John Rawls's A Theory of Justice, which was published thirty years ago, has been the most influential work of political philosophy in the second half of the twentieth century. Rawls has published a great deal since, but this later work generally advances a more guarded, less universalistic, and to my mind less interesting version of the doctrine developed in A Theory of Justice.
The earlier book remains, in any case, the crucial work because it is the source of Rawls's prestige and of the doctrine he continues to defend, albeit in modified form. Adherents of the key premises and conclusions of A Theory of Justice remain, I believe, the largest single contingent among political philosophers, normative political theorists, and legal theorists. A Theory of Justice is thus likely to remain one of the most influential works of political philosophy well into the twenty-first century. So it is incumbent upon advocates of a free society to understand its nature and appeal: to grasp its central premises, conclusions, and underlying motivation, especially Rawls's argument for his most distinctive conclusion, the strongly redistributive "difference principle."
Rawls introduces his own position by way of contrast with utilitarianism. Utilitarians hold that the best social outcome is the one with the greatest sum of happiness over unhappiness across all individuals; and that actions, policies, or institutions should aim to produce this outcome. Rawls objects that utilitarianism abrogates individualism and individual rights. This means that, in the course of his critique of utilitarianism, Rawls himself has to adopt—even if only temporarily—a strongly individualist and pro-rights position.
According to Rawls, "A person quite properly acts, at least when others are not affected, to achieve his own greatest good, to advance his rational ends as far as possible" (23). Rawls's first main objection to utilitarianism is that it makes an illicit inference from this principle of individual choice to a "principle [of choice] for society." The utilitarian argues that, just as it is reasonable for an individual to bring about certain losses within his life if this is necessary to attain greater countervailing gains in his life, it is reasonable for society to bring about losses in one individual's life if that is necessary to attain greater countervailing gains in other people's lives. Rawls rejects this utilitarian inference because its analogy between individual and society falsely "conflat[es] all persons into one," revealing that "Utilitarianism does not take seriously the distinction between persons" (27). It does not "take seriously the plurality and distinctness of individuals" (29).
According to Rawls, utilitarianism's failure to take seriously the separate moral importance of each individual implies that there are no principled moral constraints on what may be done to any individual in the course of genuinely promoting the aggregate happiness. As he puts it, "there is no reason in principle why the greater gains of some should not compensate for the lesser losses of others; or more importantly, why the violation of the liberty of a few might not be made right by the greater good shared by the many" (26). In a hallmark passage of A Theory of Justice, Rawls proclaims that:
Each person possesses an inviolability founded on justice that even the welfare of society as a whole cannot override. For this reason justice denies that the loss of freedom for some is made right by a greater good shared by others. It does not allow that the sacrifices imposed on a few are outweighed by the larger sum of advantages enjoyed by many. . . . [T]he rights secured by justice are not subject to political bargaining or to the calculus of social interests (3-4).
The separateness of persons and the propriety of seeking one's self-interest, Rawls seems to be saying, imply that the individual should reject calls to sacrifice for the aggregate good, and that individuals have a moral inviolability that forbids others from imposing such sacrifices upon them.
But how are we to identify the principles that define this inviolability? Rawls holds that these principles will reflect the separateness of persons and their rational systems of ends if and only if they are principles to which each individual would rationally agree. What political philosophy must do, if it is to take the separateness of people seriously, is to identify the set of interpersonal norms by which each and every individual, looking only to his rational self-interest, would agree to abide.
According to Rawls, before we can investigate what principles rational people would all agree to, we have to think a bit more about the nature of this crucial contract. Here Rawls's main claim is that we cannot identify the principles of justice by inspecting any actual agreement among persons. One reason is that the soundness of a principle of justice does not depend upon whether or not individuals have all agreed to abide by it. There may be no actual agreement among a set of individuals to refrain from unprovoked killing and yet it is still objectively true that such killing is unjust. A second reason is that actual agreements may be contaminated by the irrationality of some of the participating agents, or the duress under which they negotiate, or the unfairness of their background circumstances. So, to identify the principles of justice, we have to envision individuals negotiating with one another in circumstances guaranteeing that their agreement on basic principles will be rational, free, and fair. Rawls calls this imaginary situation the "original position." The principles of justice that are objectively sound, in his view, are those that individuals would adopt in this hypothetical situation. The original position is a theoretical device for investigating what principles we—qua rational, free, and fairly situated individuals—will each agree to with an eye to advancing our own individual well-being.
