Robert Brenner's critique of the market-focused theories of development which overlook class structures and relations, opening the door for third-worldist ideology.
The appearance of systematic barriers to economic advance in the course of capitalist expansion—the ‘development of underdevelopment’—has posed difficult problems for Marxist theory. There has arisen, in response, a strong tendency sharply to revise Marx’s conceptions regarding economic development. In part, this has been a healthy reaction to the Marx of the Manifesto, who envisioned a more or less direct and inevitable process of capitalist expansion: undermining old modes of production, replacing them with capitalist social productive relations and, on this basis, setting off a process of capital accumulation and economic development more or less following the pattern of the original homelands of capitalism. In the famous phrases of the Communist Manifesto: ‘The bourgeoisie cannot exist without constantly revolutionizing the instruments of production and thereby the relations of production, and with them the whole relations of society. Conservation of the old modes of production in an altered form was, on the contrary, the first condition of existence for all earlier industrial classes. Constant revolutionizing of production, uninterrupted disturbance of all social conditions, everlasting uncertainty, and agitation distinguish the bourgeois epoch from all earlier ones. The bourgeoisie . . . draw all, even the most barbarian nations into civilization. The cheap prices of its commodities are the heavy artillery with which it batters down all Chinese walls . . . It compels all nations, on pain of extinction, to adopt the bourgeois mode of production; it compels them to introduce what it calls civilization into their midst, to become bourgeois themselves. In a word, it creates a world after its own image.’
Many writers have quite properly pointed out that historical developments since the mid-nineteenth century have tended to belie this ‘optimistic’, ‘progressist’ prognosis, in that the capitalist penetration of the ‘third world’ through trade and capital investment not only has failed to carry with it capitalist economic development, but has erected positive barriers to such development. Yet the question remains, where did Marx err? What was the theoretical basis for his incorrect expectations? As can be seen from the above quotation and many others from the same period,  Marx was at first quite confident that capitalist economic expansion, through trade and investment, would inevitably bring with it the transformation of pre-capitalist social-productive relations—i.e. class relations—and the establishment of capitalist social-productive relations, a capitalist class structure. It was clearly on the premise that capitalist expansion would lead to the establishment of capitalist social relations of production on the ruins of the old modes, that he could predict world-wide economic development in a capitalist image.
But, suppose capitalist expansion through trade and investment failed to break the old modes of production (a possibility which Marx later envisaged ); or actually tended to strengthen the old modes, or to erect other non-capitalist systems of social relations of production in place of the old modes? In this case, Marx’s prediction would fall to the ground. For whatever Marx thought about the origins of capitalist social-productive relations, he was quite clear that their establishment was indispensable for the development of the productive forces, i.e. for capitalist economic development. If expansion through trade and investment did not bring with it the transition to capitalist social-productive relations—manifested in the full emergence of labour power as a commodity—there could be no capital accumulation on an extended scale. In consequence, the analysis of capitalist economic development requires an understanding, in the first place, of the manner in which the capitalist social-productive relations underpinning the accumulation of capital on an extended scale originated. In turn, it demands a comprehension of the way in which the various processes of capitalist expansion set off by the accumulation of capital brought about, or were accompanied by, alternatively: 1. the further erection of capitalist class relations; 2. merely the interconnection of capitalist with pre-capitalist forms, and indeed the strengthening of the latter; or 3. the transformation of pre-capitalist class relations, but without their substitution by fully capitalist social-productive relations of free wage labour, in which labour power is a commodity. In every case, it is class relations which clearly become pivotal: the question of their transformation in relationship to economic development.
I shall argue here that the method of an entire line of writers in the Marxist tradition has led them to displace class relations from the centre of their analyses of economic development and underdevelopment. It has been their intention to negate the optimistic model of economic advance derived from Adam Smith, whereby the development of trade and the division of labour unfailingly bring about economic development. Because they have failed, however, to discard the underlying individualistic-mechanist presuppositions of this model, they have ended up by erecting an alternative theory of capitalist development which is, in its central aspects, the mirror image of the ‘progressist’ thesis they wish to surpass. Thus, very much like those they criticize, they conceive of (changing) class relations as emerging more or less directly from the (changing) requirements for the generation of surplus and development of production, under the pressures and opportunities engendered by a growing world market. Only, whereas their opponents tend to see such market-determined processes as setting off, automatically, a dynamic of economic development, they see them as enforcing the rise of economic backwardness. As a result, they fail to take into account either the way in which class structures, once established, will in fact determine the course of economic development or underdevelopment over an entire epoch, or the way in which these class structures themselves emerge: as the outcome of class struggles whose results are incomprehensible in terms merely of market forces. In consequence, they move too quickly from the proposition that capitalism is bound up with, and supportive of, continuing underdevelopment in large parts of the world, to the conclusion not only that the rise of underdevelopment is inherent in the extension of the world division of labour through capitalist expansion, but also that the ‘development of underdevelopment’ is an indispensable condition for capitalist development itself.
Frank and Capitalist Development
It has thus been maintained that the very same mechanisms which set off underdevelopment in the ‘periphery’ are prerequisite to capital accumulation in the ‘core’. Capitalist development cannot take place in the core unless underdevelopment is developed in the periphery, because the very mechanisms which determine underdevelopment are required for capitalist accumulation. In the words of André Gunder Frank, ‘economic development and underdevelopment are the opposite faces of the same coin’. As Frank goes on to explain: ‘Both [development and underdevelopment] are the necessary result and contemporary manifestation of internal contradictions in the world capitalist system . . . economic development and underdevelopment are relational and qualitative, in that each is actually different from, yet caused by its relations with, the other. Yet development and underdevelopment are the same in that they are the product of a single, but dialectically contradictory, economic structure and process of capitalism. Thus they cannot be viewed as the product of supposedly different economic structures or systems . . . One and the same historical process of the expansion and development of capitalism throughout the world has simultaneously generated—and continues to generate—both economic development and structural underdevelopment.’  Specifically: ‘The metropolis expropriates economic surplus from its satellites and appropriates it for its own economic development. The satellites remain underdeveloped for lack of access to their own surplus and as a consequence of the same polarization and exploitative contradictions which the metropolis introduces and maintains in the satellite’s domestic structure.’ 
Obviously such a view of underdevelopment carries with it a view of development, the unitary process which ostensibly brought about both. Frank’s primary focus has in fact been on the roots of underdevelopment, so it has not been essential for him to go into great detail concerning the origins and structure of capitalist development itself. Yet, to clarify his approach, it was necessary to lay out the mainsprings of capitalist development, as well as underdevelopment; accordingly, Frank did not neglect to do this, at least in broad outline. The roots of capitalist evolution, he said, were to be found in the rise of a world ‘commercial network’, developing into a ‘mercantile capitalist system’. Thus ‘a commercial network spread out from Italian cities such as Venice and later Iberian and Northwestern European towns to incorporate the Mediterranean world and sub-Saharan Africa and the adjacent Atlantic Islands in the fifteenth century . . . until the entire face of the globe had been incorporated into a single organic mercantilist or mercantile capitalist, and later also industrial and financial, system, whose metropolitan centre developed in Western Europe and then in North America and whose peripheral satellites underdeveloped on all the remaining continents.’  With the rise of this system, there was ‘created a whole series of metropolis-satellite relationships, interlinked as in the surplus appropriation chain noted above’. As the ‘core’ end of the chain developed, the ‘peripheral’ end simultaneously underdeveloped.
Frank did not go much further than this in filling out his view of capitalism as a whole, its origins and development. But he was unambiguous in locating the dynamic of capitalist expansion in the rise of a world commercial network, while specifying the roots of both growth and backwardness in the ‘surplus appropriation chain’ which emerged in the expansionary process:  surplus appropriation by the core from the periphery, and the organization of the satellite’s internal mode of production to serve the needs of the metropolis. In this way, Frank set the stage for ceasing to locate the dynamic of capitalist development in a self-expanding process of capital accumulation by way of innovation in the core itself. Thus, for Frank, the accumulation of capital in the core depends, on the one hand, upon a process of original surplus creation in the periphery and surplus transfer to the core and, on the other hand, upon the imposition of a raw-material-producing, export-dependent economy upon the periphery to fit the productive and consumptive requirements of the core.
It has been left for Immanuel Wallerstein to carry to its logical conclusion the system outlined by Frank. Just as Frank and others have sought to find the sources of underdevelopment in the periphery in its relationship with the core, Wallerstein has sought to discover the roots of development in the core in its relationship with the periphery. Indeed, in his magisterial work, The Origins of the Modern World System,  Wallerstein attempts nothing less than to establish the origins of capitalist development and underdevelopment and to locate the mainsprings of their subsequent evolutions.
Wallerstein aims to systematize the elements of the preliminary sketch put forward in Frank’s work. His focus is on what he terms the ‘world economy’, defined negatively by contrast with the preceding universal ‘world empires’. So the world empires, which ended up by dominating all economies prior to the modern one, prevented economic development through the effects of their overarching bureaucracies, which absorbed masses of economic surplus and prevented its accumulation in the form of productive investments. In this context, Wallerstein declares that the essential condition for modern economic development was the collapse of world empire, and the prevention of the emergence of any new one from the sixteenth century until the present. Wallerstein can argue in this way because of what he sees to be the immanent developmental dynamic of unfettered world trade. Left to develop on its own, that is without the suffocating impact of the world empires, developing commerce will bring with it an ever more efficient organization of production through ever increasing regional specialization—in particular, through allowing for a more effective distribution by region of what Wallerstein terms systems of ‘labour control’ in relation to the world’s regional distribution of natural resources and population. The trade-induced world division of labour will, in turn, give rise to an international structure of unequally powerful nation states: a structure which, through maintaining and consolidating the world division of labour, determines an accelerated process of accumulation in certain regions (the core), while enforcing a cycle of backwardness in others (the periphery). 
Without, for the moment, further attempting to clarify Wallerstein’s argument, it can be clearly seen that his master conceptions of world economy and world empire were developed to distinguish the modern economy, which can and does experience systematic economic development, from the pre-capitalist economies (called world empires), which were capable only of redistributing a relatively inflexible product, because they could expand production only within definite limits. Such a distinction is both correct and necessary. For capitalism differs from all pre-capitalist modes of production in its systematic tendency to unprecedented, though neither continuous nor unlimited, economic development—in particular through the expansion of what might be called (after Marx’s terminology) relative as opposed to absolute surplus labour. That is, under capitalism, surplus is systematically achieved for the first time through increases of labour productivity, leading to the cheapening of goods and a greater total output from a given labour force (with a given working day, intensity of labour and real wage). This makes it possible for the capitalist class to increase its surplus, without necessarily having to resort to methods of increasing absolute surplus labour which dominated pre-capitalist modes—i.e. the extension of the working day, the intensification of work, and the decrease in the standard of living of the labour force. 
