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





NEVER BEFORE have so many voices shouted Adam Smith's rallying cry of laissez-faire. Never before have those aware of the liabilities of markets been so silent.

Keynesian macroeconomists who sought to improve upon free market outcomes have been sent packing in one Western economy after another, often confessing the sins of their meddling ways. Occupational health and safety economists have been stampeded by the free market juggernaut of compensating wage differentials. The notion of "comparable worth" has become almost unspeakable among economists now that we have been duly reminded that what labor markets do is precisely establish the comparable worth of different labor inputs in the only feasible way.

But most surprisingly-and to the great delight of Western economists, politicians, and journalists--even "socialist" economists are rallying to the free market bandwagon. No longer are Yugoslavs and Hungarians the only Marxists in postcapitalist economies singing in praise of markets. In China, Vietnam, and even the Soviet Union, it must appear to old-timers that the "spectre of markets" is haunting Communism! While among Western socialist scholars, proponents of planning such as Ernest Mandel are on the defensive against those such as Alec Nove who champion markets in their vision of a "feasible socialism." 1

In sum, the view of markets as cybernetic miracles, essential to providing strong economic incentives, socially efficient allocations, and necessary bulwarks for the preservation of freedom, has been restored with a vengeance. It is in this context of a worldwide, free market jubilee that we deploy our new welfare theory to reevaluate the most hallowed of economic institutions.

Our results prove considerably more critical than the findings of traditional theory, which, in turn, are modest compared to present-day euphoria. What distinguishes our critique-beyond the fact that we dare object at all in today's highly charged atmosphere-is that it rests on problems seldom emphasized by those who previously counseled that markets should be approached with caution.

Most previous critics concerned themselves with problems of adjustment dynamics and noncompetitive structures. Until recently this lead to a kind of implicit truce between market defenders and critics; critics conceded the wonders of markets' cybernetic, incentive, and efficiency properties but only under competitive equilibrium assumptions. Defenders conceded there would be efficiency problems should structures prove noncompetitive and equilibration prove slow. 2

Recently, defenders of unbridled markets broke the traditional truce by suggesting that government remedies for noncompetitive structures an slow equilibration were worse than the maladies themselves. The intuitions and historical assessments of free marketeers were converted into a kind of "impossibility theorem": no conceivable form of government interference can improve upon free market allocation whatever its faults may be. The rout was on when the "impossibility theorem" was replaced by a "perfection theorem": not only is it "impossible" for governments to improve on market allocation, so-called "market imperfections" and "nonequilibrium adjustments" are, in fact, mechanisms for achieving dynamic efficiency by continually selecting decision makers who prove themselves best at estimating consequences and assessing uncertainty. 3

In other words, the new wave of free marketeers came to the conclusion that concessions previously granted traditional critics of markets were unwarranted. It was only the "comparative statics" model that gave the appearance that nonatomistic markets and nonequilibrium adjustments were inefficient. According to the "perfection theorem," a dynamic model with uncertainty would reveal that what had been conceded as market liabilities are actually market assets.

While we will concern ourselves with the information, incentive, and efficiency properties of markets, we do not wish to redebate the issues of noncompetitive structures and adjustment dynamics. We leave aside whether "Keynesian" stabilization policies can improve upon free market business cycles; indicative planning can ameliorate "bottle-neck" and "surplus" sectors; and development planning can improve prospects of less developed economies, or more developed economies for that matter. Similarly, we sidestep uncertainty by granting the perfect knowledge assumption.

But we do all this not because we believe there are no problems with markets in these areas central to the traditional debate, but because we wish to focus on what have been presumed to be markets' greatest strengthstheir cybernetic, incentive, and efficiency properties under competitive structures and equilibrium conditions. In other words, we wish to reexamine the properties of markets in the most favorable settings imaginable.



7.1 Laying Old Debates to Rest

Before treating markets per se, we seek to alleviate confusion concerning certain claims that, in our view, stem from ideological bias. Some opponents of public ownership argue that while competitive markets coupled with private ownership generate social efficiency, when coupled with public ownership they fail to do so. On the other hand, some opponents of private enterprise argue that while competitive markets coupled with private ownership necessarily generate unacceptably inegalitarian income distributions, socially wasteful business cycles, increasing industrial concentration with consequent inefficiencies, and a malady dubbed "commodity fetishism," when coupled with public ownership they become innocuous.

Of course, the properties and quality of economic systems depend on the interaction of all their important institutions. But we find no reason to deny that according to traditional welfare criteria public enterprise market economies have as much claim to social efficiency as private enterprise market economies. Nor do we see any reason to believe that substituting public for private enterprise will miraculously cleanse market economics of all the problems critics of capitalism have long dwelled on. Claims to the contrary have been used by ideologues for private and public enterprise in turn to avoid conclusions they find inconvenient. In either case, we are diverted from a "matter of fact" treatment of market institutions, and with that purpose in mind we seek to lay these "old debates" to rest before proceeding.



