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





1. Traditional welfare theory has, to a great extent, reached the limits of what it can accomplish

2. Further progress in analysis of the labor process, preference development, external effects, and institutional structures requires a new paradigm

3. A welfare theory based on a new paradigm can be elaborated sufficiently to consolidate and integrate current work in these areas, as well as yield substantive new contributions

4. A proper view of the welfare theoretic properties of markets, private enterprise, and central planning is substantially altered in light of the new welfare theory

Part 1 of this volume assessed traditional welfare theory and its weaknesses. Part 2 elaborated a new paradigm and welfare theory including some useful abstract applications. Part 3 reanalyzed familiar economic institutions and models and reached original conclusions. Here we briefly summarize our results and the implications for designing better economic institutions.


10.1 Limits of Traditional Welfare Theory

As we saw in chapter 1 the crowning achievement of traditional welfare theory was the increasingly rigorous demonstration of what are known as the three fundamental theorems of welfare economics. Under appropriate assumptions: 1

1. Perfectly competitive private enterprise market economies have general equilibria

2. Any general equilibrium of a perfectly competitive private enterprise market economy is a Pareto optimum

3. Any Pareto optimum can be achieved as the equilibrium of a perfectly competitive private enterprise market economy with appropriate initial endowments

The obvious implication drawn from these results was that while real world problems might prevent us from implementing a perfectly competitive, private enterprise economy or might prevent such an economy from reaching its equilibrium, this kind of economic system is socially efficient and completely "flexible" regarding distributive outcomes. No alternative could be superior.

So while neoclassical theory admitted providing no compelling argument for choosing among Pareto optima, neoclassical economists nonetheless confidently placed a "cease search" sign in front of the PrEME exhibit at the "Great Economics Fair." The sign substituted for closure by implying that debates in which contractarians criticize neoclassicists' inability to "close" the "theory of social choice" can be placed on a back burner.

But these conclusions are misleading. First, the neoclassical efficiency and flexibility theorems fail as soon as we correct for the traditional paradigm's inadequate conceptualization of the labor process, external effects, and endogenous preferences. Second, under equally "permissive" (mis)assumptions, the same results obtain with equal plausibility for every other major theoretical economic model. Specifically, existence, efficiency, and flexibility theorems can be proved for public enterprise, competitive market economies of both state-managed and employee-managed varieties, and for public enterprise centrally planned economies. In an even-handed application of traditional theory, placement of the "cease search" sign at the "Great Economics Fair" is completely arbitrary. It depends only on which stall the judge visits first. 2

Traditional theory consigned the labor process to a "black box" by assuming an exogenous production possibility set that eliminated any need to analyze complexities of the labor process. But when we reviewed the work of economists whose interest in discrimination, labor market structure, and the implications of getting employees to "honor" the wage contract drove them beyond the traditional paradigm, we uncovered good reasons to question the efficiency of private enterprise production as well as the conclusion that competition for profits tends to ameliorate discrimination.

In the case of externalities and public goods, we discovered the traditional paradigm underestimated the problem and failed to provide concepts necessary for treating the problem. By continuing to champion the concept of ordinal, noncomparable preferences over individual commodity spaces the traditional paradigm stalled thinking in the theory of public finance. Labeling nonmarket allocative mechanisms "political" rather than .,economic" is hardly an incentive-compatible mechanism for economic theorists! So, not only did the traditional paradigm obstruct research in incentive-compatible mechanisms, it bears primary responsibility for perpetuating the myth that competitive markets yield socially efficient allocations in most situations.

While traditional theory never denied that tastes change, traditional theorists argued that economists need not concern themselves with changing preferences. As a matter of fact, a critical part of the traditional paradigm, the Robbins Principle, served as a methodological injunction prohibiting economists from addressing where preferences come from. In reviewing the work of the few economists who dared to violate this injunction we discovered there was every reason to consider preference development as well as preference fulfillment meaningful and every reason to assume that individual preferences depend on economic institutions as well as cultural, political, and kinship institutions. We also discovered that by ignoring preference development traditional welfare theory systematically misestimates welfare effects of different economic choices. But fundamental disagreements divided those who cared to investigate the implications of endogenous preferences for welfare theory. Gintis claimed endogenous preferences rendered the fundamental welfare theorems false. Pollak disagreed.

