Michal Polák  The Real ‘Third Way’

O autorovi

Michal Polák, slovenský levicový ekonom a sociální a politický filozof. —- Vystudoval filosofii a ekonomii na London School of Economics (1995-1999, PhD 2009). V letech 2000-2001 pracoval pro slovenskou redakci BBC World Service a v letech 2006-2007 působil jako šéfredaktor slovenského levicového politicko-kulturního týdeníku Slovo. —- Polákova doktorská dizertace pojednávala o chápaní pojmu “společenská třída” v moderních společnostech.
Díla: Hlavou proti větru (2012).


Michal Polák ve svém podrobném textu o “třetí cestě” charakterizuje kapitalismus jako ekonomiku, jejíž dominantní formou je soukromé vlastnictví. Chápe ji jako koordinační mechanismus, který se týká vzájemně soupeřících hospodářských subjektů, tvořících dohromady “trh”. “Třetí cesta” v Polákově pojetí nemá nic společného s významem, který tento pojem nabyl v britské politice 90. let, ani s “tržním socialismem”, ve smyslu kompromisu mezi tržním a plánovaným hospodářstvím. Autor se pokouší vypracovat teoretický model socialistické ekonomiky, poučený z chyb tzv. reálného socialismu a inspirovaný kritickým čtením “kybersocialismu” P. Cockshotta a A. Cottrella.

This survey concerns the literature about what used to be called ‘the third way’ (before the term was in Britain appropriated by the New Labour ideologues). In other words, I will talk about models of the economy distinct from both capitalism and what is in the West called Communism.

Capitalism is here to be understood as an economy in which the dominant form of ownership is private, and the co-ordination mechanism that relates individual economic actors to each other is the market. In other words, an economy where the present allocation of resources as well as investment for the future is achieved through the interaction of buyers and sellers who have a claim on a portion of the overall resources in the economy.

What in the Western parlance is called “communism”, what used to be called “real” or “actually existing” socialism in the countries where it was in existence, and what I prefer to call simply a Soviet-type system is a centrally administered, undemocratic, planned economy. Thus, the dominant form of ownership is state, and the co-ordination mechanism is what Kornai (1986, 1992) calls “bureaucratic”. In other words, there is an administrative hierarchy through which information flows upwards and commands flow downwards; the decisions on allocation and investment are made in aggregate form by the centre and are disaggregated by the lower levels of the hierarchy. Thus, the economy functions according to the centrally determined plan.

As I have said, the “third way” I have in mind obviously has nothing to do with how the term is used nowadays in British politics. Less obviously, it is also not what is commonly called “market socialism”. Market socialism, in the most general sense of the term, can be characterised as a “half-way house” between the capitalist market and state-owned planned economy, in that state ownership continues to be the dominant (though in many cases not the sole) form of property, while the co-ordination is achieved through the buying-and-selling interactions in the market. I will say a little more about market socialism shortly. For the moment, however, I would like to emphasise that none of the models I will describe functions according to this schema, since there is no market for (at least) the producers’ goods (and moreover, in at least two cases it is a moot point whether the form of ownership can be sensibly termed “state”).


The classical case for socialism can be considered as marrying the arguments from rationality and humanism. To the extent that it envisages the society as being governed by conscious, considered decision-making mechanism, it expresses a belief in rationality over non-rational forms of behaviour. On the other hand, various traditions interpreted socialism as abolishing exploitation, doing away with unearned incomes, redistributing resources and/or guaranteeing the basic minimum of necessary goods to everyone. In that sense, the appeal of socialism was that of humanism: its ethos was to take care even - and primarily - of the most downtrodden of the human beings.

Moreover, taken together, both strands of the argument for socialism tended to reinforce each other. First of all, the rational mechanism of allocation is taken to be by definition more efficient than any alternative (in particular, the capitalist alternative). What is meant here is not necessarilu just the traditional economic notion of Pareto efficiency, where all the resources are allocated in such a way that any re-distribution making someone better off would necessarily make at least one other person worse off. Rather, the concept of efficiency is that of being economical with resources, of achieving much with little. Socialism is meant to be more efficient than capitalism in the sense that thanks to the existence of considered, ex ante allocation it can avoid unnecessary duplication of effort and waste of resources, and thus, with the same stock of factors of production, produce a greater amount of goods. From this it follows that even if capitalism was egalitarian – perhaps thanks to government redistribution of incomes – individuals would still be better off in socialism.

Secondly, the attention that Marx, if not the whole of Marxist tradition, places on the individual freedom, is precisely what allows the rational and conscious allocation mechanism for the whole of the society to operate. It is the fact that “associated producers” together determine the desired outcome of the economy’s productive activity that allows the co-ordination mechanism to be called “conscious” and “rational”. Thus, it is the emphasis on the individual freedom and autonomy that allows the rational, society-wide decision-making.

The experience of the Soviet-type systems managed to thoroughly discredit both of these notions. The efficiency of centrally planned economy in satisfying consumer wants and needs proved to be deep below that of the Western capitalism. This is true even if one believes the Western society to be way over the optimal point of consumption, since the Soviet-type economy itself generated many needs which it could not satisfy. Thus, while there is, in my opinion, much truth in the claim that a good society should not be driven by consumerism, it seems equally true that the reasons for this are that the society possesses a different internal drive, which allows it to develop along different lines. For all the talk about the “socialist morality” (or ”New Man”), such an internal drive was markedly absent in the Soviet system.

Secondly, it is an obvious understatement to say that the system totally failed to extend the realm of freedom beyond the burgeois limits. All the evils of Stalinism and later bureaucratic dictatorship are well-known and I will not list them here. What I do want to point out, however, is that the lack of meaningful democracy, of individual and social autonomy, manifested itself also within the economy. The Czech philosopher Egon Bondy long ago stated the proposition that within the modern economy, freedom itself is one of the “means of productions” – that is, it becomes a necessary condition for creative activity which is the basis of technical and economic progress. The lack of such freedom and the accompanying alienation were, in my opinion, some of the root causes of widespread apathy and lack of responsibility for property which nominally belonged to “everyone”.

But aside from this historical experience, there is also a theoretical reason why the classical socialist project is in crisis. On the practical level, nothing could of course have more impact than the 1989-1991 collapse of the system that proclaimed itself to be socialist and that was seen to be the most consistent in implementing the classical socialist vision of a marketless, publicly owned economy. The general retreat of the state from the productive side of the economy was at the same time very much in evidence in the West.

However, raw historical data need also to be accompanied by an interpretation relating them to theoretical developments. On the level of theory, two currents converged in claiming to demonstrate the impracticability and strictly speaking impossibility of a fully socialist economy. On the one hand, there was in the West the revival of the so-called Socialist Calculation Debate, the new interpretation of which seemed to buttress the anti-socialist conclusions; on the other hand, essentially the same lessons were drawn by several East European economists (János Kornai being the most prominent among them - see e. g. Kornai 1990b) the long history of reform failures within the Soviet system.

It is within these parameters that any debate on the possible “new socialisms” must proceed. I am therefore going to first of all discuss the Socialist Calculation Debate and its modern evaluation, which will lead to the consideration of market socialism (and what is wrong with it). Next, I will turn my attention to the evaluation of the alternative, non-market models, which are my main topic.


The Socialist Calculation Debate was started by an argument of von Mises (1920), who claimed that without market exchange in a private property economy, the various goods and services do not have any price tags attached to them, and in the absence of prices, it is impossible to rationally determine the most efficient way of producing things. However, this argument as it stands had by then already been disproved by Barone (1908), who demonstrated that the conditions of efficiency are the same for any economy – in effect, solving all the simultaneous equations involving both the production and utility functions describing the economy. A fortiori, (rational) socialist economy was in principle a possibility, as long as it managed to solve these equations. However, Barone also believed that the optimality conditions were way too complex to provide a practical way of guiding the economy. This was also the argument which Robbins (1934) and Hayek (1935) employed (at least as historically understood, a point to which we’ll return later).

Such was the state of play when Lange (1938) produced his celebrated market socialist model. The significance of Lange’s contribution goes well beyond the particular context of the pre-war Socialist Calculation Debate. It served not only as an inspiration for a host of other market socialist models in the 80’s and 90’s, but moreover, in an indirect way it – perhaps surprisingly – contributed also to the development of some of the non-market models that I shall describe.

Lange’s model answers the charge of the practical impossibility of solving millions of simultaneous equations by avoiding the necessity for this altogether. Instead of drawing up a detailed plan for all economic processes, the Central Planning Bureau (CPB) simply announces prices for capital goods. The managers of the enterprises, taking these prices as given, then try to maximise their profits, employing labour supplied on a free labour market and selling their products on a free consumer market. The CPB then adjusts producer goods’ prices up or down depending on whether the quantity actually used up by the firms indicates excessive demand or supply. This iterative process goes on until perfect equilibrium is reached.

