely disabled people who are considered to have least welfare, even when they would benefit only infinitesimally from additional resources.
In this chapter I also draw a number of other distinctions between various egalitarian theories. This chapter is not intended to be a catalog of egalitarian theories;1 I focus on distinctions (p.56) that are most relevant to my argument. Many distinctions considered important by egalitarians are not particularly relevant to my argument. I note some of these distinctions, but only to show that my argument against egalitarianism is unaffected by them.
It may be useful, at the outset, to make a general tripartite distinction among egalitarian theories that is highly relevant to my argument, but that egalitarians themselves are unlikely to make. There are three kinds of egalitarianism. First, there are egalitarian theories that are so extreme that they lack even initial plausibility. Second, there are egalitarian theories that have initial plausibility, but lose that plausibility when one considers how they would deal with the disabled. Third, there are egalitarian theories that maintain plausibility in dealing with the disabled, but only because they coincide to some extent with utilitarianism, because they find some way of becoming sensitive to relative benefit.
As I discuss the problems faced by egalitarian theories, the reader may sometimes wonder: isn't this also a problem for utilitarianism? I keep this question in mind and draw some parallels to utilitarian theory.
My discussion in this chapter is thematic. Although I refer by name to a number of egalitarian theorists, I defer any detailed discussion of specific theorists until the next few chapters. As I demonstrate in those chapters, egalitarian theorists tend to compromise their theories when dealing with disability. Resource egalitarians find some way to distribute extra resources to the disabled, and welfare egalitarians find some way to halt redistribution to the disabled. Indeed, it is so rare to find a theorist who advocates unalloyed egalitarianism in dealing with the disabled that my thematic discussion here may seem a little surreal, especially with respect to welfare (p.57) egalitarianism. Nevertheless, there may be some value in exploring what would happen if we actually tried to apply egalitarian ideas to the problem of disability.
One way in which an egalitarian theory can be extreme is if it advocates what has been called “levelling down.”2 A welfare-egalitarian theory might counsel us to reduce everyone's welfare in order to make welfare more equal; a resource-egalitarian theory might counsel us to reduce everyone's resources in order to make resources more equal.3
Another way in which an egalitarian theory can be extreme is if it does not limit itself to the distribution of material resources (what I am calling simply resources), but advocates also the redistribution of body parts. Some have suggested that a consistent egalitarianism would require the redistribution of one eye or one kidney from people who have two to people who have none.4 Larry Temkin, who is an egalitarian, has suggested that a pure egalitarian theory would require that we blind sighted people. This proposal combines both kinds of egalitarian extremism: it would take away people's body parts, and it would make everyone, including the blind, worse off. Temkin does not of course advocate blinding the sighted; while he suggests it would be required by egalitarianism, he considers it wrong because it traduces other important values that he recognizes, such as liberty and utility.5
I am not too interested in the more extreme forms of egalitarianism. They do not have much intuitive plausibility, and I do not consider them to be serious competitors to utilitarianism. Thus, when I use the term “egalitarianism” in this book, I generally do not have in mind egalitarian theories that (p.58) would require leveling down or the redistribution of body parts. Egalitarians generally do not advocate such things, and it is not my job to tell them that they must do so if they want to be “pure” egalitarians. It is only my job to tell egalitarians when they have departed from their own theories, as they themselves have specified those theories, in order to deal with the problem of disability.
The “leveling down” of welfare is not of course a problem for utilitarianism; the idea of reducing everyone's welfare is anathema to utilitarianism. Arguably, there may be unusual situations where utility would be increased if everyone's resources were reduced. I do not dwell on this possibility, however, as it seems farfetched and has not in any event been raised by opponents of utilitarianism.
The opponents of utilitarianism have on occasion argued that it could require the redistribution of body parts. This is a problem for utilitarianism, but not as much as it is a problem for egalitarianism. As regards body parts that are not necessary to life, such as one kidney or one eye, there are utilitarian arguments against redistribution that may not be available to egalitarians. The prospect of having their bodies invaded and body parts forcibly removed would cause considerable insecurity among many nondisabled people. People would go to great extremes to avoid or escape the obligation to donate body parts, giving rise to heavy avoidance and enforcement costs. Utilitarianism can take such consequences of body-part redistribution into account; egalitarianism cannot take them into account as long as people with no kidneys or no eyes would still be worse off than people with two kidneys or two eyes.
Opponents of utilitarianism also sometimes conjure up examples in which utilitarianism would supposedly favor the (p.59) redistribution of body parts that are necessary to life; in the famous Transplant case, a surgeon can secretly kill one patient and use his organs to save the lives of five other patients.6 I refer to this as the Karlovian Transplant Case, because one may imagine Boris Karloff in the role of the surgeon (“What I do is for the greatest benefit of humanity. Come, Igor!”).
Once again, the Karlovian Transplant Case may be a problem for utilitarianism. But as noted below in Chapter 11, it is a worse problem for egalitarianism: egalitarianism might require a doctor to kill five patients in order to save one patient.
In any event, I am concerned in this book with egalitarian theories that do not support leveling down or the redistribution of body parts. Therefore, the most extreme form of egalitarianism I consider is a maximin egalitarianism that limits itself to the distribution of material resources. A maximin welfare egalitarianism would distribute material resources so as to maximize the welfare level of those who have the least welfare, regardless of whether those who have more welfare would benefit more from additional resources. A maximin resource egalitarianism would distribute material resources so as to maximize the resource level of those who have the fewest resources, regardless of whether those who have more resources would benefit more from additional resources.
The term “egalitarianism” has been given different meanings by different distributive theorists. In recent years some theorists have taken to using the term to refer only to a theory that would “level down,” that would reduce the welfare of the better-off without doing anything for the worse-off, or that would reduce everyone's welfare to achieve greater equality. I do not adopt this usage, as it would exclude most distributive theorists who have considered themselves egalitarians over the years.
What about egalitarian theories that are less extreme than maximin egalitarianism? Mixed egalitarian-utilitarian theories are sensitive to relative benefit to some extent. They give some preference to those who are worse off, but they still distribute resources to those who are better off if the better-off can derive sufficiently more benefit from the contested resources. I take up the contest between utilitarianism and mixed egalitarian-utilitarian theories in Chapter 9. At this point, however, and throughout most of this book, I attempt to isolate the three theories at issue—resource egalitarianism, welfare egalitarianism, and utilitarianism—so that they reflect three completely separate distributive criteria: help those who have fewer resources, help those who have less welfare, and help those who can most benefit. When I use the term “egalitarianism” in this book, I generally do not have in mind a theory that mixes egalitarian and utilitarian elements.
Another way in which an egalitarian theory can be more moderate than maximin egalitarianism is if there are intermediate claimants for resources, between the best-off class and the worst-off class. Suppose we must choose, under welfare egalitarianism, whether to raise considerably the welfare of the class with the next-to-least welfare, or whether instead to raise only slightly the welfare of the class with the least welfare. Or, to make the situation even more difficult, suppose that the only way we can slightly raise the welfare of the lowest class is to reduce considerably the welfare of the next-to-lowest class. It could be argued that welfare equality itself, aside from utilitarian considerations, can justify avoiding a large welfare loss by the next-to-lowest class at the cost of giving up a slight welfare gain by the lowest class. If so, a greater-benefit criterion has been (p.61) given sway in welfare egalitarianism, purportedly on egalitarian grounds rather than utilitarian grounds.
Unfortunately for egalitarians, there is no obvious stopping point. Under one conception of welfare egalitarianism, everyone but the person or class with the most welfare has an egalitarian complaint.7 If we combine this view with the greater-benefit criterion, we get a distributive theory that is identical to utilitarianism in distributive matters involving everyone in society except the person or class with the most welfare. Resources would be distributed to the class with the next-to-most welfare, instead of the class with the least welfare, as long as the class with the next-to-most welfare could gain a slightly greater increase in welfare from those resources. While it is unlikely that anyone who calls himself an egalitarian would advocate such a system, there is no obvious way to limit the scope of the greater-benefit criterion once it is introduced on “egalitarian” grounds. We would probably end up with a theory that is identical in most respects to the mixed egalitarian-utilitarian theories considered in Chapter 9.
