A Quiet Revolution in Welfare Economics- by Michael Albert and Robin Hahnel




In PART 1 we examined numerous failings of traditional welfare theory, tracing the origins of many of those failings to the failure to view people as conscious agents who develop within particular institutional contexts. While it might seem that the task of creating a new welfare paradigm would be monumental, if modern welfare theorists are willing to borrow from alternative social paradigms, much can be accomplished quickly. After all, welfare theorists have seldom elaborated the paradigms they employed in the past. In this area we have always exploited the advantages of "late comers" by borrowing freely from others.


5.1 People and Society

Individual people try to fulfill their needs and desires as they perceive them, frequently in cooperation with others. We are a self-conscious, social species because we define and seek to fulfill so many needs in relations with others. Moreover, in seeking to meet needs we identify today, we choose to act in ways that may change needs we perceive tomorrow.

Historically, people have created elaborate social organizations to meet their most urgent needs and desires. To satisfy their economic needs people have generated a variety of social arrangements that define and mediate frequently complicated divisions of responsibilities and rewards among participants. But we have also created intricate kinship relations through which individuals seek to satisfy sexual needs and accomplish their childrearing goals, as well as religious, cultural, and political organizations and institutions for meeting different needs. Of course, the particular social arrangements and the relations among them all vary from society to society. But what is common to all human societies is the elaboration of social relationships for the identification and pursuit of individual need fulfillment.

To develop a paradigm expressive of this view of humans and society we first concentrate on concepts helpful in treating individual people and then on concepts helpful in discussing social organizations and institutions in general. After developing means to combine the two, we distinguish between different "spheres" of social life and the possible relations between them. We conclude chapter 5 by elaborating a qualitative model for conceptualizing economic activity.

By no means do we offer an exhaustive treatment of all issues we touch upon. Nor can we offer here a full argument in favor of the paradigm we present. The interested reader should consult other works 1 Our goal in this chapter is limited to explaining the essentials of the paradigm sufficiently to use it throughout the rest of this book. As far as welfare theory is concerned, we are happy to judge the new paradigm on the basis of the economic results it proves capable of facilitating. 2


5.2 The Human Center

We are matter and alive, so we are presumably subject to the laws of matter and the laws of evolution. We are also a particular animal species, so presumably some species characteristics distinguish us from other animals. Not many insights are to be gleaned from dwelling on how humans are subject to the laws of matter, but the subjects of evolutionary laws and human nature provide a useful starting point for welfare analysis.

5.2.1 Huxley versus Kropotkin

Except for "creationists," most analysts consider "the laws of evolution" fairly straightforward and noncontroversial. Unfortunately, the dominant interpretation sprinkles more ideology over the scientific basis of Darwin's theory than most "secularists" care to admit.

T. H. Huxley inaugurated ideological interpretations of Darwin's theory of natural selection as survival of the fittest individuals in mortal combat. Peter Kropotkin responded with an alternative interpretation of natural selection in which species specific social relations contour individual competition within any species. Moreover, Kropotkin argued both logically and empirically that species specific social relations most often differ greatly from relations fostering hostile competition between individual members because effective tendencies toward what he termed "mutual aid" among members of a species was often useful in preserving species survival.3

Whether there is any more reason to assume human evolution produced an innate tendency toward intraspecies aggression than an innate tendency toward mutual aid or solidarity is an important debate. But for our purposes it is not necessary to embrace the "mutualist" view. All that is required to lend credulity to our undertaking is recognition that the view of humans as inevitably doomed to a perpetual Hobbesian joust by the inexorable laws of evolution may be exaggerated. If readers merely allow that the interpretation of evolutionary laws begun by Huxley known as "Social Darwinism" is controversial and may be more ideology than science, our arguments should be of interest. We attempt to motivate this kind of open-mindedness in the following section.


5.2.2 The Laws of Evolution Reconsidered

Human nature as it now exists was formed in accord with laws of evolution under conditions pertaining well before recorded history. Certainly conditions have since changed drastically; so if "historic time" had already lasted a million years, these changed conditions might have significantly altered human nature. But historic time has only been an evolutionary instant, making it doubtful history's altered conditions could have selected genetic characteristics significantly different from those bequeathed by our prehistory.

The obvious conclusion is that one cannot reasonably argue that human wars and pervasive hostile relationships between and within human communities during the human history known to us have caused our natures to become Hobbesian. Throughout recorded history we have been genetically essentially what we were at the outset. To believe otherwise is to believe that a baby plucked from the arms of its mother, moments after birth, five thousand years ago, and time-traveled to the present would be genetically different from babies born today. 4

Which is not to say that our history of war, oppression, and exploitation has had no impact. These aspects of our history have had important effects on our consciousness, culture, and social institutions that cannot be "willed away." But the point is that recorded history has left ideological and institutional residues, not genetic ones. Under changed institutional conditions it is possible our behavior-the combined product of our genetic inheritance and our institutional environment-need not be so Hobbesian.

This is not a conclusive case against the Hobbesian hypothesis. It merely establishes that it is unlikely the events of historic time made us innately aggressive. Of course, this only pushes the question farther back in time, and the Hobbesian hypothesis can still be formulated. But establishing the relevant time period can influence one's assessment of the plausibility of competing hypotheses. Human nature was not forged in the anvil of known human history but under the neolithic conditions of prehistory about which we know far less. Were these conditions more likely to "select" individually aggressive traits or traits that enhanced effective cooperative behavior?

If the answer to this question is not obvious, and if particular social institutions in recorded history would have elicited antisocial behavior even from people with sociable genetic dispositions, there is no more a priori reason to suppose that the "laws of evolution" have doomed us to be Hobbesian combatants than saints.


5.2.3 Human Nature

The simplest argument for the existence of a human nature is that if we are a unique species something must distinguish us from all other species. Bluntly, a human being is not a pigeon, and no conceivably rearrangement of environments can change this. Pigeon nature is not human nature. Our genetically determined physical and mental characteristics and potentials are different. We have a human nature in the strong, genetic sense. 5 Attributes that distinguish us from all other living things as a species, and they include more than simply our physical appearance are "wired in." 6

There is nothing mystical or idealistic in this. "Material" biological processes ensure that with human reproduction certain wired-in attributes remain essentially invariant over long periods of time. So we humans are not just what we eat. The human result is not merely an imprint upon a blank slate. If we were totally a product of our worldly situations, we would be perfectly and infinitely moldable. If this were the case, the effects of environmental differences would eventually mold infinitely flexible humans into shapes as different from one another as diverse patterns of clay subjected to different pressures. With no genetic structural similarities people subjected to different environments would presumably be incapable of communicating or understanding one another. If humans had unlimited behavioral flexibility, treated to "proper socialization" we would be as satisfied living in isolation as in community, repeating mindless tasks as employing creative intelligence, and hating and being hated as loving and being loved. The only cause for us to feel uncomfortable, "alienated," or "oppressed" would be a discrepancy between what we had been molded to desire and what we actually received. As long as we were socialized to expect what was coming and to desire and appreciate it, we would be as happy and fulfilled in one set of circumstances as any other.

