Exploitation in and through the Large Cooperative

In the large CooperativeJust as in the small cooperative, the workers are col­lectively the residual claimants and the ultimate decision makers. Although in theory they could elect managers directly, most discussions of market socialism envision a workers’ council selecting a manager or management team.

Capital provision in the large cooperative would be from the state, as in the smaller cooperatives. Other than sheer size, the differences between the large cooperatives and the smaller cooperatives do not appear to be that great. Indeed, in principle, smaller cooperatives could also have a workers’ council, although that possible complication was ignored in the discussion in the last section of chapter 6. Though the two types of cooperatives are simi­lar in many respects, the comparable capitalist organization, the open cor­poration, is dramatically different from its smaller cousin, the classical capi­talist firm. And that difference makes a difference in the subject of this section—the comparative evaluation of the large cooperative and the open corporation on the topic of exploitation.

Indeed, there are four important similarities between the governance structure of the large cooperative and that of the open corporation: (1) ulti­mate decision-making authority is joined to residual claimancy and both are spread widely among many different individuals; (2) there is a partial sepa­ration between these two roles on the one hand and managerial control on the other (i.e., in both types of organizations, management has only fractional residual claimant status and only fractional ultimate decision-making author­ity); (3) there is a mediating body between management and the ultimate decision-making authority—for the cooperative, the workers’ council, and for the open corporation, the board of directors—whose main task is to mon­itor management; and (4) in neither case do the members of these interme­diate bodies provide most of the firm’s capital.

As notable as these similarities are, there are important differences as well. In the open corporation, unlike the cooperative, ultimate decision-making authority and residual claimancy are joined to the provision of capital. In other words, there are equity owners, and most of the firm’s capital is pro­vided by them. By contrast, in the large cooperative there are no equity own­ers; it effectively rents all of its capital from the state. This difference means that the governing boards represent the interests of very different ultimate decision makers.

Fundamentally, the issue is about how to monitor management. How effectively could this be done in the large cooperative? In the last section, it was argued that monitoring arrangements in the small-to-medium-sized cooperative would be weakened by adverse selection and moral hazard prob­lems. Politically astute but relatively ineffective managers would be chosen and would be compromised by their status as elected officials. Managers’ sta­tus as fractional residual claimants would also affect their incentive to do a good job as manager, at least in comparison with the classical capitalist who gets to vacuum up every residual dollar.

However, it is not obvious that a large cooperative would be at a com­parative disadvantage vis-a-vis the open corporation on these points. If the workers’ council developed some expertise and experience at choosing and evaluating managers, they would be no less easily taken in by smooth-talking political types than the average board of directors is. Furthermore, if man­agement is appointed by the workers’ council, the former may not be com­promised by the democratic structure of the Hrm in the way that the manager of the small cooperative would be.

Notice, however, that if the large cooperative is going to avoid the weak­nesses of the small cooperative and do about as well as the open corporation in monitoring management, it will have to look more like a corporation and less like a cooperative in two respects. First, the workers’ council will have to have people on it with the experience and specialized knowledge to assess managerial performance, since managerial tasks or functions are, in general, hard to monitor. (More on these functions shortly.) The need for people with experience and specialized knowledge can be appreciated by considering the problems that the council faces. Although shirking by management is a always a potential problem, perhaps a more important form of opportunism is managers’ efforts to make it appear that they, the managers, are doing a better job than they really are. The inherent difficulties in monitoring the particular tasks of management are aggravated by the natural human desire to put the best face on a situation. In other words, managers (like other human beings) are prone to indulge in credit stealing and blame shifting. The monitoring problem is further compounded by the fact that much of the information needed to assess management’s performance is developed and provided by the management itself. Insofar as it concerns financial informa­tion, this problem can be dealt with in the same way that corporations deal with it: the workers’ council would hire an independent public accounting firm to audit the firm’s financial statement that management prepares. Also, as in the corporation, the cooperative’s internal auditors would report to the audit committee of the workers’ council, instead of to the managers.