An egalitarian theory that incorporates some form of the greater-benefit criterion as a way of dealing with intermediate claimants for resources can achieve some plausibility and can sometimes be hard to distinguish from utilitarianism. However, if there are no intermediate claimants, such an apparently moderate theory collapses into maximin egalitarianism. Its sensitivity to relative benefit is lost, and it yields implausible results in cases involving disability. We can then see clearly that such an apparently moderate theory is plausible only to the extent it can coincide with utilitarianism. Later in this chapter I consider an example of this type, and I return to the issue of intermediate claimants in my discussion of Ackerman's work in Chapter 8. Throughout most of this book, however, I ignore (p.62) the possibility of egalitarian theories that use a greater-benefit criterion to deal with intermediate claimants. Because theories more moderate than maximin egalitarianism lose their distinctly egalitarian character, I focus on maximin resource egalitarianism and maximin welfare egalitarianism. So not only is maximin egalitarianism the most extreme form of egalitarianism I consider; it is also, for the most part, the least extreme form of egalitarianism I consider.
At the beginning of this chapter I suggested that egalitarian theories can be divided into three types: those that are so extreme as to lack initial plausibility; those that appear plausible until one considers how they deal with the disabled; and those that maintain plausibility in dealing with the disabled by borrowing elements of utilitarianism. I have now narrowed my focus, at least temporarily, to theories of the second type: maximin-egalitarian theories.
Included in the set of maximin-egalitarian theories are leximin-egalitarian theories. Suppose that it is not possible to give any further help to members of the worst-off class. Does a maximin-egalitarian theory then proceed to maximize the welfare or resources of the next-to-worst-off class? Many theorists have thought that it should do so; they have given the name “leximin” egalitarianism to maximin-egalitarian theories that concern themselves with successively better-off classes after doing all that they can for those who are worse off.
A leximin-egalitarian theory still gives absolute priority to the worst-off class, as long as anything can be done for that class. If the treatment of better-off classes has any effect whatsoever on the worst-off class (which is usually the case), a leximin-egalitarian theory becomes identical to a maximin-egalitarian theory that has no concern at all for those who are not worst off.
Also included within the set of maximin-egalitarian theories (p.63) are theories that consider it an egalitarian improvement to reduce the membership of the worst-off class. Suppose it is no longer possible to help every member of the worst-off class, but it is possible to help some members of that class, or to prevent some people from falling into that class. Should maximin egalitarianism seek to minimize the membership of the worst-off class? The leximin issue just addressed does not precisely answer this question. Even a maximin-egalitarian theory that was indifferent to the fate of everyone but those in the worst-off class could favor helping some of those in the worst-off class if it were impossible to help everyone in the class. And while a leximin theory certainly might consider it an egalitarian improvement to reduce the membership of the worst-off class, it need not do so.
To me it seems only logical that a maximm-egahtarian theory should seek to reduce the membership of the worst-off class if it cannot further help everyone in that class. I will not impose this view on my maximin-egalitarian friends, but I will take note, in subsequent chapters, of some situations in which this issue is relevant.
Resource Egalitarianism and the Disabled
Now let us return to our two main suspects: resource egalitarianism and welfare egalitarianism. Resource egalitarianism provides too little help to some who are disabled, and welfare egalitarianism provides too much help to some who are disabled. But the analysis of both theories is quite complicated. Both can be specified in different ways.
One element that divides resource-egalitarian theories is how ambitious they are in maintaining resource equality on an ongoing basis. The least ambitious resource-egalitarian theory (p.64) is what Dworkin has called the “starting-gate theory of fairness.”8 There is an initial equal distribution and no subsequent redistribution; one could say there is egalitarianism at T1 and libertarianism thereafter. At the other end of the spectrum is a theory like Rawls's, that continually redistributes income so as to maximize the resources of the least-advantaged class. In the middle are theories that call for an initial equal distribution and redistribution upon the death of old citizens and the birth of new ones.9 Or perhaps there is a redistributive social minimum, as with Rawls, but it is not available to people who had the opportunity to earn more than the social minimum and squandered that opportunity.
There is also a time dimension to the very definition of resource equality. Probably the predominant temporal approach of resource egalitarians is to equalize total resources over a complete life. Other possible approaches are to equalize average resources over a complete life, or to equalize resources in every year.
All such resource-egalitarian theories are vulnerable to the criticism that they distribute too few resources to disabled people who could benefit greatly from additional resources. To be sure, resource egalitarianism would actually help many disabled people in comparison with current distributive regimes. There are doubtless disabled people in many countries, perhaps all countries, who have fewer resources devoted to them than what a resource-egalitarian system would provide. But many disabled people would also doubtless do worse under any resource-egalitarian system than they do now.
In general, resource egalitarianism can compensate the disabled for an inability to work, but it cannot compensate them for medical expenses. The precise amount of help available to the disabled, in a resource-egalitarian system, depends (p.65) in part on the nature of the disability and on the version of resource egalitarianism that the system embodies. I begin by considering the help that a resource-egalitarian system can give to people with disabilities that are not life-threatening.
Resource egalitarianism will provide a minimum income. The disabled poor, as well as the nondisabled poor, will receive income subsidies to raise them to the level of the minimum income. In one respect, the disabled poor will actually receive a kind of preference. If the system is sensitive to issues of opportunity, some of the nondisabled poor will not receive the full minimum income; people of working age who fail to work will generally receive no income subsidy, or will receive a reduced income subsidy. The disabled poor will generally be immune from this penalty; if someone is unable to work because of disability, he will receive the full minimum income even without working.
The problem for resource egalitarianism is that if the disabled poor are compensated for medical and other disability-related expenses, they would receive more resources than nondisabled poor people who receive the full minimum income. Any special government expenditures on behalf of the disabled would tend to reduce the funds available to pay general income subsidies to the poor. By paying special subsidies to the disabled, a resource-egalitarian system would move away from resource equality.
Some resource inequalities are permitted in a resource-egalitarian system: those inequalities that raise the income of those who have least income. I assume, with most egalitarians, that a resource-egalitarian system will not achieve, or even aim for, full resource equality. Some market-driven inequality will be permitted in order to maintain incentives for production; there will be some limit to redistribution corresponding to (p.66) Rawls's difference principle. Could subsidies for the disabled be similarly justified under a productivity rationale?
An initial problem with the productivity rationale for subsidies to the disabled is that even a resource-egalitarian system that permits market inequalities might not permit differential subsidies. In many resource-egalitarian theories, there is a strong view that the government should not encourage or discourage, by subsidy or penalty, particular uses of resources. People should be free to do whatever they like with their equal resource allotments. But I will assume that the system evaluates subsidies to the disabled under the same criterion as market-driven inequalities: if a subsidy raises the minimum income affordable to the poorest citizens, the subsidy should be given.
Even so, the scope for redistribution to the disabled would be very narrow. Any dollar paid out to support the disabled would be a dollar less available for general income subsidies. So aid to the disabled would have to more than pay for itself in the amount of money generated for general income subsidies. This is a demanding standard; it would exclude even most aid to the working disabled and would especially preclude aid to the non-working disabled. In sum, people who have non-life-threatening disabilities would likely receive little help, in a resource-egalitarian system, beyond the general income subsidy.
For life-threatening disabilities, the scope of redistribution to the disabled may be greater. A resource-egalitarian system that seeks to equalize total resources over a complete life will seek to keep people from dying before they receive a minimum lifetime income. A complete-life total resource egalitarianism will be especially solicitous of young people who have life-threatening or terminal conditions. A person who dies young has usually received, by the time of death, far less than the lifetime minimum income. A complete-life total resource (p.67) egalitarianism will seek to avoid this great inequality by channeling extra resources to young people facing death. If the young people facing death have a medical condition that can be cured with expensive care, they may fortuitously receive enough resources to obtain a cure. If the young people facing death have a medical condition that cannot be cured, they may receive a large cash grant.
Here we can see, incidentally, something of a convergence between welfare egalitarianism and the version of resource egalitarianism that seeks to equalize total resources over a complete life. While the general problem of resource egalitarianism is that it distributes too few resources to the disabled, a complete-life total resource egalitarianism would distribute too many resources to some young people with terminal conditions. Suppose that a twenty-year-old poor person is terminally ill. Nothing can be done to prolong his life. He is receiving medical care, but it is not very expensive, and further care will not be very beneficial. In such a case, complete-life total resource egalitarianism would be moved to make a large cash grant to the terminally ill young person to bring him up to the minimum lifetime income before he dies. In a rich country, such as the United States, such a cash grant could total in the hundreds of thousands of dollars. This redistribution to the terminally ill young seems excessive. Of course, it is not as excessive as the redistribution that the terminally ill young might receive under welfare egalitarianism. The limit to resource-egalitarian redistribution would be the lifetime minimum income; the limit to welfare-egalitarian redistribution would be much higher, as discussed below.