This is the view of the behaviorist school of psychology that claims its origins in a specially chosen subset of the writings of B. F. Skinner. We are what we are "reinforced" to be; we want what we have been "programmed" to want. 7 Ironically, this is also the dominant Marxist view: "Marx wholly denied any abstract human nature," holding instead that "human beings are totally malleable and derive whatever nature they possess from their concrete sociohistorical environment," that we human beings "make our own nature." 8 But this is a sorry misuse of words. To present the fact that social environment influences how people develop as evidence that no such thing as "human nature" exists is a redefinition of the ordinary meaning of "human nature," not a refutation of its existence.

One might conceivably argue that "human nature" is unhelpful in studying and evaluating the functioning of human societies. But that is not the same as saying that there is no such thing as human nature. This claim could only be valid if constraints imposed by human nature rarely, if ever, played any role in important outcomes. The remainder of this chapter and the next present our picture of human nature and use it to construct a welfare paradigm whose application will refute the view that the results of human evolution can be ignored in analyzing human societies.

5.2.4 Natural, Species, and Derived Needs and Potentials

All people, simply by virtue of being human, have various needs, capacities, and powers. Some of these needs, like the needs for food and sex, or the abilities to eat and copulate, are shared with other living creatures. We call these our natural needs and potentials. Others, such as the needs for knowledge, creative activity, and love, and the powers to conceptualize, plan ahead, evaluate complex activities, and experience complex emotions, are more distinctly human. We call these our species needs and potentials. Finally, most of our needs and powers, like the needs for Bruce Springsteen albums, to share feelings with a particular loved one, or to know the rules of baseball, and the abilities to play the guitar, repair a roof, or solve differential equations, we develop over the course of our lives (though the underlying capabilities are natural or species traits). We call these our derived needs and potentials.

In short, every person has natural attributes similar to those of other animals and species attributes shared only by other humans, both of which can be thought of as genetically "wired-in." Based on these genetic potentials people develop particular derived needs and capacities as a result of their particular life experiences.

The first important point is that derived needs and capabilities must be developed within limits set by natural and species needs and potentials that are, in turn, innately fixed. We do not create our human nature alone, as individuals, or even together, as historically connected generations elaborating changing social environments. To all intents and purposes, wiredin results of past human evolution are not subject to modification by individual or social activity. Instead, our unchanging nature provides the foundation for everything we become in our lives.

This leads to the second important point: As individuals we create our derived attributes not only within limits set by our innate natural and species characteristics but also within limits set by our historically generated social environment. While all humans have certain innate needs and potentials, the forms in which these manifest themselves are always historically mediated, as are the interrelations between them.

The absence of any conception of species needs and powers in the traditional paradigm is a major weakness we wish to redress. The traditional paradigm lumps needs, desires, and whims under the umbrella concept "preference ordering." In its effort to protect the sanctity of the individual it eschews any restrictions on people's preferences other than what it portrays as "axioms of rationality." These minimal restrictions can be justified on the grounds that no theory can reasonably be expected to say much about totally erratic behavior, or on the grounds that regardless of how erratic people's actual behavior, it is of interest to know what rational behavior would be. But rational pursuit of totally unrestricted preference fulfillment is not the same as rational pursuit of preference fulfillment within the confines of an array of innately limited potential human preferences. In other words, traditional welfare theory is less a theory of rational human behavior and more a theory of rational behavior as it might be undertaken by infinitely programmable robots.

Admittedly, specifying restrictions on the potential development patterns of human preferences can be tricky. We do not wish to violate the sanctity of the individual. Nor do we wish to assert restrictions that may prove inaccurate. But we think there are ways around these problems that permit us to develop a theory of rational choice specifically appropriate to humans. The reason for our optimism is that we do not need to defend an inclusive list of species needs and powers to develop our theory and present our argument. Instead, a rather minimal, and we believe noncontroversial, list of restrictions on potential patterns of human development will suffice.

Human Consciousness. Human beings have intellectual tools that permit them to understand and situate themselves in their surroundings. This is not to say all humans attain an accurate understanding of their physical and social environment and their position in it. No doubt, most of us deceive ourselves greatly. But an incessant striving to develop some interpretation of our relationship to our surroundings, and to act in accord with that understanding seems endemic to normally functioning human beings. Put differently, human beings have the capacity to act purposefully and a need to exercise that capacity. We call this ability and associated need "human consciousness," which has a variety of attributes and is largely responsible for complicating human systems compared to nonhuman systems.

All animals engage in activities that both change the world and themselves, but only humans consciously choose activities in light of predicted effects on both their environment and themselves. In a sense, this adds a "third dimension" to human systems. In addition to human activity and the physical and social environments exists a third "mediating" dimension of consciousness. Marx explained this crucial species difference as follows:

A spider conducts operations that resemble those of a weaver and a bee puts to shame many an architect in the construction of her cells. But what distinguishes the worst architect from the best of bees is this, that the architect raises his structure in imagination before he erects it in reality. At the end of every labor-process, we get a result that already existed in the imagination of the laborer at its commencement. He not only effects a change of form in the material on which he works but he also realizes a purpose of his own that gives the law to his modus operandi, and to which he must subordinate his will.9

In other words, consciousness allows humans to select activities in light of preconceived effects on both their environment and themselves. One kind of effect our activities have is to more or less completely fulfill our present needs and desires, but a second effect is to either reinforce or transform our needs and desires. The capacity to analyze and evaluate these latter effects is what makes endogenous preferences so much more interesting from a welfare theoretic point of view than simple habit formation, and why, along with "preference fulfillment ... .. preference development" must become a central concern of welfare theory. It is also in this sense that we humans, unlike all other species, are potentially the subjects as well as the objects of our histories and can be "self-reproductive."

The human capacity to act purposefully implies the need to exercise that capacity. Not only can we analyze and evaluate the effects of our actions, we need to exercise choice over alternatives, and we, therefore, need to be in positions to do so. While some call this the "need for freedom," it is important to note that our "need for freedom" goes considerably beyond the "need for freedom" of many animal species. There are animals that cannot be domesticated or will not reproduce in captivity, thereby exhibiting an innate "need for freedom." But the human need to employ our powers of consciousness requires "freedom" beyond the "physical freedom" other species require. People require "freedom" to choose and direct their own activities in accord with their understanding and evaluation of the effects of that activity. Later we will define the concept self-management to express this peculiarly human species need in a way that subsumes the better-known concept "individual freedom" as a special case.

