Quiet Revolution in Welfare Economics- by Michael Albert and Robin Hahnel
Our summary of traditional welfare theory is guided by three purposes:
1. To uncover the paradigm that lies behind traditional welfare theory to evaluate the further usefulness of that paradigm
2. To assess undercurrents within traditional welfare theory showing how they elucidate essential features of the underlying paradigm
3. To criticize the traditional paradigm and theory in all its forms rather than only one or another
With these goals in mind, we present classical utilitarianism, neoclassical welfare theory, and a modern contractarian formulation of welfare theory and analyze the "debate" between these competing formulations.
By utilitarianism is here meant the ethical theory that the conduct which, under any given circumstances, is objectively right, is that which will produce the greatest amount of happiness of the whole; that is, taking into account all whose happiness is affected by the conduct.1
This is Sidgwick's classic statement of the utilitarian criterion for judging social actions and institutions, a criterion prominently associated with the names of Jeremy Bentham, James Mill, and F. Y. Edgeworth.
As Sidgwick points out, the principle of the greatest sum of happiness logically presumes that pleasures (or utilities) are measurable and additive. Hence classical utilitarianism was committed by its "first principle" to the view that different pleasures suffered by the same individual are quantitatively comparable and additive as are different pleasures suffered by different individuals even of different generations.
Here as before, the assumption is involved that all pleasures included in our calculation are capable of being compared quantitatively with one another and with all pains; that every such feeling has a certain intensive quantity ... in respect of its desirableness, and that this quantity may be to some extent known: so that each may be at least roughly weighed in ideal scales against any other. This assumption is involved in the very notion of maximum happiness; as the attempt to make "as great as possible" a sum of elements not quantitatively commensurable would be a mathematical absurdity. 2
Utilitarians and their critics have been quick to note that the "greatest happiness" criterion must subsume all other notions of social "good."
The utilitarian must, in the first place, endeavor to show that the principles of truth, justice, etc. have only dependent and subordinate validity: arguing either that the principle is really only affirmed by common sense as a general rule admitting of exceptions and qualifications, as in the case of truth, and that we require some further principle for systematizing these exceptions and qualifications; or that the fundamental notion is vague and needs further determination, as in the case of justice; and further, that the different rules are liable to conflict with each other, and that we require some higher principle to decide the issue thus raised. 3
Finally, the classical utilitarians rejected any notion of a hierarchy of pleasures, or preferences, that were qualitatively superior or inferior to others. Bentham "thought it an insolent piece of dogmatism in one person to praise or condemn another in a matter of taste." 4 And, "Quantum of pleasure being equal, pushpin is as good as poetry" 5 While few today pay much attention to classical utilitarianism, we review it because
1. It is the historical antecedent to neoclassical welfare theory, and as we shall see, most "advances" in neoclassical welfare theory are represented by neoclassical theorists as weakenings of what they describe as the more dubious assumptions of classical utilitarian theory
2. It seems to us that the best explanation of the paradox that neoclassical welfare theory retains a utilitarian framework, even though the essence of traditional welfare theory is inimical to a teleological theory of utilitarian form, lies in the particular historical evolution that occurred
3. Classical utilitarianism is the clearest example of a complete theory of social choice
The essence of neoclassical welfare theory is that the performance of economic institutions can and should be judged according to whether they provide economic goods in quantities that accord with people's relative desires for those goods. High marks are given to economic systems that display a close "fit" between the relative terms on which economic goods are made available and people's relative preferences for those goods. As I.M.D Little put it, "It is a good thing that individuals should have what they want, and they themselves know best what they want." 6
As Little's statement indicates, economists in their evaluative efforts are to take the individuals' preferences as their fundamental tool. Lionel Robbins provided the strongest defense for this procedure:
Economics is not concerned with ends as such. It assumes that human beings have ends in the same sense that they have tendencies to conduct which can be defined and understood, and it asks how their progress towards their objectives is conditioned by the scarcity of means .... The ends may be noble or they may be base. Economics takes all ends for granted. 7
Herb Gintis dubbed this fundamental neoclassical welfare postulate the Robbins Principle and rephrased it:
The Robbins judgment might be more precisely stated as follows: the economist is instructed, other things being equal, to move the individual to a position chosen on the basis of his manifest preference ordering. In other words, we act as if it is a good thing that individuals have what they want, and as if they know best what they want. 8
In sum, according to neoclassical welfare theory, the individual's manifest preferences are not data because they are assumed to be knowable, psychological, verifiable "truths" concerning an aspect of reality that we refer to as human happiness. Rather, the strongest statement of the justification can be phrased in negative form typical of the classical liberalism that is its source: If the individual is not to be accepted as the best judge of his or her own needs and desires, who is? What other person, group of people, or omniscient entity would you rather trust than yourself to define your preferences?
the individual's manifest preference ordering as the determinant of whether
or not the individual is "better off," neoclassical economists proceeded to
their second basic welfare premise, usually referred to as the principle of
Pareto optimality: "The community becomes better off if one individual becomes
better off and none worse off." 9 In the
neoclassical theorist's view the great advantage of the principle of Pareto
optimality is that it is uncontroversial, whereas its disadvantage is that it
does not provide a complete theory of social choice. Although neoclassicists
were comfortable in their conclusion that Pareto optimality was a necessary
condition for a social welfare maximum, it was clear that it could not be a
sufficient condition because there are an infinity of Pareto optimal allocations.