Traditional theory certainly does not ask how economic institutions influence preference development. But different preference development effects generate crucial welfare theoretic differences between major economic institutions! Since there is precious little to choose between well-known economic models when viewed in traditional terms, the failure to explore the developmental effects that distinguish different "ideal" economic systems is even more glaring. The traditional paradigm confines comparisons to the "old welfare debate," which is essentially over, and prevents theorists from moving on to the "new welfare debate" over how economic institutions affect incentives and the development of human characteristics and preferences.


10.2 A New Welfare Paradigm and Theory 

In chapter 5 we presented a new paradigm stressing the importance of human "natural," "species," and "derived" needs, powers, and characteristics and their mutual interaction with economic roles and institutions. Using this paradigm, in chapter 6 we elaborated a new welfare theory and proved a number of important theorems that enhance the analysis of formal economic models.

The first two theorems established a preliminary case that treating preferences as endogenous has important implications for welfare theory. The first theorem, 6.1, translates a result from Gintis into our terminology.

Theorem 6.1. A neoclassical welfare theory that ignores the fact that present choices of consumption and work activities not only fulfill present preferences or desires but also generate changes in future preferences will consistently misestimate the welfare effects of economic choices.

The second theorem formally establishes a result originally intimated by Gintis, that coincidence between supplies and desires might be indicative of system stability rather than efficiency. If the coincidence is the result of the economy adjusting relative supplies to better accord with people's preferences, it is a sign of the kind of efficiency traditionally called "consumer sovereignty." But if the coincidence results from individuals' rational adjustments of their preferences to better accommodate conditions of supply, it is not "meritorious" but merely "system stabilizing."

Theorem 6.2. In general, not only will rational individuals who recognize their preferences are endogenous adjust to changes in relative terms of supply by changing the relative amounts of those goods they consume, but they will also change their preferences-reducing their preference for relatively more expensive items and increasing their preference for relatively less expensive items-by changing consumption-work choices in earlier periods to change their future human characteristics and the preferences depending on them.

Theorems 6.3, 6.4, and 6.5 reestablish the three fundamental theorems of welfare economics for private enterprise competitive market economies under all the traditional assumptions even if people have perfectly informed endogenous preferences.

Theorem 6.3. Under the usual assumptions, the existence of a general equilibrium in a private enterprise, competitive market economy is assured even if individuals' preferences are endogenous.

Theorem 6.4. Under the usual assumptions, any general equilibrium of a perfectly competitive, private enterprise economy will be a Pareto optimum even if individuals' preferences are endogenous.

Theorem 6.5. Under the usual assumptions, any Pareto optimum can be achieved as the equilibrium of a private enterprise, competitive market economy with the appropriate set of "initial endowments" even if individuals' preferences are endogenous.

These theorems establish that perfectly informed endogenous preferences, in and of themselves, do not disrupt the ability of private enterprise market economies to achieve efficiency and flexibility, Gintis' speculation notwithstanding. Although that settles one debate between Gintis and Pollak, as we discovered, the interpretation of these theorems can be highly misleading.

The possibility of misinterpretation arises because the phrase "under the usual assumptions" not only plays a legitimate technical role in the above theorems, but also hides a "multitude of sins." For Theorem 6.3, the critical technical assumption is that our endogenous well-being functions, WBi ,display the same convexity properties traditionally granted exogenous preference orderings. For Theorem 6.4, the critical assumption is that our endogenous well-being functions,  WBi
do not define any "thick" indifference classes. And for Theorem 6.5, both these technical assumptions are necessary. Since the entire difference between endogenous and exogenous preferences is contained in the difference between our well-being functions, WBi , and their exogenous counterparts in traditional theory, extending these "usual" assumptions, though not without complications, would legitimately reestablish the fundamental welfare theorems.