Consciousness implies another species trait, which although not necessary for arguments concerning welfare theory proper, is very helpful to understanding how humans and societies function. Because we see ourselves as choosing among alternatives, we frequently need to interpret our choices in some positive light. If we saw our behavior as completely beyond our own control, there would be no need to justify it, even to ourselves. But to the extent that we see ourselves as choosing among options, it can be very uncomfortable if we are not able to "rationalize" our decisions. We need an interpretation of our involvement with the world that judges our activities to be effectively oriented toward achieving our purposes and our purposes to be in some sense worthwhile. In other words, we have the need for a positive self-image.

If we could not consciously assess and control our behavior, presumably we never would have developed a wired-in need to justify it, even to ourselves. But to the extent that we came to see ourselves as choosing among options, it apparently became serviceable to survival to develop "stories" that made sense of our decisions. Evolution, we hypothesize, took this apparently useful trend far enough so that we now have a wired-in need for a positive self-image. Again, this is not to say all humans succeed in justifying their actions, even to themselves, much less that many who succeed deserve to do so! Nor do all social circumstances make it equally easy to develop a positive self-image. The point is simply that a striving to minimize "cognitive dissonance" appears to be a common human trait related to our species power of consciousness.

Finally, few other critics of traditional welfare theory have emphasized that traditional concepts fail to reflect the intrinsic importance of how and by whom choices are made. Since this would not be a problem for a welfare theory concerned with agents who were not self-conscious, this is an appropriate time to consider the issue.

According to neoclassical theory my level of happiness is a function of the quantities of material commodities I consume and the number of hours and kind of work I perform since my preference ordering is defined on commodity/work space. What if some master chooses my commodity bundle? What if I choose the same bundle myself? Neoclassical welfare theory concludes that my level of happiness, or the ranking of those two experiences by my preference ordering, would be exactly the same. But this is most unlikely for human beings.

A similar problem arises with respect to social choice. In terms of the Bergsonian social welfare function, it doesn't matter how weights multiplying each individual's level of utility are determined. Suppose a dictator chooses particular weights. Suppose the same set of weights are determined by mutual negotiation among all members of society in a "fair" environment free from information distortion. Neoclassical welfare theory concludes that if the "correct" social welfare function as defined by a given set of weights is maximized, social bliss is achieved irrespective of how the weights were determined. The more contemporary theory of social choice adds only the ad hoc principle of nondictatorship, a rather limited improvement as far as taking the decision-making process seriously.

The above problems stem from the fact that a theory of well-being relevant to agents with consciousness must account for how things get decided. Whether we are concerned with commodity bundles consumed by individuals or weights signifying the relative importance of different individual's opinions, more than just the bundles and weights matter. The process by which bundles are selected and the process by which weights are chosen matter in societies of conscious agents. Who chooses under what circumstances matters as well as what gets chosen. It is because we are conscious agents that one cannot separate ends from means.10

Not all variants of traditional welfare theory are equally insensitive to this problem. The contractarian approach expresses the importance of how choices are made both through its concept of free consent in an "original position" characterized by "fairness," and in its first principle of liberty." Later we will explain in what respects we find the contractarian solution to this problem deficient. But we hope by building on the positive parts of the contractarian attempt our paradigm can move us forward.

Human Sociability. Human beings are a social species in a number of important senses. First, the vast majority of our needs and potentials can only be satisfied and developed in conjunction with other humans. Needs for sexual and emotional gratification can only be pursued in relations with others. Intellectual and communicative potentials can only be developed in relations with others. Needs for camaraderie, community, and social esteem can only be satisfied in relation with others.

Second, needs and potentials that might, conceivably, be pursued independently seldom are. For example, people could try to satisfy their economic needs self-sufficiently, but they seldom have because establishing social relationships that define and mediate elaborate divisions of responsibilities and benefits has always proved so much more "efficient." And the same holds true for spiritual, cultural, and most other needs. Even when desires might be pursued individually, people have generally found it more fruitful to pursue them jointly.

Third, human consciousness contributes a special character to our sociability. Numerous other animal species are social in the sense that many of their needs can only be satisfied in conjunction with one another. But human beings have the ability to understand and plan their activity. And since we recognize this ability in others, we logically hold them accountable and expect them to do likewise. Our social ties, therefore, stem not only from mutual activity and dependence but also from mutual conscious understanding and expectation.

A more aesthetic expression of this social view of what it means to be human does not make it any less "scientific":

Kant called the realm of connection the kingdom of ends. Erich Gutkind's name for it was the absolute collective. My own term for the same thing is the human harvest-by which I mean the webs of connection in which all human goods (in the doubled sense of value and product) are clearly the results of a collective labor that morally binds us irrevocably to distant others. Even the words we use, the gestures we make, and the ideas we have, come to us already worn smooth by the labor of others, and they confer upon us an immense debt we do not fully acknowledge .... That is why Freud, trying to be as precise as possible, chose the word 'eros' to describe human vitality or energy. He meant by that to acknowledge the motion of all individual life toward the reality around it. We are the makers of value, all, and Eros, the force of life within the self, cannot ripen in the flesh until we find a place for it in the world. 12

It is the individualistic rather than the social view of human beings that is seen to be absurd and unscientific when examined more closely:

The individual cannot escape his dependence on society even when he acts on his own. A scientist who spends his lifetime in a laboratory may delude himself that he is a modern version of Robinson Crusoe, but the material of his activity and the apparatus and skills with which he operates are social products. They are indelible signs of the cooperation which binds men together. The very language in which a scientist thinks has been learned in a particular society. Social context also determines the career and other life goals that an individual adopts.... No one becomes a scientist or even wants to become one in a society which does not have any. In short, man's consciousness of himself and of his relations with others and with nature are that of a social being, since the manner in which he conceives of anything is a function of his society. 13

In fact, not even Robinson Crusoe was intelligible apart from the society from which he had become separated:

Production by isolated individuals outside of society-something which might happen as an exception to a civilized man who by accident got into the wilderness and already dynamically possessed within himself the forces of society-is as great an absurdity as the idea of the development of language without individuals living together and talking to one another. 14

In sum, contrary to contractarians' "hypothetical history," there never was a "Hobbesian state of nature" where individuals roamed the wilds of European forests in a "natural" state of war. Human beings have always lived in social units-tribes preceding clans and nuclear families. The roots of our sociability--our "realm of connection," or "human harvest'~-are both physical-emotional and mental-conceptual. The unique aspect of human sociability is that the "webs of connection" that connect all human beings are woven not just by a "resonance of the flesh" but by shared consciousness and mutual accountability. So it is especially true that humans do not exist in isolation from their species community. It is impossible to fulfill our needs and employ our powers independently of others. Not only have we never lived except in active interrelation with one another, we are meaningful only in this context; and no human with any sense would ever desire it to be otherwise! But before proceeding, to avoid misinterpretation we wish to clarify what we mean by human sociability.