Of course, what the parties in the original position agree upon will very much depend on what is included or excluded from the description of this hypothetical situation. Rawls may insure the selection of the principles which he antecedently favors by building that selection into his description of the original position. The most important feature of the original position is the requirement that the negotiators have no knowledge of their own particular traits, capacities, dispositions, and values. Rawls argues that people's knowledge of the particularities of their own lives or the lives of others will lead them to introduce unfair proposals tailored to their own special circumstances. For instance, if I know that I am talented, I will unfairly propose the principle that talented people should receive special governmental bonuses to be paid for out of taxes imposed on the untalented. Other people will unfairly propose principles that are tailored to their own special circumstances. The result will be either no general agreement upon principles or an agreement determined by whoever more cleverly or aggressively pushes his own unfair proposal. To preclude this, the negotiators must be deprived of all knowledge of their particular identities. The original position must include a "veil of ignorance" which eliminates all information about, for example, whether one is talented or untalented, energetic or lazy, cheerful or morose, able-bodied or crippled, materialistic or spiritual, cultured or Philistine, well-educated or ill-educated, and so on.
If the inhabitants of the original position are ignorant of who they are and of what their particular goals and values are, what do they have to bargain about? Rawls answers that, as part of their comprehensive general knowledge about human beings and the workings of society, each of these individuals knows he has some particular ultimate goals and values that will be revealed after negotiations in the original position are concluded. Moreover, each participant knows that, whatever those goals and values turn out to be, his capacity to achieve them will be a function of his supply of the instrumental goods of liberty, opportunity, and income. So each participant in the original position rationally seeks to advance his own goals and values by trying to secure principles which provide him with the most liberty, opportunity, and income. Each of the bargainers thus regards the principles of justice as rules for the distribution of the "socially primary goods" of liberty, opportunity, and income.
Rawls argues, first of all, that each negotiator wants as much liberty for himself as he can get others to agree to and each realizes that others, being similarly motivated, will not accept a less than equal share of liberty so that he can have more. So they will agree to a principle of maximal equal liberty. Similar reasoning then quickly leads each bargainer to agree to a principle of maximal equal opportunity. Things are more complex, however, when the negotiators get to the topic of the distribution of income. The bargainers, as expert social scientists, understand that if a scheme of income equality is mandated, with the most productive people heavily taxed and the least productive heavily subsidized, everyone is likely to end up worse off than if equality of income is not mandated. The taxation will discourage the talented and energetic from being productive, and the subsidies will discourage work effort in the untalented and non-energetic. Clearly the more talented and energetic will be worse off than if they were not subject to such taxation. But the less talented and energetic will be also be worse off. This is partially because they will not benefit from the economic dynamism of the more talented and energetic. But, in Rawls's mind, it is mostly because the total tax revenues gathered under mandated equality and available for redistribution will be less than the tax revenues gathered under a less heavy-handed tax-and-subsidy regime. Each bargainer will, therefore, reason that whether or not he is talented and energetic, he will end up with more income if he agrees to the principle that the more talented and energetic may receive a more than equal share of the total income.
But how much inequality will the bargainers agree to? Consider the accompanying chart, which shows the results of four distinct income regimes for two individuals, Talented and Untalented.
The regimes B, C, and D are each better for both parties than the strictly egalitarian regime. So, in the original position and in ignorance of whether he is actually Talented or Untalented, each party will prefer each of these regimes to the first one. But how will they choose among these three? Which set of possible payoffs will a rational bargainer prefer— <50, 10>, <35, 15>, or <25, 17>? Rawls argues that, under conditions of uncertainty like the original position, a rational decision-maker will choose among sets of possible payoffs entirely by looking at the worst possible payoff within the set. In these circumstances, a rational decision-maker will assume that he will turn out to be the least talented, least energetic, least fortunate group and will get the short end of the stick under any regime—10 under regime B, 15 under regime C, and 17 under the regime D. He will therefore opt for the regime with the longest shortest stick, namely regime D. Stated more generally, this decision-maker will opt for the principle: Allow and protect precisely that inequality of income which will maximize the income of the lowest-income individuals. The negotiators will agree to Rawls's difference principle, according to which there is to be as much inequality as is necessary to maximize income redistribution to the worst off. Talented may, as a matter of justice, receive as much as, but no more than, is necessary in order that Untalented be benefited as much as possible.