In any event, the more serious problem, even under complete-life total resource egalitarianism, is not excessive redistribution to those with fatal and incurable medical conditions, (p.68) but inadequate redistribution to those with life-threatening but curable conditions. Suppose that a person has a life-threatening illness that can be successfully treated. If she has already received, throughout her life, more than the minimum lifetime income, she likely would receive no help under a resource-egalitarian system. The system would pay to keep her alive only if it could use her as a means to increase the guaranteed lifetime income for everyone, if it could more than make up for the cost of treatment by collecting future taxes from her.
Under other temporal versions of resource egalitarianism, there often would be even less scope for helping those with life-threatening conditions. Take, for example, a system that tries to equalize average yearly resources over a complete life rather than total resources over a complete life. Such a resource-egalitarian system would generally not provide help even to a young person with a life-threatening illness if she had already received, on average, more than the minimum income for the years she had been alive. As before, resource egalitarianism would help such a disabled person only if it could use her as a means to increase the average lifetime income for everyone.
Under every version of resource egalitarianism, poor disabled people who are uninsured or underinsured would be forced to spend their own money on highly beneficial medical care and other aids. After paying for these items, the disabled poor would be left with less money than the nondisabled poor. Some expensive life-saving and life-improving care, such as dialysis and heart surgery, might be completely beyond the reach of the disabled poor (and “poor” here includes people who were not poor to begin with, but who exhausted their resources to pay for medical care). Under resource egalitarianism, there would be unnecessary suffering and unnecessary death.
Many, I'm sure, would find resource egalitarianism appealing (p.69) if only it could provide greater help to the disabled, and in particular if it could subsidize the medical expenses of (at least) the disabled poor. Let us therefore look more closely to see if there is any way for a resource-egalitarian system to accomplish this objective. First, however, let us examine the impulse to provide additional resources to the disabled in a resource-egalitarian system. If you are a resource egalitarian, you believe that resources should be distributed equally regardless of differences in the benefit people get from resources, and regardless of differences in their level of welfare. Either you do not care about these differences, or you believe they cannot be measured, or you believe they are irrelevant to distributive justice. Why, then, would you want to distribute extra resources to the disabled? Is it not obvious that it is because the disabled would benefit more from additional resources, or because they have less welfare, or both?. On the face of it, the impulse to distribute additional resources to the disabled is an impulse to move away from resource egalitarianism toward utilitarianism, or welfare egalitarianism, or some mix of the two.
Could a resource-egalitarian system attend to the medical needs of the disabled by giving everyone an “equal” entitlement to medical care rather than an equal amount of cash? This question is relevant to Dworkin's system of hypothetical insurance, discussed in Chapter 7, and also to the modified difference principle that Rawls advances in Justice as Fairness: A Restatement, discussed in Chapter 6. The short answer to the question is no: universal medical coverage is not a genuine resource-egalitarian scheme. Under any system of universal medical coverage, benefits will be distributed unequally. Those who obtain covered medical treatment will receive more resources than the nondisabled poor; often, they will receive more than the lifetime average income. As a system of universal (p.70) medical coverage distributes medical resources unequally, “equal entitlement to medical care” means unequal entitlement to resources.
It might be argued that while the distribution of resources in a system of universal medical coverage is unequal when benefits are received, it is equal “ex ante,” when people receive the entitlement to medical care, at birth or before. At that time, it is not known how much medical care people will need. But this argument proves too much. With such an “ex ante” perspective, we could say that any unequal distribution of resources, based on any eventuality of life, is really an equal distribution of resources.
We could say, for example, that the government maintains equality of resources when it pays a one million dollar bonus to every person who grows to be taller than seven feet. Suppose the government entered into an arrangement with insurance companies in which it made an equal premium payment, on behalf of all newly conceived persons, that guaranteed a payment of one million dollars to every person who exceeded the height of seven feet. All of this rigmarole would not change the reality that the government had instituted a resource inequality benefiting people taller than seven feet (and also, perhaps, benefiting insurance companies).
It might be objected that no one ever really has an equal chance of exceeding seven feet in height. Some people have genes that predispose them to be tall, and of course men are usually taller than women. But a similar objection could be leveled against the pretense that people have an equal chance of receiving medical resources under a system of universal medical coverage. Some people have genes that predispose them to suffer from expensive illness.
Universal medical coverage is a sensible idea, whereas (p.71) million-dollar bonuses for people who are taller than seven feet is a silly idea. But that does not mean that universal medical coverage is a kind of equality of resources; it is a sensible, rather than silly, inequality of resources.
Let us try another tack in our attempt to make resource egalitarianism more favorable to the disabled. Maybe medical care is a special use of resources, one that should not be counted when we determine a person's resource share. In Chapter 4, I discussed two important reasons why utilitarianism is inclined to help the disabled. First, medical care and other disability aids can often greatly benefit the disabled. Second, after spending some of their own money on these highly beneficial items, poor disabled people have less money than poor nondisabled people who started out with the same income; therefore, poor disabled people would benefit more from additional money.
Resource egalitarianism cannot rely on the first of these two grounds: It is committed to distributing resources equally regardless of whether some people benefit more from additional resources than other people. But what about the second? If a poor disabled person has to pay medical expenses, isn't she really poorer than a nondisabled person with the same income?
The problem is that the second ground alone does not distinguish between medical expenses and other expenses. A poor person who spends half his income gambling at a casino is left with the same resources as a poor person who spends half his income on medical care. A utilitarian system will subsidize medical care and not gambling because medical care is more beneficial to people. Resource egalitarianism cannot make this distinction.
The negative verdict on resource egalitarianism stands. Resource egalitarianism means unnecessary suffering and unnecessary death. Suffering and death will of course occur under (p.72) any system, including utilitarianism. However, utilitarianism is committed to preventing these harms if prevention can be achieved without causing greater harm. Resource egalitarianism would permit suffering and death even if the cost of avoiding them was relatively low.
I have so far neglected an argument against resource egalitarianism that is commonly put forward by utilitarians and others. Resource egalitarianism, its opponents argue, can result in redistribution to the poor that is vastly resource-inefficient and also welfare-inefficient.10 Suppose a resource-egalitarian system with market-driven inequalities that have been thought necessary to maximize the resource holdings of the poorest class. Suddenly it is discovered that with massive redistribution from the great majority of citizens, it is possible to achieve a further, very tiny increase in the resources of the poorest class. Even though the new scheme is vastly inefficient, it must be put into place: under a resource-egalitarian system, even the tiniest gain to the poorest class must be seized, whatever the cost to others.
I actually do consider this a telling argument against resource egalitarianism and in favor of utilitarianism. Utilitarianism itself would permit some resource-inefficient redistribution. Because the poor benefit more from additional money than do the rich, utilitarianism would endorse redistributive schemes in which there is some leakage of resources in the course of redistribution from rich to poor. But utilitarianism would not permit welfare-inefficient redistribution; it would only endorse redistributive schemes in which, despite the leakage of resources, the poor still gain more welfare than the rich lose (and there are no other negative effects of redistribution to offset the net gain in welfare). A redistributive scheme that is vastly resource-inefficient would probably be welfare-inefficient (p.73) as well, especially if the poor have attained a moderate standard of living.
Nevertheless, there are drawbacks to the argument conjuring up a sudden and vastly inefficient resource-egalitarian redistribution. Such arguments have a bad pedigree. They recall false predictions of disastrous effects made by opponents of greater redistribution. The most prominent recent example in the United States involved President Clinton's 1993 budget, which raised income taxes on richer Americans. Republican opponents of Clinton confidently predicted disaster, but instead the 1993 tax hike on the rich ushered in a period of great prosperity. There are issues here that go far beyond the scope of this book, but the upshot is that the scenario of a sudden and vastly inefficient redistribution is one that might possibly occur under resource egalitarianism, not one that would certainly or even probably occur.
The intuitive selling point of utilitarianism is that often it seems right to distribute resources to those who will most benefit; when opposing theories require distribution away from those who will most benefit, they often seem wrong. But to fully evoke utilitarian moral intuition against an opposing theory, we must also have the appropriate hedonic intuition: we must believe that the opposing theory really does distribute resources away from those who will most benefit. As argued in Chapter 2, this hedonic conclusion must be based on the facts of the case, not on bald stipulation.
The hypothesis of a sudden and vastly inefficient redistributive scheme purports to be a case where resource egalitarianism takes resources from those who would benefit more (the rich), destroys most of those resources, and distributes a tiny remnant to those who would benefit less (the poor). We (p.74) may not be convinced, however. Because the assumption of vast inefficiency is speculative and freighted with associations to discredited claims, it may be difficult to summon up the hedonic intuition that the rich really do lose more, in welfare, than the poor gain.