7.2 Misconceiving Public Enterprise Market Economies

It is helpful to differentiate between two different types of public enterprise market economies (PuEMEs)

1. Combining markets and public ownership with employee management yields an economic model used by many who have analyzed the Yugoslav economy since the early 1950s and championed by some who favor "workers' selfmanagement," public enterprise, employee-managed, market economies (PuEEMMEs).

2. Combining markets with public ownership and state-appointed management yields a model similar to that used by Lenin and the Bolsheviks in the Soviet Union as part of their New Economic Policy in the 192 , analyzed by Oscar Lange, Frederick Taylor, and Abba Lerner theoretically in the 1930s, 4 and championed by economic "reformers" today in many postcapitalist economies, public enterprise, state-managed, market economies (PuESMMEs).

Proponents of private enterprise have based their argument against public ownership in part on grounds that neither version of public enterprise market economy permits markets to perform their efficiency function completely. According to this view, while replacing central planning with markets in public enterprise economies improves efficiency dramatically, retaining public ownership ensures that at least some inefficiency will remain. Put differently, many champions of private enterprise argue that while the second and third fundamental theorems of traditional welfare economics hold for private enterprise market economies, they do not hold for public enterprise market economies of either the employeemanaged or the state-managed variety. First we present their argument. Then we explain why traditional welfare theory does not sustain their conclusion.


7.2.1 Employee-Managed Market Economies

A public enterprise, employee-managed, market economy (PuEEMME) is an economy in which publicly owned, employee-managed production units engage in free market exchange of goods and services with one another as well as with final consumers. Ownership is public in the sense that production units are legally owned by the whole nation and cannot be sold or shut down without permission of the state. Management is by employees in the sense that all decisions concerning enterprise functioning are ultimately controlled by a council of current employees in which each employee has one vote and majority decision rules. Of course, councils may form subcommittees or hire managers and supervisors to whom they delegate authority, but ultimately all committees and managers are answerable to the employees' council that may hire and fire them at will.

Goods exchanged on markets include consumer goods, intermediate goods, investment goods, and nonlabor primary factors. Terms of employment can be conceived in diverse ways, but it is convenient to assume that individual applicants and enterprise personnel departments negotiate wage rates and job descriptions, while implementation of agreements falls to whatever supervisory system is established by the workers' council. In a very real sense there is a labor market in PuEEMME since people are free to seek and leave employment whenever they choose, while enterprises are free to hire or fire whomever they like. But once hired, every employee is a member of the workers' council and, therefore, plays a "managerial" role as well. So while employees may receive differential wages (as negotiated with their personnel departments) they are all on equal footing as voters in the workers' council. 5

Enterprises obtain revenues by selling outputs and incur costs by purchasing inputs. If they fail to attract sufficient revenues to cover contractual obligations, they must sell bonds to the public or obtain loans from publicly owned, employee-managed banks that accept deposits (paying interest) and make loans (charging interest). Money can never be raised by sale of stock because stockholding is not permitted in a public enterprise, employeemanaged market economy. If enterprises' revenues exceed their contractual obligations, the workers' council is free to use the excess for investment in the enterprise, purchase of "collective" consumption goods for employees, payment of bonuses to employees, or saving, as it decides.

Birth and demise of enterprises is admittedly complicated in real world economies that pursue the model. But for theoretical purposes, it is easy enough to simply say that new enterprises are born when a group of workers declare themselves a new employee-managed enterprise and start selling something. The initial employees might come from other enterprises from which they resign, from among the unemployed, or from new entrants into the labor force. In the likely event that inputs needed in advance of initial sales cannot be entirely obtained on credit from their suppliers, "start up" loans would be required. While members of the initial group might provide part of the "start up" loan, this merely establishes a contractual obligation of the enterprise to them as creditors. It is as if an original employee purchased a bond from the new enterprise. Risk "capital" establishes neither ownership nor managerial prerogatives different from those of employees who arrive later.

Enterprises die by declaring bankruptcy, or by merging with some other employee-managed enterprise. When an enterprise goes bankrupt, its employees become unemployed and must seek employment with some other employee-managed enterprise, while its creditors recoup what they can just as they would from a private enterprise that went bankrupt. If two enterprises merge, all employees from both become full members on equal standing in the new enterprise's workers' council, and the assets and liabilities of both become the sum total assets and liabilities of the new enterprise.

Traditional Criticism of Employee-Managed Market Economies. Critics using traditional welfare theory have argued that employee-managed market economies are inherently less efficient than private enterprise market economies. But this charge is ill-founded. The claim is that PuEEMMEs do not achieve Pareto optimality in their general equilibria. Or, more specifically, that the second and third fundamental theorems of welfare economics do not hold for PuEEMMEs whereas they do for PrEMEs. 6

Originally, Benjamin Ward and Evsey Domar based these conclusions on short-run, partial equilibrium models of PuEEMME in which labor is the sole input in production and workers' councils are assumed to attempt to maximize average income per employee. 7 Before criticizing their conclusions, we will extend them to a more general model.