To be specific, a society can achieve increases in labour productivity leading to increases in relative surplus product/labour when it can produce a greater mass of use values with the same amount of labour as previously. Put another way, a given labour force achieves an increase in labour productivity when it can produce the means of production and means of subsistence which makes possible its own reproduction (continued existence) in less time than previously (working at the same intensity); or when, given the same amount of time worked as before, it produces a larger surplus above the means of production and means of subsistence necessary to reproduce itself than previously. This cannot take place without qualitative changes, innovations in the forces of production, which have historically required the accumulation of surplus, i.e. ‘plough back of surplus’, into production. The basis, in turn, for the operation of this mechanism as a more or less regular means to bring about economic development was a system of production organized on the basis of capitalist social-productive or class relations. As Marx put it, relative surplus value ‘presupposes that the working day is already divided into two parts, necessary labour and surplus labour. In order to prolong the surplus labour, the necessary labour is shortened by methods for producing the equivalent of the wage of labour in a shorter time. The production of absolute surplus-value turns exclusively on the length of the working day, whereas the production of relative surplus-value completely revolutionizes the technical processes of labour and the groupings into which society is divided. It therefore requires a specifically capitalist mode of production, a mode of production which, along with its methods, means and conditions, arises and develops spontaneously on the basis of the formal subsumption of labour under capital. This formal subsumption is then replaced by a real subsumption.’ (emphasis added). 
A Crucial Objection
It is the fundamental difficulty in Wallerstein’s argument that he can neither confront nor explain the fact of a systematic development of relative surplus labour based on growth of the productivity of labour as a regular and dominant feature of capitalism. In essence, his view of economic development is quantitative, revolving around: 1. the growth in size of the system itself through expansion; 2. the rearrangement of the factors of production through regional specialization to achieve greater efficiency; 3. the transfer of surplus. Thus, according to Wallerstein, the collapse of world empire made possible a worldwide system of trade and division of labour. This, in turn, determined that what for Wallerstein were the three fundamental conditions for the development of the world economy would be fulfilled: ‘an expansion of the geographical size of the world in question [incorporation], the development of variegated methods of labor control for different products and different zones of the world economy [specialization] and the creation of relatively strong state machinery in what would become the core states of this capitalist world economy [to assure transfer of surplus to the core]. (mws. p. 38.) However, as we shall show, neither the expansion of trade leading to the incorporation of greater human and natural material resources, nor the transfer of surplus leading to the build-up of wealth in the core, nor the specialization of labour control systems leading to more effective ruling-class surplus extraction can determine a process of economic development. This is because these cannot determine the rise of a system which ‘develops itself spontaneously’; which can and must continually ‘revolutionize out and out the technical processes of labour and composition of society’.
Wallerstein does not, in the last analysis, take into account the development of the forces of production through a process of accumulation by means of innovation (‘accumulation of capital on an extended scale’), in part because to do so would undermine his notion of the essential role of the underdevelopment of the periphery in contributing to the development of the core, through surplus transfer to underwrite accumulation there. More directly, Wallerstein cannot—and in fact does not—account for the systematic production of relative surplus product, because he mislocates the mechanism behind accumulation via innovation in ‘production for profit on the market’: ‘The essential feature of a capitalist world economy . . . is production for sale in a market in which the object is to realise the maximum profit. In such a system, production is constantly expanded as long as further production is profitable, and men constantly innovate new ways of producing things that expand their profit margin.’ (rfd, p. 398.)
Now, there is no doubt that capitalism is a system in which production for a profit via exchange predominates. But does the opposite hold true? Does the appearance of widespread production ‘for profit in the market’ signal the existence of capitalism, and more particularly a system in which, as a characteristic feature, ‘production is constantly expanded and men constantly innovate new ways of producing’. Certainly not, because production for exchange is perfectly compatible with a system in which it is either unnecessary or impossible, or both, to reinvest in expanded, improved production in order to ‘profit’. Indeed, we shall argue that this is the norm in pre-capitalist societies. For in such societies the social relations of production in large part confine the realization of surplus labour to the methods of extending absolute labour. The increase of relative surplus labour cannot become a systematic feature of such modes of production.
To state the case schematically: ‘production for profit via exchange’ will have the systematic effect of accumulation and the development of the productive forces only when it expresses certain specific social relations of production, namely a system of free wage labour, where labour power is a commodity. Only where labour has been separated from possession of the means of production, and where labourers have been emancipated from any direct relation of domination (such as slavery or serfdom), are both capital and labour power ‘free’ to make possible their combination at the highest possible level of technology. Only where they are free, will such combination appear feasible and desirable. Only where they are free, will such combination be necessitated. Only under conditions of free wage labour will the individual producing units (combining labour power and the means of production) be forced to sell in order to buy, to buy in order to survive and reproduce, and ultimately to expand and innovate in order to maintain this position in relationship to other competing productive units. Only under such a system, where both capital and labour power are thus commodities—and which was therefore called by Marx ‘generalized commodity production’—is there the necessity of producing at the ‘socially necessary’ labour time in order to survive, and to surpass this level of productivity to ensure continued survival.
What therefore accounts for capitalist economic development is that the class (property/surplus extraction) structure of the economy as a whole determines that the reproduction carried out by its component ‘units’ is dependent upon their ability to increase their production (accumulate) and thereby develop their forces of production, in order to increase the productivity of labour and so cheapen their commodities. In contrast, pre-capitalist economies, even those in which trade is widespread, can develop only within definite limits, because the class structure of the economy as a whole determines that their component units—specifically those producing the means of subsistence and means of production, i.e. means of survival and reproduction, rather than luxuries—neither can nor must systematically increase the forces of production, the productivity of labour, in order to reproduce themselves.
If, then, the class-structured system of reproduction in which labour power is a commodity lies behind capitalist economic development, while ‘production for profit in the market’ cannot in itself determine the development of the productive forces, it follows that the historical problem of the origins of capitalist economic development in relation to pre-capitalist modes of production becomes that of the origin of the property/surplus extraction system (class system) of free wage labour—the historical process by which labour power and the means of production become commodities. Wallerstein, like Gunder Frank, is explicit in his renunciation of this position. Consistently he argues that since ‘production on the market for profit’ determines capitalist economic development, the problem of the origins of capitalism comes down to the origins of the expanding world market, unfettered by world empire. He is at pains to distinguish the emergence of the capitalist world economy in the sixteenth century—the rise of the world division of labour which emerged with the great discoveries and expansion of trade routes—from the emergence of a system of free wage labour, and contends that the latter is derivative from the former.
II. Adam Smith and the Class Basis of Economic Development
The issues raised here were, of course, at the centre of the controversy in the 1950s over the transition from feudalism to capitalism,  as well as of subsequent controversy over the rise of capitalist underdevelopment. Indeed, it is necessary to understand Wallerstein’s position as a direct outgrowth of the arguments put forward then by Paul Sweezy, as well as of the theses advanced more recently by Frank. To grasp this line of thought, what is essential is to see that the basic theoretical underpinnings for the positions set out by all three of these writers is the model put forward by Adam Smith in The Wealth of Nations, Book 1.  The elements of Smith’s model are very familiar. The development of a society’s wealth—quite sensibly equated with the development of the productivity of labour—is a function of the degree of the division of labour. By this, Smith simply means the specialization of productive tasks—classically achieved through the separation of agriculture and manufacturing, and their assignment to country and town respectively. In turn, for Smith the degree of specialization is bound up with the degree of development of trade: the degree to which a potentially interdependent, specialized labour force can be—and is—linked up via commercial nexuses. Thus, we get Smith’s famous principle that the division of labour is limited by the extent of the market—literally, the size of the area and population linked up via trade relations.
Smith’s argument that the separation of manufacture and agriculture and their allocation to town and country, consequently upon the development of trading connections, will lead to a process of economic growth, as a result of the increased productivity which ‘naturally’ follows from the producers’ concentration on a single line of production rather than a multiplicity of different ones, has a certain plausibility—in the highly abstract form in which it is presented. The appearance of new manufacturing commodities stimulates rural production, which in turn induces the growth of urban output to supply the countryside, and so on. The fact of specialization of function, with agriculture and manufacturing now carried out by separate productive units, makes possible greater efficiency and facilitates invention. A process of self-sustaining growth appears to be entrained. Yet, when the assumptions of the model are even cursorily examined, its limited historical relevance is immediately apparent: it ‘works’ only under the premise of capitalist social relations of production, as well as the specific social forces of production with which these have been historically associated.
What precisely is taken for granted? First, that labour power can and will be transferred from rural agriculture to urban manufacturing in response to market opportunities. Second, that through the separation and specialization of productive units, labour productivity will be improved and continue to be improved. Thus, for the possibility of a developing town-country division of labour, agriculturalists must be free to leave the countryside in response to urban-industrial opportunities, while adequate sanctions must exist to prevent their remaining in the countryside in the face of such opportunities. At the same time, unless the productivity of labour—in the first instance in agriculture—is increased, it will be impossible to support the entry of increasing labour power into urban manufacturing, the sine qua non of economic development. Indeed, unless agricultural surpluses continue to grow, the urban industrial population is strictly limited; for the proportion of the population in town and country depends strictly on the productivity of labour. Yet these processes rest on certain conditions, beyond an emerging market and the desire to exploit it, which cannot in fact be assumed: 1. the potential ‘mobility of labour power’ in response to the market—which is, however, bound up with the degree of freedom/unfreedom and with that of economic dependence/independence of the direct producers; 2. the potential for developing the productivity of labour through separation and specialization of tasks—which is, however, bound up with the possibilities for developing co-operative labour in connection with growing means of production; 3. the potential for enforcing continuing pressure to develop labour productivity—which is, however, bound up with the survival and reproductive needs of the direct producers and exploiters in relation to their access to the means of subsistence and production.
The First Condition
Thus the emergence of possibilities for profitable production thanks to the establishment of commerce, classically in urban manufacturing, does not necessarily mean the movement of producers to take advantage of them. For this to occur, in the countryside there must be no substantial barriers to leaving agriculture, such as serfdom or slavery. In other words, any direct forceful controls over the movement of the direct producers, arising from the social relations by which the ruling class extracts a surplus from them, must be eliminated. Concomitantly, either the advantages to entering urban production must outweigh the incentives of the agricultural producers to remain in the countryside, or they must be subject to forcible ejection from the land. In other words, the property of the direct producers in the means of agricultural production and subsistence must be broken, or else they will not have to move towards growing industrial opportunities.
The Second Condition
At the same time, increased possibilities for profit via increased output do not automatically determine the growth of production via an increase in labour productivity, by means of a growing separation and specialization of function. To begin with, the mere separation of productive functions (e.g. industry/agriculture) and their assignment to different producers (e.g. town/country)—what constitutes ‘specialization’ in the strict sense—can only up to a point bring improvement in the productivity of labour. Smith’s argument that it does so is essentially two-fold: first, that the repetition which comes with the assignment of an individual to a single task increases the efficiency of labour; second, that such concentration leads to invention. Yet despite the undoubtedly large gains which can accrue from this sort of specialization, and although the history of the world is replete with countless examples of it (classically, the emergence of town/county division of labour), nevertheless, before the onset of capitalism, each such example was undermined within a relatively short run by the declining productivity of agricultural labour. This is because the effects of specialization, in Smith’s narrow sense of the individuation of production (the separation of previously combined productive tasks), will necessarily be restricted—unless it is accompanied by the better equipment of labour power with the means of production so as to magnify its productivity.  Yet historically, the increasing application of increased means of production to the process of labour has been inextricably tied to the emergence of co-operation—the integration of related work activities within a unit of production. Moreover, the process of developing co-operation in connection with the adoption of new and better means of production is not merely a question of individual inventions. It is especially bound up with the economy’s capability to adopt new methods of production and the necessity for it to do so—in other words, with its capacity for innovation.