In the area of disability, by contrast, the interpersonal comparisons can be immediately convincing. Resource egalitarianism would deny additional resources to some disabled people who would die or be in horrible pain without those additional resources; resource egalitarianism would distribute the resources instead to the nondisabled poor, who would benefit less. We do not have to hypothesize a vast leakage of resources in the course of redistribution; we do not have to hypothesize any leakage at all.
Resource egalitarianism would make more sense if people were basically alike and their life situations were basically similar. In that event, people would get more or less the same benefit from resources. Even if there were some minor differences, it would not be worthwhile to investigate them for purposes of varying resource holdings.
Interestingly, resource egalitarians do not argue for their theory on the ground that people are alike; they argue for it on the ground that people are very different. People are so, so, different, the resource egalitarian tells us, that it is impossible to make interpersonal comparisons of welfare among them. How can we tell whether a pacific nature lover would get more benefit from additional resources than a competitive sportsman?11 All we can measure about people, with respect to distributive justice, is how much material resources they have. Therefore, we might as well divide resources equally.
The problem for resource egalitarianism is that it breaks down when it encounters people who really are convincingly (p.75) different. When we add to the pacific nature lover and the competitive sportsman a third candidate for additional resources, one who would die or be in horrible pain without them, interpersonal comparison is suddenly quite easy. So while resource egalitarians argue for their theory on the ground that people are different, the intuitive appeal of resource egalitarianism seems to rest on our assumption that most people are really the same. When that assumption is convincingly refuted, as with disabled people who could clearly benefit more than others from additional resources, the intuitive appeal of resource egalitarianism disappears.
Welfare Egalitarianism and the Disabled
If resource egalitarianism would distribute too few resources to disabled people who could benefit greatly from additional resources, welfare egalitarianism would distribute too many resources to disabled people who could benefit only infinitesimally from additional resources. This criticism applies to all varieties of welfare egalitarianism.
The theorists that I class as welfare egalitarians do not all describe themselves as welfare egalitarians. Sen advocates equality of “capabilities” to achieve “functionings,”12 and G. A. Cohen advocates equality of “access to advantage.”13 Richard Arneson at one time advocated equality of “opportunity for welfare.”14
These distinctions do not matter much in the context of distribution to the disabled. All such theories are vulnerable to the criticism that they set no acceptable limit on redistribution to the disabled. Equality of capabilities to achieve functionings would result in excessive redistribution of resources from the nondisabled to the disabled, as would equality of access to advantage, as would equality of opportunity for welfare.
(p.76) A number of welfare-egalitarian theorists, including Sen, Cohen, Arneson, and Roemer, stress issues of opportunity and responsibility. The idea, once again, is that we will raise the welfare of people who did not have the opportunity to raise their own welfare. But of course, denying redistribution to people who are responsible for their own low welfare does not limit redistribution to the disabled, except perhaps in circumstances where we might blame the disabled for causing their own disability. Opportunity and responsibility may be very relevant to some issues treated by welfare egalitarians, but these concepts do not help welfare egalitarians to avoid unlimited redistribution to those disabled people who have very low welfare through no fault of their own. Just as a sensitivity to opportunity does not enable resource egalitarians to avoid inadequate redistribution to the disabled, a sensitivity to opportunity does not enable welfare egalitarians to avoid excessive redistribution to the disabled.
Extent of Welfare-Egalitarian Redistribution
Many theorists, including some welfare egalitarians, agree that a pure welfare egalitarianism would entail excessive redistribution to the disabled. Assessing the manner and extent of such redistribution, however, is a complicated matter. Before addressing this matter I will describe three different kinds of redistribution that now benefit the disabled in many countries, and that could have a place under any welfare-egalitarian system.
First, there is unconditional income or wealth redistribution; the disabled are given additional resources to use however they see fit. In a hypothetical initial distribution, this might (p.77) be the only kind of redistribution or (more accurately) more-than-equal distribution to the disabled. The disabled would get extra manna, or clamshells, or dollars, or whatever.
Second, there is medical redistribution, which would involve medical research and the provision or reimbursement of medical and rehabilitative goods and services. Third, there is environmental redistribution; society might spend resources to alter the public and private environment of people with disabilities, so as to improve their lives in various ways. These are doubtless rough categories.
I will now list four limits to redistribution under welfare egalitarianism. If any one of these limits is reached, redistribution will stop:
3. Equalization of welfare
To simplify an exploration of these limits, imagine that there is only one disability in society; its victims suffer constant severe pain. There is no currently known cure or alleviative treatment for the disability or the pain. But while additional resources cannot currently relieve the pain of the disabled, additional resources can improve their welfare in much the same way additional resources would improve anyone's welfare: by enabling them to get things they want. Also, a cure or treatment may be possible, though none has yet been found.
Now a maximin-welfare-egalitarian system (which, it will be remembered, I am treating as synonymous with welfare egalitarianism) would not redistribute resources from the nondisabled to the disabled if the disabled would not benefit at all from those resources. If there is a point at which additional (p.78) resources would not improve the welfare of the worst-off class at all, even though they continue to have the least welfare, redistribution would stop. This would be the point of satiation.
Would there be any point of satiation in the scenario described? In part, the answer depends on what conception of welfare the welfare-egalitarian system follows. It may be a preference-based system, under which a person's preference (or informed preference) for additional money means that additional money would increase his welfare. Under such a system, there probably would be no satiation; there are few people, of any income level, who would say they have no desire for additional money.
On the other hand, the governing conception of welfare may be a hedonic one. Under a hedonic conception of welfare, it is somewhat plausible that the worst-off disabled class could reach a point of satiation, at least with respect to unconditional income redistribution. Some social scientists, including Robert Lane, question whether present-day wealthy people in advanced societies get any increase in happiness, however slight, from a marginal dollar.15 Presumably, the same view about hedonic satiation could be advanced about disabled wealthy people as about nondisabled wealthy people.
But under a hedonic view, it is unlikely there would be any plausible point of satiation with respect to medical redistribution. Even if there is no current cure for some painful disability or disease, a cure can always be sought. All medical research projects that offered some possibility of finding a cure or ameliorative treatment would increase the expected welfare of the worst-off disabled class. Such an increase in expected welfare, however slight, would justify continued redistribution under the hedonic account.
(p.79) Satiation is not a very promising limit to redistribution under welfare egalitarianism. Let us move on to the counter-productivity limit. Possibly the redistribution of resources to the disabled will reach such a point that any attempt at further redistribution will actually decrease their welfare. This maximin-welfare-egalitarian limit to redistribution is analogous to the maximin-resource-egalitarian limit to redistribution contained in Rawls's difference principle. But the counterproductivity limit does not operate the same way under welfare egalitarianism as under resource egalitarianism. For example, it is commonly assumed that too-high taxes on the rich, in a resource-egalitarian system, would be counterproductive from a redistributive standpoint, reducing work and investment incentives. In a welfare-egalitarian system, by contrast, redistributive taxation could be so extreme as to leave most or all non-disabled people poor, as long as they had higher welfare than the worst-off disabled. A redistributive scheme so extreme that it reduced rich people to poverty might generate work and investment incentives, on balance, rather than depressing them, at least as long as nondisabled people could benefit to some extent from their work and investment. Obviously, we are now in the realm of ideal theory (a realm we have long since entered), so we need not consider how to enforce a draconian system of redistribution that could reduce the vast majority of citizens to poverty.
The third limit to redistribution under welfare egalitarianism is equalization of welfare. Possibly the worst-off disabled class will receive sufficient additional resources so that they are no longer worst off. At that point redistribution will stop, at least as between the disabled and those nondisabled people with whom they have reached welfare equality.
(p.80) It is often assumed, in discussions of disability and distributive justice, that some disabling and painful conditions are uncompensable.16 Suppose that the test of compensation is preference-based, and the issue is whether the worst-off disabled would say (if they answered honestly) that a certain amount of redistribution could make them indifferent between disability plus redistribution and a hypothetical state of no disability and no redistribution (or whatever lesser redistribution they would receive in a nondisabled state). Different disabled people might view this issue differently. The disability we are now imagining is one that causes constant severe pain. This disability of course has analogues in the real world; there are people in the real world, not just in philosophers' examples, who suffer from constant severe pain. It seems unlikely that many such people, in the real world or in our imagined setting, would consider any amount of resources to be adequate compensation for constant severe pain. Possibly no one would think such a condition compensable, at least as compared to a nondisabled state with some modicum of resources.