The above argument claims that human beings are inevitably social and that this is true regardless of the particular form of society they inhabit. But we will later criticize the relative absence of sociability in both fact and consciousness in market economies. In chapter 7 we will argue that market institutions contribute to a loss of potential well-being described by Peter Marin in the following terms: "To lose touch with the social world is to lose a part of the self, and the fatigue now at work in our own lives is directly connected to its absence, for moral activity and a life in the larger world are necessary to our own well-being." 15 The question we wish to clarify is whether these two claims are contradictory. The answer is they are not, if one adopts the view we recommend in this book of the relationship between human characteristics and social institutions.

At times we treat human characteristics and social institutions as separate entities. Often there is nothing wrong with this. It can be convenient and introduces no serious distortions regarding some issues. But regarding other matters it is critical to focus on the "interplay" between human characteristics and social institutions. Once we recognize that people derive some of their characteristics, and that the availability of particular social roles (and unavailability of others) can influence developmental trajectories, the potential connection between role offerings and derived traits becomes clear.

The paradigm we are proposing entails a view of people and society in which each affects the characteristics of the other. It is in this way we can claim, without being contradictory, both that the human world is irrevocably social and that particular social institutions might inadequately reflect this essential human quality and thereby obstruct our ability to enjoy satisfactions that derive from development of our social potentials.

Human Character Structures. People are more than their constantly developing needs and powers. At any moment we have certain personality traits, skills, ideas, and attitudes. These structural characteristics play a crucial mediating role. On the one hand, they largely determine the activities we will select by defining the goals of these activities---our present needs, desires, or preferences. On the other hand, the structures themselves are merely the cumulative imprint of our past activities on our innate potentials. What is important regarding human characteristics is to neither underestimate nor overestimate their permanence. Although we have emphasized that people derive needs, powers, and characteristics over their lifetimes as the result of their activities, we are never completely "free" to do so at any point in time. Not only are people limited by the particular menu of role offerings of the social institutions that surround them, they are constrained at any moment by the personalities, skills, knowledge, and values they have accumulated as of that moment. But even though these character structures may persist over long periods of time, they are not totally invariant. Any change in the nature of our activities that persists long enough can lead to changes in our personalities, skills, ideas, and values. In this case, the new human structural characteristics can lead to new preferences as well.

We can illustrate using a metaphor from Ilya Prigogine's theory of dissipative systems applied originally in the field of thermodynamics. For Prigogine, energy and matter are continually flowing through "dissipative systems," but usually the system remains largely unchanged. That is, the manner in which energy and matter enter, leave, and affect the system does not change appreciably over time. Occasionally, however, the whole system is transformed-the manner in which energy and matter flow through the system is altered. Thinking of human beings as dissipative systems, we can imagine people constantly "consuming" material and information "inputs" and "producing" material and information "outputs." The usual result is to "reproduce" our human character structures intact. Or, put differently, the usual result leaves our relation to material and information flows unchanged. But sometimes the character structure is altered in some substantial way, and our reaction to material and information possibility sets is changed dramatically.

In any case, a full theory of human development would have to explain how personalities, skills, ideas, and values form, why they usually persist but occasionally change, and what relationship exists between these semipermanent structures and people's needs and capacities. No such theory now exists, or is visible on the horizon. But fortunately for welfare theory, a few "low level" insights are sufficient for our purposes.

The Relation of Consciousness to Activity. The fact that our knowledge and values influence our choice of activities is easy to understand. The manner in which our activities influence our consciousness and the importance of this relation is less apparent. As discussed above, the ability and need to act "knowledgeably" implies a need to rationalize what we do. But the tendency to minimize "cognitive dissonance" can work in two directions, and creates a subtle duality to the relationship between thought and action in which each influences the other. On the one hand, we seek to choose activities harmonious with our consciousness. But we also form consciousness that put activities we find we must engage in, for whatever reason, in a good light. So when we fulfill needs through particular activities we are induced to mold our thoughts to rationalize both the logic and merit of those activities as well, thereby generating consciousness/personality structures that have a permanence beyond that of the activities that formed them.

The Possibility of Detrimental Character Structures. Individuals' needs and powers at any moment are defined and constrained by their previously developed personalities, skills, and consciousness. But these characteristics were not always "givens." They are the products of previously chosen activities in combination with "given" genetic potentials. So why would anyone choose to engage in activities that resulted in characteristics detrimental to future need fulfillment and the maximum development of future powers?

One possibility is that someone else, who does not hold our interests foremost, made the decision for us. Another possibility is that we failed to recognize important structure-producing effects of current activities chosen primarily to fulfill immediate needs. 16 But imposed choices and mistakes are not the most interesting possibilities for welfare theory.

A fully informed individual might consciously choose to develop characteristics that fail to maximize future fulfillment if the "cost" in terms of lost present fulfillment were high enough. In fact, rational choice would require such behavior. Once we recognize that the present choice of activities affects our future characteristics and that our future characteristics influence our future preference structures, what we will term "preference fulfillment" and "preference development" become competing ends, and there is every reason to expect a "trade-off' between them. In this case, people with pressing present needs may well forego opportunities to develop traits that permit greater future fulfillment, whereas people with less pressing present needs or greater overall opportunities will concentrate more on future preference development.17

On the other hand, people may develop detrimental characteristics precisely because they orient their preference development so as to increase the fulfillment they can obtain under future conditions of activity availability. If institutional structures establish conditions in which some activities are available on easy terms while other activities can only be engaged in with great difficulty, it is sensible for people to choose present activities that develop traits that enhance appreciation of the former and minimize needs for the latter. As we will see later, the conditions under which these traits can be meaningfully termed "detrimental" must be analyzed with great care, and we leave that task to chapter 6. To foreshadow the results of that discussion, if the future conditions of availability reflect nothing other than the scarcities of productive means, the limitations of known technologies, and the desires of others, it is not sensible to describe the characteristics people develop to "cope" with these conditions "detrimental." On the other hand, if institutional structures that organize social activities exert an additional influence, so that some activities are accessible only on terms that exceed their true opportunity costs while others are accessible on terms that understate their opportunity costs, traits developed to accommodate these biased conditions can be meaningfully described as "detrimental." 18

In assessing the plausibility and extent of institutional biases of this kind we need to note the subtle but important difference between a conceptualization of choice among goods and choice among activities. We can plausibly imagine an infinite continuum of items and sizes when thinking of the supply of goods. It is problematic to envision an infinite continuum of activities to choose from when most activity is socially organized.

No society can offer an infinite spectrum of social activities. In large part what distinguishes different societies is precisely which activities appear on the "menu" citizens can choose from and which activities do not. In other words, most activities come in socially structured packages, and at least as far as individuals are concerned some packages are available and some are not. 19 So, if we think of the "objects" that impact on people's well-being as goods from which they choose, it is easier to ignore the possibility of biases in the conditions of availability. Whereas once we recognize that activities are what satisfy our needs to a greater or lesser extent and that a substantial portion of our activities are socially organized by important institutions, we are far more likely to consider the possibility of institutional biases seriously.