Precisely because the increase of labour productivity is historically tied up with innovation in the means of production in relationship to the development of co-operative labour—and not merely with individuated production and individual invention coming with separation—even systematic attempts to respond to market opportunities for increasing production for profit do not necessarily entail greater productiveness of labour. This is because pre-capitalist class structures—systems of surplus extraction and property—tend to fetter the application of the means of production in relation to the development of co-operative production. This is due to the predominance of forceful relationships by which a surplus is extracted from the direct producers, and/or to the predominance of individualized production bound up with property of the direct producers in the means of subsistence and means of production.
On the one hand, where labour is organized by means of force exerted by the ruling class on the direct producers, the effectiveness of collecting labour for co-operation is muted because of the lack of interest of the direct producers in the productive process. Here, the existence of direct, non-market access of the direct producers to the means of subsistence—either in the immediate sense, as in serfdom where the producers possess their own plots, or indirectly, as in slavery where the slave-owners provide the slaves’ subsistence because the latter are their property—determines that force can be of only limited utility in affecting the quality and consistency of labour in connection with increasing, and increasingly complex, tools. On the other hand, where labour is organized by the direct producers on the basis of their property in the means of production, as exemplified in peasant freeholder production, the tendency (general among all the peasant producers) to relate their individual development of the productive forces to the goal of maintaining their family and keeping their property tends to fetter the development of co-operative labour, by keeping labour individuated and preventing the accumulation and concentration in one place of labour, land and the means of production. Small property tends to dictate individualized and unspecialized production.
It was in the context of his discussion of the difficulties of the development of co-operative labour in the face of pre-capitalist social-productive relations that Marx concluded that: ‘If, then, on the one hand, the capitalist mode of production is a historically necessary condition for the transformation of the labour process into a social process, so, on the other hand, this social form of the labour process is a method employed by capital for the more profitable exploitation of labour, by increasing its productive power.’  Thus the systematic barriers set up by pre-capitalist property forms to the development of increasing means of production in relation to co-operative labour have the end result that attempts to increase surplus in response to market opportunities under such systems tend to be ‘biased’ away from the means of extracting ‘relative surplus labour’ in favour of recourse to the methods of ‘absolute surplus labour’.
The Third Condition
Finally, even where major improvements in the forces of production are introduced in pre-capitalist modes of production—and their historical significance has, of course, been very great—they nonetheless tend to constitute ‘once and for all’ processes. In other words, the market exerts no pressure toward the continual revolution of the means of production. It is the essence of pre-capitalist social relations of production that both exploiters and the direct producers are, in one way or another, directly connected with their means of subsistence and means of production. As a result, their survival and reproduction is not dependent on the sale of their products on the market; consequently they do not have to compete in terms of their productive powers. Indeed, far from determining increased production via accumulation and innovation, such class systems tend to provide opportunities and create pressures for the exploiters and direct producers to follow other needs than the maximal expansion of their productive potential for the market: to use their surpluses for purposes other than ‘reinvestment’ in increased means of production, and/or even to avoid production for ‘maximum surplus’ in the first place.
Thus where the direct application of force is the condition for ruling-class surplus extraction, the very difficulties of increasing productive potential through the improvement of the productive forces may encourage the expenditure of surplus to enhance precisely the capacity for the application of force. In this way, the ruling class can increase its capacity to exploit the direct producers, or acquire increased means of production (land, labour, tools) through military methods. Rather than being accumulated, economic surplus is here systematically diverted from reproduction to unproductive labour. Correlatively, where the family plot forms the basis of individual peasant property, there is every incentive to direct production, and production for exchange, so that the multiplicity of labour processes and means of production which ensure the continued subsistence of the family plot can be carried out successfully. The capacity of the peasant proprietor to carry out these disparate labours for subsistence obviously hinders even the elementary steps towards the development of specialization of labour which are the crucial conditions for the development of the productive forces. The contrast, in such cases as these, with capitalist social relations—where the separation of the exploiters and direct producers from the means of subsistence enforces the use of surplus for accumulation and innovation to make possible survival and reproduction—could not be more stark.
The Structure of Capitalist Development
In sum, then, Smith’s fundamental proposition—that the rise of a trade-based division of labour will determine economic development through the growth of specialization and thereby the productivity of labour—is understandable only in terms of his individualistic methods and assumptions. It is only such premises which allow him to attribute the dynamic of the system as a whole to the qualities inherent in its individual parts—in particular, his connection of the rise of labour productivity to the individuation of production, and especially his attribution of a process of accumulation via innovation to individuals’ ‘self interest’ manifested in ‘profit maximization’ and ‘competition on the market’. This is how things appear, ‘how they really are’ under capitalism. But this is only because the specific functioning of the individual components (productive units) of the system—their ‘self-interest’ profit maximization in order to compete on the market—is structured by the system of capitalist class relations. Correlatively, Smith’s ahistorical determinations by ‘market forces’ are understandable only as a result of the failure to take into account the differential limitations and potentialities imposed by different class structures on differentially placed exploiters and producers responding to such market forces—and, further, the different sorts of interests or goals to which such exploiters and producers might attempt to subordinate exchange. It is precisely by determining such disparate and conflicting class interests that historically-developed structures of class relations (relations of surplus extraction and property) open up or foreclose different patterns of development—in particular by conditioning the structure of income distribution and social demand, and thereby the distribution of labour-power and the means of production between productive and unproductive production, while establishing the potential for developing the productive forces. Indeed, as I shall try to show, it is precisely the same class relations as those that fully open the way for the transfer of labour power from town to country, which also provide the basic conditions for the development of fixed capital in relation to co-operative labour and which furthermore generate continual pressure for accumulation by way of innovation—i.e. capitalist class relations where labour power is a commodity.
It follows, finally, that to discover the historical origins of the onset of a pattern of capitalist economic development it is not enough to refer, as Smith does, to the rise of the market. In this respect, Smith’s fundamental problem is not, as is often assumed, his attribution of trade to a ‘natural propensity in human nature to truck, and barter, and exchange’. Smith was, in fact, at pains to provide specific historical examples of ‘the original establishment of trade routes and trading connections’.  Once established, these connections of exchange set in motion, so to speak, the model of development, via the division of labour—so that for Smith both the origins and developmental pattern of capitalist production are rooted in the same process. But as I shall try to show, the rise of trade is not at the origin of a dynamic of development because trade cannot determine the transformation of class relations of production. Indeed, precisely because it does not do so, the historical problem of the origins of capitalist economic development in Europe comes down to that of the process of ‘self-transformation’ of class relations from serfdom to free wage labour—that is, of course, the class struggles by which this transformation took place.
The Line to Sweezy and Wallerstein
The parallels between the positions of both Sweezy and Wallerstein and that of Adam Smith are striking, and the defects of their arguments are the result of their adopting his assumptions. Like Smith, both Sweezy and Wallerstein, implicitly or explicitly, equate capitalism with a trade-based division of labour. They thus understand its special dynamic of accumulation through innovation as a function of the imperatives of exchange on the market and the productive effects of specialization. As a result, their accounts of the transition from feudalism to capitalism end up by assuming away the fundamental problem of the transformation of class relations—the class struggles this entailed—so that the rise of distinctively capitalist class relations of production are no longer seen as the basis for capitalist development, but as its result.
Of course, Wallerstein and Sweezy appear to differ from Smith precisely in their apparent concern for ‘class’. But, in fact, their conception of the ‘capitalist effects’ of the growth of exchange and the division of labour—the tendency to increasing output and productivity advance built into ‘production for profit on the market’—lead them to assimilate the emergence of new class relations of production to commercial development. Explicitly or implicitly, they regard the transformation of class relations as a necessary effect of continuing commercialization. They see the rise of commercial relations as forcing the individual producers continuously to develop the productive forces through the mechanisms of ‘profit maximization’ and ‘competition on the market’. At the same time, they also see that the development of the productive forces past a certain point requires the reorganization of production within the ‘productive unit’, and conclude that this will in turn require and determine the transformation of the ‘relations of production’ within that unit. The transformation of class relations, therefore, emerges as a consequence of the market-determined development of the productive forces within the individual productive units which compose the economy. Smith’s model of development is thereby ‘extended’ to subsume the transformation of class relations within the broader process of the development of a trade-based division of labour.
Thus, in the first place, both Sweezy and Wallerstein argue that the incorporation of regions dominated by feudalism—specifically, lord-peasant relations characterized by serfdom—into networks of commercial relations cum division of labour has the effect of making feudal-serf productive units function more and more like purely capitalist productive units. They are forced to accumulate and innovate. Secondly, and relatedly, both Sweezy and Wallerstein argue that once the division of labour (town/country, world economy) has been established, the ensuing process of rationalization will give rise, as an economic necessity, to a move away from traditional serf-lord relations towards the development of ‘classically capitalist’ social-productive relations of free wage labour. To develop the productive forces, at least in certain regions in certain productive lines, it eventually becomes necessary to introduce free wage labour. Thus, free wage labour arises as a techno-economic adaption within the producing unit. The class system of free wage labour emerges as a by-product of the individual actions of (de facto capitalist) producing units which reorganize production in order to maximize surplus and compete on the market. As a result, the transition to capitalism is seen to occur as a smooth unilineal process—which is essentially no transition at all. Given the rise of exchange and the techno-economic imperatives of the development of the productive forces under commercial pressures, the rise of capitalist social relations is reduced to a formality.
Sweezy and Wallerstein, like Smith, implicitly regard ‘surplus maximization’ and ‘competition on the market’ as essentially trans-historical forces, requiring only the original impetus of commerce, the rise of the market, to start working their progressive effects within the extant individual productive units. To them, therefore, as to Smith, the historical problem of the origins of capitalism becomes that of the origins of trade-based division of labour. Smith himself, as noted, was very careful to root the application of his ‘model of economic development’ in specific historical commercial/transport breakthroughs—‘a primary establishment of trading routes’. And this, indeed, is precisely the tactic followed by Sweezy and Wallerstein, both of whom found their accounts of the transition from feudalism to capitalism upon just such a primary establishment of trading routes: for Sweezy (who follows Henri Pirenne), it is the re-establishment of Mediterranean commerce after the Mohammedan invasions; for Wallerstein (who follows Frank), it is the great voyages of discovery and conquests which paved the way for the rise of the world market.