We can create a "multiperiod" general equilibrium model of PuEEMME equivalent to the model of perfectly competitive capitalism developed by Gerard Debreu in chapter 6 of his Theory of Value, by substituting the assumption that enterprises maximize profits per employee for Debreu's assumption that enterprises maximize profits. Since conclusions that hold in short-run, partial equilibrium models do not always hold in multiperiod, general equilibrium settings, this strengthens the arguments of Domar, Ward, and others critical of PuEEMMEs.

Though the elasticities and even the signs of both individual firm and aggregate supply functions might change under the new assumptions, there is no reason to expect any change in the continuity (semi-continuity) properties of the supply functions (correspondences). Therefore, Debreu's conclusions about existence of general equilibria in perfectly competitive capitalism can be directly extended to a similar version of PuEEMME. But the traditional criticism of PuEEMME does not concern the existence of equilibria, but their lack of optimality. 8 That the second and third fundamental welfare theorems will not hold in a Debreuvian model of PuEEMME is readily apparent: since the number of firms in Debreuvian models is taken as given, there is no reason they should not have different levels of profits per employee. 9 But this implies the opportunity cost of labor will vary from firm to firm in a Debreuvian model of PuEEMME, because in such a model firms must pay each employee his or her share of the enterprise profits in addition to a market-determined wage. In loose terms, this implies that firms with higher than average profits per employee will underhire labor relative to nonlabor inputs compared to firms with lower than average profits per employee, which will overhire labor relative to nonlabor inputs. If firm i has higher profits per employee than firm j, and if L stands for a labor input, NL a nonlabor input, MP for marginal product of the nonlabor inputs, W for the wage, P for price, and $ for the difference between total revenues and total costs, or profits, the first order conditions for the optimization problems of firms i and j, assuming both firms use total inputs,

Nonoptimality is immediate since the ratios of marginal products for two inputs are not equal for two PuEEMME enterprises using some of each with different levels of profits per employee. The outputs of both firms i and j could be increased by an appropriate transfer of labor from j to i and of the nonlabor input from i to j.10

A Poor Defense. In the late 1960s and early 1970s, Branko Horvat and Jan Vanek both defended PuEEMMEs from this criticism by challenging the assumption that workers' councils in PuEEMME would seek to maximize profits per worker. 11

Horvat proposed instead that PuEEM firms would try to maximize pq-[(d+d)x+cz+k], where p is the price of the firm's output, q is the amount of that output, "d is last year's or some standard personal income, d a change, normally an addition to that income per worker to be achieved in the given year," x is the number of employees, c the price of nonlabor inputs, z the quantity of those inputs, and k any fixed costs. 12 Alternatively, Jan Vanek proposes a list of nineteen possible indices a PuEEM firm might be interested in furthering. 13

If Horvat's criterion were accepted, then PuEEMMEs would achieve optimality, as he claims, since Horvat's necessary conditions for a PuEEM firm's maximization problem are precisely the same as those for privately owned firms. If, instead, PuEEM workers' councils attempt to balance a long list of goals, as Jan Vanek argues, then they do not attempt to maximize profits per employee, and the traditional "proof' of nonoptimality vanishes as well.

Both Horvat and Jan Vanek defended their hypotheses for PuEEMME firm decision-making criteria as better explanations for Yugoslav firms' actual behavior than the hypothesis of maximizing profits per employee, and evidence tends to suggest they might well be correct about the greater realism of their criteria as measured against Yugoslav practice, particularly since many of the perverse adjustment dynamics predicted by critics' modeling had not been observed. However, even if a different criterion than maximizing profits per employee proves a better approximation to actual Yugoslav behavior, this is irrelevant to the question of what rational behavior would require of the members of a workers' council in an idealized PuEEMME economy according to traditional welfare theory.

More recently, Horvat has attempted to defend the efficiency of PuEEMMEs by turning the tables on its critics. He points out that if PrEMEs. were modeled under the assumption that capitalist enterprises attempted to maximize the rate of profit on invested capital rather than the magnitude of profits, we would obtain the same "equally absurd results" critics have proclaimed for PuEEMMEs. 14 While Horvat is correct in sensing a bias in neoclassicists' choice of models begging an explanation from the theory of the sociology of knowledge, his counterattack is wide of the mark again. The problem is simple. Under traditional assumptions of no external effects, fully compensating wage differentials, and perfect "capital" markets in which firms can borrow infinitely at a market rate of interest, maximization of stockholder well-being in PrEMEs. translates into maximizing profits, whereas maximizing employee well-being in PuEEMMEs translates into maximizing profits per employee. Hence the criticisms originally formulated by Ward and Domar, which we confirmed in the more general "Debreuvian" model of PuEEMMEs, withstand the original and more recent rebuttals of Horvat and Jan Vanek.

