Guild Socialism Reconsidered

by Roger A. McCain


The subject of this essay will probably seem strange, from one point of view or another, to most readers. For start, it advocates a kind of socialist economic organization that is anti-government. As most people use the term "socialism," that would seem to be a contradiction in terms. In order to find the basic ideas of anti-government socialism, we have to go back to the period before the Russian Revolution of 1917, when some schools of socialist thought were quite close to anarchism. Guild Socialism was one of these. If that doesn't bother you, and some economic jargon about labor-managed enterprises sounds natural to you, then you can skip this preface without missing anything.

Nevertheless, as I reread the paper that follows, I realize that it is not really a very good introduction to Guild Socialism. (It wasn't written to be an introduction to those ideas, but it is the nearest thing to it that I have in my files). My purpose in this fairly extensive preface is to fill in some of those blanks and to explain "where I am coming from."

The Guild Socialists were active mostly in Britain in about 1890-1920. Their idea was that each workplace should be a more or less sovereign political jurisdiction -- and a democratic one, so that the supervisors would be elected by the people they supervise and the directors of the company be elected on a basis of one person, one vote. A workplace is a place where a group of people work together to produce some distinct good or service for sale or for the use of society. Most workplaces are small by political standards, and "workplace" implies a particular locality, not a world-wide financial empire. In making the workplace a basic political jurisdiction, the Guild Socialists linked themselves with others who wanted politics "closer to the people" in small, decentralized political units: anarchists and anarcho-syndicalists, many Christian Socialists and cooperators.

In a modern economy, a single workplace is not isolated but part of a worldwide network of cooperative production and consumption. Thus, the Guild Socialists expected that the local workplace-republics would federate both horizontally (different industries within a location) and vertically (different enterprises within an industry) at various levels. The vertical by-industry federations were the "guilds" from which "Guild Socialism" took its name.

Exchange and distribution of products would be settled by some combination of markets and plan, with the understanding that the plan would be created by agreement among the guilds, other federations, and their member enterprises, not handed down from some government. There was a range of opinion about the relative importance of plan and market in Guild Socialism, and at least a possibility that the Guild Socialist system could rely mostly on buying and selling -- that a plan might play less role in the proposed society than it realistically does in modern corporate capitalism, let alone in the Soviet society. That was left to be decided on the basis of experience, though, in that generation, there was some leaning toward the plan.

Since the workplace-republics would absorb some of the unavoidable functions of government, and their guild and regional federations would absorb others, all Guild Socialists looked forward to a reduced role of central or national government. Here, again, there was a range of opinion. Some Guild Socialists anticipated that a limited government would still be a distinctive, balancing power center in society, while some semi-anarchist extremists would have abolished traditional government completely, with the guilds and other federal bodies constituting a revocable government for the minimal tasks (such as defense) for which a central government seemed indispensable.

These ideas faded away in the face of the Russian Revolution, the fascist and New Deal developments, which made it seem that government centralization were the wave of the future. When that wave had levelled off -- if not passed -- some of the ideas of the Guild Socialists came back in the form of "market socialism." Market socialists suggested that socialist enterprises could settle their books by buying and selling in markets, rather than through a government-imposed plan. Some of these market socialists suggested that the enterprises could be organized democratically -- granting, as it were, home rule (if not sovereignty) to the work-place republics, to decide what they would produce, charge and pay. But what was missing was the politics, the determination to reduce domination, both political and economic, to the absolute minimum and to keep reducing them in the future.

And that's what I like about Guild Socialism. Of course, the author of this essay is something of a "relic of the 1960's." What happened in the 1960's (in my opinion and experience) was that the moderately conservative consensus of opinion that had prevailed in the 1950's was shaken, so that people were emboldened to think seriously about alternatives they had to some extent had in mind all along. We hadn't taken these ideas very seriously before -- it didn't seem worthwhile, since "everybody knew" there was no need for alternatives -- and only a minority were affected, since most people hadn't any thoughts about alternatives to begin from. Also, the quality of this bold thought varied depending on the odd ideas people had in their minds to start with. Some of us examined the ideas we had been repressing and discovered that they made pretty good sense. Among those were a small group of semi-anarchist economists -- Marxists called us "anarcho-liberals" -- who were "skeptical about government" before that was cool, but who didn't like capitalism, especially large-scale corporate capitalism, all that much either. But we liked democracy. Our idea was, in simple terms, that the employees ought to elect the bosses and decide what the company policies would be. That would be democratic!

When we looked at this idea from the point of view of economic theory, we found that it made more economic sense than we had expected. Nor was it untried. There had been a large number of experiments in worker-directed enterprises (mostly worker-owned, too, but with some exceptions). The evidence was that these worker-directed companies worked pretty well -- not perfectly, of course, but at about well as capitalist ones ever did, maybe better, (and far better than state enterprises) all circumstances considered. For the bibliographic references to support this, see my "Cooperation: The Proper Study of Economics,"International Journal of Social Economics, 1993, v. 20, no. 10, pp.55-78. Anyway, that was a pretty big surprise, since we had expected to pay an economic price for political righteousness. The evidence said we can be righteous for free, or even profit by it! The economic theory and statistics literature that grew up on this topic spoke of these democratically organized companies as "Labor-Managed Enterprises," "LME's," or (borrowing a pun in an early paper) "Illyrian enterprises."