From here, Wallerstein and Sweezy follow Smith in arguing for a more or less natural emergence of increased specialization, and a resulting increase in productivity due to specialization—ultimately leading to the transformation of the productive forces, and with them the productive relations. For Sweezy, it is the emergence of Smith’s town/country division, developing in early medieval Europe. This follows upon the concentration of artisan production in the towns, originally to service the needs of newly developing settlements of long-distance traders taking advantage of the opening up of trade routes. But ultimately, the artisans begin to supply the countryside with the manufactured goods it needs on a more efficient basis, while in turn offering a growing market for agricultural products. For Wallerstein, it is the division of the Atlantic World into interdependent regions, specializing in different sorts of agricultural production and/or manufacturing. Once these ‘natural’ steps have been taken from the establishment of a trading nexus to the emergence of an interdependent specialization, the authors in question consider that capitalism is either imminent (Sweezy) or already extant (Wallerstein)—in particular, that trade-induced specialization entrains a process of rationalization via accumulation and especially innovation in the socio-technical organization of production.
Now, there is perhaps nothing wrong with ‘beginning’ in this manner with such historically-specific commercial developments, for there is no denying their importance. But the fact is that such flowerings of commercial relations cum divisions of labour have been a more or less regular feature of human history for thousands of years. Because the occurrence of such ‘commercial revolutions’ has been relatively so common, the key question which must be answered by Sweezy and Wallerstein is why the rise of trade/division of labour should have set off the transition to capitalism in the case of feudal Europe? This question is pivotal because, contra Smith,  Sweezy and Wallerstein, the development of trade does not determine a transition to new class relations in which the continuing development of the productive forces via accumulation and innovation become both possible and necessary. Marx encapsulated this difficulty when he wrote: ‘on the basis of every mode of production, trade facilitates the production of surplus-products destined for exchange, in order to increase the enjoyments, or the wealth, of the producers (here meant are the owners of the products). Hence, commerce . . . all development of merchants’ capital tends to give production more and more the character of production for exchange-value and to turn products more and more into commodities. Yet its development . . . is incapable by itself of promoting and explaining the transition from one mode of production to another.’ 
Both Sweezy and Wallerstein argue, unexceptionably, that the appearance of the new products on the market tends to increase the feudal lords’ drive to increase their consumption, and that this may lead them to systematize the means of acquiring goods which can be used to buy these new products. As Marx pointed out, under a natural economy (self-sufficiency) the demands of the feudal lord is limited by the ‘walls of his castle’—supplying the immediate necessities of supporting himself and his entourage. But to state this is not to state that the process of acquiring goods in order to exchange will lead the serf-lords systematically to increase production by means of the development of the productive forces, as both Wallerstein and Sweezy do—although each in their own manner.
III. Sweezy and the Transition from Feudalism to Capitalism
Sweezy contends that the transformation of serf agricultural production was a foregone conclusion, once the basic town/country division of labour had been founded in medieval Europe: ‘the manor . . . was fundamentally inefficient and unsuited [to production for the market] . . . Techniques were primitive and the division of labour unwieldy . . . Sooner or later new types of productive relations and new forms of organization had to be found’ (emphasis added).  Sweezy seems to be arguing that the rational course for the lords would have been to commute labour services to money rents and to increase output on the demesne by farming it to a capitalist tenant, who would cultivate the land using improved methods (and ultimately wage labour).  It is now known that by the later middle ages in northwest Europe certain methods of agricultural production had been developed which would have substantially improved output. Yet, as Dobb pointed out many years ago, where serfdom existed—that is, where the lords were in a position to actually control peasant mobility and access to land—the impact of trade only induced the lords to tighten their hold over the serfs, to increase exactions (including labour rent) and, we can add, to eschew innovation in agriculture. This was as true for the areas producing for the urban food markets in England during the medieval period as it was for the East European regions producing for the world food market from the sixteenth to the eighteenth centuries.  Does this mean that the lords were ‘irrational?’
The Lords’ Demands
Now, Sweezy’s assumption that the lords’ desire for increased consumption goods would lead them to seek ways to increase the size of the product which they could appropriate from their peasants is certainly reasonable.  But why should this have led them to develop the productive forces, rather than to extract a larger surplus either by lowering the subsistence of the peasantry, thus increasing their share of the existing product, or by increasing output by making the peasantry work harder or longer—i.e. to raise what we have called ‘absolute’ rather than ‘relative’ surplus labour? Given their access to an unfree labour force, i.e. the existence of serfdom, there is no reason to take for granted that the best way to maximize the surplus available to them was to introduce new organizations of production based on new techniques, rather than to enforce more labour on the demesnes and/or increase the rent from the peasant plot—especially if the former was incompatible with the latter, given the historically developed forces of production.
In fact, the new techniques which could have substantially increased output—the revolutionary systems of ‘up and down (or convertible) husbandry’, which replaced the old ‘permanent’ two- or three-field rotations by an ‘alternation’ of animal and arable production so as to eliminate fallows, while bringing in new soil-enhancing crops—required a very carefully supervised, skilled and technically proficient agriculture.  This would have been quite difficult, if not impossible, to achieve using serf labour, for the serf worked on the lord’s demesne only because he was forced to. Labour applied to the demesne constituted a direct, forcible deduction from that applicable to the peasants’ plot, so he had no incentive to work carefully or skilfully. To put the problem another way, in order to use the new methods the lord would either have to increase substantially his outlay on the manorial supervision of production, and/or find some means to increase the rewards to—and possibly training of—the serfs, so as to elicit the necessary care and quality of labour.
Technically, for improvement to have been worthwhile, the increased surplus achievable from the increase in output arising from the adaption of new methods, allowing for the increased outlay in labour costs, would have had to have been greater than the increased surplus achievable through simply forcing the serf to work longer and harder and reducing, directly or indirectly, his subsistence (by decreasing the size of his plot or increasing the direct rent upon the plot’s output). Given the inherently forceful nature of the system of surplus extraction—the relationship between lord and serf—the method of ‘squeezing’ must have generally appeared to be the logical, perhaps the only feasible path. The alternative of improvement would have required at least an approach toward a somewhat collaborative relationship between the lords and a section of the peasantry. The lords would have had to give up precisely some of the advantage built into their class position which allowed them to extract a given level of surplus.  This was a development running directly counter to the inherently antagonistic dynamic built into the lord-serf structure.
Serfdom and Agricultural Backwardness
It is, indeed, a key confirmation of this line of argument that when the new methods of cultivation became widespread for the first time, constituting a veritable agriculture revolution in sixteenth–seventeenth-century England, it was on the basis of the emergence of an essential partnership between the landlords and richer peasants, who took over as capitalist tenant farmers supervising the introduction of innovations. By this time, serfdom had long collapsed, opening the way to an entirely different rural class structure, based on capitalist social-productive relations. On the other hand, the only examples so far adduced of the adoption of the revolutionary methods of agricultural production during the medieval period in England were in regions where the peasants had succeeded in retaining their freedom throughout the middle ages, despite the lords’ attempts to enserf them. In these instances, the lords moved to increase their surplus by first buying out the free customary peasants—whose customary rent could not be raised and who could not be simply evicted because their customary tenure guaranteed inheritance—then installing the new techniques on their consolidated demesnes using free wage labour.  Thus, it was only where it was difficult to increase their income by squeezing the peasantry because the peasantry were free (and property owners) that the lords turned to ‘improvement’. In other words, the lords sought to increase their income via relative surplus labour only where they were not, in fact, serf lords.
The point is, then, that the individual lord did not generally see a choice between relative and absolute surplus labour, because in reality he did not get to choose, on an individual basis, between production on the basis of serfdom and production using free labour. It needs to be noted, in this respect, that it is mistaken to regard the commutation of labour dues to money rents as the first step toward turning the demesne toward advanced methods of production on the basis of a capitalist tenant and wage labour. For this neglects to take into account that even after commutation, the peasant-serfs remained unfree, so still subject to the lord’s extra-economic exactions on their own plots. Thus commutation (the transformation of rents in labour into money rents) can in no way be equated with manumission (freeing the serfs so they could move, marry and buy land without the lord’s consent). In other words, the lords had no reason to free their peasants, even if they no longer wished to use them on their demesnes; for by forcing them to remain on their peasant plots on the estate as serfs, they held them in the best position to exploit, to squeeze, while reserving the possibility of redirecting their labour back to the demesnes at a later time. 
So barring actions by the peasants to free themselves—revolt or flight—even widespread commutation throughout the countryside would leave the peasantry as a whole still subject to the lords’ direct control. This would leave a formidable barrier to the emergence of a class of richer peasants who might rent the lords’ demesnes as capitalist farmers and potential improvers, as well as the rise of a free labour force. As a result, it would be difficult for individual lords to move toward a policy of estate reorganization by means of leasing and improvement even if they wished to do so—since they would still be operating within the confines of serfdom.
Finally, it needs to be realized that, even to the extent that individuals within the system (or even without) somehow moved to adopt more efficient or ‘profitable’ methods of production, it cannot be assumed that these would be generalized throughout the system, even over a relatively long run. According to Sweezy, with the development of production for exchange: ‘the inefficiency of the manorial organization of production—which probably no one recognized or at least paid any attention to, as long as it had no rival—was clearly revealed by contrast with a more rational system of specialization and division of labor’ (emphasis added).  Now, as we just have seen, there were in fact extremely few improving initiatives within the serf agricultural economy during the medieval period in England (or, as we shall see, in that of Eastern Europe between the sixteenth and eighteenth centuries).
Yet even had these been more widespread, there is no reason to assume that they would have been widely copied, as Sweezy seems to imply they must have been. This was partly because of the barriers we have already mentioned. But it was also because the serf lords were under no compulsion to produce at the highest level of efficiency. This was because they were not, in the last analysis, compelled to make a profit on the market in order to survive, since they could directly, without recourse to the market, supply their own basic (‘subsistence’) needs on their own demesnes with their peasants’ labour. The revelation of ‘inefficiency’ did not determine change and improvement. Thus, in fact, the adoption of more highly productive methods of production in the non-serf regions of medieval England, to which we earlier referred, did not apparently incite emulation in the feudal/serf English heartland. More significantly, as we shall see, the agricultural revolution which gathered strength in England from the sixteenth century penetrated few other European regions, least of all serf-bound Eastern Europe—even despite the latter’s inability to maintain its place on the world market, precisely as a result of its failure to innovate.
Capitalist Rationality under Feudalism
Sweezy’s mistake was obviously to assume the operation of norms of capitalist rationality, in a situation where capitalist social relations of production did not exist, simply because market exchange was widespread: ‘The possession of wealth soon becomes an end in itself in an exchange economy, and this psychological transformation affects not only those who are immediately involved . . . Hence, not only merchants and traders but also members of the old feudal society acquire what we should call today a businesslike attitude toward economic affairs.’  Reasoning from the part or unit of the system to the whole, Sweezy assumed that the rise of the market would lead to what he termed ‘exchange consciousness’ among the serf lords, and that this would in turn lead, by the profit-maximizing actions of individuals, to the rise of new, more efficient social-productive relations which, by virtue of their superiority on the market, would spread throughout the economy. In fact, the existing system of class relations based on serfdom largely determined what was ‘rational’ for individual lords (producing units), i.e. how they could best increase their rent. The application of these (labour-squeezing) methods, in turn, tended to foreclose the emergence of ‘post-feudal’, let alone capitalist social-productive relations, at least by way of the lords’ maximizing initiatives. And even to the extent that a more effective productive organization might emerge, it would not necessarily prevail For the serf-lords’ survival simply did not depend on their relationship to the market.