These observations are obvious enough to anyone with a basic under­standing of how large, profit-making organizations are run. Unfortunately, many socialists seem insensitive to these facts in their enthusiasm for demo­cratic governance. It is certainly possible for workers to have ultimate deci­sion-making authority without any special expertise. Stockholders in large corporations are a perfect example of this. But ultimate decision-making authority is different from actual governance, and the latter is what the work­ers’ council has to do. The problem is that it is unclear where, within the firm, all of this specialized knowledge and experience is to be found. The obvious solution would be to employ outsiders (e.g., executives from other coopera­tives) in an advisory capacity, though this means the reintroduction of wage labor, at least for these advisors. That problem to one side, the council must be careful to prevent its power and authority from being effectively ceded to these outsiders, who may have important informal ties with those being mon­itored. Though this need not lead to the reintroduction of power elites, there would be important structural continuities between governance in the large cooperative and governance in the open corporation.

The second respect in which the large cooperative would resemble the open corporation is that the workers’ council would interpose itself between the workers and management. If the management team is to be effective in preventing shirking down through the ranks, it must be at least partially insulated by the council from the white heat of democracy.

The large cooperative is, however, less likely to overpay its executives than its corporate counterpart. Boards of directors in free enterprise systems are dominated by executives who are themselves highly paid. Obviously, they have little interest in keeping down the average pay for executives. (Imagine what would happen if professors set the pay of their opposite numbers at other institutions.) If enough people with dirty fingernails sat on the com­pensation committee of the workers’ council, it is doubtful whether top man­agement would get the kind of compensation package that corporate officers sometimes get, at least in America in recent years. Conspicuous on-the-job consumption would also likely be diminished in a large cooperative. This is one case where physical proximity does make a difference in monitoring management.² On the other hand, it is not clear that executives of corpora­tions in free enterprise systems are, in general, overpaid. Those who com­plain (rightly, in this author’s view) about the fact that the compensation of some American executives has nothing to do with their performance often point to the more modest compensation packages received by Japanese and European executives. This suggests that the problem is an American prob­lem, that is, a problem with the American version of the free enterprise sys­tem and not a defect of free enterprise systems in general.

Still, it may be true that the market for managers of large corporations are not competitively efficient. It is certainly “thinner,” in the sense that there is not as much buying and selling going on as there is, for example, in whole­sale produce markets. And the product is definitely harder to evaluate. These two facts together do not guarantee that markets in executive services are not competitively efficient and that managers are generally paid more than they are worth, but it could turn out that way. However, this problem of thin mar­kets for managers (if it is a problem) would also face the large cooperatives. Indeed, it might be worse in them because of the structural impediments in all labor markets in a market socialist system, discussed previously. It is more likely, however, that managers would be underpaid for the reasons that were discussed in the first section of chapter 6. So it is not clear in the final analy­sis if the relationship between firms and their executives is inherently more exploitative in the open corporation than it would be in the market socialist worker cooperative.

Monitoring Management in the Large Cooperative

Let us turn now to the particular tasks that management is responsible for— monitoring, central contracting, and directing the firm’s product—and see how they would be monitored in the large cooperative. Let us assume that top management in the large cooperative discharges its monitoring functions by instituting performance evaluation policies for other workers and by hiring and monitoring lower-echelon managers, all with the consultation and par­ticipation of the workers themselves. Whatever the benefits of widespread par­ticipation, the problem with it, from the council’s perspective as monitor of top management, is that it entails a diffusion of responsibility for monitoring. (The same holds true for the other aspects of management to the extent that they are shared with the workers.) If monitoring is done poorly, blame is less easy to assign, and if it is done well, credit is more difficult to attribute. This provides some incentive for managers to shirk in the provision of monitoring services. More generally, this diffusion of responsibility smudges the record of managers’ contribution, thereby making it more difficult to reward them according to their contributions. The open corporation solves this problem by permitting executives to employ whatever methods of monitoring they choose (from the whip to the solicitation of advice from those being moni­tored) and then holding them personally responsible for the results. Of course, it does not always work that way: many executives are master blame shifters and credit stealers in all of their tasks. As the firm lurches from one disaster to another, they are able to convince the board that those disasters are not their fault. Credit stealing is, if anything, even more prevalent. How­ever, the same possibilities for opportunism would exist in a market socialist system. What is at issue here is the susceptibility of organizational structures to shirking and other forms of opportunism in the provision of monitoring services by top management and the relative ease with which management can be monitored. On this point, it seems that the less hierarchical, more con­sultative structure envisioned for the large cooperative is decidedly inferior to—and thus more vulnerable to exploitation than—the open corporation.