We also speculated about an economic system organized around Labor-Managed organizations, a Labor-Managed or Illyrian Economy. There was some thought that at some stages in its evolution, the Yugoslav economy might have come close, if only it had been more democratic -- but the Yugoslav economy was a moving target and the one thing that was stable was its lack of democracy. To a considerable extent, our speculations about a Labor-managed economic system were rediscovering ideas that had been developed in the nineteenth century by people who called themselves "socialists," usually with a qualifier like "Christian" or "guild." (In those days, socialists didn't have to insist that they were democrats. That was taken for granted). This paper was largely my response to discovering the ideas of the Guild Socialists in particular. I had a feeling my twentieth-century economist friends could learn something from the ghosts of the Guild Socialists -- and conversely, if ghosts could learn.

This paper was written rather gradually, over a number of years. It was never likely to be published in a great journal and persuade my dean to grant me a merit raise, so I never gave it much of my time. I pulled the bits and pieces together for a presentation at the December, 1986 meetings of the Association for Social Economy, New Orleans, La. I went to some trouble to make it sound like the sort of thing a distinguished professor writes. It would probably be more fun to read if I hadn't. have made a few revisions today to bring it up to date a little on things like the collapse of the Soviet system. If it makes sense to you, by all means drop me a line. If you want to tell me what's wrong with it, that's fine, too, but remember that I've heard the obvious points before. Either way, the times may come again (as they came without notice in the Soviet Union) when we all need to think about alternatives. If so, I hope this helps.

Reconsidering Guild Socialism

The purpose of this paper is to reassess the proposals of guild socialism in the light of developments both in the economy and in economic theory over the past seventy-five years. Guild socialism emerged from the British socialist movement in the first decade of the century, drawing on and reacting to the experience of the founders of the Fabian society and of the socialist opposition to Fabianism (Morris). This latter tendency in turn drew on the archaizing and medievalizing thought of Ruskin and others (also an important source of the thought of Tolstoy and Gandhi). Guild socialism was distinctive in its time as an attempt to synthesize socialism and anarcho-syndicalism. Its program was quite explicit and detailed, to a degree which suggests utopianism (note especially Kropotkin on this; and compare Buber). The central aim of guild socialism was expressed in the phrase "self-government in industry," with the emphasis on the factory or other workplace as the unit of self-government. However, consumers' interests were quite explicitly to be represented in industry as well, though Guild Socialists differed among themselves somewhat on the details. Perhaps most interestingly, and importantly, Guild Socialism went further than most other forms of socialism in providing directly and explicitly for the withering away of the state, which was to begin, not in the misty future, but on the morrow of the establishment of a Guild Socialist system.

These very elements of Guild Socialism must have seemed archaic in the 1930's, when it was clear that the tide of the state was rising, and that workers' councils had little part to play in the building of the New Order. By the 1970's, it seemed, however, that more fertile ground existed for Guild Socialism. This was suggested by a new interest in "self-government in industry," that is labor-management or at least labor participation in management; by the distrust of the state which seems no less deep in some parts of the left than on the right, and the corresponding value placed on decentralization of institutions; and by the growing recognition of a need for human community in the business organization. Since the 1970's, the anti-government trend has certainly continued, but with a distinctly capitalistic flavor. The Anglo-American politics of the late '70's and onwards, the collapse of the Soviet Union and the eastern European Communist governments, and the renewal of right-wing nationalism in the '90's do not invite serious consideration of socialism in any form. But this new-old capitalism has yet to improve the lives of people, and for those of us who remain just as skeptical of capitalism as we are of government, Guild Socialism offers an alternative to both that is (at least) fairly well thought out.

I shall try to summarize the economic system proposed by the Guild Socialists, as it appears in the light of more recent economic theory and development, drawing especially on G. D. H. Cole's Guild Socialism Restated. I will also try to indicate the ways in which a Guild Socialist might want to update his proposals in the light of those theories and developments. References to Guild Socialism Restated will be by page numbers only.

This book, a relatively late manifesto of the movement, thus incorporates some earlier controversies, but is an expression of a distinct, and relatively radical, tendency within the Guild movement. However, the more moderate views within the movement were just that -- moderate -- in accommodating positions which were not distinctively those of Guild Socialism but common to other tendencies of opinion. Thus, a relatively extremist exposition of the ideas of the Guild Socialists is an appropriate source for its more distinctive ideas.

Having reviewed and proposed some revisions of Guild Socialism, I will investigate the extent to which an updated Guild Socialism could be an instance of the self-managed society. This would imply two things: first, that it would provide an environment complementary to labor- management in enterprises, within which that system of enterprise governance could realize itself fully, and second, that it would realize in the society as a whole the non-economic values which some advocates of labor-management perceive in labor-management of enterprises. Accordingly, I will explore the implications of the Guild Socialist system as a whole for some of the generally acknowledged economic difficulties which would in theory attend labor-management especially in the hostile environment of capitalism.

As previously mentioned, I will propose some revisions of the Guild Socialist idea in the light of the past seventy-five years of economic thought and history. This raises a problem which arises whenever anyone attempts revision of a system of social ideas: what is essential, not to be revised, and what is inessential, and thus liable to be revised in order to preserve that which is essential? Disagreements on that are the stuff of which schisms are made, and it is no wonder that "revisionist" is an insult on the left. Nevertheless, if ideas are to be of any use at all, they must be revised from time to time in response to the emergence of human history -- as, indeed, social systems must also be revised. Accordingly, I shall sometimes be using the term "Guild Socialist" in a broad way.