The difference principle has two seemingly inconsistent aspects. First, it allows very much higher incomes for the naturally and socially most advantaged people than accrue to the naturally and socially least advantaged as long as those higher incomes are necessary for maximizing the incomes of the worst off. Suppose the authorities realize that the inhabitants of Galt's Gulch can be lured back to work only by assuring them that they will be not be subject to any taxation: the best that can be done for the worst off is to get out of the way of Galt and his companions. (In effect, the authorities realize that regimes C and D cannot, in fact, be instituted.) Then, by Rawls's argument, the authorities should assure these most advantaged individuals complete immunity from taxation and, consequently, the right to keep the vastly higher incomes that they will garner, compared to the incomes accruing to the worst off. Rawls considers this a matter of "due solicitude" for the talented and productive. Second, to put matters bluntly, the difference principle declares that the advantaged are, morally speaking, the slaves of the disadvantaged. Or, to put it somewhat more delicately, insofar as the more advantaged persons are their talents, their insightfulness, their energies, their willingness to persevere, and so on, their moral role is to be used as effectively as possible to increase the income of the least advantaged. Rawls himself says that the agreement within the original position to adopt the difference principle "represents, in effect, an agreement to regard the distribution of natural talents as a common asset" (101).
Actually, this understates the matter. For the prior acceptance of the original position as a proper device for thinking about principles of justice already renders all talents, energies, and so on the joint property of everyone. The original position device itself presumes that all economically valuable characteristics are under the rightful authority of the collectivity. The subsequent agreement within the original position to the difference principle really amounts to the transfer of ownership of all economically valuable characteristics—and, hence, of persons insofar as they are constituted by those characteristics—from the collectivity in general to the least advantaged in particular. The parties within the original position transfer ownership of all economically valuable characteristics to the least advantaged because, behind the veil of ignorance, each party expects to be among the least advantaged.
How is Rawls's advice that the most advantaged are to be treated with due solicitude consistent with my claim that the difference principle construes the most advantaged as the property of the least advantaged? The answer is that prudent owners manage their property for the sake of long-term maximal gain. If I own a goose that will lay golden eggs, but only if I treat it very well, then I will treat it very well. If necessary, I will feed it all the delicious corn which I would rather eat myself. If necessary, I will let it retain most of the golden eggs that it lays. Similarly, the difference principle advises the least advantaged (or their representatives) to be as solicitous of the talented and energetic as they have to be in order to maximize the net return that can be extracted from them. If Untalented has a very able slave, he should give that slave free rein to develop and exercise his talents, to identify new opportunities, and to retain the larger share of the wealth he thereby produces—if this is necessary to maximizing the value that Untalented can extract from the slave's talent and energy. The only reason Untalented might not follow this advice would be envy of his slave's greater absolute wealth.
But that envy, Rawls says, would be irrational. Why should Untalented be downcast by the greater absolute wealth of his able slave if he knows that the talents and energies of that able resource have been harnessed as effectively as possible to the service of Untalented's income? The difference principle advises the authorities to offer Galt and his companions tax immunity, if that is really necessary to secure their return. But, from the point of view of the difference principle, this deference is to be extended merely because they have proven to be recalcitrant slaves who will run away if not given a free rein. Under these conditions, the interests of the worst off are best served by a tax-free regime; and justice demands that Galt and his companions serve that regime.
In fairness to Rawls, we should note that redistributing income takes third place in his system after liberty and equality of opportunity. Within prosperous and institutionally advanced societies, the equal liberty principle has the highest priority, the equal opportunity principle has the next highest priority, and the difference principle is supposed to come into play only after the demands of these other principles are satisfied. Thus any proposal to fulfill the difference principle by measures which contravene people's equal liberty is ruled out by the priority of the equal liberty principle over the difference principle. Chaining talented and energetic people to their desks in order to generate more product for redistribution to the economically worst off would, Rawls insists, be ruled out as an abridgement of liberty.
Rawls' notion of liberty, however, is the impoverished notion of contemporary liberals, for whom liberty consists in the expressive or lifestyle freedom to say what one wants and have sexual relations with the species of one's choice. So, for example, being subject to a 75 percent tax on one's income or being subject to the seizure of 90 percent of one's peacefully acquired property does not count at all as an abridgment of liberty. Indeed, it is not really clear that chaining the talented and energetic to their desks should, for Rawls, count as an infringement of their liberty as long as these individuals are still permitted to express their views, cast their votes, meet with their chosen sexual partners, and, perhaps, are paroled on weekends to travel to their preferred cultural events. In any case, Rawls does not view anything the modern welfare state does in the name of income redistribution as an abridgment of liberty.
How do we get from the individualist beginnings of Rawls's theory to the conclusions I have just described? First, as I have already suggested, the individualist face of Rawls's doctrine is partially an artifact of his need to eliminate utilitarianism as his philosophical competitor. If Rawls did really take seriously the sort of criticism he makes of utilitarianism, he could not endorse the difference principle, which, after all, merely replaces the aggregate good with the good of the most disadvantaged as the end to which individuals are required to sacrifice. Second, the veil of ignorance makes it impossible for the negotiating agents to take seriously, or to adopt principles that recognize, the separateness and distinctiveness of persons. Indeed, the device of the original position requires that each agent take everything that makes him the individual that he is, and that would normally be thought to ground his rights and entitlements, to be at best morally irrelevant contingencies and at worst unfair advantages for which he must make redress.
