Iam grateful to Ernest Mandel for his thoughtful criticism of my ideas concerning ‘market socialism’ (‘In Defence of Socialist Planning’, nlr 159). By a coincidence, I received on the same day a copy of an attack on these ideas from the New Right: Crozier and Selden’s Socialism, the Grand Illusion. The authors are just as unhappy as is Mandel with any mixture of plan and market, but of course from an opposite standpoint. I mention this as a way of emphasizing that I am not a believer in laissez-faire, and am well aware of the market’s imperfections and limitations. A minimal role for the state, the untrammelled pursuit of private profit, does not ensure the welfare of society, and indeed it is necessary for these ideologists to distort the ideas of the real Adam Smith while invoking his name.

Mandel does not deny that ‘commodity-exchange may be necessary in the immediate aftermath of an anti-capitalist revolution’, during which plan and market may coexist ‘in precarious and hybrid transitional forms’. So presumably for him, as for Marx, transition to socialism involves gradually dispensing with market. This is where we disagree. Part of the disagreement may arise from what I believe to be a definitional confusion on his part. He is, of course, quite right when he notes that under ‘late capitalism’ there are giant corporations, with various degrees of vertical integration, within which hierarchical ‘direct allocation’ replaces the market. I devoted most of pages 198-203 of my book, The Economics of Feasible Socialism, to the significance of this, and also to the fact that alongside them there are many thousands of medium and small firms. We must presume that economies (and diseconomies) of scale—technological, informational, organizational—vary very widely, and would probably also vary widely in a realistically envisageable socialism—which is why it seems right to envisage several categories of producers. Where Mandel goes wrong is in drawing the line of demarcation between plan and market, between ex ante and ex post. Of course many goods are made to prior order! Surely the line between plan and market does not run between bespoke and ready-made tailoring! Mandel says that ‘it is not the market but the planned target for truck output that determines the number of (truck) bodies to be manufactured’ (p. 6). But any textbook will tell us that the demand for bodies (and other components) is derived from the market demand for trucks! Obviously, advance planning of some kind, i.e. ex ante anticipation, is the rule in capitalist market economies, whether this is based on market surveys or prior negotiated contracts. In a socialist economy too one imagines that ships and large generating equipment would be produced to special order, while shoes, skirts and cabbages would be produced or grown in anticipation of what the customer may require, an anticipation which may prove mistaken and which requires ex post verification. Surely the market, commodity production, exists when goods are made for sale, for exchange and not for use, and this remains so whatever the degree of vertical integration in the production process of this or that good.

Mandel asks: is it appropriate to use evidence culled from Soviet experience? Yes, there were specifically Russian or Soviet factors—backwardness, ‘bureaucratic misrule’. But there are lessons to be learned, concerning (for instance) scale, complexity, conflicts between partial and general interest, plan-fulfilment indicators, investment criteria, prices in theory and practice, labour incentives, diseconomies of scale in agriculture, the influence of user needs on plans and on output, the role of regional policy, and so on. While the Soviet record in handling these and other issues (including environmental pollution) may leave much to be desired, it would be foolish to ignore Soviet experience because of a prior decision to classify it as ‘not socialist’.

Thus if in the ussr today there are several million types and varieties of goods and services, produced and provided by hundreds of thousands of enterprises (industrial, construction, agricultural, transport, distribution, etc., etc.), and if the sheer complexity of marketless planning generates both bureaucracy and inefficiency, it is really not very relevant to advocate ‘democracy’ as a cure. The right of different strata to organize into pressure groups, however desirable in itself, can only make the task of planning yet more complicated. Mandel tells us that ordinary mortals do not really choose between millions of goods and services, that most people’s requirements are repetitive and largely predictable. Yes indeed, total unpredictability would make life impossible in any system! But one must ask: if there are millions of products in the ussr (as also in the West), why is this so? The point is that, while Mandel and I do not consciously consider thousands of kinds of footwear and thousands of holiday resorts, these do exist in thousands, for choices other than our own. As economies grow beyond simple subsistence levels, people ‘take pleasure in more diversified meals’ (I am here quoting Mandel’s own words), and also shoes, holidays, etc. The greater the variety of outputs, the greater also the variety of inputs. The greater also the burdens on the central planners. Mandel asks: why the central planners? Why my stress on scale in general? Do we not know that the whole consists of many parts, to which decisions can be devolved? Here I think that Mandel and those who think like him suffer from a blind spot. Let me explain of what it consists.