As noted, however, in a welfare-egalitarian system, non-disabled people need not have a modicum of resources. There is a way of achieving welfare equalization that relies not so much on compensation of the disabled as on immiseration of the nondisabled. It is possible that as more and more resources are redistributed to disabled people, they do get some slight benefit from those resources, but nondisabled people simultaneously experience enormous decreases in welfare—sometimes going hungry, for example. This would not be a process of “leveling down,” which we have ruled out as inconsistent with maximin welfare egalitarianism. Everyone would not be made worse off, and the better-off nondisabled class would not be made worse off unless doing so could increase the welfare (p.81) of the disabled class. Still, the equalization of welfare would occur not because of the slight benefits the disabled would continue to gain from redistribution, but because the nondisabled would be made miserable.
This may be an auspicious point at which to reintroduce the problem of intermediate classes. Suppose there is a class of poor nondisabled people who are made so miserable by welfare-egalitarian redistribution that their welfare declines to the same level as the now-rich disabled. But suppose that other nondisabled people in society are not poor; there is a class of rich nondisabled people.
At this point it might be argued that the immiseration of the nondisabled poor is not wrong because it produces misery, which is the utilitarian position. Rather, it might be said, the immiseration of the nondisabled poor is wrong because it creates greater inequality between the nondisabled poor and the nondisabled rich. To evaluate this argument, imagine now that there are not two or several nondisabled classes, but one non-disabled class; all nondisabled people are equally miserable. Would it be acceptable if all nondisabled people, instead of some nondisabled people, were immiserated en route to welfare equality? I think not. This exercise demonstrates again that egalitarian theories more moderate than maximin egalitarianism depend, for their apparent moderation, on the presence of intermediate classes. If there are intermediate classes, such theories can arguably coincide with utilitarianism to some extent. However, if there are only two classes, the apparently moderate theories become indistinguishable from maximin egalitarianism and subject to the same utilitarian objections.
A final possible limit to redistribution under welfare egalitarianism is sympathy. This is counterproductivity with a twist. Suppose that the disabled continue to benefit, however (p.82) slightly, from more and more resources, as far as their own personal well-being is concerned. The nondisabled poor are suffering substantial welfare losses as a result of redistribution, but they have not yet become so miserable that their welfare is no higher than that of the disabled rich. Before the nondisabled poor are actually immiserated to the level of welfare equality, the disabled rich might reach a point where further immiseration of the nondisabled poor would reduce their own welfare, because of sympathy, more than additional redistribution would increase their welfare. Even though the disabled rich still have less hedonic welfare than the nondisabled poor, the disabled rich would prefer that they not receive additional resources from the nondisabled poor, because the nondisabled poor could derive so much more benefit from those resources. Welfare-egalitarian redistribution to the disabled would then end because the disabled had partially internalized the utilitarian greater-benefit criterion.
While welfare egalitarianism clearly has counterintuitive implications for the distribution of resources between disabled and nondisabled people, the precise distributive results of a welfare-egalitarian theory are hard to specify, even in a relatively simple model. Adding some real-life complexity introduces more problems. The biggest question for welfare egalitarianism is: who has least welfare? There are in reality a number of candidates. Perhaps the people with least welfare are the ones who appear to be in the most physical pain or who are terminally ill. Perhaps the least-welfare distinction goes instead to depressed people who have seriously attempted suicide, thus revealing a (p.83) low amount of life satisfaction. Or perhaps those who are severely intellectually disabled have the least welfare.
As with resource egalitarianism, there is a temporal dimension to the definition of welfare equality and inequality. Is welfare is to be equalized across simultaneous points in time, across complete lives, or in some other way? A simultaneous-point approach could invest enormous resources on behalf of people who are now in the most pain, or are now on the point of death, even if the benefit or expected benefit to them is tiny. On the other hand, a whole-life approach could ignore the people who are now most in pain, even if they might benefit substantially from additional resources, unless and until their pain reduces their whole-life welfare sufficiently.17
There is also the issue of how narrowly the least-welfare class should be defined. In the Introduction, I suggested that young people with terminal cancer might be considered to have the least welfare in a welfare-egalitarian system. I made this suggestion because young people with terminal cancer are worse off than others in several ways: they are currently suffering, they will suffer for the rest of their lives, and their lives will be short. So the class of young people with terminal cancer is itself at the intersection of at least two classes that might claim to be worst off: people suffering from pain and young people with a terminal medical condition. We might narrow the least-welfare class further, for example by making it young people with terminal cancer who suffer from severe “breakthrough” pain (pain that breaks through medication) and also suffer from severe depression. We could in fact attempt to establish individual welfare rankings, identifying the one person with least welfare, the one person with next-to-least welfare, and so on. But I will assume a class-based approach.
(p.84) Once the least-welfare class is defined, welfare egalitarianism will lavish resources on that class unless and until one of the limits to redistribution, discussed above, is reached. As there will be several classes of severely disabled people, the equalization limit may be easier to reach in real life than in the single-disability model I presented. Possibly, after the expenditure of vast sums on the least-welfare disabled, another class will become the least-welfare disabled and will in turn receive massive redistribution. But it is also possible that the equalization limit will never be reached, and many classes of severely disabled people will never be considered worst off. It might be determined that no amount of redistribution to young people with terminal cancer can raise their welfare to the level of people with severe intellectual disability (or vice versa).
Suppose young people with terminal cancer are identified as the worst-off class. The considerations bearing on redistribution to this class would be similar to those I outlined above, with some additional complications because of the life-shortening nature of the disability. There would be massive medical redistribution. Vast sums would be devoted to medical research to find cures and treatments for terminal cancer. Vast sums would also be devoted to medical treatment to prolong life and relieve pain.
Assume that a complete cure is not found. There continue to be young people with terminal cancer, and they continue to be identified as the worst-off class. Further medical redistribution would not benefit the cancer victims significantly; the expenditure of further vast sums would only obtain for some of them tiny life extensions, or a tiny increase in the probability of finding additional cures and treatments. At this point society might still continue with further medical redistribution, since any benefit to the worst-off class, however, tiny, justifies (p.85) further redistribution. Or there might be a shift to unconditional wealth redistribution. The worst-off might receive large cash grants, over and above the cost of medical treatment. The case for cash grants is especially strong if the worst-off class is comprised of people who will die soon. In that event, it can be known that many members of the worst-off class will get no benefit at all from medical research projects, even if they are to any extent successful.
In a rich society, such as the United States, members of a relatively small worst-off class, such as young people with terminal cancer, could expect to receive cash grants in the millions of dollars. What would happen to the unspent money when they die? Possibly their relatives would be allowed to inherit; it might increase the welfare of terminally ill young people to know that their relatives will become rich.
In our one-disability model, we saw that the nondisabled poor are valuable, in a welfare-egalitarian system, only as a means to increase the welfare of the severely disabled (unless the nondisabled poor should become so miserable that they no longer have more welfare than the severely disabled). Where there are multiple classes of disabled people, this purely instrumental status extends to many disabled people as well. Those disabled people who are not considered to have least welfare would actually do worse under welfare egalitarianism than under resource egalitarianism. While resource egalitarianism can be faulted for failing to grant the disabled poor more resources than the nondisabled poor, resource egalitarianism would at least guarantee to the disabled poor the same minimum income to which everyone in society is entitled. Under welfare egalitarianism, there would be no income guarantee to anyone who is not worst off. Those disabled people who could not work might not be able to survive. If the worst-off class is (p.86) defined as young people with terminal cancer, old people with terminal cancer might receive no help at all.
Thus far I have casually assumed that every member of the worst-off class under welfare egalitarianism would be a disabled person. This assumption, though widely shared by distributive theorists, may not be strictly accurate.18 Nevertheless, any inaccuracy would not weaken the case against welfare egalitarianism.