But other assumptions are part of the "usual assumptions" granted by traditional theory. While exogenous production possibility sets and absence of externalities have nothing directly to do with endogenous preferences, they are necessary assumptions to establish Theorems 6.4 and 6.5. Those assumptions amount to assuming that nothing inherent in the functioning of the institutions of private enterprise and competitive markets prevents them from organizing production efficiently and supplying both goods and jobs in accord with people's preferences. In chapters 7 and 8 we presented good reasons for not believing that markets and private enterprise are capable of such behavior, even under the most favorable circumstances. In this context, the phrase "under the usual assumptions" hides the multitude of sins that necessarily prevent private enterprise market economies from achieving the results stated in theorems 6.4 and 6.5. But while private enterprise market economies will not achieve the above results, these failures are not in and of themselves due to endogenous preferences.

As a matter of fact, what the three theorems establish is that endogenous preferences, in and of themselves, cannot affect the efficiency and flexibility properties of any economy that fully adjusts to people's preferences. Or put differently, if there is no bias in the conditions upon which an economy makes different activities available, the fact that people have perfectly informed endogenous preferences will not prevent the system from having equilibria, or deter the system's efficiency or flexibility. 3 But this does not mean that endogenous preferences don't affect the results of welfare theory.

Theorems 6.6, 6.7, and 6.8 establish die most important formal results of our new welfare theory. In the case of any economy that contains any kind of bias in the terms upon which it makes different activities available, these theorems demonstrate that the extent of the misallocations that will result are far greater, and the nature of those misallocations are far different from what has been traditionally believed by those who treated preferences as exogenous. In other words, these three theorems establish that endogenous preferences do matter to welfare theory, but for reasons different than those previous critics believed.

Theorem 6.6. In an economy that contains a bias in the relative terms of supply of two economic activities: (1) the degree of nonoptimality will be greater than indicated by traditional welfare theory that treats preferences as exogenous; (2) the divergence from optimality will "snowball" or grow over time in a manner that can be precisely defined.

Theorem 6.7. In an economy that contains a bias in the relative terms of supply of two economic activities: (1) individuals' human development patterns will be "warped" in a manner than can be precisely defined; (2) warped human characteristics will "snowball" over time.

Theorem 6.8. The full effects of a bias in the economy will be disguised to analysts who view preferences as exogenous and to participants in the economy who adjust unconsciously and/or forget they have done so after the fact.

In sum, any economy that exhibits no bias in the conditions upon which it makes economic activities available will generate socially efficient outcomes if people's preferences are endogenous, just as it would if people's preferences were exogenous (Theorems 6.3 and 6.4). Likewise, if an economy without biases was "flexible" regarding equity assuming exogenous preferences, it will remain so assuming endogenous preferences (Theorem 6.5). However, to the extent that preferences are endogenous, any economy that contains a bias will generate more inefficiency than would have been anticipated assuming exogenous preferences, and the inefficiency will snowball over time (Theorem 6.6) so that rational individual adjustment to bias aggravates the bias and results in ever-increasing losses of potential well-being. Thus we can legitimately label the behavior "self-warping" (Theorem 6.7) and point out that it is an important example of how individual rationality can be socially counterproductive. Ironically, the process of individual rational adjustment-or self-warping-may prove socially stabilizing if its inhabitants adjust unconsciously or forget they have done so and come only to see themselves as getting what they seek (Theorem 6.8).

While these formal results presumed nothing whatsoever about what kinds of preferences people may have, beyond the fact that those preferences are, to some extent, endogenous, we also presented arguments in favor of human development trajectories characterized by self-management, solidarity, and variety. We outlined four different justifications for these welfare goals: one justification centered on fulfilling particular aspects of human nature; the second applied a modified contractarian methodology; the third was based on the criterion of species survival; and the fourth interpreted our welfare criteria as expressive of the values of the "humanist" heritage.