In other words, the CPB behaves in practice the way that Walras’ auctioneer is behaving in theory. Instead of imaginary tâtonnement, there is a real process of trial-and-error, leading to equilibrium prices. In this way, the impracticality charge dissolves; there are no millions of equations to be solved by a socialist economy.

However, it is clear despite some claims to the contrary (see e. g. Howard and King, 1992), there is no significant role for planning within Lange’s model, either. It is true that the rate of accumulation (or investment) is set by the CPB rather than by the market; but this is seen as a defect rather than as a virtue by Lange, in that it is “arbitrary”, i. e. divorced from consumers’ preferences. Even though obviously important from the long-term point of view, the central setting of the rate of accumulation is essentially a by-product of the model rather than a result of conscious emphasis on planning. The socialist element in Lange’s model is due to the fact that most means of production are state-owned and the proceeds from their activity are democratically redistributed by the state. Hence the characterisation according to which Lange’s model “combines the allocative efficiency of idealized, neoclassical, perfectly competetive capitalism with a socialist income distribution” (Adaman and Devine, 1997).

Hayek (1940) responded to Lange’s model with arguments regarding the mutual dependence of knowledge, incentives, ownership and economic efficiency. For most of the economic profession, however, these have not diminished the strength of Lange’s model. It took another forty year for Austrian economics (of which Mises and Hayek were chief protagonists) to start gaining the upper hand again.

According to the neo-Austrians, the main challenge against the “decentralized” Lange model – and a fortiori all centralized ones – lies in the nature of knowledge. It is a point worth emphasizing that the Austrians’ epistemological conception applies equally against socialist and non-socialist neoclassical economics (Adaman and Devine 1997). For Austrians, knowledge – in particular, knowledge of both technical possibilities and of consumers’ preferences – is necessarily both subjective and local. It is local, in the sense that it is scattered among a large number of individuals, none of whom forms more than a part of the “big picture” of overall knowledge. It is subjective, because individuals often possess contradictory understanding of the same matter. Obviously, this conception of “knowledge” is very different from the philosophical notion of knowledge as “justified true belief”; even though this conception has been largely abandoned, it still serves as a useful benchmark in the sense that a possibility of contradictory knowledge is hard to swallow. The Austrian conception, however, is part of the understanding of knowledge as a process. Subjective knowledge becomes objectively validated when it is confronted with reality and succeeds. That is, if an enterpreneur knows of a gap in the market and profitably exploits it, this constitutes the validation of his knowledge; similarly for an arbitrageur, an inventor, and so on. It is this process of economic discovery, of seeking out and exploiting new possibilities, on which the modern Austrian school places a great emphasis.

The two characteristics of knowledge in the Austrian conception – its localized nature and its subjectivity – are connected, but nevertheless conceptually distinct. This point seems to be often not appreciated even by those who take the Austrian challenge seriously (e. g. Adaman and Devine 1997). In a world of perfect certainty, it would still be possible for individual knowledge to be incomplete – that is, knowledge would still be local, in that it would be coming in many pieces; nevertheless it would not be subjective. On the other hand, even if everyone’s mind could encompass all the facts in the world – thus doing away with the localized nature of knowledge – it would still not necessarily mean that any two people had to agree in every respect about what those facts were. Thus, even global knowledge (in the Austrian sense of the word) could be subjective. The connection between the two characteristics stems in my view from the fact that in this world, human beings are necessarily limited; they are neither able to obtain complete certainty nor perfect insight into all the facts. Because of these limitations, the two attributes coincide; but logically speaking neither implies the other.

It will be seen that Lange’s model can comfortably cope with the fact that knowledge is localized, but not with its subjective nature and the concomitant process of economic discovery. Since no attempt is made to provide a detailed plan for all economic activity, the issue of how the Central Planning Board can acquire the knowledge of both technical possibilities and consumers’ preferences does not arise. Since no “millions of equations” are solved, it is not necessary to obtain the knowledge of the coefficients within those equations. However, the second problem is serious. There is no provision made within Lange’s model for economic innovation, discovery of new possibilities and learning, in the Austrian sense. Hence, from the Austrian perspective, “Lange’s article was seen as a rebuttal of Barone, not Mises” (Howard and King, 1992).

Considered from this perspective, there are two varieties of modern market socialism: “neoclassical” and “Austrian” (Adaman and Devine 1997). On the one hand, there are models like Nove’s (1992, first published 1983) or Roemer’s(1994). It is the hallmark of the blueprints of this kind that, like Lange’s, whilst adequate in answering the first part of the Austrian challenge, since they employ market instead of a plan, they ignore the second issue, namely the one of discovery. In the same category fall models of labour-managed firms operating in a market economy (e. g. Vanek 1970, Horvat 1982, Šik 1985, Weisskopf 1993), which have a long history based primarily on the Yugoslav experience.

On the other hand, there are also market socialist models that address the issue of enterpreneurship and discovery explicitly. A prime example of this strand is Brus and Laski’s (1989)’market socialism proper’. The model differs from the “neoclassical” ones mentioned above in that there is no attempt by the centre to monitor, regulate, much less plan the economy – that is, there is not even ‘strategic’ or ‘indicative’ planning. Instead, major investment decisions are taken by publicly owned but independently operating firms themselves, i. e. there is a real market for capital goods (just as there are labour and consumer markets).

This last model really reveals the prime source for dissatisfaction with market socialist models. Brus and Laski note that while they themselves have in the past advocated more ‘socialised’ version of the market socialist model, practical experience shows that there is an inexorable logic leading towards their ‘market socialism proper’, with the role of the state diminished to a minimum. However, once that model is specified as above, the ‘socialist’ component of the model reduces purely to formal state ownership of the assets. That ownership, however, amounts to practically nothing, given that firms must be free to have complete operational control over these assets. The retention of state ownership is thus itself “artificial and redundant”, as Adaman and Devine (1997) note. Brus and Laski (1989)themselves conclude that “the pure logic of the fully-fledged market mechanism seems to indicate the non-state (private) enterprise as the more natural constituent of the enterprise sector”.

Similar lesson is drawn by Kornai in his (1990a). The conclusion seems inescapable. Market socialism, however modelled, is not a stable formation. There is an inner logic within it which points toward independence of the decision-making units whose goal is profit maximization. The greater that independence, the more capable they are of simulating the enterpreneurship within the capitalist economy. At the same time, however, the more like the capitalist economy the system becomes. In the end, there is no real difference between the final equilibrium and the idealized private-property market models. Market socialists themselves may well become reconciled to this and simply turn into former socialists (Kornai once again being a path-breaker in this process), pointing out to this outcome as a lesser evil over the centralized planning system. That is certainly a possible choice. As a ‘third way’ between central planning and capitalism, however, market socialism is a dead end.

Even if one wants to resist this conclusion, it is undeniable that market socialism does replicate many of the evils that plague capitalism. Thus, even its supporters do not deny that their models assume a greater or lesser degree of inequality, that there is no guarantee of full employment, and that in the final analysis, the condition of the workers does not differ all that much from that under capitalism. Perhaps the best concise critique of market socialist ideas comes not from an economist, but from the writer, Kurt Vonnegut. In his book _Jailbird,_a utopian communist millionaire leaves after her death all her accumulated property “to the people”, in effect to the US government. Vonnegut compares the economy to the weather, pointing out that just like ‘ownership’ of the weather would not prevent workers from hurricanes, droughts and hailstorms, formal ownership of the firms does not protect them from all the vagaries of the market.

This is not the place to debate the relative merits of capitalism and socialism, especially since without identifying socialism with the Soviet system it is completely mysterious what the term actually means. What I do want to claim, however, is that if anyone finds capitalism repulsive, on the grounds of justice, liberty, democracy, efficiency, ecological considerations, solidarity, or any combination thereof, then market socialism cannot be a way forward.

But if this is so, then just why did market socialism become the dominant current in the socialist economic thinking? That is, what were the faults that led to such a massive disillusion with the Soviet-type, centrally planned system. Nove’s (1991) book is very important here, for it not only sketched a market socialist economy, but also showed an in-depth, critical understanding of the Soviet centrally planned experience, and forcefully attacked the ‘idealist Marxists’ (or, in Nove’s terminology, ‘revolutionary dogmatists’) who counterposed a ‘mythical’ communist society to that experience.