A Better Defense. Jaroslav Vanek voiced a much better response to traditional criticisms of the efficiency properties of PuEEMME, but never proved his intuition. Jaroslav Vanek accepted that as seen from the traditional paradigm maximizing profits per employee is the rational criterion for employeemanaged enterprises. But he hypothesized that "a multi-sectoral perfectly competitive general equilibrium of a pure labor-managed economy is Pareto optimal in the long-run provided that free entry is guaranteed." 15 In other words, the "problem" theoretically elaborated above is not in PuEEMME, but in the use of inappropriate models for analyzing the optimality properties of PuEEMMEs.

Both Ward and Domar's partial equilibrium models are short-run models where entry and exit do not equalize profits per employee. They are the PuEEMME analogues of short-run, partial equilibrium models of private enterprise market economies in which entry and exit have not had time to equalize the rate of profit between different firms and industries. In the Debreuvian model of PuEEMME we used above, the number of firms in the economy is a fixed datum. A firm is nothing more than its production possibility set. There are only so many of them, and even should the workers in one firm receive very high profits per employee, it is not possible for new enterprises to replicate their activities. It is the analogue of Debreuvian models of private enterprise market economies in which what are called different profit rates, but in fact are different quasi rents achieved due to unique technological possibilities, are not competed away.

In the case of private enterprise market economies, no damage was done to an analysis of social efficiency under the assumption of a fixed number of firms because the opportunity costs for all arguments in the production functions of all firms were the same regardless of the firms' "profit" positions. But in PuEEMMEs, the opportunity cost of labor inputs depends on the firm's rate of profit per employee, and in a model where the number of firms is fixed and, therefore, different rates of profit per employee will not be "competed away," the implications for misallocation of resources are evident.

The problem is not that the assumption of a fixed number of firms is any less realistic for PuEEMME than for private enterprise market economies, but that the simplifying assumption of a fixed number of firms is inappropriate for an analysis of the efficiency properties of PuEEMME, whereas, for very special reasons it was not inappropriate for a similar analysis of private enterprise market economies.

Thus Jaroslav Vanek was correct when he insisted that though PuEEMME must appear to be non-Pareto optimal in the simplified context of a fixed number of firms where there is no reason for profits per employee to equalize among firms, the same might not be true in models permitting free entry and exit of enterprises.

Unfortunately, at the time Jaroslav Vanek wrote, no one had developed a multisector, general equilibrium, multiperiod model of either PrEMEs or PuEEMMEs in which the number of firms in an industry is determined endogenously under the assumptions of free entry and exit. That is no longer the case since Dilip Madan, followed by William Novshek and Hugo Sonnenschein, have developed such a model for private enterprise market economies where the equilibrium number of firms in each industry as well as the uniform rate of profit in the economy are determined endogenously. 16 With such a model, it is finally possible to demonstrate Vanek's intuition rigorously. But since the adaptation of the Madan/Novshek/Sonnenschein model to PuEEMME is simple and straightforward, and since Vanek's conclusion is immediate once there is an equal rate of profit per employee for all firms-as there would be in the Madan/Novshek/Sonnenschein model of PuEEMME, just as there is an equal rate of profit for all firms in the Madan/Novshek/Sonnenschein model of PrEMEs-we leave it to the interested reader to reproduce Madan, Novshek, and Sonnenschein's work with the single modification that firms attempt to maximize profits per employee instead of profits.


7.2.2 State-Managed Market Economies

In the 1930s Oscar Lange, Frederick Taylor, and Abba Lerner all contributed to theorizing a model of public enterprise, state-managed, market economies (PuESMMEs), for the express purpose of refuting the traditional criticism that public enterprise economies could not be socially efficient. Since their economy was designed precisely for this purpose and literally incorporated necessary conditions for social efficiency from microeconomic texts as decision-making criteria, it is surprising they failed.

In some respects Lange, Taylor, and Lerner improved upon private enterprise market economies as "efficiency machines." By directing stateappointed enterprise managers to obey the rule, minimize the cost of producing any level of output always calculating the cost of inputs according to going market prices, the ability of PuESMME to achieve efficient allocations is unaffected by monopsonistic input markets. 17 By directing state-appointed managers to obey the rule, produce the level of output at which marginal cost is equal to market price, the ability of PuESMME to achieve efficient allocations is unaffected by monopolistic product markets. 18 In other words, PuESMME is socially efficient whether or not markets are competitive provided managers obey the above rules.

To those who reasoned managers would maximize profits rather than obey the rules whenever there was a discrepancy between the two courses of action, champions of PuESMME replied there was no reason for them to do so. First, enterprises in PuESMME do not retain any "profits" they generate. And second, since they were state appointed, presumably the state would fashion incentives for enterprise managers to reward ruleobeying behavior and punish profit-maximizing behavior in the event of discrepancies.