It seems the "problem of rational choice" has somehow been turned around. One view is people choosing goods from consumption possibility sets in light of given preference orderings. The other view is people choosing what preferences to develop in light of the terms of availability of different activities that are "given" by the institutions they face. 20

Summary: People are self-creative within the limits defined by human nature, but this must be interpreted carefully. At any moment individuals are constrained by their previously developed personalities, skills, and consciousness. Moreover, individually we are powerless to change the social roles defined by society's major institutions within which most of our activity must take place. So as individuals we are largely powerless to affect the kind of behavior that will mold our future character traits. Hence, these traits and any preferences that may depend on them remain to some extent beyond our reach, and our power of self-generation is effectively constrained by the social situations in which we find ourselves. But in the sense that these social situations are ultimately human creations and individuals have maneuverability within given social situations the potential for self-creation is preserved. We humans are both the subjects and the objects of history-as we might have expected. We define the concept of the Human Center to incorporate these conclusions.

THE HUMAN CENTER is the collection of people who live within a society including all their needs, powers, personalities, skins, and consciousness. This includes our natural and species needs and powers-the results of a human evolutionary process that we assume to be complete for purposes of studying known human history. It includes all the structural human characteristics that are givens for the individual at any moment, but are, in fact, the accumulated imprint of previous activity choices on innate potentials. And it includes our derived needs and powers, or preferences and capacities, which are determined by the interaction of our natural and species needs and powers with the structure of human characteristics existing at the time.

5.3 The Institutional Boundary

We have seen that people "create" themselves only in closely defined settings that limit their options beyond the limits of genetic human potential. In addition to the natural environment, and man-made objects or "artifacts," social institutions also "structure" people's "self-creative" efforts. We will refer to all three of these together as the "boundary" that "surrounds" the human center both defining and limiting its potentials. But we will also distinguish between our world's physical endowments (the natural environment), the physical transformations of the natural environment that take the form of both useful artifacts and environmental deterioration (the built environment), and social institutions that establish patterns of expectations that mediate activity (the institutional boundary), without implying a firm dividing line between them. 21

Social institutions are conglomerations of interrelated roles. At the risk of belaboring points that are commonplace for generations of "institutionalist" economists and the subject matter of introductory sociology courses, we clarify these matters that most economists seldom consider.

A factory's buildings, assembly lines, raw materials, and material products are part of the "built" environment. Ruth, Joe, Helen, and Sam, the people who work in or own the factory, are people and part of society's human center. The factory as an institution, however, is the roles and the relationships between those roles: assembly-line worker, foreman, supervisor, plant manager, pay clerk, union steward, minority stockholder, majority stockholder, member of the board of directors, etc.

Similarly, the market as an institution comprises the roles of buyers and sellers. It is neither the place where buying and selling occurs nor the actual people who buy and sell. It is not even the actual behavior of buying and selling (or, in the case of the factory, riveting, accounting, managing, etc.). The actual behavior belongs in the sphere of human activity, or history itself, and is not the same as the social institution that produces that history in interaction with the human center. Precisely stated, the market institution is the commonly held expectation that the social activity of material exchange will take place through the patterned activities of "freely" agreed to buying and selling. Insofar as such expectations are codified in rules, these rules are also part of the institution. But nothing more.

Please note, we have defined roles and institutions apart from whether or not expectations that establish them are actually fulfilled or rules that define them are actually obeyed. To think of roles or conglomerations of roles as fulfilled expectations or obeyed rules lends them permanence they may not deserve. A social institution only achieves structural continuity if the commonly held expectation or codified rules concerning potential behavior are confirmed by repeated actual behavior. But if institutions are defined as fulfilled expectations or obeyed rules, it becomes difficult to understand how they might change.

In order not to presume stability, we define institutions as commonly held expectations or codified rules of behavior and leave open whether or not these expectations or rules are or will continue to be fulfilled or obeyed. That is, whether or not particular institutions will maintain their structural continuity or be transformed should not be prejudged.

But why do institutions exist? Why must we add to the sphere of activity and the sphere of consciousness, a sphere of commonly held expectations about the behavior patterns of others? Once again, specifically human potentials and limitations enter our paradigm where there is a void in the traditional "choice theoretic" approach.

If we were all mind readers, or had infinite consultative time, or if decision making were infinitely fast compared to acting, human societies might not require mediating institutions. But if there is a "division of labor," and if we are not blessed with infinite time to search out and consult with countless others, we must act on the basis of expectations about the behavior of others. If you make a pair of shoes to sell in order to pay a dentist to fill your daughter's cavities, you expect others to play the role of shoe buyer, and dentists to render their services for a fee. You neither read the shoe-buyers' and dentist's minds nor take the time to arrange and confirm all these coordinated activities before proceeding to make each pair of shoes; so you must act on the basis of expectations about others' behavior in a market economy.

Stated differently, institutions are the necessary consequence of human's lack of omniscience, and we might well have listed this limitation as a species characteristic of human nature. Consequently, the relevant question about institutions is not whether they should exist but, as we shall see, whether particular institutions foster our development and fulfillment or pose unnecessarily oppressive limits. 22 If one insists on asking the question "where can I find the institutional boundary," the answer is that beyond codified rules of behavior which may be written down, commonly held expectations about individual behavior patterns are mental phenomena. But because our definition of roles and institutions locates these features where we have also located consciousness does not mean the two are the same. Expectations are different from consciousness. Consciousness makes it possible to change expectations, roles, and thus society's institutions. Animals cannot change their institutions. They do not create or adapt them, but simply encounter their institutions as part of their "wired-in" genetic inheritance. In contrast, humans inherit only the necessity of creating some social institutions as well as potentials for and limits on our abilities to do so. Our sociability and lack of omniscience impose the need for institutions as a condition of survival and fulfillment. But the specific institutions are, within the limits of our potentials, ours to design. 23

THE INSTITUTIONAL BOUNDARY is any society's particular set of social institutions understood as conglomerations of interconnected roles or commonly held expectations about appropriate behavior patterns. These roles exist independently of whether or not the expectations they represent will continue to be fulfilled and apart from whatever incentives might or might not exist for individuals to choose to behave in their accord. The institutional boundary is a necessary sphere in any human society because we lack omniscience, but is distinct from both the sphere of consciousness and the sphere of activity. While the institutional boundary necessarily constrains individuals' behavior, the sphere of consciousness makes possible purposeful transformations of the institutional boundary through social activity.