But with remarkable ingenuity neoclassicists used convex set theory and fixed
point theorems to refine the relationship between Pareto optimality and the
equilibria of competitive private enterprise market economies in ways transcending
untidy treatments of local versus global maxima, corner solutions in which not
all agents utilize positive amounts of all goods, and endogenous determinations
of which goods would become "economic" rather than 'free'.
Having eliminated the social welfare function explicit in classical utilitarianism, all that could be claimed for the theorem that under "traditional assumptions" any general equilibrium of a competitive private enterprise market economy was a Pareto optimum was that it kept such economies "in the running" for ideal economic system "after the trial heats." 11 More precisely, competitive private enterprise market economies satisfied one necessary condition for a social optimum.
However, the proof of a second theorem, that under "traditional assumptions" any Pareto optimum could be achieved as the equilibrium of some competitive private enterprise market economy with an appropriate set of initial endowments, seemed to provide a solution to the lack of closure in the Pareto principle as a welfare theory. This second theorem appeared to have refined the relationship between the equilibria of competitive private enterprise market economies and Pareto optima to a state of functional equivalence. That any Pareto optimum desired could be achieved as the equilibrium of a competitive private enterprise market economy with the appropriate initial endowments meant it was possible for neoclassical welfare theorists to conclude that we could cease our search for desirable economic institutions with private enterprise market systems. It appeared, in theory, such systems could deliver any result we might desire.
Of course, this was not the same as being able to declare which set of initial distributions economists recommended for achieving social bliss. But that was probably of lesser concern than the "cease search" announcement anyway. Koopmans spelled out the approach as follows:
To duplicate any imagined Pareto optimum through a competitive equilibrium will therefore require some particular distribution of rights to the income from ownership of resources and from profitable production-or some redistribution of such income flowing from given ownership and participation arrangements through quite possibly discriminating head taxes and subsidies. 12
And Arrow drove the point home:
Any complaints about its [competitive private enterprise market economy's] operation can be reduced to complaints about the distribution of income, which should then be rectified by lump sum transfers. 13
The conclusion implied by the fundamental theorems seemed perfectly clear to most neoclassical theorists. Although various real world problems might prevent realization of a perfectly competitive, private enterprise market economy, such a system is, nonetheless, the ideal economic arrangement. It always delivers Pareto optimal outcomes, and by proper manipulation of initial conditions, it can deliver any particular Pareto optimum we might prefer on equity grounds.
Since the late 1950s when they abandoned efforts to blaze a trail from the set of all Pareto optimal allocations to the bliss point using only the "hatchet" of efficiency in the form of compensation tests, neoclassical economists have had to content themselves with an unspecified "Bergsonian" social welfare function of all individuals' utilities, W(U(1), U(2), ... U(N)), or, of a weighted sum of individual utilities.
with the weights, k(i), unspecified. However, neoclassicists point out that specification of the function, or weights, is sufficient to make neoclassical welfare theory a complete theory of social choice, as was its predecessor, classical utilitarianism. We are also assured that for any specification the optimum result can be achieved by an appropriate private enterprise, competitive market economy.
The resurgence of contractarianism has only recently spread from social philosophers to economists. The tradition, of course, dates back to Kant, Rousseau, and Locke and was largely concerned with defining the rights of the governed by considering a transition from the "state of nature" to "civil society." Whether this transition from a nonsocial arrangement without duties, obligations, or rights, to a society in which individuals have contractual obligations was viewed as an historical or hypothetical process was often unclear. The principal concerns of modern contractarians have been to clarify:
(1) the explicitly hypothetical interpretation,
(2) the nature of the justificatory power of the contractual view, and
(3) the kind of major social institutions that would conform to the principles that would be agreed to in a hypothetical, contractual situation. We confine our attention to John Rawls, the most important figure in the resurgence, and Robert Nozick, a leading exponent of a kind of libertarianism that has gained popularity in certain economic circles. 14
Rawls identifies three fundamental differences between the contractarian and utilitarian approaches:
It has seemed to many philosophers, and it appears to be supported by the conviction of common sense, that we distinguish as a matter of principle between the claims of liberty and right on the one hand and the desirability of increasing aggregate social welfare on the other; and that we give a certain priority, if not absolute weight, to the former .... While the contract doctrine accepts our convictions about the priority of justice as on the whole sound, utilitarianism seeks to account for them as a socially useful illusion ... [because] even the excessive zeal with which we are apt to affirm these precepts and to appeal to these rights is itself granted a certain usefulness, since it counterbalances a natural human tendency to violate them in ways not sanctioned by utility.