The creators of PuESMME also sought to improve upon the ability of real market economies to reach equilibria and, thereby, avoid the inefficiencies that nonequilibrium outcomes imply. In Lange, Taylor, and Lerner's economy a planning board manipulates prices to eliminate excess demands and supplies for intermediate goods rather than leaving price adjustments to the "free market mechanism"--which the Western economies of the 1930s did little to recommend! But whether or not there is any reason to believe the Lange-Lemer-Taylor system would equilibrate any more quickly than free markets generated remarkably little discussion. 19

Traditional Criticism of State-Managed Market Socialism. Criticism of PuESMME on efficiency grounds centered on its presumed misallocation of society's resources over time. What proportion of society's resources would be used for present consumption and what proportion used for investment? As designed by Lange, Taylor, and Lerner, all enterprise profits were to be turned over to the state. Out of this pool of profits the state first pays bills of enterprises for which costs exceed revenues. 20 The remainder is then divided into two parts. One part is sent to a state investment bank, and the other part is distributed as bonuses to workers.

Lange believed the bonuses should be paid as a percentage of workers' wages to avoid distorting incentives to offer services in different labor categories. Ironically, while this bonus system accomplishes Lange's aim, it distorts the incentive to work rather than use one's time in leisure. 21 The key to distributing bonuses without disturbing efficiency is to hand out bonuses totally unrelated to work time or wage rate. Equal bonuses to all members of society (working and nonworking) is the simplest system of 41 efficient" bonuses. But this does not solve all problems.

The part of the state's profit pool that was turned over to the state investment bank was to be loaned to state enterprises for investment. The interest rate was to be set just like prices of all intermediate goods in PuESMME - raised in case of excess demand for loans, and lowered in case of excess supply. Although there was no explicit discussion, presumably Lange, Taylor, and Lerner would not have objected to permitting individual citizens to deposit some of their income in this same bank and receive the same rate of interest. But the question is how the division of the profit pool between a part earmarked for bonuses and a part earmarked for investment was determined in the first place. In Lange, Taylor, and Lerner's description, the division was arbitrary. But an arbitrary division may bias society's present versus future consumption.

Critics argued as follows: suppose society's citizens have a strong preference for present as opposed to future consumption, but the arbitrary division of the state profit pool is heavily weighted toward the investment bank and away from citizen bonuses. In this case, a large proportion of society's resources would go to investment-since the large supply of funds would generate low interest rates and greater loans to enterprisesdespite society's preference for greater present consumption. The point was well taken. To our knowledge few advocates of public enterprise market economies of this type have bothered to reply beyond noting that the alternative notion that the rate of time preferences of the wealthy in private enterprise economies should govern the rate of investment and, therefore, growth and development prospects for all of society is obscene. While this argument is compelling on its own grounds, it is hardly an adequate response to critics of PuESMME

A Simple Defense. There is a simple way to salvage the efficiency properties of PuESMME as judged by traditional welfare theory 22 . If all of the state profit pool is distributed as bonuses, and citizens are free to deposit whatever part of their total incomes they wish in the state investment bank in return for interest payments, society's "rate of time discount" will be left to determine the division between present and future consumption. Alternatively, the state profit pool can be arbitrarily divided between bonus payments and deposits in the state bank as long as citizens are free to take out consumer loans at the same rate of interest enterprises take out investment loans. In other words, the profit pool can be arbitrarily divided between the state bank and bonuses, but then the state bank cannot be an investment bank in the sense that its only loan customers are enterprises. In any case, citizens must be free to make deposits and receive interest and take out loans paying interest, and the rates of interest paid cannot be different for loans out of state funds and loans from consumer deposits, and the rates of interest charged cannot be different for enterprise and consumer loan customers if PuESMME is to achieve social efficiency according to traditional welfare theory.

This concludes our analysis of public enterprise market economies from the point of view of traditional welfare theory. When corrected and analyzed using appropriate models, both PuEEMME and PuESMME prove just as capable of generating efficient outcomes as private enterprise market economies. While we have not taken the time to generate every analogous theorem, the demonstrations are all perfectly straightforward; the first, second, and third fundamental theorems of welfare theory all hold for both PuEEMME and PuESMME just as for private enterprise market economies, according to the norms of traditional welfare theory. If we wished we could even extend the theorems under the assumption of endogenous preferences provided there were no reasons to suspect PuEEMME and PuESMME of biasing conditions of supply of different goods and/or work roles. And according to the traditional paradigm, there are no more reasons to suspect PuEEMME and PuESMME of any such biases than their private enterprise counterparts. 23 In the eyes of traditional welfare theorists, therefore, PuEEMME and PuESMME should appear equally meritorious as private enterprise market economies-provided those eyes are not blinded by ideological prejudice against public ownership. In fact, from the point of view of welfare theoretic results, according to traditional theory the three systems are indistinguishable!


7.2.3 An Overzealous Defense of Public Enterprise Market Economies

While both versions of public enterprise market economies can be defended from charges that they are inherently less efficient than private enterprise market economies, the claim that they are not subject to a familiar catalog of other complaints voiced by leftists about private enterprise market systems cannot be defended. In our opinion, there is every reason to believe public enterprise market systems suffer from many of the maladies leftist critics have long cited in arguments against capitalism. In this case, what is bad for the goose is bad for the gander as well.