THE INSTITUTIONAL BOUNDARY is any society's particular set of social institutions understood as conglomerations of interconnected roles or commonly held expectations about appropriate behavior patterns. These roles exist independently of whether or not the expectations they represent will continue to be fulfilled and apart from whatever incentives might or might not exist for individuals to choose to behave in their accord. The institutional boundary is a necessary sphere in any human society because we lack omniscience, but is distinct from both the sphere of consciousness and the sphere of activity. While the institutional boundary necessarily constrains individuals' behavior, the sphere of consciousness makes possible purposeful transformations of the institutional boundary through social activity.

5.4 Complementary Holism

We want a paradigm more appropriate to specifically human social conditions and more sensitive to the relations among institutional structures and between institutional structures and human characteristics than the paradigm economists have traditionally relied on. Therefore, it is sensible to pay close attention to the possible relations between different kinds of social institutions and between the whole institutional boundary and the human center. Elsewhere we have called our treatment of such issues "complementary holism." 24 Here we outline enough of this approach to inform a paradigm relevant for welfare theory.

5.4.1 Four Spheres of Social Life

The economy is not the only "sphere" of social activity. In addition to creating social institutions to organize their efforts to meet material needs and desires, people have organized complex relationships for addressing their cultural and spiritual needs, intricate "sex-gender systems" for satisfying their sexual needs and discharging their parental functions, and elaborate political arrangements for mediating social conflicts and enforcing social decisions. So in addition to the "economic sphere" of social life we have what we call a "community sphere," a "kinship sphere," and a "political sphere" as well. While we are concerned here with evaluating the performance of the economic sphere, the possible relationships between that sphere and other spheres of social life deserve consideration.

A "monist" paradigm presumes some form of dominance, or hierarchy of influence among the spheres, while a "pluralist" approach studies the dynamics of each sphere separately and then attempts to "sum" the results. We espouse a "complementary holist" approach that assumes any form of dominance (or lack of dominance) among spheres will be historically contingent and is a matter to be determined by studying particular situations. Since all four spheres are socially necessary and can take many different forms, any patterns of dominance that may (or may not) result in a particular society and time cannot be determined by theory alone. Instead of a priori presumptions of dominance, complementary holism holds that a number of possible kinds of relations can exist among spheres. The possibilities are somewhat limited, but which possibility pertains in a particular situation can only be determined by empirical investigation.

5.4.2 Relations between Center and Boundary and between Spheres

The human center and institutional boundary, and the four spheres of social life are fundamental conceptual building blocks of the complementary holist paradigm. The human center and institutional boundary concepts include all four kinds of social activity, but distinguish between people and institutions. The concepts of spheres of social activity encompass both the human and institutional aspects of social activity but distinguish between different primary functions of different activities. The possible relations between center and boundary and between different spheres is obviously critical.

It is evident that if a society is to be stable people must generally "fit" the roles they are going to fill. Actual behavior must generally conform to the expected patterns of behavior defined by society's major social institutions. People must choose activities in accord with the finite number of role offerings available, and this requires that people's personalities, skills, and consciousness be such that they do so.

In our terms, there must be conformity between society's human center and institutional boundary for social stability. We must be capable and willing to do what is required of us. Suppose this were not the case. For example, suppose South African whites shed their racist consciousness overnight but all the institutions of apartheid remained intact. Unless the natures of the institutions were also changed, rationalization of continued participation in institutions guided by racist norms would eventually regenerate racist consciousness among South African whites. On a smaller scale, suppose one university professor eliminates grades, makes papers optional, and no longer dictates course curriculum or delivers monologues but instead awaits student initiatives. If students arrive conditioned to respond to grading incentives and wanting to be led and entertained by the instructor, elimination of authoritarianism in a single classroom in context of highly authoritarian expectations in the student body might well result in very little learning in the "participatory" classroom.

Whether the final result of any "discrepancy" between the human center and institutional boundary be remolding the center to conform with an unchanging boundary, or elimination of aspects of the boundary that prove incompatible with an unyielding human center, the point is stabilizing forces within societies usually act to bring the center and boundary into conformity, and lack of conformity is a sign of social instability.

But this is not to say that all societies' human centers and institutional boundaries are equally "stabilizable." While we are always being "socialized" by the institutions that confront us, this process can run into more or fewer obstacles depending on the extent to which particular institutional structures are compatible or incompatible with innate human potentials. In other words, just as "stabilizing" forces are always at work in societies, to date there have always been "destabilizing" forces as well resulting from institutional incompatibilities with fundamental human needs. For instance, no matter how "well oiled" the socialization processes of a slave society, a fundamental incompatibility remains between the social role of slave and aspects of innate human potential. That incompatibility is a constant source of potential instability in societies that seek to confine people to slave status.

The possible relations among spheres are somewhat analogous. It is possible that the organization of social activity in one sphere is highly compatible with the organization of social activity in another sphere. For example, it might be that the functioning of the nuclear family produces personality structures that are compatible with economic role requirements, and vice versa. On the other hand, it is possible for the activity in one sphere to be disruptive of the manner in which activity is organized in another sphere. For instance, the educational system in the kinship sphere might graduate more people seeking a particular kind of economic role than the economic sphere can provide under its current organization. This would produce "destabilizing" expectations and demands in the economic sphere and/or the kinship sphere. Some argued this was the case during the 1960s and 1970s when expansion of college education produced "too many" with higher level thinking skills for the number of positions permitting the exercise of such potentials in the U.S. economy, giving rise, in part, to student rebellion. In any case, at the broadest level, we might have conformity between spheres or destabilizing relations. But within the general category of conformity there are possibilities worth distinguishing further.

One kind of interaction between spheres consistent with stability is accommodation wherein critical features of one sphere accommodate to requirements of other spheres. For example, assignment of people to economic roles accommodates prevailing racial hierarchies in the community sphere if the income of minorities is less than whites and thereby consistent with their extraeconomic subordination. But it is also possible for the definition of roles in one sphere to reflect dynamics emanating from another sphere. For instance, if the economic role of secretary includes tending the coffee machine as well as typing, filing, and taking dictation, it is likely the role of secretary is defined not only by purely economic dynamics but by kinship dynamics as well. When compatibility is stronger than simple accommodation, we say the relationship between spheres is one of codefinition, and the activity of one sphere co-reproduces defining features of the organization of activity in another sphere. Thus, whereas social stability implies compatible relations between spheres as well as between human center and institutional boundary, the fact of stability does not tell us whether spheres merely accommodate or more actively coreproduce one another's defining features.

Finally, we should also note that in any society "underlying" stabilizing and destabilizing forces that exist between center and boundary and among different spheres may, of course, be complemented by more conscious efforts of particular social groups who seek to maintain or transform the status quo in light of whether they feel benefited or prejudiced by existing institutional arrangements.