10.3 New Results

In part 3 we used our new welfare theory to analyze the major economic institutions used today to organize production (private and public enterprise) and allocate goods and resources (markets and central planning). In every case we came to a number of conclusions at odds with, traditional beliefs.


10.3.1 Markets 

Proponents of markets have silenced their opponents in ingenious ways. External effects are so pervasive that open debate was unthinkable. Yet failure to finesse the issue threatened over two hundred years of economic theory. The solution was to bury the problem of external effects within a traditional paradigm regarded as necessary for generating any welfare theoretic conclusions whatsoever. The message was: those who ask just how "exceptional" external effects really are had best be prepared for early retirement.

Those with other employment opportunities who argued that the emperor had no clothes, such as E. K. Hunt, were simply ignored. Pure theoreticians, who tackled problems as they proved susceptible to new theoretical toys, such as Theodore Groves and Co., were potentially more dangerous since they have special status within the establishment. But theorists are seldom politicians. They do not devote much time to debating the importance of the mind games they play and are easily distracted by new toys that appear with a new holiday season which lend themselves to a different set of gedanken experiments. 4

There are other reasons why market failures have gone so remarkably undernoticed. Losses due to "market failure" have, until very recently, been much less visible than other kinds of waste. Market failure does not appear in the form of 1,000 lb. nails or people working one hour out of eight on the job, which provide vivid images of the inefficiency of central planning. And it is the inefficiency of market disequilibria, not market failure, that is stamped in people's minds by unemployment lines coincident with stockpiles of unsellable goods. Market failure does not even show up as slow economic growth rates, which current Soviet developments demonstrate can shake even long-established economic orders. Losses due to market failure are not subtracted from GNP estimates. Vivid reminders such as Three Mile Island, Chernobyl, and Valdez, Alaska, are still recent, and the ecology movement that seeks to publicize the most heinous examples is still in its political infancy. But irrespective of its relative lack of notoriety, the fact remains that in postfeudal history a plausible case can be made that no economic failure has contributed more to the waste of productive resources than misallocations of markets uncorrected for external effects.

Our formal contribution to a critical reevaluation of markets is contained in Theorems 7.1 and 7.2.

Theorem 7. 1. Markets overcharge purchasers of goods with greater than average positive external effects.

Theorem 7.2. Not only will markets misallocate resources initiallyundersupplying goods with greater than average positive external effectsbut there will be a cumulative divergence away from optimal allocations over time as individuals "rationally" adjust their personal characteristics to diminish their needs for goods with greater than average positive external effects.

In verbal argument we detailed the reasons for expecting most activities to diverge from the average degree of external effects, thereby establishing the prevalence, rather than exceptionality, of the problem. We also explained how the cybernetic and incentive characteristics of market institutions aggravate the allocative bias and how individually rational adjustment compounds the social irrationality.

We also clarified how an elementary confusion between "Pareto optimal " and "Pareto improvement" had blinded public finance theorists about "solutions" to the "free rider" problem. We pointed out that simple proportional taxation is as likely to yield truthful reporting and efficient outcomes as the more complicated taxing schemes discussed in the "incentive-compatible" literature. Although the traditional objection to proportional taxation was really on the grounds of equity rather than efficiency, we explained why the "incentive-compatible" literature is, nonetheless, important as well as innovative. Besides efficiency and alternative notions of equity at stake in the choice of taxation schemes is what kind of incentives will be established for individual preference development. We argued that, in the end, the prospects for diversity or conformity in people's tastes are at stake. And we showed how external effects may imply a fundamental trade-off between short-run efficiency and diversity of tastes. In sum, we presented a number of criticisms of markets not usually emphasized and suggested a novel interpretation of the "incentive-compatible" literature. We also provided a rigorous explanation of why commonly expressed fears among noneconomists that markets are "socially alienating" are well founded despite disclaimers from "professional economists" that markets are ideal allocative institutions.