If the worst-off class is defined as those who are in the most physical pain, then every member of the worst-off class would indeed be a disabled person; under my broad definition of disability, physical pain is always a disability. Similarly, if the worst-off class is defined as the terminally ill young, every member of that class would be disabled; under my broad definition of disability, illness is always a disability. However, if the worst-off class is defined as people with the least life satisfaction, or people who are on the point of committing suicide, not every member of that class would be a disabled person. Those who are on the point of suicide because of a broken hedonic regulatory system (that is, chemical depression) would be disabled, but those who are on the point of suicide because of a broken heart would not necessarily be disabled.19
In any event, under any sincere definition of the worst-off class, welfare egalitarianism would end up distributing resources to people who derive hardly any benefit from those resources, instead of to people who could benefit greatly from additional resources. Either people in the worst-off class can be raised out of that class, or they cannot. If they cannot be raised, welfare egalitarianism will concentrate all its redistributive energy on them, as long as they can benefit at all from additional resources. If they can be raised, welfare egalitarianism (p.87) will raise them and then move on to the next candidates for redistribution, ultimately settling on a class of people who cannot be raised to equality of welfare with the class above them, but can continue to benefit from additional resources. Thus, for example, when welfare egalitarianism is first instituted, the worst-off class might include both people who are dying of starvation and young people with terminal cancer. Assuming that society is at least moderately wealthy, welfare egalitarianism will raise starving people out of the worst-off class by providing them with the means to obtain food. Welfare egalitarianism will then concentrate all its redistributive energy on young people with terminal cancer, who cannot be so easily helped.
In Chapter 2, I claimed that those who summon up the specter of the utility monster as an argument against utilitarianism are making a deceptive and illegitimate argument; they are in reality evoking utilitarian intuitions and turning those intuitions against utilitarianism. We do not really believe that a person could be a utility monster, that he could gain more from all sacrifices of others than those others lose. So when we decide that the supposed utility monster should not receive all the world's resources, this intuitive moral judgment is a utilitarian one; it is based on the assumption that sacrificing everyone to the utility monster would lead to a massive reduction in welfare, rather than (as unconvincingly stipulated) an increase in welfare.
The true and legitimate home of the utility monster is not utilitarianism, but welfare egalitarianism. Under welfare egalitarianism, those who are worst off have a massive claim on the resources of others. This claim does not depend, as it would under utilitarianism, on the unconvincing stipulation (p.88) that the worst-off would actually benefit more from all the sacrifices of others than those others lose. The claim of the worst-off under welfare egalitarianism is good as long as the worst-off derive any benefit from the sacrifice of others, while still remaining worst-off. It is quite easy to believe that the worst-off would derive some benefit from massive sacrifices of those who are better off. Hence, welfare egalitarianism turns the worst-off into credible utility monsters.
The implausibility of welfare egalitarianism is so great that one is moved to modify its strictures—not in order to render it plausible (that is impossible), but in order to lessen its absurdity. Possibly a welfare-egalitarian system will not choose between contending conceptions of welfare, but will compromise between them. If some say that young people with terminal cancer have least welfare and others say that those with severe intellectual disability have least welfare, perhaps the system will split its massive redistribution between these groups.
Perhaps also the absolute priority of maximin egalitarianism will be relaxed, at least as to people who are severely disabled. The system could extend some help to classes of disabled people other than those with least welfare, and could adopt something like a utilitarian greater-benefit criterion to make tradeoffs. The system could decide, for example, that it is more important to give a significant benefit to older people with terminal cancer than to give only an insignificant benefit to younger people with terminal cancer. The farther a welfare-egalitarian system travels in the direction of utilitarianism, the more plausible it will be. But welfare egalitarianism can travel very far in the direction of utilitarianism and still remain implausible. A welfare-egalitarian system could extend help to many classes of disabled people and still maintain the nondisabled poor in a state of destitution.
A common resource-egalitarian critique of welfare egalitarianism is that welfare egalitarianism would not deal satisfactorily with the happy disabled. This is actually a resource-egalitarian comeback to the welfare-egalitarian critique of resource egalitarianism. Welfare egalitarians join utilitarians in complaining that resource egalitarianism would not allocate any more resources to the disabled than to the nondisabled. Resource egalitarians respond: “Aha! What about disabled people who are happier than nondisabled people? Welfare egalitarianism would not allocate additional resources to those happy disabled people; it would allocate fewer resources to them.” A common example involves Tiny Tim and Scrooge.20 Tiny Tim is happier than Scrooge, even though Tiny Tim is poor and disabled and Scrooge is rich and nondisabled. Doesn't welfare egalitarianism therefore have to redistribute resources from Tiny Tim to Scrooge instead of the other way around?
I would not dismiss the interpersonal comparison in this example as something that could only be dreamed up by Wiseacre (a character I introduced in Chapter 2). Although the disabled have on average less welfare than the nondisabled, there are undoubtedly many disabled people who have more welfare than many nondisabled people. On the other hand, it seems unlikely that in any practical welfare-egalitarian system (if such a thing could exist) the purposes of welfare egalitarianism would be served by seeking to determine which disabled people were temperamentally happy and should therefore receive fewer resources.
In any event, however much the happy disabled are a problem for welfare egalitarianism, they are less of a problem (p.90) for utilitarianism. Utilitarians are concerned with how much welfare can be raised, not how low it is. The violinist Itzhak Perlman is a victim of childhood polio. He can walk only painfully and with difficulty, using elbow crutches.21 Yet I would not claim that I have more welfare than Itzhak Perlman; if I had to guess, I would say it is probably the other way around. I would claim, however, that anything which enabled polio victims to move around more easily would benefit them greatly, even if they already enjoyed a high level of welfare. Utilitarians can assume with some confidence that if distributive policy leads to a cure or substantial amelioration of disability, virtually every affected disabled person will experience an increase in welfare.
In the Tiny Tim example, utilitarianism would distribute additional resources to Tiny Tim, even though he has more welfare than Scrooge, because he can also benefit more from resources than can Scrooge. To be sure, it is possible to suppose that a rich nondisabled person (such as Scrooge) can benefit more from resources than a poor disabled person (such as Tiny Tim), just as it is possible to suppose that a rich nondisabled person is less happy than a poor disabled person. But the first supposition is far less plausible than the second. We could easily be convinced, after meeting a rich nondisabled person and a poor disabled person, that the poor disabled person is happier; it is unlikely that we could be convinced that the poor disabled person would benefit less from additional resources.22
The analysis of resource egalitarianism and welfare egalitarianism has proved to be complicated, but the negative verdicts on both these theories stand. All versions of resource egalitarianism give too little to some disabled people, and all (p.91) versions of welfare egalitarianism give too much to some disabled people. Nevertheless, we have uncovered some surprises. A thoroughgoing welfare egalitarianism is very unfavorable for disabled people who are not considered to have the least welfare, as well as for the nondisabled poor. A resource egalitarianism that seeks to equalize total resources over a complete life actually gives too much to some disabled people.23 All this is an example of how egalitarianism, lacking a sensitivity to relative benefit, oscillates between inadequate redistribution and excessive redistribution to various classes.
It is also interesting that of the three contending theories—utilitarianism, resource egalitarianism, and welfare egalitarianism—only utilitarianism can even potentially support national health insurance. In the previous chapter, I noted that there is a strong utilitarian argument for national health insurance, and that utilitarianism would at the very least subsidize the medical expenses of the poor. Neither resource egalitarianism nor welfare egalitarianism can even get as far as subsidizing the medical expenses of the poor. Resource egalitarianism, as we have seen, cannot in general allow the disabled poor more resources than the nondisabled poor. Welfare egalitarianism lavishes medical care on those disabled people who have least welfare, but cannot in general provide for the medical needs of others. Both egalitarian theories can help people outside their favored groups only as a means to help people inside the favored groups. This instrumental approach can justify some subsidized medical care, on productivity grounds, but it cannot come close to a general subsidy for medical care, or even a medical subsidy for the poor.24 Utilitarianism, by contrast, treats every person (or at least, every person's welfare) as an end in itself, not solely as a means.
Before leaving behind the taxonomy of egalitarian theories, we should consider marginal egalitarianism, a theory that would not be acknowledged as egalitarian by everyone. Marginal egalitarianism tells us to distribute equally whatever resources are available for redistribution at a given point in time. I take the term “marginal egalitarianis” from Douglas Rae, who is one of the few scholars to analyze it.25
Suppose that we have two dollars to distribute among two claimants: Donald Trump and a homeless person (this is assuming that Donald Trump has not himself been reduced to homelessness through the vicissitudes of the real estate market). The three theories with which I am most concerned—utilitarianism, resource egalitarianism, and welfare egalitarianism—would all counsel us to give two dollars to the homeless person and nothing to Donald Trump. Marginal egalitarianism, however, would counsel us to give one dollar to each.
Marginal egalitarianism is closer to resource egalitarianism than to welfare egalitarianism. What I am calling marginal egalitarianism is actually marginal resource egalitarianism. There could theoretically be a marginal welfare egalitarianism; such a theory would tell us to increase, by equal amounts, the welfare of all those whose welfare we can increase. If we are distributing money between a rich person and a poor person, marginal welfare egalitarianism would tell us to give more money to the rich person, because the poor person can get a larger welfare increase from any given amount of money.