Sketch of Guild Socialism

"The Factory to the Workers, the Mine to the Miners -- and the dust to the dustmen, I suppose."

-- Beatrice Webb

What, then, is essential to Guild Socialism? Certainly one essential is representative democracy, but representative democracy in a rather specific form: functional pluralism. "It follows that there must be, in the Society,as many separately elected groups of representatives as there are distinct essential groups of functions to be performed." (p. 33) Each factory,workshop, mine, etc, defines one such group, and so must have its own representative council and internal autonomy. The administration of an industry is to be a federation of such groups -- in some cases, a rather loose federation, though this would vary from industry to industry (and by time and place) according to convenience. Some industries will need considerable centralization, but there is a prejudice for decentralization. These federations of factories or workshops are the guilds; they, in turn, federate themselves for the administration of industry as a whole. (p. 124) This federation is not simple, however: the guilds federate themselves to form an economic administrative body for each each city or rural district; the regional administration is at once a federation of industries and of towns and districts; the national administration is at once a federation of industries and regions. (Compare Abad de Santillian, Baldelli). Moreover, this interlocking multiple federalism in industrial administration is a model for society and for the state or "commune" as a whole. In a more ordinary federation, an individual is a citizen of only one or another of the constituent states or provinces. But this cannot be so in a functionally constituted society. The functional groups in society cut across one another,and an individual is a member at the same time of several of them. He is a worker in an industry, a resident of a town or district, a consumer of this and that, and he must be a citizen in each group, represented in each, and thus the society must be at once a federation of overlapping groups of many different kinds, with their own respective competencies and powers.

Thus, in addition to the guilds, there are other functional bodies concerned with economic administration. At the least, there are consumers' councils or boards for the public utilities and for other private consumption. (p. 87) There may be many such councils, perhaps one for each industry. (p. 93) At the level of town or district, region, and nation, the various functional bodies federate themselves to form a "coordinating body." This "coordinating body,"the "commune," is the minimal state of the guild socialist society. (Cole claims that it is not a state in the most crucial sense, p. 121, but his reasons for this choice of terminology -- that is explicitly what it is -- may now be thought obsolete. In any case the commune disposes of whatever remains of the coercive power of the state, pp. 128-9, 149, 152). The decision-taking body of the commune is a joint board of representatives from the guilds, the consumers' boards, and other functional or territorial bodies. The guilds (workers) and the consumers are roughly equally represented. (p. 125) The decision-making process of the communal council is majoritarian. In addition to the communal council, however, there is also a guilds congress at each level, local, regional, and national. The guilds congress is the joint council of the guilds. Like the communal council, it is a representative body with coordinating functions as among the guilds. It does not have powers which ordinary usage would attribute to the state, but it may appoint the guilds' representatives to the communal council; alternatively they may be appointed directly by the guilds. A simplified table of organization is provided in figure 1 for the reader's convenience.

You'll just have to load this one.

(I offer this table of organization with some hesitation, as the utopian literature of the last two centuries has had a certain weakness for such diagrams. But fascism, too, liked tables of organization, and it can be alarmingly difficult to distinguish between the libertarian ones and the fascist ones! It must be understood that there is no such thing as a "blueprint" for a new society and that the essential things cannot be represented by a graphic.)

Some other functional bodies are explicitly proposed; more detail is offered on the character of the consumers' councils. Before proceeding to that,however, we might notice in passing how commodity prices and some other economic magnitudes are proposed to be determined -- both for their intrinsic importance (for economic theory, at least) and as an illustration of the working of the proposed system. The consumers councils have the power to consult with the guilds about the prices, quantities of output, and qualities of the consumers' goods which the guilds produce. These things are decided, in the first instance, by negotiation (pp. 125, 128, 142) with the communal council arbitrating in case of impasse. Thus, we have something in the nature of worker-consumer codetermination. (McCain 1980a) A French proposal for nationalized industries under joint worker-consumer control is "not guild socialism" though very closely allied. (p. 214) The prices of intermediate goods are determined by direct bargaining between the guilds concerned, with arbitration to the guilds congress in the case of impasse and further arbitration by the communal council if the guilds congress cannot arrive at a decision. Interlocking directorates may help closely related guilds to coordinate their decision processes, at their discretion; (p. 68) perhaps mixed worker-consumer boards would be similarly helpful in the negotiation of consumer prices.

The determination of prices may lead to operating cross-subsidies, for, as Cole remarks (p. 143) "It might be considered desirable ... to sell a particular commodity at ... less than the natural price based on its cost of production." This involves a system of direct lump-sum transfers among the guilds, to cover the costs of production of some goods for sale below their "natural" prices. These transfers are decided like other economic magnitudes under guild socialism -- by direct negotiations among the guilds with arbitration by the next, more inclusive body in the case of impasse. The allocation of investment is to be decided and financed in the same way, (pp.143-145) as are wage schedules. (p. 74) There is to be no profit-sharing; (p.198) rather, profits revert to the commune. (p. 147) Incomes are independent of the state of the market. (p. 74) The Guilds Congress thus also concerns itself with "the adjustment of supply and demand." (p. 48, 60) Thus, we seem to see the Guilds Congress generating something in the nature of a central economic plan, by negotiation. The plan comprises prescriptions for outputs and prices by industries, for wages and cross-subsidies, investments and gross taxes, among other things. Negotiation impasses are arbitrated by the communal council, which must in any case ratify the result.