A second contrast is that whereas the utilitarian extends to society the principle of choice for one man ... a contract view assumes that the principles of social choice, and so the principles of justice, are themselves the object of an original agreement. There is no reason to suppose that the principles which should regulate an association of men are simply an extension of the principle of choice for one man. On the contrary: if we assume that the correct regulative principle for anything depends on the nature of that thing, and that the plurality of distinct persons with separate systems of ends is an essential feature of human societies, we should not expect the principles of social choice to be utilitarian.
The last contrast ... is that utilitarianism is a teleological theory whereas justice as fairness is not .... Justice as fairness is a deontological theory [that] ... does not interpret the right as maximizing the good .... For if it is assumed that the persons in the original position would choose a principle of equal liberty and restrict economic and social inequalities to those in everyone's interests, there is no reason to think that just institutions would maximize the good .... The question of attaining the greatest net balance of satisfaction never arises injustice as fairness; this maximum principle is not used at all .... This priority of the right over the good in justice as fairness turns out to be a central feature of the conception. 15
Nozick points out that contractarian approaches to social philosophy reflect Kant's imperative: "Act in such a way that you always treat humanity, whether in your own person or in the person of any other, never simply as a means, but always at the same time as an end." 16 In Nozick's interpretation this means "there are only individual people, different individual people, with their own individual lives," and "using one of these people for the benefit of others ... does not sufficiently respect and take account of the fact that he is a separate person." 17
Rawl's refined the social contract through the concept of unanimous choice in an explicitly hypothetical "original position" behind a "veil of ignorance" designed to filter out all information that would allow the discussants to know what their personal positions would be in the actual society agreed on. The content of the discussion is then divided into (1) what principles would be agreed to by individuals in this situation, which is designed to embody the notion of fairness, and (2) what major social and economic institutions would conform to these principles.
Rawls argued that "rational" individuals in the original position would agree to a liberty principle of maximum individual liberty consistent with equal liberty for all and a distributive principle referred to as the "maximin" or "difference" principle, whereby inequalities in the distribution of "basic economic goods" are acceptable only if they serve to increase the value of basic economic goods available to the worst-off member of society.
Nozick disagreed vehemently with Rawls' second principle, arguing that the difference principle is unfair to the more "productive" members in the social cooperative process. Whereas Rawls extends the veil of ignorance to exclude knowledge concerning one's own "productivity," Nozick would permit individuals in the original position to know what he terms their "entitlements." The debate hinges on what information would be "fair" for people to have in the "original position," since Nozick argues convincingly that the "difference principle" would hardly be agreed to unanimously if people knew what their productivities were going to be. 18 Rawls argued further-and in this regard Nozick would no doubt agree-that the liberty principle would be given absolute priority over the maximin principle by inhabitants of the original position.
The major political institutions Rawls considered compatible with the two principles are a constitutional democracy and Bill of Rights similar to those of the United States. The economic institutions Rawls favored are private property and markets. 19 These institutions are judged to be compatible with the two principles provided (1) Markets are sufficiently competitive to generate the social efficiency necessary to satisfy the difference principle, and (2) lump-sum transfers of initial endowments conform to the distributional exigencies of the difference principle. Essentially, the new contractarian approach has supported private enterprise, competitive market economies, and political liberalism.
The important point for our purposes is that the form of the justification is quite different from that of neoclassical welfare theory. Whereas neoclassical welfare theory has retained a utilitarian form and justifies competitive private enterprise market economies on the grounds that any social welfare function could be maximized by an appropriate version of such an economy, the new contractarian approach justifies private enterprise market economies on the grounds that these institutions are compatible with individual liberty and are capable of yielding results that satisfy the distributive principle agreed to by those in the original position.
In either major version, the new contractarian approach is a complete theory of social choice. In Rawls' interpretation the initial endowments that should be established by lump-sum "redistributive" transfers are those that will achieve a final distribution of economic goods that maximizes the value of the bundle of the worst-off group. In Nozick's view redistributive transfers violate individual "entitlements" and cannot be justified. But in either case "closure" is achieved as one Pareto optimum is selected above all others.