To focus on what we believe to be the most theoretically significant and least recognized deficiencies of markets, we will say little about a number of problems others have dwelled on. Yet to claim these problems would plague private enterprise market systems but not public enterprise market systems is not our intent. Before proceeding to our main argument, we wish to disassociate ourselves from such overzealous defenses of the use of markets in public enterprise economies.

We leave the question of whether or not there is reason to believe public enterprise market economies would or would not be plagued by unacceptably inegalitarian distributions of income to chapter 8 where we examine ownership and income distribution in detail. And we defer the question of whether or not public enterprise market economies would generate the kind of "false consciousness" Marxists call "commodity fetishism" to follow our treatment of the cybernetic characteristics of markets later in this chapter. Here we disassociate ourselves from the notion that public enterprise market economies would be immune to business cycles and increasing industrial concentration.

Economists have long been concerned with the obvious inefficiency associated with the kind of gross disequilibria typical of alternating cycles of boom and bust. Economists have also long been aware of the obstacles industrial concentration pose for efficiency. Since the "anarchy of capitalist production" and the "law of ever-increasing concentration of capital" are fundamental precepts of Marxist economics it is hardly surprising that Marxists associate these maladies with private enterprise market economies. But there was never any compelling theoretical reason to believe public enterprise market economies would be immune to these maladies. And empirical evidence is now accumulating to the contrary. While influenced to some extent by ownership, business cycles and increasing industrial concentration are fundamentally market phenomena.

This is not the place to review competing theories of the business cycle. While there are "political economy" theories in which the struggle over capital and labor's share play a critical role, 24 the self-feeding nature of aggregate market expansions and contractions has received increasing theoretical attention as well. 25 But more convincing is the clear empirical evidence of a business cycle in the Yugoslav economy over the past thirty years. 26 And for those who think PuESMMEs might be less susceptible to cycles than PuEEMMEs, the Hungarian experience is proving sobering. In any case, real world public enterprise market economies have not proven immune to the cycles that plague their private enterprise counterparts, making a strong prima facie case that cycles are associated with markets per se.

One might argue that industrial concentration is irrelevant to at least some versions of PuESMMEs since it does not interfere with their efficiency. 27 But such is not the case for PuEEMMEs. Nor are increasingly popular PuESMME "reform" models in which enterprises are encouraged to maximize profits, at least part of which they are permitted to retain, rather than obey the Lange/Lerner/Taylor "rules" immune to inefficiencies should monopolistic market structures develop. In these cases in which enterprises do compete against one another for profits, or profits per employee, it strikes us as naive to presume they would not avail themselves of tactics that have always proven highly successful in such endeavors. Arrangements in which a small number of sellers face a large number of buyers have seldom proven damaging to sellers' profits. We see no reason for public enterprises competing with one another to find differently.

The empirical evidence here is also overwhelming. There has been a lively debate in Yugoslavia since the early 1950s over how to combat the effects of monopolistic behavior where some recommend price controls, others favor increased foreign competition, and still others call for breaking up large enterprises. 28 And it is common knowledge among Soviet economic historians that monopolistic factors played an important role in the famous "scissors crises" that followed implementation of the NEP in the 1920s. 29

In sum, while we will overlook problems with market dynamics and market imperfections to focus on other matters, there is no reason to believe substituting public for private enterprise solves these very real problems.



7.3 New Criticisms of Markets: Snowballing Individualism

We are now ready to examine the cybernetic, incentive, and efficiency properties of markets with our new valuative theory. As we will see, a very different picture develops than the one celebrated by the free market jubilee. Instead of cybernetic miracles, optimal incentive devices, and efficiency machines, we will find that markets bias and obstruct the flow of essential information, promote antisocial incentives over equally powerful motivations that need not be socially destructive, and generate increasingly inefficient allocations of resources. In sum, we will find that markets promote snowballing individualism that is demonstrably nonoptimal regardless of whether they are combined with private or public enterprise.



7.4 Markets as Cybernetic Disasters

In both production and consumption, a group of people aided by physical implements act on inputs to generate outputs. In the process they transform not only the tools they use, but the individual and group characteristics of the people who undertake the activity as well. The qualitative model of the economic sphere we developed in chapter 5 was designed to highlight all this. But the inputs and outputs of economic activities we usually conceptualize separately actually connect each activity to others. No activity is possible without inputs that must be generated as outputs by other activities. And the meaning of an activity hinges on the uses to which others put the outputs it generates.

Another way of saying this is that the economic activities of different groups of people at different times are parts of an extensive network of activities that extends through space and time. Each activity is connected to countless others implying extensive relationships between both things, the inputs and outputs passed, and people-those who generate outputs for others and use inputs others produce. This has been referred to casually as "the division of labor," which Adam Smith celebrated as the engine of evergreater productivity. And it has been analyzed with great care under the title of "commodity fetishism" by Karl Marx, who was concerned that the very people related to one another might be blinded to the nature and importance of those human, social relationships under particular arrangements for carrying them out.