Although, in theory, the same problem arises when the council seeks to monitor management’s work on the input and output interfaces between the firm and the market, in point of fact, this problem probably would not arise. There is some empirical evidence from what used to be Yugoslavia support­ing the proposition that workers in large cooperatives in a market socialist system would be relatively uninterested in these matters, at least in compar­ison to matters such as income policies and work rules.³ Indeed, that makes sense, since the issues that management must deal with on these fronts are more remote from the workers’ knowledge and immediate interests and require experience and more specialized knowledge to deal with. Should the firm challenge the Wordperfect Corporation in the desktop publishing mar­ket? Should it sell semifinished products to competitors? Do changes in the regulatory environment necessitate the redesign of products or production processes? Do changes in the tax code require changes in maintenance pro­cedures or investment strategies? It is facile to suppose that workers can eas­ily acquire the requisite specialized knowledge and the sheer experience required to have meaningful input on issues and questions such as these.

How, then, would the workers’ council monitor these aspects of manage­ment, since the council is composed of workers? As indicated before, they would likely hire other executives in an advisory capacity before ratifying major entrepreneurial decisions proposed by management. The question now becomes, Would the workers’ council do about as well as (or better than) a corporation’s board of directors in monitoring in management’s entrepre­neurial activities? Putting aside the principal-agent problems posed by the outside advisors, it seems that the council would have an incentive to moni­tor management as well as or better than corporate boards. After all, mem­bers of the workers’ council are all much more substantial residual claimants —substantial in the sense that their own financial well-being is intimately connected with the fate of the cooperative—than the average board mem­ber in the open corporation. Though the incentive to monitor might be superior, the interests being protected are systematically different, and it is for this reason that there are distinctive opportunities for systematic exploita­tion in a market socialist system that are not found in a free enterprise sys­tem. These opportunities are traceable to the peculiar structure of owner­ship rights to the firm’s capital and the nature of the workers’ status as residual claimants.

As in the case of the small cooperative, three facts are of decisive impor­tance in this connection: (1) neither members of the workers’ council nor the workers themselves own any of the firm’s capital; instead, the state effectively owns the cooperative’s capital; (2) the workers’ claim on the residuals is lim­ited to their tenure at the firm; (3) the workers’ claims on the residuals are inalienable—that is, their claim on the residuals cannot be sold to outsiders or indeed to anyone. What this means is that the workers’ council, unlike the board of directors, represents only current residual claimants—not the cap­ital providers and not the future residual claimants. There is no reason to suppose that the council will be any more solicitous of the interests of future residual claimants (i.e., future workers) and the capital providers (i.e., tax­payers) than the managers would be. Indeed, they could all conspire to exploit these voiceless citizens of market socialist society. Like the small-to- medium-sized cooperative, the large cooperative would be characterized by poor maintenance, dubious formulas for calculating replacement costs of cap­ital, and income shifting from the more distant future to the present.

By contrast, the board of directors of the open corporation represents, albeit imperfectly, the interests of the equity owners (the shareholders). The latter not only own have a claim on the firm’s residuals that extends indefi­nitely far into the future but can also sell their ownership shares at any time. Those shares represent a claim on all the assets that the firm owns (net of its debt obligations), as well as future residuals that these assets can be used to generate. Debt holders are protected because in the event of bankruptcy, they get paid off before the equity owners do, which means that the board cannot protect the initial investment of equity owners without also protecting the investment of those who hold corporate debt instruments. In the cooperative, there are no equity owners at all to protect the major debt holder—the state. The state has no equity owners to suffer the blows of financial misfortune. Moreover, to the extent that a corporation’s board of directors represents the interests of the shareholders as opposed to those of (merely) current residual claimants, it is concerned with the long-term consequences of management’s decisions for the equity value of the firm. In other words, it protects the inter­ests of the capital providers and the future residual claimants.

This is not to say that the board always does its job. Sometimes incompe­tent management is protected, top managers effectively choose the board that oversees them, and various other forms of self-dealing take place. But putting principal-agent problems to one side for a moment, it is clear that the open corporation’s board of directors protects the interests of future residual claimants and capital providers better than the group whose institutional function is to represent only the interests of the current residual claimants. Through their agents—management and the workers’ council—the ultimate decision makers in the cooperative (i.e., the workers) can appropriate the quasi-rent value of the firm and, to a lesser extent, the quasi-rent value of the assets that future residual claimants might bring to the firm (i.e., the quasi­rent value of future worker’s human capital). By contrast, the ultimate deci­sion makers in the open corporation cannot exploit the capital providers because they are the capital providers. And they can sell their claims on the future residuals. As in the case of the classical capitalist firm, equity owner­ship closes down avenues of exploitation that are wide open in a market socialist system.