Precisely because the rise of trade in an economy of serfdom did not necessarily create pressures to develop the productive forces in order to increase income, let alone enforce the generalization of innovation, it could not determine a pattern of economic development, let alone a transformation of social-productive relations away from serf labour, in the direction of free labour and eventually free wage labour. Indeed, the serf social relations, under the impact of trade, tended to entrain a stagnant, often regressive, pattern of overall societal development, making a mockery of the optimistic Smithian model largely taken over by Sweezy, which was built around the assumption that an abstractly-conceived town/country division of labour would lead to productivity increase via specialization. Sweezy is, therefore, far from the mark in his fundamental contention that trade is external to feudalism—in his postulate that ‘trade can in no sense be regarded as a form of feudal economy’.  Precisely because trade developed as an expression of feudal class relations, in relation to needs which were structured by these relations, the specific pressures set up by the rise of commerce and of urban production to increase surplus output in the countryside determined a tendency towards declining productivity in agriculture—which in turn meant a stunted development of the division of labour itself.
Because the lords could not easily improve the productive forces under serfdom, they were largely confined to increasing their incomes via the increase of absolute surplus labour. They could, specifically, increase output only within the definite limits of the available land (subject to transport costs), population, intensity of labour and minimum subsistence level. They thus had little incentive to ‘accumulate’: to reinvest surplus in improved means of production. On the contrary, unproductive expenditures on military equipment or conspicuous consumption could make possible the attraction and equipment of followers. The resultant enhancement of military capability could make possible the improvement of the individual lord’s productive potential—that is, through the outright seizure of lands and labourers in warfare. Indeed, precisely because the potential for the development of the productive forces was so limited, development of military strength might be the most promising means to increase the productive powers of the individual lords.  That the rise of commerce in large part took the form of a market in luxuries and military goods was thus obviously conditioned by the needs of the feudal-serf order—while it undoubtedly intensified these needs and encouraged increased surplus extraction to fulfil them. Yet to the degree that surplus was spent on military/luxury goods, it meant a subtraction from the society’s resources available for developing the means of production or means of subsistence. Because the development of commerce was therefore not external to feudalism, determining the development of new needs in the abstract, but was rather an expression of demands emanating from feudal class relations, especially the feudal ruling class, it ended up by ‘determining’ a form of division of labour which turned in upon itself.
The impact of feudal class relations on the peasants’ ‘own’ production, on their plots, only tended to exacerbate the foregoing tendencies. The lords’ control over mobility of labour and land, deriving from serfdom, hindered the emergence of markets in labour and land, and thus the ability of peasants to accumulate the means of production requisite to improvement (a difficulty further exacerbated, as we shall see, by the general unwillingness of the peasantry to part with their land). At the same time, because the peasant plot was responsible for producing the labour power (peasant labourers) and means of production (tools) which produced the lords’ surplus (via direct labour on the demesne or levies in kind or money), there was every tendency on the part of the lords to undermine the labour power and means of production of the whole system, through undermining the peasants’ long-term reproductive power via short-term surplus extraction. Especially because the peasant plot operated as a productive unit beyond the lord’s direct supervision, there was little means and/or incentive to gauge the destructive effect of lordly levies on the peasant plot’s potential for reproducing the labour power and means of production.  In particular, there was a tendency to exhaust the soil through failure to allow the peasantry enough land, labour time, and means of production (enough of the economic surplus) to support the animals needed to plough and fertilize the land adequately. The resulting tendency to declining productivity meant that demographic crisis was a normal, if not inevitable outcome. 
Serfdom and Urban Underdevelopment
In light of the foregoing, it may be seen that the Sweezy/Smith notion of the urban centres of industry as radiating foci of a nascent capitalism—the source of pressures for progress and models for innovation—is, in the context of serfdom, misconceived. Indeed, in the last analysis, the social relations of serfdom in the rural productive sector not only circumscribed the potential for urban industrial development, but imparted to urban industry an essentially parasitic and conservative character. With declining rural productivity, the number of urban producers was naturally limited in the immediate sense by the potential food supply. On the other hand, by dramatically reducing the purchasing power of the rural producers—by first limiting their productive power, and secondly their ability to keep what they produced—the social structure of serfdom under the pressures of exchange in fact tended to prevent the emergence of a mass market for urban manufacturing, either for consumer goods (especially clothing) or for producer goods (tools).
As a consequence, there was little pressure for productive innovation in urban industry—innovation which would have cheapened the means of production or consumption in the rural-productive sector. On the contrary, effective rural demand arose largely from the landlords’ desires for limited numbers of expensive luxury products, goods which could in no way enhance rural production. It was, indeed, the character of this demand which provided the rationale for the guilds, which tended to dominate manufacturing, limiting entry and output, and determining productive methods. There was only a limited market. Moreover, what was required was highly-crafted goods—which to some extent justified the apprenticeship that was the primary method by which the guilds limited entry. Ironically, then, the development of exchange, as it operated on production, through the prism of serf class relations, tended precisely to strangle the very development of the division of labour which it made possible in the first place.
In this context, we can see the difficulty with Sweezy’s final argument that the towns not only provided the incentives for change, but also the pressures to do so, through serving as a magnet for fleeing serfs—and thus the ultimate cause for the dissolution of serfdom. This argument, in the first place, simply begs the major question of the power of the rural lords to keep the peasants on the land by force—how could they leave? On the other hand, it was not just that guild-organized artisans limited access to industrial opportunities; these opportunities were, as we have seen, sharply limited by the character of rural production. 
In sum, Sweezy’s entire account of the transition from feudalism to capitalism is based on the implicit assumption that capitalism already exists. This occurs because Sweezy mistakenly believes that trade/towns constitute a sort of capitalism in embryo. The expansion of trade/towns will transmit to the economy as a whole, even one dominated by serfdom, a tendency to self-transformation by means of processes of accumulation and innovation which will inevitably lead to the decline of feudalism (and ultimately the adoption of wage labour), due to the exigencies of the development of the productive forces. By virtue of what might be termed a historical functionalism, the relations of production are thus seen to change as a result of the needs of development of the social forces of production. Sweezy can apparently in this manner assume away the central problem of the transformation from a serf to a free labour force, as a result of a classical form of economic determinism: attributing a universal significance to capitalist motivations and mechanisms—‘profit maximization’ and ‘competition on the market’—given only the existence of a ‘system of exchange’, but not capitalist social-productive relations.
In the last analysis, Sweezy’s error is two-fold. It is to posit that the producers’ relationship to the market determines their operation and development and, ultimately, their relationship to one another—rather than vice versa. Correlatively, it is to locate the system’s potential for development in the capacities of its component individual units (thus, the emphasis on motivations), rather than in the system as a whole—specifically, in the overall system of class relations of production which determine/condition the nature of the interrelationships between the individual units and, in this manner, their operation and development. For Sweezy, then, it is the market relation which gives rise to new needs, engenders a ‘profit motive’ leading to specialization and the development of production, and which forces competition for survival.
Now under capitalism, of course, this is the case, at least from the viewpoint of the individual producer. But this is because, under capitalism, the producers’ relationship to the market merely expresses their fundamental relationship to one another as individual commodity producers—i.e. as producers who must sell their products at a profit in order to be able to survive, since they also must buy their means of production and subsistence in order to reproduce. This relationship, which enforces competitive production at the ‘socially necessary’ level upon each producer, in order to make it possible for them to sell and exchange, in turn expresses the fundamental class structure of capitalists and free wage labourers. It is only with free wage labour—with the producers separated from their means of subsistence and means of production—that not only labour power, but also the means of subsistence and means of production can and must appear as commodities—as forms of capital (variable capital and constant capital). Without this separation, on the one hand, there are the strictest barriers to the accumulation of capital: large masses of means of production and means of subsistence (use values) are not ‘free’, subject to be combined at the highest technological level. Use values cannot take the form of exchange values. On the other hand, with the means of production and especially the means of subsistence in the hands of the direct producers, there is no compulsion to exchange in order to reproduce, no pressure/necessity to compete, thus no requirement to accumulate especially by way of innovating in order to survive.
Thus in Marx’s words, only ‘when this [free] labour has been released from its objective conditions of existence through the process of history . . . does it also encounter the possibility of buying these conditions themselves.’  Also, ‘What enables money-wealth to become capital is the encounter, on one side, with free workers; and on the other side, with the necessaries and materials etc., which previously were in one way or. another the property of the masses who have now become object-less, and are also free and purchasable.’ (Marx’s emphasis).  At the same time, ‘Capital proper does nothing but bring together the mass of hands and instruments which it finds on hand. It agglomerates them under its command. That is its real stockpiling; the stockpiling of workers, along with their instruments, at particular points.’ (Marx’s emphasis). 
‘Production for Exchange’
It is Sweezy’s topsy-turvy conception of an abstractly considered market as determining the operation of the ‘units’ which make up the class system of production, notably serfdom, which leads him logically to see the fundamental break in the transition to capitalism as between production for use and production for the market. According to Sweezy, ‘the root cause of the decline of feudalism was the growth of trade’ and ‘the important conflict in this connection is . . . between production for the market and production for use’.  From here it is a natural step to understand the changes in the serf class system of production as emerging as an outcome of the rise of exchange, which, as we have seen, determines in Sweezy’s view a development of the productive forces that ultimately—due to the needs of technical progress—calls free labour into being.
Sweezy quotes Marx to the effect, that ‘in any given economic formation of society, where not the exchange value but the use value of the product predominates, surplus labour will be limited by a given set of wants which may be greater or less, and that here no boundless thirst for surplus labour arises from the nature of production itself’. The logical corollary, as Sweezy goes on to point out correctly, is that, on the contrary, only with the predominance of exchange values do we get the ‘pressure which exists under capitalism for continued improvement in the methods of production’.  Yet the question is, under what conditions does exchange value predominate. Sweezy, of course, gives the straightforward answer: with the rise of production for the market. Here we get a ‘system of production for exchange’, ‘the existence of exchange value as a massive economic fact’, and thus an erosion of ‘the system of use value’, the pressures to improvement, and the slow evolution away from feudal in the direction of capitalist social-production relations.