The essence of the new contractarian approach is that each individual has a distinct conception of "the good," and the task of social institutions is not to maximize "the good," but to allow individuals maximum freedom to pursue their own concepts of the good without interference. So according to the new contractarian approach the institutions of private enterprise and competitive markets are justified because they respect individual liberty and are capable of yielding distributive justice, not because they maximize some concept of social welfare. As a separate issue, disagreements between Rawls and Nozick over whether or not the veil of ignorance should be permeable to information about "entitlements," and consequently, whether lump-sum redistributions should be made at all, much less according to the difference principle, will be of interest when we discuss "distributive maxims" in chapter 8.
According to the protagonists, the evolution of the modern neoclassical approach from classical utilitarianism was marked not only by great successes, but by major disappointments as well. The step-by-step transition from cardinal utility to the choice theoretic framework, along with the perfection of the fundamental theorems relating the equilibria of private enterprise, competitive market economies with Pareto optima were accomplishments of great pride to modern neoclassical theorists. However, attempts to specify a social welfare function or even establish the possibility of a reasonably desirable social welfare function have ended in frustration. All attempts to date to make neoclassical welfare theory a complete theory of social choice have failed to do so despite painstaking work and, in some cases, brilliant insight. We discuss the successes first and the disappointments second.
Starting from the classical notion of knowable, additive, interpersonally comparable, "cardinal" utilities, neoclassical welfare theorists have finally arrived at the resting grounds of the choice theoretic framework. The pilgrimage was difficult, passing through ordinal utilities, indifference curves, revealed preferences, and finally the axioms of "rational choice." Each of these stations of the cross have been portrayed as a significant advance for what seem to us a frequently confused combination of reasons.
First, it was argued that the "hedonistic calculus" was an outdated theory of actual human behavior and, therefore, an unsound foundation for welfare theory. But this was no reason to reject cardinal utility as the basis of welfare theory. Some classical utilitarians staked a triple claim:
1. Social arrangements should be made in accord with the principle of the greatest happiness for all
2. People do, in fact, behave in accord with the dictates of the hedonistic calculus of maximizing their own individual happiness
3. Private property and market institutions are such that when individuals behave in accord with the hedonistic calculus in the context of these institutions the greatest happiness for all is in fact achieved, provided certain lump-sum transfers (of a generally egalitarian nature) are made
However, the first claim is logically independent of the truth of the latter two, so that a demonstration of the falseness of either of the latter propositions need not compel one to abandon the first.
Rejection of the second proposition might be good reason not to base an analytic theory of consumer demand on the hedonistic calculus.
Likewise, a rejection of the third proposition might lead us to abandon private property market institutions and search for other ways of organizing economic activity that generate more utility. But the truth of the latter two propositions has no bearing on whether the classical utilitarian principle is the most appropriate evaluative criterion.
A second reason given for abandoning cardinal utility for the choice theoretic framework was the principle of Occam's razor. It was the supreme accomplishment of neoclassical theorists to discover that the fundamental theorems of welfare theory did not require the assumption of cardinal utilities. In the choice theoretic framework it was discovered sufficient to assume that the set of consumption bundles at least equally preferred to a given bundle was convex in order to establish the fundamental theorems of welfare economics. 20
Of course, the principle of Occam's razor could not be used to replace the assumption of cardinal utilities with the assumption of "rational choice" with regards to the problem of selecting one Pareto optimum from the infinity of Pareto optima as the "bliss point." 21 But since neoclassical theorists concerned themselves less and less with this problem, the gain from whittling down the assumptions necessary for equating competitive equilibria with Pareto optima was considered to outweigh the loss of a complete theory of social choice.
Neoclassicists have pointed out that the principle that the right course of action is to maximize the total sum happiness of all not only permits but requires blanket interference in the affairs of individuals whenever total happiness can be increased by doing so. For example, an individual's opinion regarding his or her happiness would have to be rejected if a reasonable argument could be made that a "scientific test," or more knowledgeable person had a higher probability of estimating the individual's welfare correctly. And changes that yield an increase in sum total happiness would have to be implemented even if they resulted in individual sacrifices deemed grossly unfair and undesirable by many systems of ethical standards. Examples mentioned include further immiserization of destitute peasants from an early generation to provide the primitive accumulation necessary for the affluence of succeeding generations and redistribution from poorer "uncultured" members of society to wealthy epicureans.
The objection is that classical utilitarianism might sanction not only totalitarian tendencies that prohibit individuals' freedom to make their own mistakes but gross violations of elementary justice as well. Neoclassical welfare theorists have claimed that the Robbins Principle prevents neoclassical welfare theory from making the first kind of mistake, and their extreme reluctance to consider interpersonal comparisons of utility prevents instances of the latter.