It is impossible to summarise all the arguments and counterarguments Nove makes, but the principal focus of the book is unambiguously information. Nove stresses again and again the vast amount of data that the centre needs to collect in order to be able to plan efficiently, and discusses the measures that the Soviet planners took in the face of practical impossibility of this. The impossibility stems from two different sources. On the one hand, there is the question of incentives, on the other, the issue of computational ability. The information that the centre collects needs to come necessarily from lower levels, all the way down to the actual point of production. Now, the lower ranking officials are held responsible for fulfilling the plan at their level, as they must be, if there is to be any sort of check on what they are doing. However, for this very reason, they have an incentive in providing inaccurate information to those higher up, in order to make sure that the plan they will eventually have to fulfil contains a substantial amount of slack. This slack in shows in practice in the phenomena of labour- and capital hoarding, and associated shortages and bottlenecks in other parts of the economy.

Secondly, even if the centre could collect accurate information, it seems impossible that it could take it all into account. This is where Hayek’s and Robbins’ theoretical objection comes fully to the fore. The centre simply does not possess the computational capacity to deal with the millions of products in their entirety. Instead, it plans only in general, highly aggregated terms, leaving the precise composition of the aggregates to the lower levels. As a result, the plan is full of (hidden) contradictions and loose ends, which in practice means that it can never be completely fulfilled. Priority then is accorded certain sectors considered strategic (in the actual history, the military and heavy industry), while others are relegated to a less important category (historically, light industry and consumer goods). The more important sectors are the ones where additional effort is undertaken to secure the fulfilment of the target, while the less important ones suffer.

To be sure, the Austrian “informational” (as opposed to the “enterpreneurial”) critique is not completely vindicated. It is not that the centrally planned system is totally irrational and incapable of great progress. Rather, it is unable to fully account for everything that calls for attention; it can do rather well for a few well-defined priorities. It has – as has been well noted in the past, notably by Lenin – many signs of a capitalist war economy, which is also suited to achieving a few clear objectives. The rapid growth in the initial stages of the Soviet economy was not an illusion, even if the statistics were often exaggerated. Nevertheless, the informational difficulties were real, and must be noted by anyone seriously considering a possible non-market economy.


To summarise, then, the problems that any future or even imagined non-market, socialist economy must face, are as follows:

  1. Computation. The model must provide a realistic, well-specified description of a mechanism through which, without the benefit of the market, the society can decide on the basis of masses of available data (presumed given at this point) what, how, and when to produce.

  2. Collection of information. The model must specify how the data on which the information is to proceed should be collected. This involves two sub-issues:

a) incentives. Why should those who possess the localized knowledge of their particular circumstances pass it on to anyone else? What mechanism is there for making sure that they in fact do it? How is it possible to avoid introducing perverse incentives which cuase people to distort or hide the information required?

b)technology of collection. Assuming that people do in fact want to give correct information when it is required, how can they do this? In what way can they actually practically let others know what needs to be passed on? What is the technical equipment that allows this?

  1. Innovation. There needs to be a dimension of the model which allows for the evolution of the economy over time – and rational evolution at that, that is, economic progress. This obviously involves some way of determining what economic progress is. Also, there needs to be a specification of what conditions there are for producing innovations, that is, how is research and development encouraged, and on the other hand, how can innovation be implemented. This second issue involves, similarly as the problem of information gathering, both motivational and practical questions. How is it possible to ensure that there are no interests vested in the old technology, or if there are, that they can be overcome? And assuming this problem is solvable, who and how introduces the new technology that has been developed?

These are issues whose importance seems clear from the discussion so far. Nevertheless, there are also other questions which the concentration on the issue of the market has obscured, although they are no less essential for any economy that wants to call itself socialist:

  1. Democracy. What is the extent to which individuals have a say over various political and economic questions, and why should the limits be drawn where they have been drawn? It is assumed that the level of democracy achieved by the existing Western capitalist societies is the minimum that any socialist society must start from; although specific institutional arrangements may be different, the questions that can be addressed within capitalist democratic framework it must be also possible to address in a socialist one. In this context, what are the institutional arrangements that allow groups and individuals to express and act on their opinions? What is to make sure that no one individual or group gains political and/or economic power over the society? And finally

  2. Socialist Co-ordination Mechanism. Parts of this have of course been covered by all the points above. Nevertheless, the really crucial question of a socialist society needs to be posed over and above those partial issues. The question is, in what way does society govern itself? What is the mechanism through which an aggregate of individuals can together, socially take decisions? How can sectional interests be legitimately expressed and yet the society find solutions to its problems as a whole?

These are the questions that I think non-market, public ownership models need to answer. Let me now turn to discussing the three well-developed models that – to my knowledge – are today in existence. All of them are conscious attempts to address Nove’s challenge, which, in a short, sharp form, states: “it is clear that someone (some institution) hat to tell the producers about what the users require. If that ‘someone’ is not the impersonal market mechanism it can only be a hierarchical superior. There are horizontal links (market), there are vertical links (hierarchy). What other dimension is there?” (Nove 1991, p.245)


The first model I want to consider is that of Devine (1988). Devine identifies the following three main objectives for the “political economy of a self-governing society”: socialisation of production, democracy and abolition of the social division of labour. However, these are not objectives in the sense of abstract goals which admit many different ways of implementation. As will be clear from the discussion, the relation between these “objectives” and the rest of the model is not one of ends and means, but more of different ways of describing both ends and means.

With respect to the socialisation of production, Devine identifies social ownership as distinct from state ownership, on the one hand, and also from sectional ownership on the other hand. The former distinction is necessary to differentiate his concept from both the form of state ownership in the capitalist and “statist” (i. e. Soviet-type) economies; the latter, to emphasise the difference between his understanding of the term and the content which it was given for example in the official ideology of Yugoslavia, where ‘labour self-managed’ firms competing in a market were deemed to be “socially owned”. Devine also wants to contrast it with the understanding of “social ownership” as any ownership form distinct from the capitalist or individual property (“state, local authority, worker self-managed and cooperative”, Devine 1988, p.133).

Social ownership, then, is none of the above. What is it, then? Devine follows Brus (1975), according to whom there are two conditions for social ownership: “(1) the means of production must be employed in the interests of the society, and (2) society must have effective disposition over the means of production it owns.” The two conditions are “inseparable”, since without a democratic, popular participation in determining the course of development of the economy, it is impossible to ascertain what the “interests of the society” actually are. Devine is at pains to emphasise the direct, face-to-face nature of the “popular control” that is required in order to achieve real social ownership. There is, however, superficially a contradiction between the committment to decentralisation and participation on the one hand and the necessity of centrally taken decision in any society which wants to call itself “planned” – or, to put it differently, how can groups and individuals feel real control over resources without those resources thereby becoming a property of sectional rather than societal interests? Devine of course provides an answer to this seeming contradiction in the last part of his book; but it is worth underlining the existence of this issue, which must be in some way or other resolved by any non-market socialist model.

Since without voicing the concerns and interests of all individuals and groups in a society it is impossible to achieve social ownership in Devine’s sense, it is not surprising that he devotes much attention to democracy, both political and economic. Devine envisages both the continuation of the system of representative democracy and party pluralism. However, as he says, his preferred structure “displays clear continuity with, and yet would constitute a revolutionary transformation of, existing systems of representative democracy” (Devine 1988, p.142). What does this transformation consist of? Devine, following Beetham (1981), identifies two dimensions in which a “balance between centralization and decentralization” must be achieved: functional and vertical. In the first respect, the question is essentially, who runs public services? Is it one structure of authority, or is there plural responsibility for different activities? In the second respect, the question is how much responsibility resides in the centre and how much is devolved to lower ranks of the hierarchy. Devine’s preferred answer to both these questions “involves, on the one hand, separating out responsibility for different functional activities and assigning it to those affected, and, on the other hand, within each functional activity decentralizing responsibility vertically downwards to the most local level consistent with efficiency and equity” (Devine 1988, p. 141-142).

Secondly, Devine envisages a system in which there is a continuous and creative interaction between the state and the “self-governing groups”, i. e. the civil society. These groups would have to be taken into account when making decisions, and gradually would themselves take over the roles at present fulfilled by the state – “directly by involvement in running their immediate communities and local services, indirectly by making demands on the various levels of government for resources and desired policy changes” (Devine 1988, p. 145).

In the sphere of economic democracy, Devine rejects pure workers’ self-management, advocating instead a kind of two tier-structure. The overall decision-making body should consist of “all affected interests” – that is, consumers of an enterprise’s products, inhabitants of the geographical area in which it is placed, workers themselves, suppliers of inputs, etc. – while the internal structure actually implementing the decisions should be on the basis of self-management only by the workers.

The question that needs to be posed in a model attempting to put self-government at the heart of the society is how can self-government become a reality for those who do not at present possess the skills, the intellectual tools and the ability to meaningfully participate? Devine here follows Bahro (1978) with a radical suggestion that the only way this can be achieved is through the abolition of the division of labour.