However, as we have seen, the mere fact of production for exchange ensures none of these things. Indeed, the opposite tendencies toward retrogression are likely, so long as serfdom predominates, and the producing units thus retain their ability directly to supply their own means of production and especially means of subsistence. It is only with the emergence of free wage labour, labour power as a commodity, that there is the separation of the producers from the means of subsistence and production; that production must be marketed to make possible reproduction; that there is, in a true sense, production for exchange. Only then is there predominance of exchange value, leading to systematic pressure to accumulate and thus develop the forces of production. As Marx puts it, ‘the domination of exchange value itself, and of exchange-value-producing production, presupposes alien labour capacity itself as an exchange value—i.e. the separation of living labour capacity from its objective conditions; a relation to them—or to its own objectivity—as alien property; a relation to them, in a word, as capital.’ (emphasis added). 
Thus the correct counterposition cannot be production for the market versus production for use, but the class system of production based on free wage labour (capitalism) versus pre-capitalist class systems. There may be trade, exchange on the market, in both; but the significance of trade in each is fundamentally different, for its effects on the system’s operation and development are fundamentally different. Thus in pre-capitalist modes of production, there is always production ‘for use’, in the sense that the individual production units have direct (non-market) access to the means of subsistence, even if a large amount of what is produced is traded on the market. In essence, it is surpluses above necessity (possibly proportionately large surpluses) which are traded. Since it is not means of production and means of subsistence required for reproduction which are being traded (circulated), the market—specifically other competitive producers on the market—can have only a limited impact on production, its character or the amount produced.
It is this access to the means of subsistence and production which, from the point of view of the individual unit of production, provides the ultimate barrier to the operation of capitalist pressures for surplus maximization and competition on the market. From the point of view of the whole system, it is the institutionalized social relations by which the ruling class extracts a surplus from the direct producers, which prevents the ‘factors of production’—the use values in the form of labour power, land and means of production which are ‘already there’—from being united by money capital. So whatever the level of trade in pre-capitalist societies, the fact that free wage labour does not predominate meant, in Marx’s words, that the ‘instrument itself [and the means of subsistence] is still so intertwined with living labour . . . that it does not truly circulate’. (emphasis added).  Because labour power and the means of production are not separated from one another (and thus ‘free’), neither are fully commodities. As a result, money cannot circulate, does not have to circulate, so as to bring about via exchange the combination of use values in the form of labour power and the means of production, so as to bring about production at the ‘socially necessary’ rate. Money, in other words cannot buy, invest in, the commodities labour power and means of production, so as to bring them together in production, in order to make more money (m–c–m’). 
The capitalist circulation of commodities, therefore, has a meaning radically different from trade under pre-capitalist modes. Here the exchange of commodities (circulation) is a direct and necessary expression of the class structure of the economy as a whole. Because under capitalism the immediate producers (capitalists and workers) do not and cannot produce their own means of production and means of subsistence (that is, the subsistence goods and tools for their own labour process/production), but on the other hand must produce goods necessary for others’ productive processes, their continued production and reproduction depends on circulation—which therefore forms a necessary phase in the total process of production. As Marx summarized the entirely different significances of ‘exchange’ in pre-capitalist and capitalist modes of production, its sharply contrasting position and role: ‘Money and circulation can mediate between spheres of production of widely different [pre-capitalist] organization whose internal structure is still chiefly adjusted to the output of use values. This individualization of the circulation process, in which spheres of production are connected by means of a third, has a two-fold significance. On the one hand, that circulation has not established a hold on production, but is related to it as a given premise. On the other hand, that the production process has not as yet absorbed circulation as a mere phase of production. Both, however, are the case in capitalist production. The production process rests wholly on circulation, and circulation is a mere transitional phase of production, in which the product created as a commodity is realized and its elements of production, likewise created as commodities, are replaced’. 
Exchange Value and Capitalism
To put it simply, the ‘predominance of exchange value’ is nothing less than the predominance of free wage labour, where labour power is a commodity.  But this should come as no surprise. For after all, Marx’s theory of capitalist development begins from the notion of exchange value as merely a form of value. As to value itself, it arises as an expression of a productive system based on ‘abstract labour’; abstract labour, in turn, reflects an economy of individualized, private producers, where every producer must exchange in order to re-produce—so that labour power can and must ‘move’, or be moved, through the action of capital, into productive lines, so as to get the ‘socially necessary rate’ or ‘the average rate of profit’. Such an economy exists only where the direct producers have been separated from the means of production and especially the means of subsistence—i.e. under a system of free wage labour, where labour power is a commodity, in other words capitalism. 
It is precisely the separation of labour power and the means of production, and their appearance as commodities in the individualistic system of private production, which determines the full development of the function of money in circulation ‘as an agent of productive capital’. This separation has this result by determining at once the radical opposition (polarization) between use values and exchange values—since exchange of commodities must take place in order to make possible their use in production—and between the commodity money and the commodities labour power and the means of production—precisely since commodities cannot directly appear as values, ‘marked’ as it were with their values (labour-time embodied), but must instead find their place in production through being circulated. The money commodity can and must function so as to circulate to ‘re-combine’ labour power and means of production, to facilitate production for the highest rate of profit (m–c–m’).
Indeed, it is the achievement of capitalism to bring with it, as it were—albeit in an unconscious and uncontrollable, ‘anarchic’ manner—the interdependence of all producers, in a way which requires that each must produce to the hilt for every other. This interdependence is manifested in, and achieved through exchange on the market; but the market neither creates this interdependence, nor determines its operation. This is a product of the class system of individualistic production, based on separation of the producers from the means of production and means of subsistence—the same separation which enforces accumulation via innovation by way of the exchange of money capital for free labour power and the means of production, in order to make the ‘socially necessary rate’ or average rate of profit.
In the foregoing context, we can see that the original historical process by which the predominance of exchange value emerged is precisely the same as that by which there arise the social productive-relations of capital and free wage labour: they are one and the same. As Marx points out, for capitalist production, ‘we suppose historical processes [of dissolution] which transform a mass of individuals . . . if not perhaps immediately into genuine free labourers, then at any rate into potential free labourers, whose only property is their labour power and a possibility of exchanging it for existing values.’ He then goes on to include among these historical processes of dissolution the destruction of serfdom, the separation of the peasantry from their means of subsistence, the breakup of guilds, and the separation of the artisan from his means of production. As Marx concludes, ‘It will be seen on closer inspection that all these processes of dissolution mean the dissolution of relations of production in which use value predominates.’  We have tried to show that the mere rise of trade cannot, in itself, determine the processes of dissolution. They are understandable only in terms of the conflictual processes, processes of class transformation and class struggle, which tend to emerge from the contradictory character of the pre-capitalist social relations themselves.
IV. Wallerstein and the Modern World Economy
In Wallerstein’s The Modern World System, the Smithian theory embedded in Sweezy’s analysis of the transition from feudalism to capitalism is made entirely explicit, and carried to its logical conclusion. It is in the light of our discussion of Sweezy and Smith that we can begin to locate the central problems with Wallerstein’s approach. Thus Wallerstein straightforwardly defines capitalism as a trade-based division of labour, and it is here that he locates the dynamic of capitalist economic development. ‘Leaving aside the now defunct minisystems [tribes, etc.], the only kind of social system is a world system, which we define quite simply as a unit with a single division of labor, and multiple cultural systems’ (emphasis added) (rfd, p. 390) . . . ‘It is a “world” system, not because it encompasses the whole world, but because it is larger than any judicially defined political unit. And it is a “world economy” because the basic linkages between the parts of the system are economic’ [i.e. trade/exchange, in contrast with a world empire where the basic linkages between the parts of the system are political, via an all-encompassing, over-arching, tax-collecting bureaucracy] (emphasis added) (mws, p. 15) . . . ‘Capitalism and a world economy (that is, a single division of labor, but multiple polities and cultures) are obverse sides of the same coin’ (emphasis added) (rfd, p. 391).
As with Sweezy, although more explicitly, the mainspring of the developing division of labour is simply the ‘profit motive’, which is induced by trade and the market and which, in turn, induces accumulation (plough-back of surplus) and innovation. Capitalism, says Wallerstein, is ‘a mode of production, production for profit in a market’ (emphasis added) (rfd, p. 399). Wallerstein draws the logical consequences of this position, which in Sweezy were left unstated: trade in itself will lead to accumulation and innovation via the profit-motivated development of the division of labour;  therefore, it logically follows that any region which is part of the apparently interdependent system of exchange which constitutes the world division of labour is capitalist, whatever its methods of ‘labour control’ and of ‘rewarding labour power’. Once embedded within the world economy/world market, the productive regions based on serfdom (what Wallerstein calls ‘coerced cash crop labour’), in particular the grain-exporting regions of the Eastern European ‘periphery’, cease to be one bit less capitalist than the regions whose production for the market is based on free wage-labour.
Once, says Wallerstein, ‘[the] so-called reciprocal nexus we identify with feudalism, the exchange of protection for labour services . . . is contained within a capitalist world economy, its autonomous reality disappears. It becomes rather one of the many forms of bourgeois employment of proletarian labor to be found in a capitalist mode of production, a form that is maintained, expanded or diminished in relation to its profitability on the market’ (emphasis added) (ffc, pp. 278–9). So that, for Wallerstein, ‘Capitalism thus means labor as a commodity to be sure. But in the era of agricultural capitalism, wage labor is only one of the modes in which labor is recruited and recompensed on the labour market. Slavery, coerced cash cropping (. . . the so-called “second feudalism”), share-cropping and tenancy are all alternative modes’ (rfd, p. 400). Indeed, it is precisely the specialization of ‘capitalist’ systems of labour control/reward to labour by region, made possible by trade, which constitute the basis of the capitalist world economy, and account for its ability to develop. Specifically, ‘The emergence of an industrial sector [in the core] was important [in the rise of the world capitalist division of labour], but what made this possible was the transformation of agricultural activity from feudal to capitalist forms. Not all these capitalist “forms” were based on “free” labor—only those in the core of the economy. But the motivations of landlord and laborer in the non-“free” sector [in the periphery] were as capitalist as those in the core’ (mws, p. 126).
The World Division of Labour
For Wallerstein, then, the growth of the world division of labour is the development of capitalism. Not surprisingly, therefore, he can forthrightly state that the rise of free labour is merely an aspect of the development of the world division of labour, determined by the technical requirements of the development of the productive forces in given types of production and specific regions. Sweezy could not have come explicitly to this conclusion, for he seems to accept Marx’s massive emphasis, in both Capital (especially Part 8 on ‘So-called Primitive Accumulation of Capital’) and the Grundrisse (especially the passages on pre-capitalist economic formations), on the rise of free wage labour/labour power as a commodity, presented as the fundamental basis for the capitalist mode of production—for the accumulation of capital. But Wallerstein states: ‘“proletarianization of labor” and “commercialization of land” . . . do not involve the transformation of feudalism into capitalism, but, are aspects of the development of the capitalist world economy’ (Wallersteins emphasis) (ffc, p. 277). Specifically, with the rise of a trade-based division of labour, free wage labour as a system of labour control/reward to labour emerges for productive tasks using greater amounts of capital and requiring more skills. As Wallerstein succinctly puts it, ‘Free labor is the form of labor control used for skilled work in the core countries, whereas coerced labor is used for less skilled work in the peripheral areas’ (mws, p. 127).