An obvious solution to the problem of the workers’ limited stake in the firm is to include representatives of the capital providers on the workers’ council. This proposal bears affinities to the codetermination movement, which seeks state-mandated representation of labor interests on corporate boards in free enterprise systems (Pejovich 1978). However, this body cannot properly be called a workers’ council. One of the most important rights of workers, the right of self-management, has been attenuated. In other words, the workers no longer collectively have ultimate decision-making authority in the firm, since representatives of the provider of the firm’s capital (i.e., state officials representing the public) now have a say in who the managers will be and how the firm will be run. But perhaps that is not such a bad thing. There appears to be a serious weakness in the structure of the market socialist system, and the most natural way to deal with it would be to put state officials on the council. Besides, there is no reason for socialists to recoil in horror from par­tial state control of the cooperatives. That is an anathema only to those on the Right.

Various forms of state ownership of the cooperatives will be discussed in chapter 8. For present purposes, it is sufficient to point out that the public officials who serve on such boards or those who appoint them may well have an even shorter time horizon than their counterparts in the cooperative, at least under the assumption that the political system is democratic. Since polit­ical actors can be turned out of office at the next election, their time horizons tend to be fairly limited. This does not mean that it never pays for political authorities to take the long view, but at least in recent decades, democrati­cally elected politicians and those who answer to them have not been known for their patience and farsightedness on economic matters. In sum, it is far from clear that appointing political authorities to the governing council of the large cooperative would provide much protection for the interests of the capital providers, the taxpayers (not to mention the future residual claimants).⁴

Equally important, the systematic difference in interests between the residual claimants and the providers of capital would probably make this arrangement ultimately unstable. Representatives of one or the other group would likely achieve a position of dominance (perhaps by changing the laws that set the ground rules) from which it could serve its own interests at the expense of the losing group. If, for example, the workers’ representatives maintained hegemony, the problem of the exploitation of taxpayers would remain. On the other hand, if the state achieved dominance, the system would be one of effective state ownership of most the means of production, which is incompatible with the type of market socialism under discussion.

Returning to the workers’ council as it was originally described, it is not even clear that it offers current residual claimants (the workers) better pro­tection against exploitation by bad management than a board of directors offers its equity owners against the same threat. The reason why has to do with the nature of the assets that the respective groups put at management’s disposal. For the workers, it is their knowledge and skills, and for the equity owners, it is their capital. The quasi-rent value of the capital contributed by equity owners tends to be relatively less than the quasi-rent value of the work­ers’ assets because it is easier for equity owners to withdraw from the firm by selling their equity shares on the stock market and redeploying those assets elsewhere.⁵ Indeed, an important consequence of the creation of the modern

corporation and the associated stock markets in which equity shares are freely traded is that the quasi-rent value of investors’ capital is significantly reduced; this makes these investors less vulnerable to exploitation than they would be if their capital contribution were less fungible. By contrast, workers’ assets (their respective capacities to labor and specialized skills) are, in general, more difficult to withdraw and reposition. If managers make some bad deci­sions to the point where a worker wants out, the latter cannot easily withdraw some or all of her labor assets from the firm the next day and park them somewhere else. She does, of course, have other avenues of redress in the large cooperative, but it is not obvious that they are more effective, from her point of view, than the investor’s option of selling stock. Selling one’s stock is a way of sending a message; if the message does not get through, the sender is at least no longer exposing her assets to dissipation by poor managers. It is the economic equivalent of voting with one’s feet; while it carries costs with it, those costs are generally much less than the costs workers incur in chang­ingjobs.....

The stock market in a free enterprise system is an extremely sensitive— albeit imperfect—device whereby the value of a firm is constantly assessed and reassessed. It represents the collective judgment of an enormous num­ber of people who have a vital personal interest in making a correct assess­ment of the equity value of large corporations. If poor managers start to dis­sipate that value, the market tends to register that fact and thereby invites equity owners to withdraw. This outside monitoring of management, assisted by securities analysts, cannot be provided in a market socialist system because there is no equity ownership, and for obvious reasons there could be no stock market in workers’ labor assets.