Devine recognises that this aim has largely been considered impractical and/or utopian even by most Marxists. He argues against this by emphasising the division between the functional and the social division of labour. It may be necessary and efficient to separate various activities into distinct kinds of work, which corresponds to the functional division of labour. However, in Devine’s opinion, in modern society it is no longer necessary that whole groups and classes of people should spend their life-times doing only certain types of jobs. “It is a measure of the power of ideology that the inhumanity of people spending their entire working lives on [unskilled and repetitive drudgery] should for so long, and still today, be regarded as natural, normal or, however regrettable, necessary” (Devine 1988 p.172)

Devine summarizes the discussion in Bahro (1978), which justifies the aim of the abolition of social division of labour in terms of the need to do away with the differential development of consciousness. Bahro applies to social consciousness certain classical Marxist distinction, which yields the division of consciousness into “necessary” (or “absorbed”) and “surplus”. Absorbed consciousness is that which is used up in producing the basic conditions of human existence – i. e., the consciousness necessary for routine production and reproduction tasks. Surplus consciousness is what remains when these have been “done with”, a “free” consciousness, to be used “for good or ill according to the possibilities offered by the way in which society is structured” (Devine 1988 p.164). Surplus consciousness can be used either for compensatory or emancipatory activity, i. e. either in reaction to the way society constrains individual development, or for self-activation and self-growth. It is the compensatory aspect of surplus consciousness that for Bahro explains the phenomenon of consumerism, in which the satisfaction of ever more sophisticated material wants substitutes for real personal development.

Devine goes along with this analysis, adding that if particular human experience consists almost wholly of absorbed consciousness – i. e., of concentration on routine, tedious, unskilled and unrewarding tasks – they develop subaltern consciousness, an inability to think in terms of the social whole and to function at higher levels of abstraction, to fully develop as a human being. The abolition of the social division of labour is, according to Devine, needed precisely to overcome this “intellectual pauperism”.

Thus, Devine argues that productive and reproductive economic activities can be divided into five rough categories:

planning and running – accepting responsibility for a larger whole

creative – coming up with new ideas, concepts and art in various context

nurturing – taking care of other people’s needs, practical and emotional

skilled – actual “hands-on” (but not the same as manual!) work requiring particular skills

unskilled – rote, tedious, undemanding type of work

It is argued that what is both necessary and possible for the kind of society Devine envisages is that everyone is involved in all of these types of activities. Note the word “types” – it is not claimed that “everyone can do everything”. The claim is, rather, that everyone (apart from a small minority of handicapped people, perhaps) has potential to execute certain tasks within each of the above categories. This has implications for the educational system, organisation of work, gender division of labour in the workplace and at home. Devine claims that “although a long way off, the abolition of the social division of labour, the division between people who perform mental and those who perform manual labour, between men’s work and women’s work, is an entirely realistic objective.” In a society where this has been achieved, it can be expected, according to Devine, that people’s interests will diverge from the orientation on growth of material satisfaction, and thus economic growth in the conventional understanding of the term, and towards internal growth of psychological and creative needs, thus steering away from ecologically unsustainable development.

Lastly, Devine pays attention to the fact that whatever the formal structure of the society, the necessity for expertise may nevertheless generate anew the distinction between the rulers and the ruled, the subjects and the objects of the society. His proposed solution is “the abolition of the social division of labour between experts and non-experts. We cannot all be experts in everything but we can all be experts in something”. Devine’s belief is, simply put, that people tend to give respect and consideration to others on the same level of functional expertise, even if it is in an area different from their own, whereas they tend to either look up or down on those in the same area, but on a different functional level. Thus “skilled manual workers, for instance electricians and engineers, tend to be more at home with one another, irrespective of their trade, than they are with … for instance professional electrical and mechanical engineers” (Devine 1988 p.182). Thus, what is needed is divorcing the “personal dependence” (associated with the division into experts and non-experts) from “technical dependence” (which will continue to exist as long as there is distinct quality of expertise).

Such, then, are Devine’s “objectives”. The rest of his model is devoted to the discussion of those institutional underpinnings that have not as yet been fully described.

As regards the co-ordination mechanism, Devine is unequivocally opposed to the possibility of market forces constituting a foundation for a socialist society (Devine 1988, 1992, Adaman and Devine 1997). Given the nature of his objectives, this is hardly surprising. However, he is actively in favour of using market exchange in the running of a socialist society. What does this distinction mean? Devine says that market forces make themselves felt in a process

whereby change occurs in the pattern of investment, in the structure of productive capacities, in the relative size of different industries, in the geographical distribution of economic activity, in the size and evn the existence of individual production units, as a result of atomized decisions, independently taken, motivated solely by the individual decision-makers perception of their individual self-interest, not consciously co-ordinated by them in advance (Devine 1988, p.23)

In other words, market forces are that which guides major changes within a market economy – the drive for expanding the profitable and shrinking the unprofitable economic activities.

By contrast, market exchange refers to the simple act of buying and selling. There is no suggestion that this automatically has any implications for the changes in productive capacities. In fact, the simple exchange of goods in itself does not even imply profit maximisation with respect to the use of current capacity. This is precisely at the root of Devine’s model.

Devine proposes to replace market co-ordination by what he calls negotiated co-ordination. Perhaps the best way to clarify this notion is to compare it to the idea of internalising externalities, mentioned above. Negotiated co-ordination involves not only internalising externalities identified as such by neoclassical economic theory, but an attempt to institutionalise all major interdependencies that exist within the economy. That is, decisions on major investment, employment of labour, moving industries, using up fossil fuels, raising or lowering wages all affect many more people than just two parties in a market exchange. For this reason, they should not be takne simply by those two parties. All affected interests should be represented when such decisions are taken.

In Devine’s model, enterprises, whose governing boards include all such parties, engage in market exchange with one another and also the consumers. That is, they buy and sell inputs and consumer goods. They do not, however, attempt to maximize profits. They simply respond to the demand, as long as their stocks and capacity allow for this. If there is persistent excess demand or excess supply by an enterprise, this indicates that capacity needs to expand or shrink, just like in a market economy. However, unlike in the case of a market, this is not enough for the enterprise to take such actions. Quantitative information about stocks and sales must be supplemented by qualitative information from the persons affected – suppliers, consumers, workers, etc. – before changes in capacity are undertaken.

Negotiated co-ordination is thus different from both market and bureaucratic co-ordination. Unlike in the market case, market forces do not on their own shape the economy. However, neither is it bureaucratic instruction from above that determines the allocation of resources. Rather, the overall strategic priorities, as well as prices of primary inputs, including wages, are determined on a national level by a democratic assembly, taking into account the advice of a consultative “House of Interests”. The disaggregation then proceeds through ever lower levels, each of which involves negotiated co-ordination bodies. The issues that are decided on a particular level are determined by the principle that all decisions are taken on the most decentralized level consistent with taking into account all relevant interests.

As regards the question of ownership, clearly there is none in the usual sense of the word. That is, there is no “residual claimant” who also has the residual rights. The rights of control are distributed across different levels, with various interests represented on the negotiated co-ordination bodies, while there is no residual income as such, since both current production and investment are planned.

How does Devine’s model square up to the requirements I have mentioned above? First of all, there is no centre to undertake the computation process for the economy as a whole, other than in terms of broad priorities. Similarly to the Soviet system, these are disaggregated at lower levels. The difference of course is that the way the plan is implemented is not determined by “the above”, but by a negotiated interaction between higher-level and lower-level interests. As far as the collection of information is concerned, Devine does not describe any incentive/sanction mechanism for this. He assumes that “people who have been involved in the construction of the collectively determined social interest would actively want to implement it” (Devine 1988, p.203) He does claim that introducing incentives and sanctions is possible, but does not place emphasis on them, preferring to stress the “transformatory dynamic” that is involved in the process of negotiation, which is meant to broaden people’s viewpoints and lead them away from narrow self-interest. The model is thus open to an extent open to the charge that it assumes a “New Man” (or Woman) who selflessly wants to work for the good of the society, and to that extent is utopian. However, two points need to be noted. First, Devine is possibly right in claiming that the right incentives could be introduced into the model. And second, one does not need to believe in Utopias to think that material incentives are not always necessary for the required work to be performed. It is imaginable, even if by no means certain, that the process of negotiated co-ordination, of being involved in the decision making and actively participating in the shaping of a particular structure, may well create a certain sense of “extended ownership” over that structure. What I have in mind is a certain “this is mine” sentiment, expressing the sense that one is intimately concerned with the succes of the structure, even though one does not gain any great material benefit from it (other than what could be expected pretty much anyway). Such identification with projects, workplaces or communities occurs frequently even within the present-day society, and is indeed a powerful source of motivation (an example comes to mind of a librarian who deeply cares about the books she/he is reponsible for; this is an easily recognised type, at least in my own home country).