Marginal welfare egalitarianism has no intuitive appeal. However, marginal resource egalitarianism (what I am calling simply marginal egalitarianism) does have some intuitive appeal, (p.93) even perhaps in extreme examples such as the Trump/homeless person example, and more so in examples that are less extreme. When it is aligned with another theory of distributive justice, marginal egalitarianism can make the result favored by that theory more appealing. The most frequent beneficiaries of alignment with marginal egalitarianism are resource egalitarianism and welfare egalitarianism. But as demonstrated by the following example, marginal egalitarianism can be aligned with utilitarianism against one of the other egalitarian theories.
Suppose there are two patients suffering from persistent pain. Patient A is in pain 15 hours a day, and patient B is in pain 20 hours a day. We have two pills that will, somewhat magically, provide permanent but partial relief from pain to both patients, to a different extent. If patient A receives 1 pill, he is relieved of 10 hours of pain per day; if he receives 2 pills, he is relieved of an additional 30 minutes of pain per day. If patient B receives 1 pill, he is relieved of 1 hour of pain per day; if he receives 2 pills, he is relieved of an additional 1 hour of pain per day.
Here, the welfare-egalitarian course is to give both pills to patient B, for 2 hours of pain relief from 20 hours of pain per day. The utilitarian course is to give 1 pill to each patient, so that patient A gets 10 hours of pain relief from 15 hours, and patient B gets 1 hour of pain relief from 20 hours (This is the utilitarian course if we assume that patient A benefits more by being relieved of 10 hours from 15 than patient B benefits by being relieved of 1 hour from 19).
The marginal egalitarian course is in this case the same as the utilitarian: 1 pill to each patient. Therefore, if we think it right to distribute 1 pill to each patient, we cannot be sure whether we are endorsing utilitarianism, or marginal egalitarianism, or some combination of the two.
(p.94) It is important to notice when marginal egalitarianism is in the picture. For example, resource egalitarians often begin their argument by hypothesizing the initial distribution of unowned material resources to people who have no other material resources. In this setting, resource egalitarianism is aligned with marginal egalitarianism: both tell us to distribute the unowned resources equally. If we think this approach is right, we may be responding to the intuitive appeal of marginal egalitarianism rather than, or in addition to, the intuitive appeal of resource egalitarianism.
Whence the surprising intuitive appeal of marginal egalitarianism? In part, it may be due to laziness of moral thought. An equal division of whatever is being distributed is a very prominent and easy-to-administer solution. If we seize upon that solution, we do not have to concern ourselves with difficult and controversial assessments of contending claims. Another possible reason for the appeal of marginal egalitarianism is that it draws on libertarian intuitions, or on intuitions related to the security of existing holdings. If everyone is assumed to be entitled to whatever she currently has, it is easier to ignore the current pattern of holdings in making distributive decisions.
Marginal egalitarianism is not very prominent in core debates over distributive justice, except as a hidden ally of one of the other egalitarian theories. Marginal egalitarianism is more prominent in the field of bioethics. In that field, it may seem more reasonable to consider resources on hand and available for distribution (such as a heart for transplant) as subject to special rules. Even in the bioethics field, however, theorists sometimes ignore the distinction between marginal egalitarianism and other egalitarian theories.26
One of my favorite examples in moral philosophy comes from Peter Singer's book Practical Ethics. I will refer to this example as the Singer Earthquake Case, and I will present a number of variations on Singer's original example in this chapter and in Chapter 9.
After an earthquake, Singer imagines, we come upon two victims, one more severely injured than the other:
The more severely injured victim, A, has lost a leg and is in danger of losing a toe from her remaining leg; while the less severely injured victim, B, has an injury to her leg, but the limb can be saved. We have medical supplies for only one person. If we use them on the more severely injured victim the most we can do is save her toe, whereas if we use them on the less severely injured victim we can save her leg.27
Most people, I hope, will agree with Singer that here we should distribute the scarce medical supplies to the person who can most benefit from them, even though the other candidate is worse off. In other words, we should follow the utilitarian course rather than the welfare-egalitarian course.
We can modify the example so that utilitarianism is opposed not only to welfare egalitarianism, but also to a version of resource egalitarianism. Suppose that person A (who has already lost a leg and risks losing a toe on her remaining leg) is obviously poor, while person B (who risks losing a leg, but has no other injury) is obviously rich. I will refer to this as the Modified Singer Earthquake Case. Most people, I hope, will (p.96) persist in thinking that in the Modified Singer Earthquake Case, we should still distribute the scarce medical resources to person B, who will most benefit from them, even though she already has more welfare and more resources.
The Modified Singer Earthquake Case does not necessarily bring utilitarianism into conflict with all versions of resource egalitarianism. Resource egalitarians are often unwilling to apply their theory in small-scale triage situations. Rawls, for example, tells us that “the maximin criterion is not meant to apply to … how a doctor should treat his patients or a university its students.”28
Resource egalitarians might have no theory for triage cases, or they might incorporate elements of other theories for application to such cases. Resource egalitarians might, for example, apply utilitarianism to small-scale distributive problems. While any application of utilitarianism would certainly be a step in the right direction, there are other possibilities as well. Resource egalitarians might think that the only resources which should be equalized, in a case like the Modified Singer Earthquake Case, are the specialized medical resources that both candidates need. In other words, resource egalitarians might apply marginal egalitarianism to triage cases.
In both the Modified Singer Earthquake Case and the original Singer Earthquake Case, it is impossible to divide the scarce medical resources: they must be used on one candidate or the other. We can modify the example further by stipulating that the medical resources can be divided in half, in which case they will be half as effective: Person A will lose half a toe and person B will lose half a leg. So, in what I will call Modified Singer Earthquake Case #2, we now have the following choices:
2. We can use all the medical supplies on person B, a rich person who will lose a leg if we do not help her, but has no other injury. Then person B's leg will be saved.
3. We can divide the medical supplies in half. Then person A will lose half a toe, and person B will lose half a leg.
Most people, I hope, will agree that we should not divide the medical supplies, but should instead use them all on person B, who will benefit most. In this example, then, utilitarianism seems right even though it is opposed by not one, but three different egalitarian theories. The golden mean of utilitarianism is not yet tarnished, and egalitarian theories seem implausible even when they are aligned together against utilitarianism.
Egalitarian critics of utilitarianism sometimes claim that utilitarianism imposes too great a demand on people at the bottom of society. If helping people at the bottom means that those who are better off must forgo greater benefits, no help will be forthcoming under a utilitarian system. Rawls, T. M. Scanlon, and Thomas Nagel, among others, have expressed such sentiments. Nagel appeals to a “fundamental psychological fact” that “it is easier to accept sacrifices or forgo advantages for the sake of those worse off than you than for the sake of those better off than you.”29
Nagel's view has some initial plausibility. It does seem that it must be easier to accept sacrifices or forgo advantages (p.98) for the sake of those who are worse off than you than for the sake of those who are better off than you. Of course, utilitarianism only requires worse-off people to accept sacrifices and forgo advantages for the sake of better-off people if the better-off people would thereby obtain greater advantages or avoid greater sacrifices. There is also some ambiguity in the terms “worse off” and “better off.” Do we mean worse off in terms of resources or worse off in terms of welfare? While Rawls is a resource egalitarian, Nagel is more of a welfare egalitarian.
The Modified Singer Earthquake Case can help to test the view that utilitarianism imposes more unreasonable demands on people than do various forms of egalitarianism. Do we think, in this example, that it would be easier for person A (poor person who has already lost a leg and is about to lose a toe) to forgo treatment in favor of person B? Or do we think it would be easier for person B (rich person who is about to lose a leg, but has no other injury) to forgo treatment in favor of person A? Doubtless it would not be easy for either person to forgo treatment. But as to the relative ease or difficulty of sacrifice, the answer, to me, is obvious: it would be easier for the person who is worse off, in both welfare and resources, to forgo an advantage (saving a toe) so that the person who is better off can obtain a greater advantage (saving a leg).
I contend that egalitarian critics of utilitarianism hold false views about the strains of commitment and the relative burden of sacrifices under egalitarianism and utilitarianism. It is not true, as a general matter, that it is “easier to accept sacrifices or forgo advantages for the sake of those worse off than you than for the sake of those better off than you.”30 The initial plausibility of this claim depends on its partial coincidence with utilitarianism: often those who are worse off than you would benefit more from your sacrifice. But when utilitarianism (p.99) diverges from egalitarianism, as in the examples here discussed, the egalitarian claim loses its plausibility. Contrary to Nagel's view, the most accurate simple statement we can make about the relative burden of sacrifice is that it is easier to accept lesser sacrifices so that others can avoid greater sacrifices, and it is easier to forgo lesser advantages so that others can obtain greater advantages.