Cole groups the consumers' representatives into two broad bodies: one for public utilities, the "collective council," and one for all other personal consumption, the "cooperative council." (p. 87) This clearly reflects the role of the consumers' cooperation in the working class movement of Cole's Britain, and Cole welcomes the emergence of far more specialized (hence potentially far more numerous) consumers' bodies. (p. 93) Perhaps, in the pure case, there would be a consumers council corresponding to each guild which produces consumers' goods, and such general consumers' bodies as the collective council and the cooperative council might then be federative. This reinforces the picture of the system as one of worker-consumer codetermination. There is, however, another possible reason for the limited number of consumers' bodies which will be explored later: it is the problem of political participation.

Nothing has yet been said of the administration of what are called public services. Some of these are taken over by the guilds: each guild is to have its own small civil service (p. 141) but the residuum includes education and health care, which are to be free and public. It is here that we find one of the most interesting and far-reaching of the guild socialist innovations. The production of civic services is also to be carried on by guild organizations,the civic guilds. Corresponding to each civic guild or major grouping of civic guilds is a specialized citizens' council, which is to be consulted in the administration of the civic services. The civic guilds and the specialized citizens' councils are represented (about equally) on the communal councils;presumably the civic guilds are represented on the guilds congress. Cole proposes just two citizens' councils, and corresponding groups of civic guilds:a health council and a cultural (education and arts) council. It is not clear to me where firefighters, coast guardsmen, census statisticians, and Beatrice Webb's dustmen fit into these two, or if they are subsumed to other guilds or guild groupings. (pp. 114, 108) (The navy, army, and police are to be organized along guild lines but are under the authority of the commune; no mention of citizens' councils is made in this connection, though that might be an oversight. A preference against a professional military is expressed. pp.151-2.)

The civic services are funded by direct transfers between the industrial and civic guilds. (p. 146) "The Guild system has thus a perfectly easy and flexible instrument of taxation. It does not tax the individual, save perhaps in exceptional cases, such as enterprises continuing to be conducted outside the guild system. It taxes at the source and draws the sums approved by the Commune in the form of an agreed claim on the labor-power of the Guilds.... Let me again emphasize the fact that, in practice, the greater part of this detailed financial work would not be done by the Commune at all, but directly and in consultation by the various functional bodies. Only moot points needing settlement and general questions of principle would normally come before the commune." (pp.146-7) That is, the transfers are decided like other economic magnitudes in Guild socialism. The difference between the civic guilds and other guilds, some of which receive operating cross-subsidies and sell their outputs at less than cost, is strictly a difference of degree. There is little or no distinction between public and non-public finance. Since individuals are not taxed, the commune is indeed not a state as Schumpeter uses the term.(Capitalism, Socialism and Democracy, p. 169)

What about entry into the guilds? In Cole, this is rather vague: The guilds are "open associations which any man may join" but "this does not mean, of course, that any person will be able to claim admission, as an absolute right,into the guild of his choice." (p. 75) Then what does it mean? There may be training requirements, and "a man clearly cannot get into a Guild unless it needs fresh recruits for its work. He will have free choice, but only of the available openings." (p. 75) There does not seem to be much novelty in the new society in these matters, except that one of the functions of the guilds congresses is to assure that the guilds do not restrict entry "from any ulterior motive." The ulterior motive would, of course, be the exploitation of monopoly power visavis other groups in society.

We shall have to fill in some blanks. For any person to be excluded from a livelihood through not being a member of a guild is clearly inconsistent with Cole's equalitarianism, (pp. 16-20, 75) which is that of any socialist. Thus membership in some guild, and entitlement to a job and an income through that membership, must be a matter of right in the guild socialist society. Cole does not mention hereditary membership, but some Guild Socialists did defend it, and it has the merit of assuring that membership in some guild would indeed be a right to a young person, presuming that an initial generation has been universally enrolled. But this would be likely to produce an inefficient pattern of employment, and inequalities.

It is, as noted, a responsibility of the Guilds congress to assure that membership and employment in the guilds is not restricted in any antisocial way. Here is one possibility. We might suppose that the number of new recruits to each guild would be negotiated within the Guilds Congress along with prices, investment and transfers. This, too is subject to ratification by the communal council. The recruiting quota is then disaggregated by each guild and "sent down" to its member enterprises, each of which must then open a corresponding number of memberships but, within that guideline, may "pick and choose" among applicants. On that basis, a guild which is to expand might negotiate, as a condition of its expansion, the allocation of sufficient investment to supply its new recruits with tools. Guilds faced with reductions in force might have the transfer of some definite number of their members negotiated.

Since the guilds are themselves federative, and in some cases highly decentralized, something may be said about the division of function among their levels. The national guilds, as federal bodies, specialize in those industrial functions for which considerable economies of scale are plausible, including marketing and procurement, and would "organize research" and regulate quality.The latter two functions presumably subsume product development between them. Production, however, is carried on by the local factories and workshops,subject to the qualification that the industry total output is determined by the guilds through negotiation. This qualification seems to me to be very strong, as it seems to imply that the plan is disaggregated by the national guilds, establishing output quotas by factory and workshop, thus leaving very little really for the self-governing workshops in their autonomy to decide. None of this, however, is mentioned. I will return to the point later.