10.3.2 Private Enterprise 

Our major goal in chapter 8 was to solidify the claim of the "conflict school" that private enterprise production will be inefficient and aggravate economic discrimination and to make clear this conclusion obtains even in conditions of perfect knowledge and competitive labor markets. But before treating this debate we clarified the conditions under which the distributional flexibility theorem holds in both private and public enterprise environments.

First, we pointed out that the flexibility theorem only holds in either setting in a Debreuvian world of perfect foresight where taxing authorities are free to enact any handicaps they may choose, in all time periods, once, and only once, before any races begin. Second, we pointed out that for both political and ideological reasons different kinds of economies will gravitate toward particular distributional "maxims." We showed that while flexible in theory, PrEMEs have practical tendencies to distribute according to the market (mis)judged productivities of individuals' human and nonhuman capital. And that while equally flexible in theory, PuEMEs will tend to distribute according to the market (mis)judged productivities of individuals' human capital. After pointing out that neither principle is the equivalent of distribution according to effort and sacrifice, we proceeded to clarify the most important insights of the conflict school of private employment. Formally our summary of the logic of the conflict school is contained in Theorem 8.1.

Theorem 8.1. Unless there is 100 percent labor turnover each time period, profit maximization under competitive conditions with perfect knowledge implies that any kind of laboring activity that generates employee empowering traits will have an actual market wage that is less than the socially optimal wage and be undersupplied. And any kind of labor activity that weakens employee empowering traits will be paid more than the socially optimal wage and oversupplied.

Various writers in the conflict school have argued that work opportunities under participatory, cooperative, and fair conditions constitute laboring activities of the first kind-which will be undersupplied according to Theorem 8.1-whereas work opportunities under discriminatory conditions or in situations with artificial hierarchies constitute laboring activities of the second kind-and will, therefore, be oversupplied. So the common thread running through the conflict school is that socially counterproductive practices such as wage and employment discrimination, exaggerated hierarchies, and de-skilling are part of profit maximization.

Theorem 8.1 clarifies the basis for such behavior and its implications for social efficiency. These practices are all cases of working conditions that tend to erode employee bargaining strength. The primary mechanism through which this occurs is that working under such conditions affects both the individual and group characteristics of employees in future time periods. But while it would be irrational for profit-maximizing employers to ignore these effects when considering technologies and reward structures, by doing so they establish wage rates that diverge from "efficiency prices" for labor services and misallocate productive resources. The social inefficiency is transparent once the divergence between private benefits and social benefits is spelled out.

Theorem 8.2 demonstrates that the fact that people's preferences are endogenous will compound the problem.

Theorem 8.2. Not only will production under private enterprise fail to deliver optimal job mixes in some initial time period--oversupplying work conditions that empower employers vis-a-vis employees-but there will be a cumulative divergence away from optimal allocations in future time periods as individuals "rationally" adjust their personal characteristics to diminish their preferences for work opportunities that are underpaid and enhance their preferences for work opportunities that are overpaid.

In this case, individual rationality by employees proves socially counterproductive by causing inefficiencies generated by employer rationality to grow over time. In conclusion, we pointed out that according to our analysis the traditional model of "perfectly competitive capitalism," or what we prefer to call PrEMEs is "doubly" flawed and can be expected to increasingly diverge from socially efficient allocations for two separate reasons--one associated with each of its defining institutions.