In any case, those who carry out economic activities and who are integrally related to one another may be more or less aware of those relations and their complexities. And depending on the allocative mechanism used, people may receive more or less appropriate and useful information for developing such an awareness. One important question to ask concerning any allocative mechanism is whether or not it facilitates awareness of the interconnection of economic possibilities between groups of people who carry out activities in different times and places.

It might be objected that this is not an appropriate question for all economies. While different allocative mechanisms have different cybernetic qualities, they also have different cybernetic needs. The ability of groups of separated producers and consumers to sensibly select and coordinate their efforts obviously hinges on their knowledge of the interrelations between those efforts. But if instead of producers and consumers deciding how to organize their activities, a central planning agency is going to make those decisions for them, then it is not the producers and consumers but the central planners who need to understand the interconnections. 30 And if the coordination is going to be achieved by market interactions without conscious planning and coordination on anyone's part, then there is no need for anyone to have the information. 31

While it is obvious that different allocative mechanisms have different cybernetic needs as well as qualities, this does not preclude asking evaluative questions about the cybernetic qualities of allocative mechanisms regardless of their internal information needs. The evaluative theory developed in chapter 6 awards high marks to decision-making processes characterized by self-management and solidarity, ceteris paribus. So regardless of who needs to know what for any particular decision-making framework to function, it bears asking what people would have to know to engage in self-managed decision making, what information would facilitate solidarity among economic actors, and to what extent any particular allocative mechanism provides that information to the economic actors themselves. In sum, what economic actors need to know to do what is required of them is a valid analytical question about any decision-making system. Whether economic actors will receive the information necessary for self-managed, solidaritous decision making is a no less valid valuative question and we can reasonably pose it for any allocative mechanism.

The market mechanism coordinates economic activities taking place in different places and times by providing all individual units with the opportunity to offer the material outputs of their activities in exchange for the material outputs of other units' activities, with a general assumption of noncoercion. 32 This exchange of material inputs and outputs between units is an expression of the fact that people in those separate groups are actually engaged in social activity with one another. Although the two units involved in exchange are separated by distance and/or time and likely lack information about the nature of each other's roles in their shared activity, neither of their activities is possible or makes sense without the other. We have no difficulty understanding that workers at the beginning and end of an automobile assembly line are engaged in the social activity of making a car. And even though the workers in a blast furnace are not connected by a physical assembly line to the workers in the rolling mill, we see their activities as integrated and part of the joint endeavor of making steel. But in a market system it is more difficult to recognize that workers in a steel mill and workers in an automobile assembly plant are similarly involved in a shared activity.

The reason for our blindness is that within individual units the activities of different individuals are consciously planned and coordinated to achieve a conceived goal. But as we have seen, the great miracle of market economies is precisely that the activities of different units are not consciously coordinated and planned by anyone. Those working in auto plants do not appear to be engaged in a social activity with steel workers, or bound by a specific set of production possibilities with steel workers. Instead it appears they are engaged in isolated productive activity and have relations only with other things, i.e., the sheet metal they use as inputs and the automobiles they create as outputs. The relations between people and things-people consuming inputs and people producing outputs-are readily visible, as are work relations between people in the same unit. But in market systems, the relations between groups of people working in different places and times can easily escape notice because people do not consciously plan and coordinate these relations. Instead they take the form of an exchange of things. 33

But the "Problem" of "commodity fetishism" as it has been treated by Marxists is overly abstract and philosophical. If one were concerned to eradicate this particular misconception among producers in "commodity producing societies," the simple solution would appear to be including introductory price theory in the high school social science curriculum. Then people could be taught that commodity exchange is the form in which important economic social relationships are expressed in market economies, just as people are taught now that the earth revolves around the sun although it appears otherwise. 34

The more important question is whether market cybernetics would help separated producers and consumers discover much about the complexities of their actual interrelations in market systems should they realize there was something to be discovered! Having argued that self-management and solidarity are desirable welfare goals, it is only logical to inquire whether or not the cybernetic qualities of markets are conducive to these ends. For our auto workers to evaluate their work in human terms they would have to know the human-social implications for others of providing them with the steel they need to work with, the human-social consequences for them of their own work process, and the human-social consequences for others of consuming the cars they produce. Let's assume our auto workers know (perhaps more than they care to) about the human-social effects of their own work process. How can they discover the human-social implications of supplying them with steel, rubber, and the human-social consequences of driving the cars they produce?

The only information the market provides is the prices of steel and cars. Even if these prices accurately represent the total human and social costs and benefits associated with supplying rubber, steel, and cars-and we will argue shortly that this is not the case-this information is totally insufficient if auto workers are to develop the kind of understanding that can serve as the basis for a realistic and empathetic evaluation of the full consequences of their activity. Market institutions "efficiently" delete all this information about the concrete relations between separated activities that is necessary for developing a humanized awareness of interconnected possibilities. And while a humanized awareness may not be necessary for some kinds of decision making, it certainly is necessary if separated groups of economic actors are to develop a sense of solidarity for one anothers' aims and predicaments.