The technical collection of the information occurs simply in the negotiation process, with most of it never reaching the centre (as it is not necessary). Finally we come to the issue of innovation. Adaman and Devine (1997) claim that the model, being decentralized, provides just the right environment for innovation to occur, employing as it does the local knowledge of people involved; in fact, the system can be claimed to be more efficient than capitalism, in that the enterpreneurial abilities of everyone are employed, not just of a narrow class of capitalists. However, it is not clear how would the process of innovation and implementation work in practice. While the model as such does not seem to preclude it, the “transmission mechanism” of innovation from the first vague idea to a finished product is not specified.

On the issues of democracy and socialisation, Devine’s model does score very highly, chiefly due to its emphasis on participation, devolved and criss-crossing pattern of power, and the abolition of social division of labour. It is, however, important to consider that the practice may turn out rather different from the intentions. Especially in the area of the abolition of division of labour, it is more than likely to evoke a sceptical reaction even among many who consider themselves socialists. Nevertheless, I believe the objective is worthwhile and in the long run necessary; such an admission does not detract from the difficulties that need to be solved before the objective can be achieved.


Cockshott & Cottrell (1993) propose a model of what could be called ‘cybernetic socialism’. Despite its reliance on sophisticated technology, however, one of its basic concepts go way back even beyond Marx: namely, the idea of a perfectly moneyless economy in which accounting and distribution to consumers is based on labour time. In other words, workers are paid in labour tokens in proportion to their hours of work. (Cockshott & Cottrell allow also for differential effort, which could alter the number of labour tokens awarded per hour.) The basis for this, of course, is the Marxian labour theory of value. However, as will become clear later, Cockshott & Cottrell do not think that the labour theory of value is in itself sufficient as a basis for economic mechanism.

Accounting in terms of labour time leads the modellers straightaway to the heart of the informational problems of socialism. For if their system is to work, they must provide a way through which the labour content of the millions of products in an economy can be calculated, involving both direct and indirect labour used in their production. The fundamental basis on which Cockshott & Cottrell rely is the Leontieff input-output analysis, i. e. the matrix in which all inputs going into the production of each output are specified. They then confront head on the Nove’s (and of course Robbins’ and Hayek’s) challenge that the computation on the scale required is simply impossible. Referring to the results of the complexity theory, Cockshott & Cottrell state that solving the input-output table through the canonical method for solving simultaneous equations – the Gaussian elimination – would involve running time proportional to N 3, where N is the number of the distinct types of output. For an economy in which the number of products is of the order of a million, a supercomputer of the 1993 standard would take at least 16 000 years, which seems to vindicate the Nove-Hayek argument.

However, Cockshott & Cottrell provide an alternative solution, relying on the fact that the input-output matrix is very sparse – its coefficients are mostly zero, since most products’ inputs do not run into more than hundreds (e. g., ships are not used in the production of sealing wax which in turn is not used in the production of cabbages). Thus, one can use a method of successive approximation, in which first only direct labour expended is used in the calculation of labour values; in the next round, the indirect inputs enter as well, on the basis of their labour values as calculated in the first round, and so on. This procedure will lead to any desired degree of accuracy – though it won’t be absolutely correct. However, neither does it need to be, for as Cockshott & Cottrell state, it would (on the same assumptions as above) take only a few minutes to complete a calculation the result of which would be valid to four significant decimal digits – “which is better than the market can achieve” (Cockshott & Cottrell 1993, p. 59) In other words, once one realises that what is at stake is not complete optimality, but only an improvement on the operation of the market, the problem simplifies considerably.

If – and this is a big if - we know the technical coefficients of production (i. e. the amount of each input used in the production of an output), then a similar iterative technique (called the Jacobi method) can be also used for calculating the detailed requirements of a near-consistent plan. That is, for a given set of outputs, a set of necessary inputs can be calculated so that the discrepancy between the actual solution of the simultaneous equations and the calculated solution is smaller than a given ‘small’ number. The complexity of this technique is hardly excessively large, given the capacity of modern computers.

If there are, say, 10 million products with an average of 200 direct inputs each, and 100 iterations are needed, then 2×1011 calculations must be performed. Suppose each of these requires 10 computer instructions. In that case we have a total of 2×1012 instructions, and a computer with a speed of one billion instructions per second could do the job in 2×103 second, or a little over half an hour. (Cockshott & Cottrell 1993 p. 90)

However, Cockshott & Cottrell worry also about an additional problem, namely that a specified list of final outputs may call for the amount of capacity and human labour that simply is not in existence. For such stock-constrained plan, they provide a new algorith, based on techniques developed for neural net simulations in Artificial Intelligence. The time order of this algorithm is “roughly nm where n is the number of industries and m the mean number of inputs per industry” (Cockshott & Cottrell 1993, p.98) Again, based on the computing standards of 1993, the authors estimate that “the plan for a major economy would be computable in about 10 minutes” (Cockshott & Cottrell 1993, p.99). Moreover,

in addition to arriving at a feasible set of target outputs, the algorith would also produce as a side effect the correct allocation of capital goods and raw materials between industries. This is exactly the detailed information that one needs for a plan (Cockshott & Cottrell 1993, p.99)

It is instructive at this point to recall an article by Lange (1967). Lange here considers his earlier market socialist model as superseded by the technical developments (all quotations are from Lange 1967, page numbers refer to Nove and Nuti eds. 1972):

The market process with its cumbersome tâtonnements appears old-fashioned. Indeed, it may be considered as a computing device of the pre-electronic age. … The market mechanism and trial and error procedure in my essay really played the role of a computing device for solving a system of simultaneous equations. The solution was found by a process of iteration which was assumed to be convergent. … The market may be considered as one of the oldest historical devices for solving simultaneous equations. (p. 402)

Interestingly, Lange sees as one of the important functions of the computers the long-term planning, i. e. for decisions about investment and accumulation:

long term investments have to be taken out of the market mechanism and based on judgement of developmental economic policy. This is because present prices reflect present data, whereas investment changes data by creating new incomes, new technical conditions of production and frequently also by creating new wants (…) planning of long term economic development as a rule is based on overall considerations of economic policy rather than upon calcluations based on current prices. However, the theory and practice of mathematical (linear and non-linear) programming makes it possible to introduce strict economic accounting into this process. After setting up an objective function (…) and certain constraints, future shadow prices can be calculated. These shadow prices serve as an instrument of economic planning in long-term development plans. … Here, the electronic computer does not replacce the market. It fulfils a function which the market never was able to perform. (p. 404-5)

Lange’s optimism clearly turned out to be misplaced. This should introduce a cautionary note into overly-enthusiastic discussions of ‘cybernetic socialism’. Of course, there were political as well as material constraints in the Soviet type systems which did not allow the full utilisation of what is possible now in the West (Cockshott & Cottrell 1993, p. 90-92). However, it is important to bear in mind that unforeseen problems do appear after any major change, and therefore to not place too much of an emphasis on purely technical possibilities, which, after all, are not the same as social possibilities (for example, as Cockshott & Cottrell 1993 themselves point out – p. 49 – while the Romans knew of the technical possibilities, they never moved to widespread application of mechanical power, simply because the structure of the slave-owning society did not allow for it).

Secondly, it is interesting to note that it is investment that Lange sees as such a strong point in favour of electronic calculation. For the question it poses is partly precisely the question of enterpreneurship and discovery on which the modern Austrian school places such an emphasis, and which it claims to be the decisive failing of any possible socialism. Indeed, as socialist theorists who take the Austrian challenge seriously, Adaman and Devine (1997) believe that “while Cockshott and Cottrell’s model effectively disposes of the argument that direct calculation is practically impossible … it does so on the basis of the neoclassical epistemological assumption and nowhere addresses the issues of discovery and entrepreneurship”. Their reasons are that in order for their planning procedure to work, Cockshott & Cottrell require “lists of products being produced and regular updates on the technology used in each production process” (Cockshott & Cottrell 1993, p.126) and much other information that on Austrian view does not exist in the form ready to be tapped by the central authority.

The two issues of the local dispersal and the “subjectivity” (or potentiality) of knowledge need to be separated here. First, let us deal with the fact that the knowledge of things like technical coefficients is necessarily local. Cockshott & Cottrell effectively argue that this does not constitute a major problem, for “although it may seem an immense task to gather information about every product technique used in the economy, we should recognise that this information is already being recorded” (Cockshott & Cottrell 1993, p. 131, emphasis in the original) – namely, in the spreadsheet packages of capitalist firms, through which the purchase orders of inputs are nowadays organised. These purchase orders thus constitute a mirror-image of the company’s technology.