10.3.3 Central Planning 

The "old debate" on central planning should have been closed long agoWe demonstrated that under traditional assumptions public enterprise, centrally planned economies have as much claim to social efficiency and distributional flexibility as their major competitors, and this remains true under the assumption of endogenous preferences as well. We also argued that central planning fares better in the "modern debate" than usually admitted. When information and incentive problems are taken seriously, rather than assumed away by traditional assumptions, we showed how proponents of central planning might still claim efficiency properties for their system. Ironically, their ability to do so is largely due to work on -iterative planning mechanisms" by Western economic theorists, who are by no means proponents of central planning, but who nonetheless developed credible procedures central planners could use to gather information about the technological capabilities of production units. Unfortunately, these information gathering procedures are not -incentive-compatible." But is it naive to think Western theorists, armed with principal-agent models who began work on incentive-compatible mechanisms only recently, may come up with procedures to ameliorate this problem as well? In any case, proponents of PuECPEs can already point out that the same "material incentives" that private enterprise and market economies rely on can be deployed to ensure that the plan calculated is carried out. In sum, a plausible case can be made that, even in terms of the -modern debate," central planning can lay claims to efficiency and flexibility just as a variety of market systems have done.

Yet our major purpose in chapter 9 was not "to praise Caesar." We proposed a "new debate" over the welfare theoretic properties of central planning, highlighting the fundamental flaw in this allocative mechanism even under the most favorable assumptions. Our formal argument is summarized in our third imperfection theorem, Theorem 9. 1, and our last snowballing theorem, Theorem 9.2.

Theorem 9. 1. Centrally planned economies will "charge" individual job seekers (in the form of lower wage rates) more for self-managed work activities compared to other activities than is socially optimal resulting in less self-managed work being performed than is socially optimal.

Theorem 9.1 holds because even the most democratic system of central planning is not capable of providing greater than average decisionmaking authority to those working in a particular production unit concerning the fundamental choices about how work will be done in their unit. Instead, - democratic" central planning gives everyone equal say in determining the social welfare function and then mechanically translates that social welfare function into specific production plans for each unit in the economy. This implies that workers in other plants have as much say in determining how work will be done in my work place as I have, and vice versa. Since the effects on those who work someplace and on those who do not are unequal, equal decision-making input fails to approximate self-management.

This is a more fundamental criticism of central planning than those voiced by opponents in the "old" and the "modern" debates. We have shown that even if a Central Planning Board had complete knowledge of the production possibilities of all individual units, and knew all primary resource availabilities and initial stocks of capital goods, even if the social welfare function were determined by democratic "point" voting and voters made no mistakes in evaluating both the fulfillment and development effects of different goods and activities, and even if incentives guaranteed that the plan would be carried out perfectly, central planning would still have a bias against provision of self-managed work activities. Moreover, our new welfare theory demonstrates that when we combine the inherent bias of central planning against self-managed work with recognition of endogenous preferences, the inefficient organization of work "snowballs":

Theorem 9.2. In a centrally planned economy not only will self-managed laboring activities be underpaid and undersupplied at some initial point in time, but the degree of divergence from optimality will be greater than indicated by traditional welfare theory and grow, or "snowball," over time. Accompanying the snowballing nonoptimal allocations will be a "snowballing" apathy consisting of "warped" human characteristics which undervalue self-managed work activity.

But workers' individually rational adjustments are not only socially irrational in that they increase social inefficiency, their adjustments reinforce the authoritarian characteristics of central planning. In general terms, the problem with central planning as seen from every vantage point of our new welfare theory is that it generates authoritarian dynamics. Central planning minimizes producers' knowledge of how they fit into the rest of the economy and generates a monopoly of technical information in the hands of central planners. Central planning embodies command relationships between the planners and production units, which, in turn, promote similar command relationships within production units. And no matter how democratic the process of choosing the social welfare function may be, inherent bias against provision of self-managed work activity "snowballs" as people "rationally" lower their desires for self-managed work.

So, unlike traditional welfare theory that should find central planning equally as "efficient" and "flexible" as public and private enterprise market systems, our evaluation provided a critique of central planning on theoretical as well as practical grounds.