The regulation of quality and the determination of price and industry output are determined by the guild in consultation with the corresponding consumers' council. The competence of the consumers' councils is, however, limited to these three things. These are the aspects of production which directly concern consumers, and thus, for Cole, the only matters in which they are entitled to a say (p. 88). Since these are functions of the national guilds, then, it seems that consumer- consultation is dispensable for the local units of production, at least in so far as they serve national markets.

In assessing Guild Socialism, we may well ask, "What is the point of all this negotiation?" Efficiency is not central to it. In a world without transaction costs, negotiated allocations of resources must be Pareto- efficient, but so must those which arise from market interactions. (Coase, Demsetz, McCain1986a) When we allow for transaction costs, we must at once observe that negotiation is costly: Guild Socialism begins with a large overhead of fixed transaction costs built into its economic system. Now, it may be that this investment reduces variable transaction costs sufficiently so that Guild Socialism is more efficient on total cost grounds (McCain 1986b). The variable transaction costs are those of procuring "public goods" (McCain 1986a) for, i.e. defending the collective economic interests of the functional groups. However, in short, the relative efficiency of Guild Socialism and market capitalism is simply a moot question, at best, in a world of positive transaction costs.

The answer is rather to be found in the Guild Socialists' distinctive concept of democracy. We ordinarily associate "democracy" with one man, one vote and majority rule. This is not the only possible interpretation. The historical connection between Guild Socialism and Anarchism has been mentioned. R. P.Wolff claims that Anarchism is "unanimous direct democracy," i.e. unanimity rule. (Note also McCain 1972, 1974, 1980b) This is not free-marketeering (Reubens, 1980, somewhat to the contrary notwithstanding) but the difference is that the mutually voluntary outcomes of free-market trading are mutually voluntary only subject to some given initial distribution of goods and rights. Unanimous direct democracy imposes much narrower requirements -- formally,those of the non-envious allocations of equity theory. (Varian, Baumol, McCain1980b) That is, the initial allocation is not given but must itself be subject to unanimous agreement.

No doubt unanimous direct democracy is an extreme position: the Guild Socialist position is more moderate. It is also founded in a view of what is practical. "It is clear that direct coercion of such a group by means of the economic boycott ... would be possible, but extremely undesirable. ...Coercion of a powerful group is inevitably not far removed from civil war, if the will to resist is present in the group; and Society ought therefore to be so ordered that such coercion shall only be invoked in the last resort, and shall scarcely ever, if at all, be required. The best way of making this provision is by ensuring for every reasonable claim the fullest possible amount of social consideration ...." (pp. 156-7) Thus social decisions should be usually based on the unanimous consent of the functional groups. This provides a practical approximation of "unanimous direct democracy" or of equity in the sense of non-envy.

By way of overview, then, the guild socialist society erects collective bargaining into the major principle of social organization, so that both the constitution of the state itself and economic planning are products of collective bargaining. (In the phrase "the constitution of the state," the word constitution is a transitive verb, not a noun). In this bargaining process, impasses are resolved by arbitration upward, the disputes being arbitrated by the next more inclusive body. Thus the national communal council is the final arbiter, but the supposition is that it will be called upon to do very little, since little will be "bucked" all the way up to it. The Guild Socialist society relies to countervailing power (Galbraith) to bring about accommodation among groups and to protect freedom and enhance efficiency. Countervailing power presupposes that the contending interests are all organized. This is why the functional organizations of Guild Socialism are both so all-encompassing and so complex (but compare pp. 64, 65, 114-5). This complex, interlocking system of social groups constitutes society, as the Guild Socialist conceives society. It is a democratic society, in so far as each of these groups is itself a representative democracy; but perhaps not a democracy if democracy means one man, one vote in every decision. Who is entitled to a vote on which matter depends very much on who is supposed to have in interest in which matter, and by "interest," material interest is most usually meant. While the commune may be described as a state, its functions are the minimal ones of coordination and arbitration among the major functional groups in society, and thus it may be thought of as a natural and minimal state. Nevertheless the Guild Socialist society is a socialist society if, with W. A.Lewis, (p. viii) we consider democracy and the classless society as the essentials of socialism.

Problems of Guild Socialism

At least in the form that Cole gave it, Guild Socialism was not, as I have said, market socialism. But most of the "Socialist Controversy" lay in the future as Cole wrote, as did experience of the shortcomings and difficulties of planning in Soviet-type economies, the Yugoslav experiment, codetermination in West Germany (p. 205) and a couple of generations of experiments in worker-ownership in capitalist countries. Cole certainly did not think that material incentives would be necessary to assure efficiency: workers were to be "put 'on their honor' to do their best." (pp. 88-89) I believe that most advocates of labor-management would now hold that material incentives to effort are by far the least evil means of maintaining effort over any considerable period of time. Apart from that, I shall say nothing to justify the following comments to one who might believe today as Cole did in 1920, not because there is nothing to say, but because there is too much, and it is too familiar.