Of course, the early utilitarians, Bentham in particular, were anxious to employ their theory to just the opposite ends. They were dedicated social reformers. By combining the principle of the greatest happiness with the not implausible postulate of declining marginal utility of income and a firm rejection of elitist conceptions of pleasure, they thought they had forged a powerful argument for radical egalitarian programs. Bentham would no doubt find it ironic that his theory is assailed by modern neoclassicists for failure to adequately protect the rights of the disadvantaged, as in the hypothetical examples above. But it is undeniable that classical utilitarianism is open to the criticism of leaving individual liberties and elementary justice protected by the, at least questionable, factual proposition that observance of such liberties and justice will always coincide with maximization of sum total happiness. Although Rawls is, of course, not an adherent of neoclassical welfare theory, in the following quotation he provides an insightful interpretation of classical utilitarianism and an unsurpassed formulation of the objection that it does not reflect the sanctity of the individual, Neoclassical theorists would be fortunate to state the problem so eloquently. Whether or not their own theory is as fully insulated against this problem as they believe, we shall see presently.
We may note first that there is indeed a way of thinking of society which makes it easy to suppose that the most rational conception of justice is utilitarian. For consider: each man in realizing his own interests is certainly free to balance his own losses against his own gains .... Now why should not a society act on precisely the same principle applied to the group and therefore regard that which is rational for one man as right for an association of men? Since the principle for an individual is to advance as far as possible his own welfare, his own system of desires, the principle for society is to advance as far as possible the welfare of the group .... The principle of choice for an association of men is interpreted as an extension of the principle of choice for one man. Social justice is the principle of rational prudence applied to an aggregate conception of the welfare of the group .... On this conception of society separate individuals are thought of as so many different lines along which rights and duties are to be assigned and scarce means of satisfaction allocated in accordance with rules so as to give the greatest fulfillment of wants .... Here we may note a curious anomaly. It is customary to think of utilitarianism as individualistic, and certainly there are good reasons for this. The utilitarians were strong defenders of liberty and freedom of thought, and they held that the good of society is constituted by the advantages enjoyed by individuals. Yet utilitarianism is not individualistic, at least when arrived at by the more natural course of reflection, in that, by conflating all systems of desires, it applies to society the principle of choice for one man .... It is this conflation, and the principle based on it, which subjects the rights secured by justice to the calculus of social interests. 22
Neoclassical theorists admit to disappointments as well as proclaim successes in the evolution from classical utilitarianism to modern welfare theory. In their view the main problem still plaguing them is their inability to close their system and make it a complete theory of social choice. First attempts at closure took the form of "compensation tests," while later work followed paths pioneered by Arrow's famous impossibility theorem.
The compensation tests turned out to be a vain search for a path from the set of all Pareto optimal outcomes to the bliss point of maximum social welfare eschewing interpersonal utility comparisons and using only the tool of efficiency.
The compensation tests all spring from a desire to see what can be said about social welfare ... without making interpersonal comparisons of well-being .... They have a common origin in Pareto's definition of an increase in social welfare-that at least one man must be better off and no one worse off-but they are extended to situations in which some people are made worse off. It is here that the possibility of compensation tests becomes relevant. 23
The search for an appropriate compensation test that would allow the ranking of certain Pareto optima above others on the basis that the gainers from movements from one Pareto optimum to another would be able to compensate the losers and still retain net positive benefits began with Kaldor in 1939 and extended through Hicks, Scitovsky, Samuelson, and Little 24 But while the debate raged around whether or not different compensation criteria were equivalent, and if not, which was more compelling, the real problem lay elsewhere. When someone finally noticed the difference between compensation actually paid and compensation that was discussed but never handed over, it was apparent the compensation tests did not really leave the profession any "better off 'than the Pareto principle we had begun with. In fact, we were not left any "better off' than the medieval Japanese townspeople of Titipu in their efforts to justify who should be favored:
how the fates their gifts allot, For A is happy, B is not. Yet B is worthy,
I dare say, Of more prosperity than A.
B more worthy? I should say, He's worth a great deal more than A. Yet A is happy! Oh, so happy!
Laughing, Ha! Ha! Chaffing, Ha! Ha! Nectar quaffing, Ha! ha! ha! Ever joyous, ever gay, Happy, undeserving A!
If I were fortune-which I'm not B should enjoy A's happy lot And A should die in misery. That is assuming I am B.
But should A perish? That should he (Of course assuming I am B.) B should be happy! Oh, so happy!
Laughing, Ha! Ha! Chaffing, Ha! ha! Nectar quaffing, Ha! ha! ha! But condemned to die is he, Wretched, meritorious B!
But condemned to die is he, Wretched, meritorious B!
-W. S. Gilbert and Arthur Sullivan, The Mikado
Just as a detailed examination of compensation tests was unnecessary to our argument, we need not delve into the intricacies of investigations parting from Arrow's impossibility theorem. Arrow's results, as well as those of Sen, Pattanaik, and others who followed, demonstrate that in context of a traditional conception of preferences even a small number of intuitively appealing characteristics cannot be incorporated into a minimally satisfactory social choice rule without contradicting one another. While the impact of this "negative" conclusion has been weakened by subsequent demonstrations that particular social choice rules can comply with something less than Arrow deemed "minimally acceptable conditions," the fact remains, no "positive" conclusion has emerged from that permits traditional welfare theory to elevate one Pareto optimum above all others. On the contrary, developments in the theory of social choice clarify that traditional welfare theory is farther from "closure" than had been assumed.