10.3.4 Summary

Until very recently only a small number of abstract economies had been seriously analyzed by welfare theorists. The traditional names for these economies are Perfectly Competitive Capitalism (PCC), Centrally Planned Socialism (CPS), and two versions of Market Socialism (MS). Since each of these economies is simply a combination of particular institutions for organizing production and allocation, we proposed naming them according to their defining institutions. This not only had the advantage of not prejudging what was and what was not a "socialist" economy, but might defuse prejudice and improve the quality of debate. In any case, we simply called the economies: private enterprise market economies (PrEMEs), public enterprise, centrally planned economies (PuECPEs), public enterprise, employeemanaged, market economies (PuEEMMEs), and public enterprise, statemanaged, market economies (PuESMMEs).

We argued that under traditional assumptions the "old debate" about the welfare theoretic properties of these different economies has been laid to rest. According to the norms of traditional welfare theory, provided they are not clouded by ideological prejudice, each of the above economies can be shown to be socially efficient and flexible. Under the usual assumptions, an appropriate existence, efficiency, and flexibility theorem holds for each of the above systems even when we allow that people's preferences are endogenous.

However, this means traditional welfare theory cannot distinguish between these models in terms of their welfare theoretic properties-a small return for a not inconsiderable investment in technique. Recently, what we called the "modern debate" has relaxed the traditional assumption of perfect information. How central planners might gather information about productive unit capabilities and the quality of such information has been examined. The effects of different kinds of contracts in light of asymmetric information between private employers and their employees has also received attention. As a result, to some extent the modern debate is better able to distinguish between the different systems, which do have different information and incentive properties and, therefore, different problems in these regards.

Moreover, we completely support such efforts. A more realistic treatment of information is long overdue, and this means relaxing the "perfect information" assumption of the traditional paradigm. But in our view important welfare theoretic differences can be noted without dropping the traditional perfect knowledge assumption. In what we called the "new debate," the project of analyzing the relation between fully informed individual rationality and social rationality in different institutional contexts does not fail to draw important distinctions between the major economic systems of the day. Thus before relaxing the perfect knowledge assumption, we recommend clarifying the important welfare theoretic differences between economic systems assuming decision makers are fully informed. As we summarized previously, those differences can best be understood as the inherent biases different institutions display regarding the terms of availability of different activities and the developmental consequences of individually rational adaptation to those circumstances.  



10.4 Toward a New Economic System

Traditional welfare theory disguises, rather than clarifies, the fundamental flaws of well-known economic models. The same new paradigm and welfare theory that highlights the deficiencies of traditional economic models, be they private enterprise or public enterprise, market or centrally planned, can also indicate the necessary characteristics of an economic system that improves upon known alternatives.

One step toward elaborating a new model is to clarify the reasons existing models are deficient, as we have attempted to do in this book. Another step is to conceive ways to organize production and consumption that avoid the production inefficiencies caused by "conflicts" over payments and effort and the consumption inefficiencies caused by incentives for individuals to "ride for free." But the implication of the foregoing analysis of allocative relations is that improved organization of production and consumption will only be subverted by market or centrally planned allocations. So, in addition to elaborating new ways to organize production and consumption, we must define compatible new means of allocation that promote solidarity, selfmanagement, and variety. In other words, we must continue the search for allocative mechanisms that allow informed individual rationality to be fully consistent with social rationality.

Adam Smith thought he had detected in competitive markets a shortcut, but he was wrong-the answer is not so easy. Marxists thought the answer lay in public ownership and central planning-translation of the principles of individual goal maximization subject to constraints to social goal maximization subject to constraints. But collective self-management is not so simple, and it is now abundantly clear that this, also, is not the solution. The failures of central planning arise not only from poor implementation; the model is flawed.

But the answer is not to deny the simple truth that for a social species whose individual members are blessed with consciousness, the ultimate interest of each individual lies in consciously attempting to promote the interests of all others as well as his or her own. Instead, we economists should admit we have not yet done our job. We have not yet fashioned allocative mechanisms that allow individuals to behave in a manner consistent with this simple truth. Once we see clearly what must be done and get down to the task, there is every reason to believe better allocative models will be forthcoming. Implementation is another matter, since material interests and not just intellectual obstacles must be overcome. 5