Guild socialism could be, and should be, revised in the direction of market socialism. Each enterprise in Guild Socialist industry should be permitted to keep any residual of the revenues from selling its products over the cost of producing them, and in routine circumstances, the wage bill of an enterprise would be no more than its revenues net of payments to other enterprises for raw materials, rents for its capital goods, and contributions for social services.(Previously negotiated operating cross-subsidies are included in the revenues). In other words, the individual enterprise would function as the "Illyrian" labor-managed enterprise (Ward, Vanek) is supposed to do, deciding its own output subject to parametric prices. These prices, in turn, might be decided by supply and demand (where competition is feasible) or by collective negotiation between consumers' representatives and those of the workers. Thus the economic plan would not determine wages or outputs, and any plan magnitudes for those variables would be indicative in character. The same would be true of competitive prices, though not, perhaps, of monopoly prices.

Another difficulty in Guild Socialism, at least equally fundamental, is the problem of inducing a sufficient degree of political participation in the selection of the various representative bodies. This would be especially difficult for the numerous consumers' and civic councils, both because they would be so numerous, and because each consumers' good and individual civic service would be such a relatively small concern for the individual consumer. Cole is aware of this difficulty: I suspect that it is part of the reason for his proposing rather broad grouping of consumers' and civic council representation. The problem, as I see it, is this: The case for representative democracy arises from the obvious cost and inefficiency of direct democracy. It is very costly for each person to inform herself fully about each issue; and it is not at all likely that very many will do so. Some will participate ignorantly, and some will decline to participate at all. This is likelier as the issues have less significance for the person, as, e.g, deliberations on the price of salt will elicit less participation from me than will deliberations on the salary schedule of my employer. Representation permits some persons to specialize in the consideration of the issues, and thus economizes on effort, and, one may hope, produces decisions which are technically better. It does not altogether solve the problem, though. If the representatives are elected, then the voter must inform herself about the candidates, and this is far more complex than informing oneself about any single issue. A representative is a mixed bag, with whom I may agree in some issues and disagree on others. (p. 33) Candidates can dissemble, while an issue can hardly lie. An awareness of these difficulties lies behind Cole's preference for indirect election and his demand for powers of recall. (pp.133-5)

These difficulties may be overcome to some extent when a single representative is to be elected to an omnicompetent unicameral legislature. In that case, the election touches in the widest possible way on the person's interests. However, the more functional representatives are chosen, and the narrower the competence of each, the less participation there will be in each election, and the less expert that participation will be. Only about half of the electorate turn out for American presidential elections. The history of the consumers' cooperative movement will illustrate how serious and chronic this difficulty can be, with respect to consumer representation. An election for an American consumers' council, the Health Services Board, in my area seems to have been decided by something like ten or twelve total votes for the winning contestant in a population of something like a hundred thousand. (Personal information from one of those elected). Elective boards simply do not, in practice, represent consumers' interests.

A relatively recent proposal (with very old roots) is that instead of electing representatives, we might have them chosen at random. (Mueller, Tollison, and Willet, Dahl) This has been called Randomocracy. Modern statistical theory assures us that such a body would indeed be representative in a rather precise sense. It would, with a high probability, have about the same proportions of blacks, women, butchers, bakers, candlestick makers, city and country people, old and young people, radical and conservative people, and residents of Texas and Delaware, as the population as a whole. It would not require periodic redistricting, and would not admit of Gerrymandering and rotten boroughs. It would avoid the arithmetic paradox of representative democracy, under which measures might be passed although supported by as few as twenty-six percent of the voting population even if the system worked perfectly. (Only 51% of the representatives are needed, and each of them needs the support of only 51% of voters. Thus, winning voters could be 51% of 51% =26% of the voters).

Of course, Randomocracy has its shortcomings. First, its advantages are only probable, and quite nonrepresentative bodies could be chosen. (But the same is true of elected representatives). Second, it relies on the Law of Large Numbers, so, for most purposes, the bodies selected at random would need to be fairly large. Finally, while it may be a satisfactory means of making collective decisions, it may not be an acceptable means of choosing leaders. Where an authority relationship exists between the person chosen and the electors (e.g. between a work group and a foremen), the question is one of responsibility rather than representation. In such cases, the people subject to the authority should be able, if it is their will, deliberately to reject him and choose another instead. This means voting rather than random selection. But for representative decision-making bodies, such as consumers' and citizens' councils, randomocracy seems appropriate.

This, I believe, goes some distance toward solving the problem of participation. Thus, the members of consumers' councils would not be elected but would be chosen at random from appropriate lists of consumers. Those selected for (let us say) the consumers' council for local telephone service would be able to specialize in that, and acquire knowledge of the conditions of telephone service, without wasteful duplication of effort and without neglecting the special knowledge needed in the bus council and the local distributive trades council, for example, each of which would be somebody else's responsibility. On this basis, I would suggest the following further revision of Guild Socialism: One consumer board for each industrial guild, and one citizens' board for each civil guild. The workers' and consumers' council together then constitute a bicameral board which could speak for the industry as a whole in its relations with other industries, with respect to such things as the allocation of capital and cross-subsidies, which affect the interests of both workers and consumers. This admits of more symmetry in the treatment of workers and consumers within the commune.

The history of the past sixty years requires mention of two other problems: the status of otherwise unemployed homemakers and the problem of environmental preservation.

Any more or less syndicalist system disenfranchises those who do not hold jobs, if holding a job is a condition of membership in the syndicate. On the other hand, nonworking members of a syndicate compromise its functional character. Since homemakers commonly do not hold jobs, (this was even more true in the beginning of the century, of course) such a system could be quite thoroughly sexist, and so unacceptable. Several possible solutions, all compromises, suggest themselves. The least unpromising one would seem to be this: otherwise unemployed homemakers would be regarded as full working and voting members of their spouses' guilds, since they depend on those guilds for their incomes, while retaining the option to return to their own guilds when they wish. Every adult would be a member of a guild with a right to employment and income in that guild.