"Information economy" means deleting information that is not needed by decision makers. Deleting information about the drudgery and unpleasantness of work in steel mills and auto factories and deleting information about the injuries from accidents in unsafe cars is not "economy" of information if the goal is a self-managed decisionmaking process characterized by solidarity, because such information is needed for informed collective self-management. 35 As cybernetic devices markets are systematically destructive of the development of solidarity based on understanding and concern for the situation of others.

We can summarize using the qualitative economic model of chapter 5. Assume actors understand their own activity possibility set, {A (i,t)}. How can they come to understand

{A (j,t)} ji, t=0,1 .... T, that is, the activity possibility sets for all the other actors and times their activity is, in fact, related to? Do they receive sufficient information? Is it organized in ways that are usable? For fully informed self-managed decision making based on solidarity with others' problems and desires, we would ideally wish every producer-consumer to know as much about others' situations as they do about their own. But not everybody can work-consume everywhere and everything. And while people can read novels by Upton Sinclair, this is not the point. The question is how well the market itself serves the information needs of a self-managed, solidaritous decision-making process by economic actors. Perhaps we need not have belabored the answer so; at best markets systematically delete qualitative human information necessary for solidaritous decision making. At worst, as we will explore further, they provide inaccurate summaries of overall consequences. The cybernetic "miracles" of markets looks more like a disaster from the perspective of our welfare criteria.

Moreover, we see no reason to believe the information deficiencies of markets should be any less serious in public enterprise than private enterprise settings and wish to disassociate ourselves from contrary claims.

For example, in one of the more sophisticated defenses of employeemanaged, public enterprise market economies Arthur DiQuattro specifically argued they would not suffer the malady of commodity fetishism that plagues their private enterprise cousins. According to DiQuattro:

Not the market, but class structure and class-determined technological misdevelopment, are responsible for alienated labor. It's true enough that in market socialism workers produce for exchange and not directly for use, but why must production under such an arrangement result in alienated labor, false consciousness, and the estrangement of individuals from one another? ... If workers succeed in supplanting capitalist control with socialist productive relations, they would at the same time abolish the class basis of "commodity fetishism."


With the transformation of work and other spheres of social life (e.g., education, community and political organization), is there any reason to believe that market operation need generate "false consciousness,", casting a fog over the social character of production, and subjecting producers to its sway? In market socialism, the working community freely chooses, with a thorough and reasoned comprehension of the automatic workings of market arrangements, to rely on those arrangements for the purpose of allocating economic resources. The idea that production for exchange must engender false consciousness is a metaphysical "must," based on a confusion between productive relations and modes of allocation, ruling out a priori the distinct analytical factand historical possibility of diminished alienation in market socialism. 36

In our view, DiQuattro's defense of "market socialism" from the charge that it too would be plagued by commodity fetishism is another example of an overzealous defense that cannot be justified. 37

First DiQuattro conflates two distinct Marxist concepts, "commodity fetishism" and "alienation," arguing that if alienation is not present then commodity fetishism will not be either. While this line of reasoning may appeal to many Marxists who do not believe alienation exists in public enterprise economies, it is a non sequitur. Even if employee-managed market economies were immune to alienation, this would not imply they were necessarily immune to commodity fetishism. For the question remains whether markets hinder or aid the process of expanding one's understanding of the full effects of one's economic activities from {A (i,t)}, to {A (j,t)} for j i and t = 1,...,T. DiQuattro offers no direct rebuttal to our argument that markets hinder actors' abilities to achieve a full understanding of their economic activities. And his indirect rebuttal hinges on a proposition he does not demonstrate, namely that alienation is a necessary condition for fetishism. In this, he assumes his conclusion.

Second, DiQuattro seems to believe that as long as institutions are "freely" chosen "with a thorough and reasoned comprehension of their automatic workings," any deficiencies in those workings disappear! DiQuattro presumes that if workers are aware of the cybernetic deficiencies of markets when they choose to use them for allocative purposes, they will not suffer from those cybernetic deficiencies. Hardly a compelling argument, when recognized for what it is. And while we might agree with DiQuattro that the more thorough and reasoned our understanding of the automatic workings of market arrangements the better off we will be, the same could be said about slaves' comprehension of slavery! This would not be because slaves' greater awareness would dull the negative effects of slavery, but because the more thorough their understanding of slavery the more capable slaves become of overthrowing the institutions that oppress them! In the same vein, a thorough comprehension of the workings of markets does not lessen the negative effects of market institutions, but makes more likely their replacement by a superior cybernetic mechanism.

Finally, it is DiQuattro who suffers from a "metaphysical" confusion not uncommon to Marxists that only "productive relations" can generate effects as profound as "false consciousness." But there is no a priori reason "allocative relations" cannot have consequences as profound as those of "productive relations," no matter what a majority of Marxists may believe. In any case, a successful argument downgrading the importance of allocative relations would merely diminish the importance of the criticisms voiced in this chapter rather than alter any of its conclusions.