5.5 A Qualitative Model of the Economic Sphere

A paradigm provides an orienting vision of what we intend to study. So an economic paradigm must project an image of what an economy is, what its components are, and what basic rules guide their interactions. We propose thinking of the economy as groups of people engaging in economic activities thought of, in turn, as economic processes, where for any period of time, t, each economic process is usefully characterized by six aspects:

1. All material objects and human efforts that enter the process as inputs

2. All material objects and human fulfillments that emerge from the process as outputs

3. The initial state of the individual human and group characteristics of the people carrying out the process

4. The final state of the individual human and group characteristics of the people who engaged in the process

5. The initial state of the tools (plant and machinery) used

6. The final state of the tools used

5.5.1 Flow Variables Described

First, we can usefully distinguish between material outputs of other human economic processes and materials that are not humanly produced, or natural resources. The purpose here is to make our paradigm useful for studying ecological relations between the natural environment and the economic sphere of social life. 25 We wish to distinguish between steel produced by economic processes in the present and oil "produced" by natural processes long ago. The latter exist in some ultimately limited supply regardless of our changing skills. 26

Second, when we refer to human inputs we are talking about mental and physical human efforts applied in carrying out particular tasks. Traditionally, such efforts are measured in terms of hours of different laboring activities expended. To account for intensity as well as duration, it is important to note that if effort is measured in hours, we implicitly assume average intensity and that above and below average intensity are possible. We also note that human inputs enter "consumption" processes as well as "production" processes, since consumption activity necessarily entails effort of a given intensity over a given time. 27

Third, by our definition no material natural resources can be produced as outputs of human economic activities, but there certainly can be material outputs such as steel and shirts, or services such as nursing care and research studies. 28 In the traditional conceptualization of "production" the primary, focus is precisely on material outputs. In our conceptualization there are also human outputs of all economic activities in the form of needs that are met more or less satisfactorily. In other words, whether it be laboring or consuming activity, all economic activity either succeeds or fails to fulfill the preferences of those engaging in that activity to some particular extent. We conceptualize these levels of fulfillment as the human outputs of economic activity. 29


5.5.2 State Variables Described

We find it useful to distinguish between inputs and outputs of "flows" that enter and leave an economic process and the initial and final values of "state" variables that characterize the human and physical center of the economic activity.

First, the physical state variables consist of the "tools" that the people who carry out the activity use, which in modern economies include buildings and various forms of frequently incredibly complicated machinery. Since the initial and final conditions of these physical state variables may differ as a result of the economic activity in which they are used, we identify initial and final values for their status.

Second, each of the individual people who participate in an economic activity have initial characteristics such as personality traits, talents, or skills, knowledge about technical and social aspects of the economic process they are presently engaged in as well as broader knowledge about how the world works in general, and attitudes or "values" toward what they understand to be going on in their proximity or farther away. As a result of carrying out their assignments, people's various individual characteristics may be transformed and the final values of these human state variables may differ from their initial values. This is a second human consequence of economic activity. In addition to generating well-being or preference fulfillment, economic activity can transform or reinforce any of a variety of human characteristics. 30

Finally, some human state variables we do not define individually because they are more appropriately conceptualized as characteristics of the group of people carrying out the activity. For example, suppose the group carrying out the activity is characterized by a particular kind of relation between its male and female members, or between members of different races or religions, or between those who have and those who do not have certain kinds of knowledge and information concerning the activity they jointly carry out. One possible consequence of carrying out the activity according to a particular assignment of duties and responsibilities might be to erode or reinforce any of these possible group characteristics. Therefore, we include social characteristics of the group of people engaged in an activity as one of our human state variables whose initial and final values we chart in our alternative paradigm.

5.5.3 Flow Variables Formalized

Any group of people, i, in time period t, can be thought of as engaging in a particular (kth) economic activity, Ak(i,t). A full description of any Ak(i,t) includes all the material and human inputs and outputs that "flow" into and out of Ak(i,t), as well as values for all the physical, human, and social state variables at the beginning and end of the period t.

If we can imagine a list of all available natural resources, then we can let

Ri = ( r1, ... rR )i represent the quantity of each resource "consumed" by the activity as an [i]nput.

Likewise, while its entries would all necessarily be zero, we can let

Ro = ( r1, ... rR )o represent natural resources "produced" by the activity as [o]utputs.

If we imagine a list of all producible materials (and services), we can let

Xi = ( x1, ... xX )i represent the quantity of each produced object or service "consumed" by the activity as an input, and

Xo = ( x1, ... xX )o represent the quantities "produced" as outputs.

We must be careful that the list of produced inputs includes all inputs "consumed" whether or not their "consumption" precludes their consumption by other activities and whether or not those who carry out the activity are in some sense held accountable for their "consumption." We must also be careful that the list of produced outputs includes all "bads" along with all "goods" irrespective of whether or not those who carry out the activity are held accountable for their production. With these provisos, our descriptive catalog of producible materials will be exhaustive.

If we imagine a list of all possible kinds of human economic efforts we can let

Ei= (e1 [%] , ...eE [%] )i represent the quantity of each kind of human input measured in units of time applied and degree of effort expended.

For what are traditionally thought of as production processes, these effort inputs would be hours of different kinds of labor at specific intensities. For example, if those performing a particular operation in an assembly line involving the jth type of effort were working at average intensity, we would record their hours on the job as ej [ 100%]. If they were working at an intensity level 25 percent above average, their human input would be recorded as ej [ 125%], while if working at 25 percent below average intensity we would record their input as ej [75%].

But what if the kth activity of group i during time t refers to "consumption" rather than "production"? While it takes time to drink a six-pack while consuming the Monday night football game, recording these twoplus hours of "effort" is unimportant. But in other cases "consumption" does require time and effort important to track. The time it takes to fill and check out your grocery cart can be as important as what objects are in it. The intensity of four hours of toy shopping the Friday after Thanksgiving as compared to four hours early in January matters to many with experience in such matters. A paradigm capable of approaching efforts of homemakers with the same seriousness as efforts of factory workers might prove useful in an area feminists have exposed as a theoretical cul de sac for others.31

Human outputs are degrees of fulfillment that individuals attain from carrying out their particular assignments in activities. Since there is a long tradition of treating different individual's satisfactions as incomparable, and since for many purposes there is no need to draw such comparisons, we need to track the fulfillment individuals obtain from participating in activity which in many cases is undertaken in concert with others. A long tradition in welfare theory treats individual fulfillment as a "ranking" of different possible activities according to whether they yield more or less overall satisfaction. This approach explicitly conflates different needs and the different degrees to which they are fulfilled under the traditional conception of individual preference orderings. We believe that much may be lost by such a conceptualization because individuals undoubtedly have different needs that are fulfilled to different extents by particular activities or bundles of activitiesand tracking all this may be important for some purposes. 32 But with some misgivings we follow the traditional lead of conceptualizing need fulfillment in terms of overall preference orderings.