Moreover, in their critique of Hayek Cockshott & Cottrell (1994), the authors argue that while some local knowledge would indeed be too difficult to centralise, there is no need for this, since it can be simply used locally. “Consider the knowledge, at the level of the enterprise, of which particular workers are best at which tasks, who is the fastest worker and who the most relaible and so on (…) Why shouldn’t such knowledge just be used locally in drawing up the enterprise’s own detailed schedules for meeting an output plan given from the ‘centre’?” (Cockshott & Cottrell 1994) They also point out that the alleged simplicity of the price mechanism as the means through which local knowledge is made public is not a robust characteristic of the market system, by analysing the associated costs and comparing them to a fictional planned economy.

These arguments do go some way towards demonstrating that the collection of local knowledge is in fact possible. What is missing, however, is some way of dealing with the fact that new products, new techniques, and new forms of organisation to a large extent depend on what the Austrians term the “subjective knowledge”. Cockshott & Cottrell envisage strategic planning of the future development of the economy as something done by the experts:

It is not clear to what extent this level of planning can be democratic. The knowledge of which technologies are likely to be relevant in 10 to 20 years time will initially be concentrated in a small research community, and it is difficult for people without specialist knowledge to make judgements on the matter. It may be possible, however, for the technical specialists to draw up a number of feasible options for future industrial development, which could then be canvassed in public debate (Cockshott & Cottrell 1993, p.71)

But if the course of development is decided by people with such specialist knowledge, then even this democratic selection of options is redundant – as Lange’s (1967) description above points out, once the objective function is specified, the rational decision is a simple matter of economic calculation with the help of computers. The problem is however precisely the fact that such a procedure entirely bypasses the knowledge of the people actually ‘on the ground’ of where innovations are needed and what they could be, which new products they would like to see and what production techniques should be encouraged or progressively phased out according to how much they enrich or conflict with ‘good life’. Surely the structure of the economy ‘in 10 to 20 years’ should not be determined by the views of a small group of scientists, who are no more expert in determining the wishes of Joanna Bloggs than she is herself (in fact, rather less so). Granted, most people will have neither capability nor any great desire to concern themselves with something so abstract as the ‘structure of the economy’; they will, however, be interested in whether the specific ideas of improvements that they have are implemented or not. The function of the enterpreneur in the Austrian conception is precisely to search out such possible improvements and realise them. In a non-market socialist society, this function is likely to be performed in a different way; but, if that society is to remain democratic in a real, meaningful, ‘lived’ sense, performed it must be. Whether the other elements of Cockshott & Cottrell’s system could be combined with some sort of generalised ‘socalist enterpreneurship’, in whatever form, remains an open question.

Lange, however, was the predecessor of Cockshott & Cottrell’s model also in a different sense, the one I believe they referred to in their characterisation mentioned above. For although the model they propose is not a market one, there is a role for the market in the distribution of consumer goods. This requires some explanation.

Cockshott & Cottrell envisage consumer goods as being marked with the labour content that went into their production. However, the price – i. e. the amount of labour tokens needed for their purchase – is determined by supply and demand. In other words, a good which takes 10 hours to produce may well be selling for 12, or 8, depending on whether there is excess demand or supply. The justification Cockshott & Cottrell offer for this is that in an essentially egalitarian economy, it is unobjectionable for the good going to those who desire it most (and are willing to back up their desire with their labour tokens). And just like in Lange’s model, the information provided by the market – in this case, the ratio of the market price to the labour value of a product – is used in current decisions on expanding or lowering production. Cockshott & Cottrell believe that this mechanism provides an answer to those who point to the inability of the Soviet-type systems to satisfy consumer wants and needs and generalise it to all possible types of socialism.

Finally, let us discuss the social and political system Cockshott & Cottrell envisage as a complement of their economy. First of all it should be noted that there seems to be no intention to eliminate differences between people, other than those between owners and workers (and to some extent, also men and women – but only on the basis of socialisation of the work that today is largely gender specific, rather than on the equalisation of the share undertaken by both sexes). Thus, words such as ‘planners’, ‘engineers’ and ‘managers’ recur throughout the book, as well as ‘workers’. Thus, the contrast between ‘workers by hand’ and ‘workers by brain’ is presumably set to continue under their preferred system.

Secondly, Cockshott & Cottrell come up with the interesting idea of ‘urban communes’, whose main role is “to replace the family” (Cockshott & Cottrell 1993, p.166). This somewhat totalitarian-sounding (and certainly Utopian) statement is made more palatable by explaining that a) it is the economic functions of the family that are at stake, and b) no compulsion is envisaged – “communes will flourish only if they can demonstrate their desirability as an alternative to burgeois family life” (Cockshott & Cottrell 1993, p.166). The point of communes is to take over the functions such as housing, food preparation, childcare, some leisure activities and care of the elderly. It can be seen that the communes are seen by the authors as the main means through which the sexual division of labour is eradicated, with the commune taking over most of the work that is at present largely up to women. The commune can do this either by employing outsiders who perform the activities for it, or by rotating the tasks among its own members.

If such an idea sounds radical, Cockshott & Cottrell’s views on democracy are perhaps even more so. Generally they can be characterised as advocating the ‘classical’, Athenian type of democracy, as described by Aristotle. This has several dimensions. First of all, they reject elections as “an institution of class rule”. The reason is that, in accordance with Aristotle, they consider elections to be a principle of aristocracy, that is, selection of ‘the best’ (the aristoi). The reason according to Cockshott & Cottrell is that it is not long before ‘the best’ become synonymous with the wealthy. This is somewhat puzzling as an explanation for why elections should not be used in their envisaged socialist system, as there would not be presumably any large-scale material inequality. However, one could view their opposition to elections as an implicit recognition that privileged access to information, education and ‘cultural goods’ could constitute a basis for stratification just as much as possession of material wealth – in other words, that formal representative democracy would not prevent a formation of a class of ‘planners’ and generally under-represented ‘workers’

Cockshott & Cottrell therefore advocate a form of direct democracy, in which the major strategic decisions “(the level of taxes, the percentage of national income going towards investment, health, education, etc.)” (Cockshott & Cottrell 1993, p.186) are decided by the ‘assemblies’ of all citizens. These assemblies are of course not real, but what we would call today ‘virtual’ – people would vote electronically after seeing a TV debate by a ‘representative studio assembly’. As for the day-to-day running of the state, Cockshott & Cottrell follow Burnheim (1985)– as well, of course, as the ancient Athenians – in advocating decision-making by committees “drawn by lot from among those with a legitimate material interest in the subject under consideration” (Cockshott & Cottrell 1993, p. 184) In this system of demarchy, the functional bodies would have their own ‘juries’, selected at random from the citizens affected by their work (e. g., everyone in the case of the health service, inhabitants of a particular town in case of a local sport facilities’ centre), which would subject their proposals to a vote. Thus, “planning, for example, is not under government control but under a supervisory committee of ordinary citizens, who, since they are drawn by lot, will be predominantly working people [sic]” (Cockshott & Cottrell, 1993, p. 188) In other words, Cockshott & Cottrell envisage a kind of society where there is essentially no central government, and power resides instead with the demarchic councils drawn by lot.

Whether one finds this vision appealing or not, it seems to be in sharp contrast with the technocratic spirit of the rest of their system. The problem of economic discovery notwithstanding, it seems to me that the most pressing worry about Cockshott & Cottrell is not technical (even though their model would surely encounter difficulties just as much as any other). The problem, rather, is political. Despite their insistence on a radically different version of democracy, the logic of their position seems unequivocally centralizing. The fact is that it is a single body – namely, the state – which has a complete monopoly over both production and accumulation decision-making. Of course, as Cockshott & Cottrell point out, their state is ‘acephalous’ (without a head), devolved into various demarchic councils. But the question remains to what extent such democratic institutions can be consistent with the centralizing logic of their economy. Nove (1991) quotes Radovan Selucky as making a similar point about Marx himself: his “economic model of socialism was highly centralised, the political model highly decentralised; yet on Marx’s own showing, how could such a political superstructure rest upon this economic base?” (Nove 1991, p.128) Indeed, Nove argues that regardless of the Bolsheviks’ original intentions, the centralised structure of the Soviet economy necessarily called into being a bureaucracy that was administering it. And clearly, one does not have to accept full-blown “base and superstructure” account of the society to feel that there may well be real tension in other models with contradictory influences in the economic and the political sphere, and that one sphere may well eventually subvert the other (without presuming to know when, and which one it will be).