It can immediately be seen that, like Sweezy, Wallerstein takes it for granted that ‘profit maximization’ and (implicitly) ‘competition on the market’ will lead to the accumulation of capital and innovation. Not surprisingly too, Wallerstein, like Sweezy, falls back in this regard upon the subjective motivations of the exploiting classes, in the face of the market, as if the need or desire to increase their surplus will automatically lead to the increase in production, and even the improvement of the forces of production. Yet, as I have argued, such mechanisms as profit maximization and competition on the market are ‘effective’ only insofar as they express capitalist class relations. They will operate to bring about a tendency to accumulation by way of innovation only under a system of free wage labour, where labour power is a commodity. They cannot be assumed to do so, for example, under serfdom. Thus, a historical transformation of class structures, which the market itself cannot induce, is at the centre of the feudalism-capitalism transition.
It is necessary to emphasize that Sweezy did not, explicitly, reject the foregoing viewpoint. Indeed, it was no doubt his position that a system of free wage labour is a precondition for a built-in tendency to capital accumulation and the development of the productive forces. However, in arguing that the pressures of market production would lead to an evolution away from serfdom toward capitalism due to market-induced needs of the ruling class to increase production and thus to adopt new productive forces inoperable under the old mode, he ended up contradicting this viewpoint. For the latter argument implicitly entails the idea that serfdom itself will develop a tendency to socio-technical innovation under market pressure (bringing with it ultimately a change to free labour)—so that free labour becomes a consequence rather than a condition of capitalist development.
Wallerstein attempts to cut through this contradiction by banishing it. If one contends that labour power as a commodity is the essential condition for economic development via accumulation and innovation, it is illogical to argue that trade will induce processes of development via accumulation and innovation within the old mode of production which will bring about the transformation of the old mode itself—towards free wage labour. In that case, the dynamic of development clearly resides in trade, not in the class relations of labour power as a commodity. Thus Wallerstein simply denies from the start that free wage labour is a condition for accumulation via innovation, so that he can consistently argue that a trade-based division of labour is not only responsible for the origins of capitalism, but also the source of its dynamic of development. Thus various forms of ‘labour control/reward to labour’—free wage labour included—emerge merely to facilitate the market-induced processes of economic development (and underdevelopment). Yet, as we have already seen with regard to Adam Smith, the general consequence of such a position is an ahistorical, non-class conception of the division of labour, which fails to notice that the very development of the trade-based division of labour can only be a product, not the source, of the development of the productive forces (the productivity of labour), which in turn are dependent upon and limited by the class relations in which they evolve.
A Quantitative Model of Development
The fact is that in order to see the growth of capitalism as an expression of the development of the world division of labour, Wallerstein must end up by sketching a conception of the development of the productive forces which does not really incorporate qualitative advance, specifically by way of the growth of labour productivity. The picture of development which Wallerstein lays out is essentially quantitative, for it does not actually specify the development of the productivity of labour as a regular, if neither continuous nor permanent, feature of the capitalist system, the product of capitalism’s built-in tendency to accumulate by means of innovation.
Wallerstein’s systematic refusal to integrate innovation and technical change as a regular feature of capitalist development may seem hard to credit. Yet Wallerstein is himself quite explicit. He emphasizes that there have been two, and only two, types of world system: world empires and world economies. What distinguishes economic development within the world empire is the domination of a single, surplus-extracting bureaucracy. In contrast, the world economy, consisting of a multitude of polities, is not burdened by one: ‘thus far there have only existed two varieties of . . . world systems: world empires, in which there is a single political system over most of the area however attenuated the degree of its effective control; and [world economies], in which such a single political system does not exist over all, or virtually all, of the space’ (mws, p. 348).
Wallerstein makes clear that the economic superiority of the world economy over the world empire is not really ‘positive’, to be found in its superior system of production; but ‘negative’, located in its superior system of distribution—that is, in the non-existence of a surplus-absorbing bureaucracy. Thus, says Wallerstein, ‘It is the social achievement of the modern world, if you will, to have invented the technology that makes it possible to increase the flow of the surplus from the lower to the upper strata, from the periphery to the center, from the majority to the minority, by eliminating the “waste” of too cumbersome a political structure’ (emphasis added) (mws, pp. 15–16). He makes a further specification: ‘I have said that a world economy is an invention of the modern world. Not quite. There were world economies before. But they were always transformed into empires: China, Persia, Rome. The modern world economy might have gone in that same direction—indeed it has sporadically seemed as though it would—except that the techniques of modern capitalism and the technology of modern science, the two being somewhat linked as we know, enabled the world economy to thrive, produce, and expand without the emergence of a unified political structure’ (mws, p. 16).
Wallerstein could hardly be more straightforward in asserting that the modern world economy contains no inherent dynamic toward technological innovations. It is only one in a long succession of world economies (trade-based divisions of labour with multiple polities); and none of the previous ones succeeded, by virtue of their organization of production, to transform the productive forces. The modern world economy could easily have gone the way of its predecessors, for its social organization of production was not essentially different—except for the unexplained appearance on the scene of ‘the techniques of modern capitalism and the technology of modern science’. Technical advance and innovation, to the extent they have a place in Wallerstein’s system, function as a deus ex machina.
Wallerstein’s failure specifically to incorporate innovation flows from his attempt to substitute for a qualitative model of the development of the productive forces based on ‘the accumulation of capital on an extended scale’, a quantitative model based on the expansion of the division of labour, which makes possible specialization. Specialization, as a natural result of an expanded world economy made possible by trade, thus becomes for Wallerstein the key to the development of the productive forces—in particular, specialization by ‘methods of labour control/reward to labour’. As Wallerstein states, ‘The world economy was based precisely on the assumption that there were in fact these three zones and that they did in fact have different modes of labor control. Were this not so, it would not have been possible to assure the kind of flow of surplus which enabled the capitalist system to come into existence’ (emphasis added) (mws, p. 87).
Thus, for Wallerstein, it was the specialization of methods of labour control by region which was behind the development of production, and which made possible, in turn, the growing division of labour. Since the different methods of labour control were most appropriate for (suitable for) the specialized productions which they governed, it was precisely their adoption under the pressures of the world market which made possible the growth in production, and thereby the further growth in the division of labour: ‘Why different modes of organizing labor—slavery, “feudalism”, wage labor, self-employment . . . Because each mode of labor control is best suited for particular types of production’ (mws, p. 87).
Yet this line of argument breaks down, as a result of the contradictory character of Wallerstein’s conception of ‘systems of labour control/reward to labour’. On the one hand, Wallerstein’s argument that systems of labour control are best-suited to particular types of production—indeed that the world economy was precisely made possible by matching the most suitable mode of labour control to each region—suggests a view of these systems as essentially technical methods of production. These are chosen like any others, according to the exigencies of maximizing production (output), given the eco-demographic character of the region. On the other hand, modes of labour control are obviously, for Wallerstein, methods by which a ruling class extracts a surplus from the direct producers. Under this conception, a method of labour control is most suitable if it maximizes the surplus for the ruling class of that region. The problem is that maximizing output is not the same thing as maximizing surplus to the ruling class: each involves a different ‘principle’ and different, mutually contradictory, processes.
Thus Wallerstein’s conception of economic development via necessary specialization by the most suitable mode of labour control makes it appear that the adoption of different modes of labour control will follow something like a Ricardian law of comparative advantage. By this reasoning, which can only be termed ‘ultra-Smithian’, as the world market expands, the eco-demographic characteristics of an area determine its specialization, what will be produced and the most appropriate method of production. This carries with it, in turn, a system of labour control and reward to labour. The result is maximal output everywhere, maximal growth for the system as a whole.
It is not difficult to get to the bottom of this deterministic system. For its logical premise is the extra-historical universe of homo oeconomicus, of individual profit maximizers competing on the market, outside of any system of social relations of exploitation. It is a universe in which any apparent structure of social relations which emerges in production is merely a technically determined outcome of individual choices by free individual ‘producers’ who have access to different, relatively scarce factors of production, and who have a given range of alternative productive techniques at their disposal. Above all, it is a universe where payments or rewards go to ‘factors’ according to their relative scarcity, not to classes by virtue of their exploitative capacities. Nevertheless, to be consistent, Wallerstein must in fact reason largely according to these premises. This is because he must be arguing that each mode of labour control is somehow ‘productive’, that it contributes to maximizing production for the market. For it is the superiority of the mode of labour control for a given region which explains at once its adoption and the resultant maximal contribution of this region to the development of the world division of labour. Each method of labour control emerges and is maintained because it is the most competitive for that region on the world market; each method contributes maximally to the development of the world economy because it ensures the region’s maximization of output.
Yet to say that the mode of labour control is thus in a real sense productive is also to say that the exploiters’ labour control function is a contribution to production, indeed a necessary ‘labour’. By this logic, the ‘exploiters’ (come to) occupy the position they do because they contribute a factor which is competitive on the market (their method of labour control is more competitive than others). Correlatively, they receive a part of the ‘surplus’ because they supply a factor (labour/function) which is relatively scarce and is therefore rewarded by exchange. This is, of course, perfectly in line with neo-classical economic assumptions. But it leaves little room for exploitation or class. It also leaves little room for the social realities of the period which Wallerstein is analysing. For example, as we shall see, it would be difficult to argue that the landlords’ ‘contribution’ to what Wallerstein calls the ‘coerced cash crop’ method of labour control (serfdom) was necessary to peasant production, indeed made possible its highest productive development in serf-dominated Eastern Europe of the early modern period; for peasant production ‘alone’, would have been far more productive than a peasant production organized via serfdom. Yet this is the necessary conclusion of Wallerstein’s argument concerning the maximization of production and the development of the world economy as a result of the emergence of regions specialized in terms of the ‘most suitable’ method of labour control. It is also the result of a line of thought which must implicitly assume the mechanistic notion that the ‘needs of the forces of production’ (the labour process or social organization of work) can determine the structure of exploitation (the social relations of property and force by which a surplus is extracted from the direct producers).
To avoid misunderstanding, I am not seeking to deny that exploiting classes at times perform productive functions; rather to affirm that their productive function derives from (but does not determine) their position as exploiters. As Marx put it, ‘It is not because he is a leader of industry that a man is a capitalist; on the contrary, he is a leader of industry because he is a capitalist.’  The correlative point is that it is not the exploiters’ contribution to production—their response to the abstract needs of production—which allows them to extract a surplus from the direct producers. No function they perform can explain why they receive part of the product. It is, on the contrary, their position as exploiters which allows them to carry out any productive function they do, in a manner which accords with their needs and capacities (strictly limited by the character of their relationship of exploitation with the direct producers) to extract the maximum surplus.