A related liability of a market socialist system is a weak or nonexistent market for corporate control. For example, members of the Yezhov cooper­ative and their workers’ council might not realize how poorly their manage­ment is doing relative to the firm’s potential. Corporate raiders from the Beria cooperative cannot buy control of the Yezhov cooperative, oust the lat­ter’s ineffective management, and bring the Yezhov cooperative up to its full potential. The most they can do is offer to join the Yezhov cooperative and run for office, in exchange for which they would get an equal share of the residuals, if elected. The prohibition on a market for corporate control serves to insulate ineffective managers and their workers’ councils from market­based accountability. This permits the exploitation of the other residual claimants (namely, the other workers), not to mention the capital providers.

It is tempting to think that the workers’ council, because it is composed of workers, would serve faithfully the interests of the cooperative’s members. While their hearts might be in the right place (the incentive alignments are, after all, quite good), there is no guarantee that they will do a good job at monitoring management. Monitoring very large profit-making organizations is difficult to do, and good intentions are not enough. Outside monitoring by people who specialize in that task and who stand to make a great deal of money if they discover that a firm is being poorly managed is a better dam­age-control mechanism than whatever internal recall mechanisms might be developed in a large cooperative to deal with poor management or poor oversight from the workers’ council. One reason for this is that to the extent that hierarchy exists in the cooperative, disaffected workers would be at an obvious disadvantage in launching an effort to replace management or coun­cil members. (Imagine corporate raiders on Wall Street being employees of their takeover targets.) In addition, disaffected workers would also be easier to buy off than corporate raiders. Of course, hostile takeovers in the open corporation are often costly and not easy to execute. Proxy fights, whereby dissident shareholders try to oust incumbent management through votes of the equity owners, tend to be even less successful, though how much of this is an artifact of the legal system is unclear. Nevertheless (subject to a few exceptions), in neither case are dissidents going up against people who can take their jobs away from them. It is naive to suppose that managers of large cooperatives would not develop mechanisms to neutralize dissidents within the firm, whatever the cooperative’s charter says.

Because it is difficult to monitor management of large, complex orga­nizations, opportunistic managers can get away with much more exploita­tion than can workers down on the assembly line, where monitoring is easy and effective. This fact might explain some of the resentment that has been historically directed at those at the top in business organizations; it might also explain why many people find management jobs so attractive. Both large cooperatives and open corporations must face this fact, which is inher­ent in the institutional role. However, at least in the open corporation, though the ill-gotten gains to individuals may be substantial, the overall losses are probably not too serious. The reason is that if losses were signif­icant, it is likely that the open corporation would have succumbed, like the dinosaurs, to its smaller, warm-blooded rival, the classical capitalist firm. Whether a similar fate would befall the larger cooperative at the hands of its smaller rival is less clear since both types of firms have the same organi­zational structure.

Exploitation along the Firm-Market Boundary

Before closing this section, it is worth discussing briefly some distinctive opportunities for exploitation that could occur along the firm-market bound­ary in a market socialist system. This requires a relaxation of some assump­tions made in the last chapter, namely, that transactions cost efficiencies would play about the same role in drawing the boundary between the firm and the market in a market socialist system as in a free enterprise system and that both systems would have about the same mix of small, medium, and large firms. One reason the latter might be false is that there could be an ideolog­ical bias for smaller organizations in a socialist society. In addition, an endoge­nous (i.e., economic) reason is that in this type of system, it would be much more difficult to effect mergers and acquisitions than it is in a free enterprise

system, which means that average firm size would likely be smaller. The rea­sons for this, and its implications for exploitation, warrant some elaboration.