In sum, neoclassical welfare theory admits to incompleteness. Yet, all in all, we believe it fair to characterize neoclassicists' attitude toward the problem of selecting from different Pareto optima as "indifference."
Some neoclassicists argue that in the modern world of specialization economists are only responsible for clarifying concepts related to efficiency anyway. In this view arguments concerning the relative merits of different efficient allocations should be left to moral philosophers who are, presumably, experts in such matters.
Other neoclassicists, with a more technocratic self-image, observe that the choice between different Pareto optima is a strictly distributional matter that will be settled on the battlefield of power politics rather than in the ivory towers of moral philosophy or welfare economics. These self-styled pragmatists console themselves (if not their victims) with the thought that if economists contribute to making Pareto improvements at least nobody is made any worse-off by their interventions.
A more virulent strain of neoclassicists centered around the "Chicago School" combine a strict interpretation of individual liberty, a religious devotion to the marginal productivity theory of entitlements, and a definition of the individual that encompasses his or her entire line of ancestors so that "just" initial endowments become identical with historical distributions. Or, "just" endowments would be identical with historical endowments had these never been tampered with by "meddling" governments who unwittingly confirm they are little more than expressions of "mobtyranny" by implementing progressive taxes. 25 Obviously once "just" initial endowments have been located, Pareto optimality becomes a sufficient as well as necessary condition for maximizing social welfare. But this justification already takes us across the bridge between utilitarian and contractarian philosophical frameworks, whether self-styled neoclassical economists of this ilk realize it or not. While only some of these economists who persist in labeling themselves neoclassicists seem aware of their contractarian status, we must classify them properly as genus, traditional welfare theory; species, contractarian, age, modern.
Regardless which rationalization one prefers, all can rejoice that the "cease search" sign planted in front of the private enterprise, competitive market economy exhibit at the "Great Economic Institution Fair" can substitute for closure for neoclassical welfare theory.
Contractarians' objections to neoclassical welfare theory are simple and to the point. Neoclassical welfare theory remains a utilitarian theory subject to weaknesses that plague any evaluative theory of utilitarian form; it "does not take seriously the distinction between persons." 26
No matter how much more successfully neoclassical welfare theory has managed to protect the sanctity of the individual than classical utilitarian theory, as a utilitarian theory neoclassical theory itself continues to depart from the view that "the principles which should regulate an association of men is simply an extension of the principle of choice for one man.' 27
Neoclassical welfare theory remains a teleological theory committed to the view that the right social institutions or actions are those that maximize the good defined as a weighted sum of individual welfares. As such, the basic design of neoclassical theory cannot express the view that "individuals are ends and not merely means; they may not be sacrificed or used for the achieving of other ends without their consent; individuals are inviolable." 28
Put differently, if we believe that "as a matter of principle each member of society is thought to have inviolability founded on justice which even the welfare of everyone else cannot override, [and] that a loss of freedom for some is [not] made right by a greater good share by others," 29 we shall have to look beyond a utilitarian theory, which neoclassical theory remains, for an account of the principles ofj ustice. The new contractarians admit that neoclassical welfare theory protects the sanctity of the individual in ways that classical utilitarian theory did not, just as they have nothing but praise for neoclassical efforts toward elevating the sanctity of the individual. But new contractarians point out that the protection afforded by neoclassical theory is inadequate because it fails to discard a philosophical form expressive of the principle of rational choice for one person for a form expressive of the inviolability of each individual.
That the Robbins Principle of neoclassical welfare theory guarantees the individual's right to define his or her own conception of "the good" against any benevolent (or otherwise) political or scientific dictatorship is recognized by contractarians. And it is recognized that an almost fanatical commitment to eschew "interpersonal welfare comparisons" creates an environment where "blanket interference in the affairs of individuals" is less likely than in an environment where different individual's utilities are presumed to be not only comparable but additive. But contractarians point out that no matter how much more protective of the sanctity of the individual these features of neoclassical welfare theory are than classical utilitarianism, the Bergsonian social welfare function continues to hang ominously over the individual's head.