Finally, there is the question of environmental preservation. In a system based on countervailing power, it would seem that the preservation of the environment would itself have to be represented by some organization. Now, the interest in preserving the environment is the interest of people who live in the same region or locality, so this is to say that territorial units are also functional (via the biological environment) and thus territorial representation is not altogether to be dispensed with. Cole's Guild Socialism does not, however, provide for territorial representation apart from the other functional bodies. Thus, local territorial councils with particular responsibility for environmental preservation might be added to the complex of councils and congresses which Guild Socialism proposes, and empowered to consult with local enterprises on the choice of techniques and related matters, and with consumers' cooperatives on stocking of products which might threaten the environment. Because of the association of territorial representation with omnicompetence, it is probably best that there be no territorial representation above the local level, since such representation would compromise the functional character of the social organization.

These brief comments are meant only to indicate how the problems might be dealt with, not how they must -- as indeed is the purpose of the entire exercise, and of Cole's. I see no reason to doubt that they may indeed be resolved within the fundamental principles of functional representation, negotiation with arbitration upward, and self-government in industry.

Problems of Illyria

The time has come to turn the issue around and consider some acknowledged theoretical difficulties of the Illyrian system of "market syndicalism" and inquire whether Guild Socialism offers any potential answers to them. The difficulties I refer to are the Ward-Vanek effect and the Furubotn-Pejovich effect. (McCain 1977) These refer, respectively, to alleged tendencies to recruit too few members (so that there will be fewer heads among whom to divide any economic profits) and to allocate too little to "plowback" investment out of current proceeds. A further problem is that the principles of entry of new enterprises under labor-management are unclear, and this is particularly crucial, since entry would be expected to mitigate or eliminate the Ward-Vanek effect in the long run. (R. Martin) Now, the Ward-Vanek and Furubotn-Pejovich effects have been the subjects of large and often contentious literatures, and there is reason to believe that both have been exaggerated. (McCain 1977,1985, 1986c; Bonin, Conte, Dreze. I have made proposals for financial arrangements that seem to me to reduce those problems considerably; McCain 1977 and a paper on this server). For present purposes, however, I will take both as being fully valid, in the context of realistically imperfect markets (positive transaction costs).

In Ward's "Illyria" the enterprise has no net worth, capital being hired just as labor is. This means that there is no question of plowback investment, and the Furubotn-Pejovich effect does not apply. Moreover, economic profits (and losses) arise only as transitory results of unanticipated fluctuations in price, so that the restriction of membership is a short run phenomenon, expressing itself as a backward-sloping or (more generally) excessively inelastic "supply curve." Furubotn and Pejovich criticize the zero net worth or "pure rental" assumption as unrealistic, but that assumption has a normative significance in a socialist context. For an enterprise to have a positive net worth in nonhuman means of production is to violate the norm of "social property," which holds that the rents to nonhuman means of production (rent and interest) should return, in some way, to society as a whole. Thus, Vanek proposed a "National Labor-Management Agency," a financial body to which rents for capital would have to be paid, regardless whether the investment was financed from plowback or not. This would at least mitigate both of the effects, perhaps removing the Furubotn-Pejovich effect. (Note also McCain 1977). Presumably the Guild Socialist society would include some such institution. The negotiation of cross-industry flows of investment funds would also offer an opportunity to rectify any shortcomings in the investment decisions of the guilds and their members.

With respect to the restriction of membership, the Ward-Vanek effect, we can say at least that Cole was aware of this danger long before Ward had written (p. 75) and regarded the prevention of such restriction as a function of the commune. I have suggested that this function might be discharged through negotiated quotas for new entrants. For existing members of a guild, continued membership, employment and income would be a right (subject to "due process" as any right is in civil society, p. 56). The guilds are in the position of the labor-managed enterprise in Vanek's "social short run;" they make their allocative decisions subject to fixed membership, and thus are not able to restrict membership. The Ward-Vanek effect is thus eliminated.

Conditions of entry for labor-managed firms have remained somewhat mysterious in the literature. Probably this is inherent in the logic of labor-management: before the organization of an enterprise, there is by definition no body which can take decisions in the name of its future members. This is particularly troubling, since it is known that free-entry equilibria in labor-managed market economies are fully efficient. (Dreze) In an important contribution, Martin has analyzed entry subject to finite but positive entry costs in capitalism and labor-management. He finds that the "backward-bending" supply curve of Ward and Vanek is eliminated.

This is suggestive, but it is likely that entry costs would be higher in a labor-managed market society than in a capitalist economy; that, i.e., bringing about agreement among a group of otherwise unorganized potential members is simply more costly than hiring some capital and labor is to a capitalist entrepreneur facing similarly profitable prospects. (McCain 1986a) The Guild organization, however, could provide an answer to this hypothetical difficulty. A guild, faced with the responsibility of employing such and such a number of members with such and such tools and funds to spend on new tools will presumably choose between expanding older enterprises ("sending down" employment quotas) and forming new enterprises according to which would be more productive in the interests of all of the members. Thus, initiative for entry could be expected commonly to come from the guilds themselves. There is no reason why it should not come from other functional organizations as well, and the local communal councils might be especially productive in this way. We should note that Cole explicitly accepts the organization of new enterprises outside the Guild structure, by the individuals who then comprise them (p.65-66).