Of course, egalitarian critics of utilitarianism probably do not have in mind examples like the ones here discussed. They probably have in mind a tableau of fat and satisfied rich people31 claiming that nothing should be done to alleviate the misery of the poor; if taxes are raised, it will merely reduce the work and investment incentives of those satisfied rich people. In other words, egalitarian critics of utilitarianism have in mind a situation in which better-off people claim that they would have to forgo greater advantages to help the poor, but in which that claim is not particularly convincing. As argued in Chapter 2, that is not the way to do moral philosophy.
Maximally Demanding Theories
A number of moral theorists have asked whether morality allows individuals a zone of partiality, and how wide such a zone might be.32 Both utilitarianism and egalitarianism could accommodate a zone of partiality; both could allow individuals, in some or most circumstances, to advance their own interests rather than the distributive goals of the theory. On the other hand, both utilitarianism and egalitarianism could be maximally demanding; both could deny individuals any zone of partiality. A maximally demanding utilitarianism would tell people to give up their resources to others who could benefit more from those resources (assuming no offsetting reduction (p.100) in welfare). A maximally demanding egalitarianism would tell people to give up their resources to others who are in some way worse off than them.
Mostly, the demand of both utilitarianism and egalitarianism is for the those who are better off to give resources to those who are in various ways worse off. But in unusual circumstances, a maximally demanding utilitarianism can require those who are worse off to give their resources over to people who are better off, as long as those who are better off can also benefit more. Leaving aside the question of whether any theory should be maximally demanding, is such a maximally demanding utilitarianism less plausible than a maximally demanding egalitarianism?
The Modified Singer Earthquake case can once again shed some light on this question. Let us modify that example further by dispensing with the disinterested, benevolent, uninjured person who happens to have medical supplies. Let us suppose, alternately, that each of the two earthquake victims herself possesses the scarce medical supplies, and faces the question of whether to give those supplies to her fellow victim instead of using them herself. I will refer to this version of the example as Modified Singer Earthquake Case #3.
A maximally demanding utilitarianism would say that person A (poor person who has already lost a leg and is about to lose a toe) has a moral obligation to give, or perhaps sell, her medical supplies to person B (rich person who is about to lose a leg, but has no other injury). A maximally demanding egalitarianism would say that person B (rich person who is about to lose a leg, but has no other injury) has a moral obligation to give her medical supplies to person A (poor person who has already lost a leg and is about to lose a toe). Put another way, maximally demanding utilitarianism would say that person A (p.101) has a moral obligation to give up a toe so that someone better off can avoid losing a leg, while maximally demanding egalitarianism would say that person B has a moral obligation to give up a leg so that someone worse off can avoid losing a toe.
Possibly we think that both maximally demanding theories are implausible. Possibly neither earthquake victim would have a moral obligation to give up scarce medical supplies that she needs. Surely, however, maximally demanding egalitarianism is far less plausible than maximally demanding utilitarianism. Even if it is sometimes wrong to say that people should give up their resources to others who would benefit more, it is certainly wrong to say that people should give up their resources to others who would benefit less.
I am inclined to think that utilitarianism should not be maximally demanding, that there should be a zone of partiality in utilitarianism. However, I do not try to justify this view here, and I do not venture to suggest how wide the zone of partiality should be. Throughout most of this book I assume away issues of partiality; I assume that a decision has already been made to view distributive issues from an impartial perspective.
(1.) For general discussions of various aspects of egalitarianism, see Larry Temkin, Inequality (Oxford University Press, 1993); Douglas Rae et al., Equalities (Harvard University Press, 1989); and Amartya Sen, Inequality Reexamined (Harvard University Press, 1992).
(2.) Derek Parfit, “Equality and Priority,” in Ideals of Equality, ed. A. Mason (Blackwell, 1998), p. 10.
(3.) Technically, the term “leveling down” also refers to a decision to reduce the welfare (or resources) only of those who are better off, leaving the welfare (or resources) of those who are worse off unaffected. But it is hard to think of even a hypothetical example in which those who are worse off will be completely unaffected, for good or ill, by what we do to those who are better off; it is far more likely that harming those who are better off will either help those who are worse off (in which case it will not be leveling down) or harm those who are worse off.
(4.) Robert Nozick, Anarchy, State, and Utopia (Basic Books, 1974), p. 206; Ian Shapiro, Democratic Justice (Yale University Press, 1999), p. 157; Jan Narveson, “On Dworkinian Equality,” Social Philosophy and Policy 1, no. 1 (1983): 1–23, p. 16.
(5.) Temkin, Inequality, pp. 279–280.
(6.) See, for example, Judith Jarvis Thomson, The Realm of Rights (Harvard University Press, 1990), pp. 135–148.
(7.) Temkin, Inequality, p. 23.
(8.) Ronald Dworkin, Sovereign Virtue: The Theory and Practice of Equality (Harvard University Press, 2000), p. 87.
(9.) This describes one aspect of Bruce Ackerman's theory. See Ackerman, Social Justice in the Liberal State (Yale University Press, 1980), pp. 201–221.
(10.) See, for example, Harvey S. Rosen, Public Finance, 6th edition (Irwin, 2002), pp. 145–146.
(11.) The reference is to an example from John Rawls, “Social Unity and Primary Goods,” in Utilitarianism and Beyond, ed. Amartya Sen and Bernard Williams (Cambridge University Press, 1982), pp. 159–185. For similar resource-egalitarian views, see Ackerman, Social Justice in the Liberal State, pp. 46–48; and Dworkin, Sovereign Virtue, pp. 11–64.
(12.) See Sen, Inequality Reexamined, pp. 39–56.
(13.) See G. A. Cohen, “On the Currency of Egalitarian Justice,” Ethics 99, no. 4 (1989): 906–945.
(14.) Richard Arneson, “Equality and Equal Opportunity for Welfare,” Philosophical Studies 56 (1989): 77–93. Arneson has more recently modified his position somewhat. Arneson, “Luck Egalitarianism and Prioritarianism,” Ethics 110, no. 2 (2000): 339–349; Arneson, “Equality of Opportunity for Welfare Defended and Recanted,” Journal of Political Philosophy 7, no. 4 (1999): 488–497.
(15.) Robert E. Lane, The Loss of Happiness in Market Democracies (Yale University Press, 2000), pp. 59–76.
(16.) Dworkin, Sovereign Virtue, p. 79; Ackerman, Social Justice in the Liberal State, p. 269.
(19.) Someone with a broken heart has a metaphoric disability, not necessarily a real one.
(20.) Dworkin, Sovereign Virtue, p. 60.
(21.) “It is one of the most moving experiences in classical music to attend a performance by the wonderfully talented Itzhak Perlman. The Israeli, crippled by polio as a child, limps painfully on elbow crutches on to the platform. He slumps gratefully into a chair, stowing his sticks by his side, and reaches for his violin. Then, as if remorseful Fate is trying to atone for his disability, he plays the great works with a sublime and inspired facility.” Roger Watkins, “Classical Choice,” Times of London March 6, 1999.
(22.) For similar reasons, the resource-egalitarian argument that welfare (p.281) egalitarianism allocates too many resources to those with expensive tastes is far less of an argument against utilitarianism.
(23.) Terminally ill young people who, not needing expensive medical care, must be given cash grants so that they can receive the lifetime minimum income.
(24.) Resource egalitarianism and welfare egalitarianism might support national health insurance as an alternative to some current systems, but national health insurance would be very far from the ideal in either theory.
(25.) Rae et al., Equalities, p. 50.
(26.) This distinction is, I think, insufficiently appreciated in Erik Nord's book Cost-Value Analysis in Health Care: Making Sense of QALYs (Cambridge University Press, 1999).
(27.) Peter Singer, Practical Ethics, 2nd edition (Cambridge University Press, 1993), p. 25.
(28.) John Rawls, “Concepts of Distributional Equity: Some Reasons for the Maximin Criterion,” American Economic Review 64, no. 2 (1974): 141–146, p.142.
(29.) Thomas Nagel, Equality and Partiality (Oxford University Press, 1991), p. 78.
(31.) Or, since this is the twenty-first century, trim, fit, and satisfied rich people.
(32.) Two of the contributions to this discussion are Shelly Kagan, The Limits of Morality (Oxford University Press, 1989), and Nagel, Equality and Partiality.