The unquestioning acceptance of "manifest" preferences and an extreme aversion to "interpersonal utility comparisons" establishes an environment more respectful of the sanctity of the individual, but this notion of inviolability remains unincorporated into the basic framework of neoclassical welfare theory because the theory retains a social welfare function as the final arbiter. The specification of social welfare, W, as a function of individual utilities, W(U(1),U(2),...U(N)), not only permits, but insists on, interpersonal utility trade-offs whenever they would increase the value of W. Contractarians rightfully point out that an unspecified Bergsonian welfare function is potentially as dangerous to the sanctity of the individual as the classical utilitarian principle. But while contractarians may be sympathetic to neoclassicists' desire to protect the sanctity of the individual-just as neoclassicists were sympathetic to classical utilitarians' goal of establishing individual well-being as the yardstick of social worth--contractarians remain critical of the failure of neoclassical theory to recognize "that the plurality of distinct persons with separate systems of ends is an essential feature of human societies." 30
As Rawls points out, classical utilitarian theory is a complete theory of social choice totally consistent with a "conception of society [where] separate individuals are thought of as so many different lines along which rights and duties are to be assigned and scarce means of satisfaction allocated in accordance with rules so as to give the greatest fulfillment of wants." 31 And according to this conception, just as "each man in realizing his own interests is certainly free to balance his own losses against his own gains ... why should not a society act on precisely the same principle applied to the group? ... The principle for an individual is to advance as far as possible his own welfare, his own system of desires, the principle for a society is to advance as far as possible the welfare of the group." 32
In sum, classical utilitarian theory expresses the concept of social rationality as an extension of individual rationality and, ironically, provides the ideal justification for public enterprise, centrally planned economies rather than private enterprise, market econornies-a matter we will return to in chapter 9.
Neoclassical welfare theory, on the other hand, is neither a complete theory of social choice nor a consistent reflection of a particular view of society. Neoclassicists, impressed by their ability to deduce fundamental theorems relating the equilibria of competitive, private enterprise market economies with Pareto optimal outcomes without assuming cardinal or interpersonally comparable utilities, jettisoned the classical utilitarian principle. As a result, neoclassicists drifted into the uncharted sea of incomplete evaluative methodologies, which proved far more difficult to cross than they had supposed.
The incompleteness of neoclassical welfare theory appeared to be alleviated when the two fundamental welfare theorems relating the equilibria of competitive, private enterprise market economies and Pareto optima were interpreted as reason to post a "cease search" sign. But as we shall see in part 3, under similarly generous assumptions, equally compelling proofs establish the ability of a number of other "ideal" economic systems to achieve whatever Pareto optimal outcome may be desired. 33 As a result, the placement of the "cease search" sign becomes arbitrary. Which exhibit in the Great Economic Institution Fair will display the prize depends only on which booth the judges visit first. In this light, not only is there no alleviating the liability of incompleteness, but neoclassical welfare theory may have very little to tell us about the relative advantages of different economic systems.
While abandoning cardinal utilities for the choice theoretic approach and eschewing interpersonal welfare comparisons appeared to more accurately reflect the notion of the sanctity of the individual, as we saw previously, in the end neoclassical welfare theory also fails to adequately express the view "that the plurality of distinct persons with separate systems of ends is an essential feature of human societies," 34 due to its retention of a utilitarian form in general and the Bergsonian social welfare function in particular.
Although it might seem harsh, we are forced to conclude that neoclassical welfare theory is a philosophically unsophisticated halfway ground that retained a utilitarian format while attempting to express a notion of individual sanctity inimical to teleological theories of utilitarian form. Moreover, neoclassicists jettisoned the completeness contained in classical utilitarianism in the name of Occam's razor-a principle that applies only to jettisoning unnecessary equipment at embarkation, not cargo at disembarkation.
In contrast, the new contractarian approach is both a complete theory of social choice and a consistent reflection of a particular view of society. When combined with either lump-sum, redistributive transfers cum Rawls' difference principle, or "stand-pat" historical distributions cum Nozick's theory of entitlements, the institutions of private enterprise and competitive markets provide a complete answer to the economic problem for new contractarians. Furthermore, the justification is not that these institutions maximize some notion of social welfare, but that individuals with independent conceptions of "the good" would freely consent to these economic arrangements in a situation characterized by fairness.
In our opinion, the claim that only a contractarian doctrine fully expresses the view "that the plurality of distinct persons with separate systems of ends is an essential feature of human societies" 35 must be taken seriously. In sum, contractarianism seems a more appropriate expression of the sanctity of the individual so critical to the traditional welfare theory research project.
Although we have taken great pains to distinguish between the different versions of traditional welfare theory as they evolved, we now return to the primary task of this chapter: formulating the essence and principle strengths of traditional welfare theory when viewed as a whole. 36
It seems to us the most powerful statement of the desire to defend the sanctity.of the individual lies in its implicit challenge to alternative views: If the individual is not to be accepted as the best judge of his or her own needs and desires, then who is? Or, as John Stuart Mill put it:
(The individual's] own good, either physical or moral, is not a sufficient warrant (to interfere with preferences]. He cannot rightfully be compelled to do or forbear because it will be better for him to do so, because it will make him happier, because in the opinions of others, to do so would be wise, or even right. These are good reasons for remonstrating with him, or entreating him, but not for compelling him or visiting him with any evil in case he do otherwise. 37
Classical utilitarians tried to express this notion by denying that there was any hierarchy of pleasures, their intuition being that the greatest threat to the sanctity of the individual came from a paternalistic view that the " masses" were ill-equipped, through lack of education and cultural exposure, to enjoy "higher" pleasures.