First, there is the centralizing logic, in a modern interconnected industrial economy, of planning on the basis of conscious assessment of need by the ‘associated producers’. The resultant production and allocation decisions must reflect the priorities decided (by whatever means) by ‘society’ or its representatives. Decisions once made must be implemented, which involves the commitment of resources produced in many parts of the country or outside it. Unless ‘abundance’ is assumed, in the sense that there is enough for everyone and so no problem of mutually exclusive choices, some body (somebody) must allocate resources between alternative uses. Yes, the market does this too, and does it imperfectly. But the existence of innumerable freely-negotiated horizontal contractual links removes an otherwise impossible burden from the centre, and it is therefore not surprising that such is the recommendation of those Soviet reformers closest to Gorbachev.

Mandel criticizes me for seeing only administered allocation and market (purchase-and-sale) as alternatives. He insists: tertium datur. He also stresses that Man is not only a consumer but also a producer, with freedom to make choices in both capacities. Yes, of course, but let us think through the implications. Some activities even today are of their nature not decentralizable. Take the electricity grid and the rail network as two examples. ‘Self-management’ at local level (one power station, one segment of line or train) must for obvious reasons be strictly limited: if peak demand for power or for passengers is to be discerned and acted upon, this must be at a higher level in the hierarchy, whether today or in an imagined socialist society. This is indeed why I classify such sectors as necessarily centrally planned. However, I do advocate that the bulk of the economy be decentralized, so that people be as free as possible both as consumers and as producers. It is astonishing to read (on page 17) that, in Mandel’s view, ‘the simplest and most democratic way of adapting material resources to social wants is not to interpose the medium of money between the two, but to find out people’s needs just by asking them what they are.’ Yet the snags are obvious enough. How by this route can one discover the relative intensity of people’s wants, which is signalled (however imperfectly) by willingness to pay? How is one to deal with the predictable outcome—that total desires exceed the means of satisfying them? How about quality: who will opt for the cheaper cuts of meat, for instance? Finally, what remedy is available to those citizens who find that what they ask for is not supplied? And what powers are available to the planners to ensure that what they have decided is in fact implemented? It is no answer to say that some urgent needs are clear priorities: food for the hungry and medication for the sick should take precedence over luxuries. Agreed, but we are somewhat beyond that sort of simple world; to recite Mandel, we ‘take pleasure in more diversified meals’.

But I have strayed from my basic point. Let us look closely at the logic of decentralized decision-making, whether the decision is that of a consumer, a producer, a would-be innovator, a commune, or for that matter a local planning office. They all have one thing, one requirement, in common: the need for a number of material inputs. Some of these can indeed be centrally provided in fairly predictable quantities: water, for instance, or electricity. Others, many thousand others, need to be made and delivered by numerous other enterprises. These, in turn, can only provide them if they can obtain the means of production required for the purpose, and if they have sufficient degree of control over what goods and services they will be producing so that they can supply what the customer actually requires. In the present context we can set aside the question, important in its own right, about the way decisions are actually arrived at, the degree of self-management or participation within the production unit. The essential point is that it is a decentralized decision, which calls for material inputs if it is to be implemented. How, in Mandel’s world, are they to be obtained? By application to the ‘associated producers’ or some planning office? Bearing in mind that each decision involves several different inputs, provided by a number of different enterprises each of which in turn requires different inputs, can he really not see the overwhelming complexities involved? The use of computers can speed up calculations and help achieve material balance. But it will be human beings, not computers, who administer priorities and judge whether some new proposal or innovation ought to receive material support.