Imagine the situation facing the management of two Hrms contemplating a merger or one firm contemplating an acquisition. Putting one of these together would be much more difficult and problematic in a market socialist system than it would be for an open corporation in a free enterprise systems. This is so for three reasons. First, in the case of an acquisition, the manage­ment of the acquiring company need only seek the approval of the board of directors and not the stockholders, whereas a majority of both cooperatives would presumably have to agree in a market socialist system. In an open cor­poration, the acquiring company simply acquires controlling shares of the stock. The situation is a little more complicated in a merger, but the merger of cooperatives would still require a broader consensus. Second (and perhaps more important), one common result of either type of change is that jobs are eliminated because of the attending efficiencies that the acquisition or merger realizes, and it is often unclear at the outset which jobs will go. Most—indeed nearly all—of the equity owners of the firms in a free enterprise system have no interest whatever in preventing a shrinkage of the workforce when a merger or acquisition goes through. By contrast, the ultimate decision mak­ers of cooperatives would have every interest in world in preventing just that.⁶ Finally, since the cooperative seeks to maximize net income per worker rather than net income, some mergers and acquisitions that would go through in a free enterprise system would be blocked in a market socialist system.

The fact that some mergers and acquisitions do not take place that from an economic point of view should take place has some important implications for exploitation in market socialist system. As pointed out in chapter 5, one of the principal costs of organizing transactions across markets rather than within firms is that transactors can have their quasi-rents exposed to their trading partners. If the size of firms in a market socialist economy were held down by ideological bias or by a bias traceable to the structure of coopera­tives, expropriation hazards (i.e., opportunities for exploitation) induced by asset specificities would exist that otherwise would not.

The same considerations apply in the other direction. Sometimes organi­zations are too large. The major costs of the firm, as opposed to the market, are the costs associated with bureaucratic inefficiency that is endemic to hier­archy.⁷ The manager of a division within a large multidivisional corporation or of a unit in a conglomerate may take his division or unit private through a leveraged buyout. This cuts out some of the cross-subsidization (and exploitation) that is inevitable in these large organizations; it also introduces the superior incentive alignments of the classical capitalist firm or, at least, shortens and narrows the bureaucratic hierarchy responsible for the original inefficiencies.

This would be a much more problematic undertaking in a large cooper­ative. Presumably, a unit could not secede without the concurrence of the larger cooperative of which it is a part; otherwise, some units would face seri­ous expropriation hazards, in the form of threats to secede, from other units within the cooperative. Suppose that an unusually efficient unit believed that it was responsible for more “value added” than were other units. If that were true and widely known, workers in other units would be reluctant to let it go. In a market socialist system, large cooperatives might well be composed of factions held together by a strongman or by an uneasy truce—Titoism at the level of the firm.

These observations do not presuppose that free enterprise systems always draw the line between the firm and the market optimally. Rather, this type of system permits experimentation with organizational forms within very broad limits (e.g., no slavery); the competitive process tends to select the efficient experiments and reject the inefficient ones. This openness to experimenta­tion in free enterprise systems, which would be curtailed in a market social­ist system, also plays a role in the evolution of new and different contractual arrangements. In recent years, a hybrid of the classical capitalist firm and the open corporation has become an increasingly important organizational form in free enterprise systems: the franchise arrangement. Individuals enter into contractual agreements with large corporations (e.g., McDonald’s) to market and sell the latter’s products. The franchisee is the residual claimant on the franchise, but he is significantly constrained in how he operates his business by policies and procedures imposed by the franchiser. This type of business arrangement, which has never, to this author’s knowledge, been investigated and evaluated by socialist theoreticians would create a real problem for a market socialist system. Must the franchisees be ultimate decision makers in the parent company, or would that role be limited to the members of the lat­ter? And who, after all, is to count as member of the cooperative? Some of the most interesting recent work in transactions cost analysis explicitly recognizes that the boundary between the firm and the market is often blurred or arbi­trary. The study of the economics of organizations is being transformed into the study of the economics of contracts. As they scrutinize novel and exotic contractual arrangements for conformity to the principles of market social­ism, socialist theoreticians run the risk of looking like Islamic fundamental­ists trying to devise a civil code for a modern state by applying the Koran.

To summarize, opportunities for exploitation will exist along the firm­market boundary in both free enterprise and market socialist economic sys­tems. And free enterprise systems do not draw that boundary optimally so as to minimize, in any absolute sense, those opportunities. (For a better alter­native, imagine an economic system in which God is in charge of determin­ing the boundary between the firm and the market.) However, free enterprise systems encourage experimentation on where and how to draw the bound­ary, whereas market socialist systems prohibit or discourage experiments for structural, legal, and/or ideological reasons. Because of this, a free enterprise system tends to do better than a market socialist system in removing or min­imizing opportunities for exploitation along the boundary between the firm and the market.

218

The Philosophy and Economics of Market Socialism