As we have seen, neoclassical welfare theory's main bulwark for protecting the sanctity of the individual is the Robbins Principle, sometimes approvingly referred to as the principle of ethical neutrality. This principle directs economists to accept the individual's manifest preferences as part of the fundamental data for judging economic systems and policies and to leave unexplored and unquestioned the origins and merits of these preferences themselves.
The new contractarians insist that adequate protection of the sanctity of the individual must not only recognize the individual's inalienable right to define his or her own conception of "the good," but also guarantee the individual's freedom to pursue that conception without being overruled in the interests of society-at-large, unless the individual would have consented to do so in a fair situation. Contractarians argue further that all forms of utilitarian theories are incapable of protecting the sanctity of the, individual-whether the danger be from classical utilitarians' first principle in the form of individual pleasure machines, primitive accumulation, or wise rulers, or from neoclassicists' unspecified Bergsonian social welfare functionand insist that only contractarian theories are suited to this task. 38
There is no social entity ... that undergoes some sacrifice for its own good. There are only individual people, different individual people, with their own individual lives. 39
This view is common to all three variants of traditional welfare theory. As Rawls points out, classical utilitarians "held that the good of society is constituted by the advantages enjoyed by individuals,'-40 even if they are willing to permit the welfare of some to be sacrificed to the greater welfare of others. Neoclassical theorists are equally insistent that only individuals can experience welfare and are much more hesitant than classical utilitarians to discuss the merits of particular trade-offs between individuals' welfare.
Stated in negative form, all three variants of traditional welfare theory adamantly eschew the concept of "collective values" or "a social good" that is not derived from individual well-being.
The evolution of traditional welfare theory has helped sharpen the distinctions between different forms of evaluative theories. Rawls enumerates four disjoint and exhaustive categories of evaluative theories: utilitarian, contractarian, perfectionist, and intuitionist.
Utilitarian theories justify choices according to whether they maximize "the good," or in our case, some aggregation of individual welfares. Contractarian theories justify on grounds that choices would have been freely agreed to under circumstances characterized by fairness. Perfectionist theories judge social institutions and decisions according to whether or not they achieve some particular form of human or societal "excellence" without reference to the consequent well-being of individuals or whether or not individuals would have freely consented to abide by their rules. Intuitionist theories refuse to commit to any single form of justification and reserve the right to apply different forms of justification in different circumstances as determied by their "intuition."
Traditional welfare theorists are unanimous in rejecting all forms of perfectionist theories. Examples of perfectionist theories that express some form of "individual" excellence as their goal can be found in Aristotle and Nietzsche. A variety of fundamentalist Moslem, Christian, and Jewish sects are examples of "modern" perfectionist theories. Marxism has been characterized as "socially" perfectionist as well. And the label is apt for Marxists who make evaluative judgments based simply on the efficiency of policies in achieving the social "goal" of "communism" without justifying such an achievement in terms of the well-being of individuals, or on grounds that such a society would be chosen by fully informed individuals in a fair situation.
In any case, clarification of the nature of different justificatory procedures brought to traditional welfare theory principally by new contractarians is a positive contribution. And the unanimous rejection of perfectionist forms of justification is one of the strengths of traditional welfare theory as a whole.
The problem for traditional welfare theorists has been how best to secure a particular view of the sanctity of the individual while generating a complete theory of social choice. We have examined all three variants of traditional welfare theory to elucidate this problem. In the process we discovered that both classical utilitarian theory and neoclassical welfare theory fail to solve the problem.
While the new contractarian approach solves the problem posed by traditional welfare theory better than its predecessors-thereby better expressing the traditional welfare paradigm-it is just as incapable as its predecessors of resolving the remaining problems of welfare theory. We are now ready to demonstrate this point, which will be our task for the remainder of part 1. The reason is simple to state: the traditional paradigm itself is wanting as a vehicle for addressing the remaining issues.
The problem is no longer with the means of expressing the traditional paradigm. Neoclassical theorists honed the meaning of social efficiency and its relation to competitive equilibria. Then the philosophical incongruities of the inherited utilitarian format were corrected by a new generation of contractarians. The result is a technically and philosophically refined theory that fully expresses the traditional welfare paradigm. If traditional welfare theory has reached the end of the problems it can solve, it is not due to lack of refinement but to inadequacies of the underlying paradigm itself.