This said, we can represent human outputs of any particular activity, Ak(i,t), by a vector each component of which represents overall satisfaction obtained by an individual member of group i. These are to be interpreted as selfjudged, interpersonally incomparable, ordinal rankings under the assumption of a given set of assignments in all other activities in which the individual participates. We thus have

U0 = (u1, ... uI)0 as the vector of human outputs for activity Ak(i,t), where the group i has I participants.

5.5.4 State Variables Formalized

Imagine a list of all the "machines" that might conceivably be utilized in any possible economic activity in our economy.

We can let the vector Mb = ( m1 [%] , ... mM [%] )b represent the "beginning" status of these machines in the locale of the economic activity.

The process begins with a stock of machine j in a given state of repair that will be recorded by mj [%]b where me is the number of machines of type j in the stock and percent is an "efficiency" rating for the machine ranging from 0 percent for a machine that has become completely useless to 100 percent for a new tool just successfully "broken in."

Likewise, we can let the vectori Mf = ( m1 [%] , ... mM [%] )f represent the "final" status of all machines associated with the activity center. Additions, deletions, or changes in the stock of machine j will be indicated by a difference between mj [%]b and mj [%]f.

We also want to treat individuals' human characteristics as state variables whose beginning and final status may differ. If we can list all the components of personality traits, skills, knowledge, and values toward what we understand to be going on, we can represent initial human characteristics of individuals participating in an economic activity with the vectors:

and his or her final characteristics with the vectors

Any additions, deletions, strengthening, or eroding of characteristic pl would be noted by a difference between plb and plt tracking the human development effects of participating in different economic activities. When combined with our above defined Uo we have both fulfillment and development human consequences of economic activities.

While tracking development effects for particular individuals is of primary concern for some welfare purposes, for other purposes it need not concern us precisely which members of group i undergo transformations. While our formal welfare theory will leave evaluation of development effects to individuals themselves and not treat them as interpersonally comparable, in less formal analyses we will "aggregate" these effects over all members of group i participating in Ak(i,t). Since for human "state" variables the difference between the final and beginning values is qualitative and directional, summing the difference over all individuals in the group provides a rough indication of the number of people enhancing or reducing a particular human characteristic as a result of the activity carried out by the group.

Finally, the representation of group human characteristics presents no individualization or aggregation problems. If we imagine an exhaustive list of social characteristics pertaining to a group of humans rather than distinct individuals, then the vectors

list the beginning and final status of these characteristics for the group.

And our overall image of any economic activity, for any group of people i, carried out during period t, would be:

This way of looking at a particular economic activity or process for any period is easily extended to a view of the whole economy at any period or to a view of the development of the economy over time. At any time, t, what is actually happening in the economy consists of Ak(i,t) with k running over all activities and i running over all groups carrying out activities. The development of the economy over time would be seen by observing Ak(i,t) with k running over all activities, i running over all groups, and t running over all time periods.

5.5.5 Possibility Sets, Choice, Information, and Incentives

As incredibly cluttered as our symbolic representation already appears, we have yet to incorporate most of what we need to know about the economic sphere. No matter how many actual economic activities we may chart for groups and time periods, all we will note is what actually happens. Presumably, for every economic activity any particular group i undertakes in period t, there exists a much larger set of activities the group might have carried out with alternative sets of inputs. 33 And while not all activities that are individually possible are mutually compatible, there is presumably a set of all economic activities compatible with one another at a point in time that is considerably larger than the actual outcome in the economy for that period. And while not all outcomes for the economy in different periods are mutually compatible, there is presumably a set of all mutually compatible economic histories that is considerably larger than the single economic development trajectory that occurs historically. In other words, since what actually happens is only one element from the set of all possible economic histories that could conceivably have occurred, so far we have dealt with only the tip of the formal representational iceberg.

Let us represent the set of all possible activities i can carry out in period t by {Ak(i,t)} Let us further represent the set of all possible mutually compatible activities for all groups in the economy at time t by {A(t)}. Finally, let us symbolize all the mutually compatible histories of economic activities for all groups in the economy over a span of time encompassing many periods by {A}. Since mutual compatibility requires that any vector of inputs to be used be available, clearly the set {A(t)} is considerably smaller than the cartesian product of the individual possibility sets {A(i,t)} for all i, and the set {A} is considerably smaller than the cartesian product of the sets {i(t)} for all t.

Nonetheless, the sets {A(i,t)}, {A(t)} and {A} are all multielement sets. And this poses the dilemma of how a specific element of {A(i,t)} comes to be chosen for each group in each period; how a specific element of {A(t)} gets chosen for the whole society for a single period; and how a specific element of {A} gets chosen for the whole society for a sequence of many periods. Conceptualizing this "planning problem" finally brings us to grips with the real complexities of interconnected human economies in which a great variety of possible combinations of conscious human and institutional decisionmaking processes choose among multiple possibilities at every juncture. All of which brings us to the importance of viewing a human economy not only as realized human activity-as we have above-but as an elaborate decisionmaking system, mixing conscious choices of individuals and groups with unconscious institutional limitations, processing vast stores of information to settle on particular outcomes from an array of possibilities.

Thus a complete understanding of any economy requires more than knowing all inputs and outputs and changes in material and human state variables that occur. We also need to know how, and by whom, or by what institutionalized process some Ak(i,t) was selected from {A(i,t)} , some A(t) was selected from {A(t)} and some A is selected from {A} . In some cases it might be relatively easy to identify a particular person or group who selected a particular option for the economy from a set of possible options. In other cases it might be clear that decisions were more or less "automatic" results of particular institutional requirements. But generally, we should be prepared to find complicated and subtle combinations of institutional requirements and groups of interested people together determining what occurs.

In dealing with these complex decisions, sometimes we will focus on the selection of a single element from a set of possibilities. Other times we will find it more useful to envision the "choice" process as elimination of subsets of various Ak(i,t)'s from consideration. By seeing how certain possibilities are "whittled" away, as well as how final options are selected from remaining possibilities, we will better understand how institutions influence economic choice.

To the extent individuals and groups make conscious choices, they do so based on goals they conceptualize and information they possess. If economic institutions influence what goals people formulate and what information does or does not come into their hands, they influence choices indirectly as well as directly. One way to make sure we do not ignore such matters is to think of an economy not just as a conglomeration of actual activities, but as an incentive system and cybernetic system as well. Even if we conceptualize individuals' overall goals as maximizing fulfillment over their lives according to preference orderings they develop, the institutions that organize social economic activity can powerfully influence how this goal is translated into a more specific incentive system. And even if we frequently grant the same "perfect knowledge" assumptions so common to the traditional welfare paradigm, at times we will be at pains to track what information comes to whom far more carefully. We hope a welfare paradigm that insists that the economy be seen as an incentive system and an information system as well as a network of activities will yield new insights by forcing more explicit treatment of assumptions and issues long overlooked.