It is, in fact, precisely this sort of consideration that informs the vision behind the third model I want to consider, that of “Participatory Economics” put forth by Albert & Hahnel (1991a, b). These authors argue that central planning necessarily subverts self-management, which they define as “decision making input in proportion to the degree one is affected” (Albert & Hahnel 1991b, p. 11). They reject the hierarchical organisation, in which up comes information and down go commands. However, at the same time they also reject markets as the type of economic co-ordination mechanism. Their reasons for this are set out in full in their (1990), which sets out to transform the traditional welfare economics on the basis of assumptions of endogenous preferences. In particular, it is argued that individuals consciously shape their preferences in order to avoid the frustration of not having their needs satisfied. Albert & Hahnel prove several “new welfare theorems” on the basis of this assumption. They conceptualise the opportunities for self-management, empowerment and solidarity as having positive externalities, which of course means that the market provides sub-optimal amounts of them. The novel twist is that given the endogeneity of preferences, people adapt to this feature of the market, which in turn lowers the demand for the socially desirable activities, thus causing fewer resources being allocated to them, and so on – in other words, the non-optimality “snowballs” over time, and proper human development is “warped”.

Having rejected both markets and central planning as possible co-ordination mechanisms, Albert & Hahnel (1991a, b) then offer a model based on _de_centralised planning. The interesting thing about the model is that its origins, like those of others, trace back to Lange’s inspired contribution. Recall that in Lange’s (1938) model, price adjustment is worked out ex post through the operation of the market, with the Central Planning Bureau simulating the Walrasian auctioneer. However, this called for disequilibrium trading, which is absent from the (purely imaginary) process of Walrasian tâtonnement. Arrow and Hurwicz (1960) later designed a new version of the model, in which the equilibrium prices are arrived at ex ante, before production starts (see also Arrow, Hurwicz and Uzawa, 1958). It is this ex ante version that Albert & Hahnel start from.

Before the procedure can be described in detail, however, it is necessary to devote some attention to other features of the model. First of all, there is the issue of class and how to transcend it, which for Albert & Hahnel is not simply a question of property ownership. They believe that class divisions exist partly also because of differential means of access to information and decision-making; the class that monopolizes these means is termed the ‘coordinator class’. It is precisely this class which they believe was in the saddle in Soviet-type systems (which they accordingly refer to as ‘coordinatorism’ rather than socialism). Therefore, abolishing class distinctions necessarily requires, in their view, also the abolition of the division into the managers and the managed.

Although they do not refer to it in these terms, Albert & Hahnel thus fully subscribe to the Bahro/Devine aim of the “abolition of social division of labour”. However, the emphasis they place on this goal is even greater; the description of the means through which it is meant to be achieved is much more detailed, and it plays an organic role within their whole model.

The basic element of doing away with class divisions is the creation of “balanced job complexes” in place of traditional employment. These should combine empowering, skilled, intellectual, desirable tasks with the ones which are onerous, tedious, unskilled and unpleasant, so that (on average) everyone takes part in an equal share of each type. The job complexes need to be balanced not just within a workplace, but also between workplaces. Of course, as pointed out in the discussion of Devine’s ideas, this does not involve the impossible prescription of everyone doing everything: “The point is simply that people should rotate in some reasonable time period through some sequence of tasks for which they are adequately trained so that no one enjoys consistent advantages over others” (Albert & Hahnel 1991b, p. 25).

Given the existence of balanced job complexes, there is no reason for differential payment on the basis of differential disutility of work. Obviously, within a participatory economy, the payment on the basis of private ownership of the means of production is also impossible. Albert & Hahnel discuss three other possible “distributive maxims”: payment according to personal contribution, according to effort, or according to need (Albert & Hahnel 1991b, p.45). The last one they consider to be “beyond equity” – “It is not a conception of equity as much as a (fortunate) situation in which equity is no longer an issue” (Albert & Hahnel 1991b, p.45). They reject the first one on the grounds that talent or luck are not grounds for reward. Neither is training as such, unless what counts is the sacrifice involved in it – which, however, reduces to rewarding the effort that went into it rather than the skills acquired themselves. Thus, Albert & Hahnel propose for their economy rewards based on effort. This is to be judged by one’s workmates, who can best observe how much effort has been put in. They allow for variously fine systems of grading effort; their proposals thus gradually blend into those of Cockshott & Cottrell above, who, while taking labour time as their unit of account, also consider differential payment for effort.

Albert & Hahnel believe that it is only the balanced job complexes that make meaningful economic democracy possible. The democratic economicpower is exercised through the institution of workers’ councils. These councils exist on different levels, i. e. there are various councils including various groups of people. The organising principle is that the council for a particular level should involve those people who are affected by the decisions taken at that level.

This is of course quite similar to Devine’s conception of self-government, but with an important difference: interests other than workers’ are not directly involved in the running of any particular enterprise. In just a second we’ll come to the provision that is made elsewhere in the model for representing the consumers’ interests; but the concerns of, say, inhabitants of the place where an enterprise is situated, are not explicitly recognised in Albert & Hahnel’s model. Perhaps the expectation is that they are not fundamentally economic interests and as such would be taken care of within the political structure of the society, but this is not openly stated.

The consumers’ interests are represented by consumers’ councils, in a kind of mirror-image of the workplace organisation. These councils also exist in various levels, and are nested, so that, for example, a neighbourhood consumption council is part of a larger district, or town council, and so on. The role of these councils is to receive and consider consumption requests and to provide for collective consumption of public goods. This last issue is one of the simple, yet very interesting features of the model. The collective provision is decided in the council before the individual requests are considered. In other words, what the individuals spend is what is left over after all the public goods already taken care of. This has potentially huge consequences which Albert & Hahnel themselves do not emphasise. Given the importance of framing (investigated in many works by Kahneman and Tversky),it makes a great difference indeed whether public consumption is seen to be financed from taxation, i. e. what is perceived as being in one way or another being taken away from the individual, or whether on the contrary not only there is no taxation, but the private consumption is financed by the residual income, i. e. what is perceived as being given to the individual. It can be confidently expected that the contradiction which bedevils the fiscal decision-making of the existing Western societies – i. e. the demand for better public services expressed simultaneously with apparently insatiable desire for lower taxation – would never become an issue under such a system.

Coming back to the private sphere, Albert & Hahnel want their model to take care of externalities in consumption as well as in production, and hence assign to the consumption councils a role which is highly controversial. The councils receive individual consumption requests which are then up for debate by any member. The authors do postulate that the lists of goods are submitted anonymously, and that no low-to-average consumption requests can actually be rejected by a council. Nevertheless, council members are free to criticise consumption requests submitted by others, and this is in fact one of the main points of having the council.

The other point, of course, is to serve as one of the parties in the interactive planning process. The essence of the process can be put simply: it is an iterative procedure which through repeated rounds of receiving proposals and sending out information eventually balances supply and demand (with some built-in constraints designed to ensure convergence). The starting point is for every individual to submit their proposal concerning how much and how hard they are willing to work in the next planning period. Since everyone has a job complex balanced to the same level of desirability, the former proposals are implicitly suggestions for how what should be the work input for the whole economy. At the same time, consumption requests for the next planning period are also submitted, based on the value of the goods in “indicative prices” from the last accounting period. This, of course, is a potential weak spot of the model, for how am I meant to know how many ice-creams, books, pencils, scarves, cars, or earplugs I will consume next year? However, Albert & Hahnel point out that once the procedure is in place, all one needs to consider are marginal changes. To imagine that I will decide on whether I want more ice-cream (or books, or…) than last year is much easier.

After the first proposals are in, it is necessary to compare the aggregate productive activity offered and the aggregate consumption requested, with the overwhelming likelihood that they are mutually inconsistent. Given the implied supply and demand, however, it is possible to adjust the indicative prices up or down for every good. Information on this is then sent back to the individuals, who make changes in their work proposals and consumption requests accordingly. Both are subject to “peer” review by other individuals (and the same goes for councils – the demands and supplies of each are subjected to the review of all the rest), with pressure towards not consuming more than one’s fair share, i. e. the national average (but requests for being excepted may be granted). Next, the process is repeated. To ensure convergence, it may be required that in later stages the proportion of goods demanded can be adjusted only by a certain percentage, or that only higher-level councils change their proposals, with the individual suggestions now fixed. Eventually, several versions of the plan may be produced, which are then voted on by everyone.

To allow for unexpected increases in aggregate demand, Albert & Hahnel envisage “slack planning”, i. e. higher stocks of goods as well as higher productive capacity than implied by the finished plan. This allows for updating the finished plan in the light of actual development through the planning period (Albert & Hahnel 1991a, pp. 79-81)

To the above description, it must be added that aside from the indicative prices, Albert & Hahnel envisage also large amount of qualitative information to play a role in the planning process. This information, concerning condition of production of various goods, is recorded in publicly accessible databanks, which any individual may consult when making their work and consumption proposals.