Rawls's deep motivation for eliminating the influence of the particular facts of people's identities and lives is his belief that these facts are all a matter of brute luck and that, as rational beings, we ought not to submit to social outcomes resulting from such "arbitrary" contingencies. According to Rawls, no social order that is shaped by these contingencies can be rational. Rational social action will design and construct a world that overcomes the influence of these factors that are "arbitrary from the moral point of view."
While the bare facts that make people different from one another are neither just nor unjust, says Rawls, it is unjust to let them govern our economic and social lives. Thus Rawls's "principle of redress" asserts that "undeserved inequalities call for redress; and since inequalities of birth and natural endowment are undeserved, these inequalities are to be somehow compensated for" (100). The principle of redress, not the prudent calculations of bargainers in the original position, is Rawls's fundamental reason for advocating redistribution of income.
Why does he say that the talents and other attributes that make us unequal are undeserved? Because, in Rawls's mind, they are entirely determined by causes outside our control, such as genes and environment. They are the results of natural and social "lotteries" in which we are entirely passive. We cannot choose or change (or take credit for) our own identities. If we resign ourselves to these contingencies by allowing them to determine how well individuals fare in their lives, we will be perpetuating our own unfreedom. The only way that we can really be free is to rise above the brute forces of fortune and misfortune by turning to our rational selves and to the principles which those rational selves would legislate for themselves.
Rawls is saying, in effect, we can achieve a free and rational existence only by living in accord with principles that disembodied beings would unanimously accept. Only by reconstructing our existences in accord with an unworldly, depersonalized Reason can we rid ourselves of unfreedom and irrationality. Rawls himself invokes the notion of an otherworldly self, in the form of Immanuel Kant's distinction between a "phenomenal self," which exists in the natural world and has particular interests, talents, and values; and a "noumenal self" that exists in some realm beyond nature.
My suggestion is that we think of the original position as the point of view from which noumenal selves see the world. The parties qua noumenal selves . . . have a desire to express their nature as rational and equal members of the intelligible [i.e., noumenal] realm. . . . They must decide which principles . . . most fully reveal their independence from natural contingencies and social accident (255).
This is the deepest premise in Rawls's complex argument: that living in accord with the principles agreed to in the original position is advantageous to each of us, not as worldly selves, but as noumenal selves.
And it leads to one of the deepest problems in his system. Fully disembodied beings would not care a bit about securing a useful supply of liberty, opportunity, and income. They would, instead, express their unworldly freedom by disdaining worldly liberty, opportunity, and income. They would reveal their independence from natural contingencies and social accident by focusing on some fully unworldly and, hence, non-political end, such as Kant's categorical end of doing one's duty for duty's sake. If the bargainers in the original position are noumenal selves who are detached from all earthly concerns, then they will not agree upon any worldly political norms. Hence, the original position will completely fail as a device for grounding political philosophy.
On some level, Rawls is aware of this problem. His veil of ignorance blinds the inhabitants of the original position to all the particular "natural contingencies and social accident[s]" that constitute their particular identities, but does not blind them to the contingency that they all have some specific (albeit unknown) ends to which worldly liberty, opportunity, and income are means. Making the deliberations of the bargainers in the original position dependent on the contingency that we all are beings with specific worldly identities, for example, with some specific traits and goals, enables Rawls to depict these bargainers as negotiating about the distribution of liberty, opportunity, and income, and, perhaps, as agreeing upon some worldly political principles.
This retreat from his view of the original position as the point of view from which noumenal selves see the world contravenes his deep conviction that all "arbitrary" contingencies must be transcended—that rational social action and freedom require a world shaped only by facts that have been validated by a de-personalized Reason standing outside of the merely phenomenal world and decreeing how it should be. If Rawls is to be true to this deep conviction, he must insist that individuals in the original position deliberate in a way that is independent even of the contingency that they are beings with determinate (albeit unknown) worldly identities. This means that he must resolutely conceive of the parties within the original position as purely noumenal beings. But this, we have seen, means that no worldly political principles whatsoever can be grounded upon the original position.
Although powerful objections can be launched against Rawls's detailed arguments for his specific principles—especially the enslaving difference principle—I have maintained that his deeply anti-worldly aversion to what he calls the contingent and the arbitrary features of human life actually precludes Rawls from arriving at any political principles at all.
This article first appeared in the July/August 2001 issue of Navigator magazine.