The key point is this. The shortcomings of "market syndicalism" are the shortcomings of a system of local organizations isolated from one another, with no one speaking for the industry as a whole. But the industry as a whole is a functional grouping, and the first premise of Guild Socialism is that each functional grouping needs a common framework of representation and organization, with the guilds designed as such a framework for the workers in an industry. Thus on the one hand, the shortcomings of Illyria offer evidence for the functionalist position, and on the other, Guild Socialism speaks directly to those shortcomings.


A revised and updated Guild Socialism would rely on a kind of planning mechanism for the allocation of most investment and for new entry of labor in to particular industries. The plan would be based on negotiation among autonomous Guilds and other economic and functional organizations. However, it would leave to the enterprises the determination of the outputs and most of the details of production, relying on material incentives to maintain effort. Thus it would adopt the aspects of the Illyrian system where theory and empirical work indicates that the Illyrian system is effective, but relying on the Guild organization to remedy the theoretical shortcomings of Illyria.


Abad de Santillan, Diego, (1937) After the Revolution (New York:Greenberg).

Baldelli, Giovanni, (1971) Social Anarchism (Harmondsworth, England:Penguin).

Baumol, William (1986), Superfairness: Applications and Theory(Cambridge: MIT Press)

Bonin, John, (1980) "On the Theory of the Competitive Labor-Managed Firm under Price Uncertainty: A Correction," Journal of Comparative Economics, v.4, no. 3, pp. 331-37.

Cole, G.D.H, (1980) Guild Socialism Restated (New Brunswick, N.J.:Transaction Books).

Conte, Michael, (1980) "On the Economic Theory of the Labor-Managed Firm in the Short Run," Journal of Comparative Economics v. 4, no. 2, pp. 173-183.

Dahl, Robert A, (1970) After The Revolution (New Haven: Yale University Press).

Dreze, J, (1977) "Some Theory of Labor-Management and Participation," in Intriligator, M. D, Ed., Frontiers of Quantitative Economics,v. IIIA(New York: North-Holland).

Galbraith, J. K, (1952) American Capitalism (Boston: Houghton Mifflin).

Kropotkin, Peter, (1970) "Must We Occupy Ourselves with an Examination of the Ideal of a Future System?" in Selected Writings on Anarchism and Revolution (Cambridge, Mass.: MIT press) pp. 46-117.

Lewis, W. A, (1969) Principles of Economic Planning, (London: Allen and Unwin).

Martin, Robert, (1986) "Long Run Supply in Competitive Labor-Managed Industries," unpublished, forthcoming, Advances in the Economics of Participatory and Labor-Managed Firms.

McCain, Roger A, (1972) "Post-Scarcity Anarchism," by Murray Bookchin (book review) Annals of Regional Science v. 6, no. 2, pp. 160-162.

McCain, Roger A, (1973) "Recent Anarchist Theory," (review essay) the Match, v. 4, no. 7, pp. 7-8.

McCain, Roger A, (1977) "On the Optimum Financial Environment for Worker-Cooperatives," Zeitschift fur Nationalökonomie, v. 37, nos.3-4, pp.355-384.

McCain, Roger A, (1980a), "A Theory of Codetermination," Zeitschift fur Nationalökonomie, v. 40, nos.1-2, pp. 65-90.

McCain, Roger A, (1980b) "Anarchy as a Norm of Social Choice," in Economics of Public Choice, Ed. by Leitner and Sirkin (New York: Cyrco Press).

McCain, Roger A, (1985) "The Economics of the Labor-Managed Firm in the Short Run," in D. Jones and J. Svejnar, Eds., Advances in the Economics of Participatory and Labor-Managed Firms v. 1, pp. 41-53.

McCain, Roger A, (1986a) "Transaction Costs and a Theory of Public Policy,"Review of Social Economy

McCain, Roger A, (1986b) "Transaction Costs and National Economic Planning,"unpublished.

McCain, Roger A, (1986c) "Increasing "Alienation": The Work Environment and the Direction of Technical Progress under Alternative Forms of Enterprise Organization," Advances in the Economics of Participatory and Labor-Managed Firms.

Morris, William, (1912) News from Nowhere, the Collected Works of William Morris (London: Longmans, Green) v.16.

Mueller, D, R. Tollison and T. Willet, (1972) "Representative Democracy via Random Selection," Public Choice v. 12, pp. 59-68.

Ruebens, Edwin P, (1980) "Comments on 'Anarchy as a Norm of Social Choice'," inEconomics of Public Choice, Ed. by Leitner and Sirkin (New York: Cyrco Press).

Schumpeter, J. A. (1950) Capitalism, Socialism and Democracy, 3d. Ed.(New York: Harper).

Vanek, Jaroslav, (1970) The General Theory of Labor-Managed Market Economies (Ithaca: Cornell University Press).

Varian, H. R, (1974) "Equity, Envy and Efficiency," Journal of Economic Theory v. 9, pp. 63-91.

Varian, H. R, (1975) "Distributive Justice, Welfare Economics, and the Theory of Fairness," Philosophy and Public Affairs, v. 4, pp. 223-247.

Ward, Benjamin, (1967) The Socialist Economy (New York: Random House).

Wolff, R. P, (1970) In Defense of Anarchism (New York: Harper).