How feasible is the model of participatory economics? As regards incentives for passing on the correct data, the model has one great advantage: the information that goes into the planning process is going to be truthful almost by definition. For what is being collected is not the data on technical possibilities, but on the individual wishes – how much work and consumption people want. The problem of incorrect information arises only in aggregating these wishes into a coherent whole, which is the work of the Iteration Facilitation Boards. These Boards could in theory introduce false elements into the plan, thus biasing the outcomes. However, “facilitation boards merely manipulate data and project hypothetical outcomes, which then serve as further data that producers and consumers use however they like. Facilitators do not decide anything” (Albert & Hahnel 1991a, p.92) Even the slight possibility of personal gain from such manipulation “could be trivially undone by guaranteeing that individuals don’t handle data bearing on their own economic activities” (Albert & Hahnel 1991a, p.92).

More worrisome is the actual amount of information that passes through the hands of the IFBs. Although these boards “do not decide anything”, they nevertheless collect and need to work on as much information as any central planning bureau – for they are responsible for drawing up hypothetical plans involving the description of the work activity and consumption of each and every individual in an economy. If the task of a central planner was impossible, the task of the facilitation boards is equally (or more) so.

On the other hand, this does not take into account the advances in computer technology, so ably exploited by Cockshott & Cottrell. If it is possible to devise a computer package with standardised entries for characteristics of work activities and for types of consumer goods, then the work of the IFBs is obviously greatly simplified. But this points towards the conclusion that participatory economics is a model only suited to at least medium-developed countries with access to modern information technology. Albert & Hahnel nevertheless intend their model even for Third World countries, such as Cuba (Albert & Hahnel 1991a, p.39). It is a little unclear how they imagine the necessary computation would be done under such conditions, with presumably few technical resources.

Adaman and Devine (1997) charge that Albert & Hahnel’s model does not deal with investment, and thus cannot provide an answer to the Austrian challenge of the lack of enterpreneurship in socialism. It is true that the authors do not discuss investment in much detail, and there is certainly no detailed description of the process of technical innovation. However, it does not seem impossible for there to be specialised workplaces dedicated to R&D, much as they exist now – of course, with the difference of their workers also being involved in less conceptual and more tedious activities, as part of their balanced job complexes.

But this, of course, does not entirely do away with the point about local and subjective nature of knowledge. Like in the case of Cockshott & Cottrell, it is the discovery rather than the localised character of knowledge that is at stake. Local information is supplied willingly within the model of participatory economics, simply because it is expected to benefit the supplier themselves (through providing insight into their working conditions and thus an opportunity for sympathy and solidarity of others). The question is more whether the ideas of individuals can be mobilised for the social good.

One way in which this could be done is of course through the various workers’ councils running the enterprises. The slight difficulty is that any technical change that is introduced to improve the quality of work, efficiency etc. will immediately change the balance of tasks in the concerned workers’ job complexes. In order for the whole of society being on the same standard, these complexes would have to be re-balanced, perhaps to include more tasks elsewhere, thus taking the strain off others. In general, any such technical change will have effects on everyone, since the smaller incidence of unpleasant/boring/unskilled tasks will make it possible for everyone to do less of them. However, at the same time this means that the technical change in any particular workplace will not really have any great effect on that particular workplace. Thus, there is very little incentive for workers to try to realize the innovations that occur to them in the course of their employment. Perhaps a way of overcoming this would be to re-balance the job-complexes only in discrete intervals. In other words, workers would enjoy the fruits of their technical innovations until the society-wide re-evaluation of the job complexes, by which time it could be reasonably hoped that many innovations have also been introduced elsewhere, thus making a general improvement possible.


The three models examined provide, I think, ample reason to believe that an investigation into possible non-market system is worthwhile. I am not suggesting that they are without problems; and indeed, I have pointed out some of these problems myself. Doubtless there are more. Nevertheless, I find them a fertile foundation on the basis of which the next round of designing non-market systems could proceed.


Albert, Michael, and Hahnel, Robin (1990): Quiet Revolution in Welfare Economics, Princeton University Press, Princeton

– (1991a): Looking Forward: Participatory Economics for the Twenty First Century, South End Press, Boston

– (1991b): The Political Economy of Participatory Economics, Princeton University Press, Princeton

Adaman, Fikret, and Devine, Pat (1996): ”The Economic Calculation Debate: Lessons for Socialists”, Cambridge Journal of Economics, 20, pp. 523-537, repr. in Boettke (2000)

– (1997): “On the Economic Theory of Socialism”, New Left Review, 221, January/February, pp. 54-80

Arrow, Kenneth; Hurwicz, Leonid, and Uzawa, Hirofumi: (1958): Studies in Linear and Nonlinear Programming, Stanford University Press, Stanford

Arrow, Kenneth, and Hurwicz, Leonid (1960): “Decentralization and Computation in Resource Allocation”, in: Arrow, Kenneth, and Hurwicz, Leonid eds., Studies in Resource Allocation Processes, Cambridge University Press, Cambridge 1977

Bardhan, Pranab K., and Roemer, John E. eds. (1993): Market Socialism, Oxford Univesity Press, New York

Barone, Enrico (1908):”The Ministry of Production in the Collectivist State”, repr. in Nove and Nuti (1972)

Boettke, Peter J. ed. (2000): Socialism and the Market: The Socialist Calculation Debate Revisited, vol. I - IX, Routledge, London

Brus, Wlodzimierz, and Laski, Kazimirz (1989): From Marx to Market, Clarendon Press, Oxford

Burnheim, James (1985): Is Democracy Possible?, Polity Press, Oxford

Cockshott, W. Paul, and Cottrell, Allin (1993): Towards a New Socialism, Spokesman, Nottigham

– (1994): ”Information and Economics: A Critique of Hayek”, http://th.ac.uk/~wpc/reports/hayek/hayek.html

Devine, Pat (1988): Democracy and Economic Planning, Polity Press, Cambridge

– (1992): “Market Socialism or Participatory Planning?”, Review of Radical Political Economics, Vol. 24, no. 3 & 4, Fall & Winter 1992

Hayek, Friedrich (1935): “The Nature and History of the Problem”; “The Present State of the Debate”, in Hayek, ed. (1935)

Hayek, Friedrich ed. (1935): Collectivist Economic Planning, repr. as vol. II of Boettke (2000)

– (1940): ”Socialist Calculation: The Competetive ‘Solution’”, Economica, Vol. 7, No. 26, May 1940, pp. 125-149, repr. in Boettke (2000) vol. V

Horvat, Branko (1982): The Political Economy of Socialism, Martin Robertson, Oxford

Howard, M. C., and King, J. E. (1992): A History of Marxian Eocnomics: Volume II, 1929-1990, Macmillan, London

Kornai, János (1986): “The Hungarian Reform Process: Visions, Hopes, and Reality”, Journal of Economic literature

– (1990a): “The Affinity Between Ownership Forms and Coordination Mechanisms: The Common Experience of Reform in Socialist Countries”, Journal of Economic Perspectives

– (1990b): The Road to a Free economy: Shiftng from a Socialist System; the Example of Hungary, W. W. Norton & Company, New York

– (1992): The Socialist System: The Political Economy of Communism, Oxford University Press, Oxford

Lange, Oscar (1938): “On the Economic Theory of Socialism”, in: Lippincot ed. (1938)

– (1967): ”The Computer and the Market”, repr in Nove and Nuti (1972)

Lippincott ed. (1938): On the Economic Theory of Socialism, The University of Minnesota Press, Minneapolis

Mises, Ludwig (1920): “Economic Calculation in the Socialist Commonwealth”, repr. in Nove and Nuti (1972)

– (1949): Human Action: A Treatise on Economics, Yale University Press, New Haven

Nove, Alec, and Nuti, D. M. (1972): Socialist Economics, Penguin Education, Harmondsworth

Nove, Alec (1992): The Economics of Feasible Socialism Revisited, Second Edition, HarperCollins Academic, London

Robbins, Lionel C. (1934): The Great Depression, Macmillan, London

Roemer, John E. (1994): A Future for Socialism, Verso, London

Šik, Ota (1985): For a Humane Economic Democracy, Praeger, New York

Taylor, Fred M. (1938): “The Guidance of Production in a Socialist State”, in: Lippincott ed. (1938)

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

Weisskopf, Thomas (1993): “A Democratic Enterprise-Based Market Socialism”, in Bardhan and Roemer eds. (1993)