Class Structure and Class Conflict
Yet, if one is to begin in this manner with class, with a historically evolved system of surplus extraction ultimately sanctioned by force, which cannot therefore be seen to arise from the immediate needs of production itself, one is compelled to admit that the rise of a market-based division of labour cannot determine the ‘optimum’ development of the productive forces through eliciting the ‘most suitable’ method of labour control for each region (for competitive production on the world market). This is, to begin with, because the historical evolution or emergence of any given class structure is not comprehensible as the mere product of a ruling-class choice and imposition, but, as we shall see, represents the outcome of class conflicts through which the direct producers have, to a greater or lesser extent, succeeded in restricting the form and extent of ruling-class access to surplus labour. Furthermore, once given the establishment of a structure of class relations, the manner in which any ruling class can or does best maximize its surplus may not at all correspond with the objective requirements for the development of the productive forces, i.e. output. Finally, even if an established pre-capitalist ruling (or propertied) class does not maximize its surplus, there is generally no economic necessity for it to be surpassed by more effective ‘surplus maximizers’ (more ‘suitable’ methods of labour control); for its constituent members (units), having direct access to their means of subsistence and reproduction, do not have to compete on the market to survive and reproduce. We have, indeed, already made these latter points vis-`-vis Sweezy’s theory, by reference to the case of medieval serfdom. Here the ruling class sought to ‘use’ the class relationship of serfdom to maximize its surplus for the market; this in fact led to the degeneration of the productive forces, and even to the undermining of the surplus going to the ruling class. On the other hand, a direct passage to capitalist class relations by individual serf-lord profit maximizers would have been generally unfeasible, indeed counter-productive for the lords. Yet this general inability to adopt more effective forms of surplus extraction did not determine the replacement of the particular class system of production by a more productive one.
Correlatively, as I shall show, precisely because the grain agriculture of Eastern Europe was based on serfdom, it could in no way be considered the ‘most suitable’ form of production for subsidizing—through maximizing output available for export—the development of the world division of labour. Indeed, precisely because it was not (despite Wallerstein’s assertion) ‘capitalist’, it could not develop the productive potential successfully to underwrite the long-term development of the intensive agriculture and manufacturing of the ‘core’. Nor could its inability to do so, indeed its inability to maximize surplus or compete on the market, dictate the adoption of capitalist social-productive relations. On the other hand, I shall argue that it was in fact the ‘classically’ capitalist social relationships of landlord-capitalist tenant-free labour, which arose in certain regions of the European core, which were most instrumental in making possible the development of labour productivity in agriculture—indeed, a tendency to the continuing development of the productive forces. This was essential to break through the age-old barrier to economic development represented by declining productivity in basic food production, and correlatively to underwrite the growth of production in manufacturing—through ‘freeing’ the labour force from labour for subsistence, and through providing a market for manufacturing products. It was this development which made possible a qualitative breakthrough in the development of the division of labour.
Wallerstein cannot put forward the foregoing viewpoint because, as has been noted, his entire theoretical edifice is designed to buttress a fundamental conclusion: that capitalist development and underdevelopment are the opposite sides of the same coin. The logic of this position, as Wallerstein is fully aware, is that capitalist underdevelopment is as much the cause of capitalist development, as capitalist development is the cause of capitalist underdevelopment. Such an argument is not compatible with the view of capitalist economic development as a function of the tendency toward capital accumulation via innovation, built into a historically developed structure of class relations of free wage labour. For from this vantage point, neither economic development nor underdevelopment are directly dependent upon, caused by, one another. Each is the product of a specific evolution of class relations, in part determined historically ‘outside’ capitalism, in relationship with non-capitalist modes. To take the view that development and underdevelopment are indeed directly mutually determining, Wallerstein resorts to the position that both development in the core and underdevelopment in the periphery are essentially the result of a process of transfer of surplus from periphery to core. He must thus end up by essentially ignoring any inherent tendency of capitalism to develop the productive forces through the accumulation of capital, in favour of a view which sees such development in the core as a result of a ‘primitive accumulation of capital’ extracted from the periphery, and which sees underdevelopment as a result of ‘lack of capital’. Capitalism thus appears to be essentially one more system based primarily on the extraction of what we have called absolute surplus labour.
The ‘Transfer of Surplus’
Wallerstein seems to have two modes of explaining the putative transfer of surplus from core to the periphery: one directly ‘economic’, the other ‘political’. Thus, he states: ‘The division of the world-economy involves a hierarchy of occupational tasks, in which tasks requiring higher levels of skill and greater capitalization are reserved for higher ranking areas. Since a capitalist world-economy essentially rewards accumulated capital, including human capital, at a higher rate than “raw” labor power, the geographical maldistribution of these occupational skills involves a strong trend toward self-maintenance. The forces of the marketplace reinforce them rather than undermine them’ (mws, p. 350). At the same time, Wallerstein argues that the system of labour control/rewards to labour gives rise to strong states in the core and weak ones in the periphery. As a consequence, the strong states are able to assure, ultimately by force it appears, an unequal economic relationship between the core economies and those of the periphery. ‘In [the core] states, the creation of a strong state machinery . . . serves . . . as a mechanism to protect disparities that have arisen within the world system’ (mws, p. 349). ‘Once we get a difference in the strength of the state-machineries, we get the operation of “unequal exchange” which is enforced by strong states on weak ones, by core states on peripheral areas. Thus [agricultural] capitalism [of the early modern period] involves not only appropriation of the surplus-value by an owner from a laborer, but an appropriation of surplus of the whole world-economy by core areas’ (rfd, p. 401).
Neither of these arguments is perfectly clear. But both seem rooted in the same sort of economic determinism that pervades all aspects of Wallerstein’s theoretical framework. As to the economic argument, the first question which must be asked is what determines Wallerstein’s ‘hierarchy of tasks’, such that some productive tasks are carried on with more capital and skilled labour than others. Wallerstein goes far in the direction of arguing that it is actually the tasks themselves which determine the amount of capital and skill which is used to carry them out. Thus, he states: ‘Given the great expansion of the geographic and demographic scope to world commerce and industry, some areas of Europe could amass the profits of this expansion all the more if they could specialize in the activities essential to reaping the profit. They thus had to spend less of their time, manpower, land, and other natural resources on sustaining themselves in basic necessities. Either Eastern Europe would become the “breadbasket” of Western Europe or vice versa’ (mws, pp. 98–9). According to this account, food production apparently necessitated less capital and skill, hence less of a share in the total surplus, than manufacturing. Wallerstein can thus conclude that the ‘development of underdevelopment’ is the ‘result of being involved in the world economy as a peripheral raw-material-producing area’ (rfd, p. 392).
It is obvious that such a techno-determinism would fit perfectly with Wallerstein’s argument for the dominating position of the world division of labour in the development of capitalism: the assignment to a raw-material-producing role means the assignment to a role producing with less capital and skilled labour, and thus to an unfavourable position vis-`-vis the world’s market system for distributing surplus. Yet, equally clearly, such a logic cannot work. It is not necessary to resort to such modern comparisons as that of American export agriculture with third-world export manufacturing, the former often using far greater amounts of capital and skill than the latter. For during the early modern era itself, by Wallerstein’s own testimony, it was not just core manufacturing which was more capital-and skill-intensive than peripheral agriculture; all core productive activities had those qualities. And this includes, as Wallerstein fails to emphasize, basic food production, where for the same products English agriculturalists applied far more capital and skill (to much greater effect) than did their Polish counterparts. 
Clearly, the product itself could not determine the skill and capital used to produce it. Yet, on the other hand, if Wallerstein is contending that it is the fact of presence in the core which itself determines superior equipment by capital and skill in all productive lines, he must explain why this should be so, especially in order to avoid the tautologous conclusion that what determines a region’s place (core or periphery) is the capital and skill applied to the productive tasks there. As we have seen already, the world market cannot determine the type of production carried out in any area, especially the level and character of the productive forces applied, except insofar as its impact is, in turn, determined by the region’s class structure. On the other hand, it would contradict Wallerstein’s whole line of reasoning to contend that indeed it was not a region’s position in the world market that determined the level of its productive forces; but that it was the level of development of its productive forces—labour productivity—which determined its place in the world market. This must, in turn, be referred back to the class structure in which those forces of production do or do not, could or could not, be developed.
At the same time, it should be clear that whatever the cause of the distribution of capital and skill applied in production—its concentration in the ‘core’—this distribution in itself cannot be assumed to determine a process of surplus transfer in the early modern period. Wallerstein specifically invokes Arghiri Emmanuel’s conception of ‘unequal exchange’ to account for such a process. Yet whatever one thinks of this notion, it will certainly not apply to Wallerstein’s purposes. Unequal exchange, in Emmanuel’s view, occurs when production for exchange is carried out by different regions (lines of production) with different wage rates (and equal organic compositions of capital). Given the premise that there is equalization of profit rates in all regions and lines of production—as is indeed the tendency under capitalism—part of the surplus appears to be lost from the lower wage region to the higher wage region.  Yet this theory, whatever its validity, explicitly requires in order to be applicable the free mobility of capital throughout the system, precisely so as to equalize profit rates—to make the premise of profit equalization a realistic one. However, as we have noted in discussing Sweezy, there was no free labour (nor free land) under the serf mode of production, so there could be no ‘free capital’. The ‘use-values’ in land, labour and capital were not free to be combined at the ‘average rate of profit’. Put another way, it was simply impossible to ensure a flow of investment in order to equalize profit rates—either into the serf-dominated regions if profitability was higher there than elsewhere, or out of those regions if it was lower. Indeed, precisely because there could be no such mobility of investment to increase output, the general outcome of growing demand for East European products produced under serfdom in the seventeenth century was simply a rise in their price (rather than their output), so that the relative prices of eastern agriculture and western industry shifted in favour of the former. As a result, the market did facilitate a certain ‘transfer of surplus’, but from the western ‘core’ to the eastern ‘periphery’, rather than vice versa.
The Determination of State Structures
What then of Wallerstein’s notion that surplus transfer was assured politically, by the strong states of the core against the weak ones of the periphery? Again, it is necessary to back up one step to a prior question: that of the distribution of strong and weak states themselves. Wallerstein says: ‘In the sixteenth century, some monarchs achieved great strength . . . Others failed. This is closely related . . . to the role of the area in the division of labor within the world economy. The different roles led to different class structures which led to different politics’ (mws, p. 157). Yet again, Wallerstein sets up a strictly economically determined structure, which breaks down over his contradictory conception of systems of labour control/rewards to labour. On the one hand, Wallerstein wishes to see class structures (systems of labour control/reward to labour) as determined by the world economy, by the organization of work in a particular line of production, in the last analysis by a region’s role in the division of labour—i.e. as a product of market-determined technical-economic exigencies. On the other hand, it is also class structure, now considered as a relationship of ruling-class exploiters to labouring exploited, which, for Wallerstein, in turn determines the character of the state: ‘the modes of labor control greatly affect the political system (in particular the strength of the state apparatus), and the possibilities for an indigenous bourgeoisie to thrive’ (mws, p. 87).
As I have argued, however, to view the labour control system as a class structure of exploitation precludes its being conceptualized in essentially technical-functional terms. Since, from this viewpoint, the region’s class structure conditions the very development of its productive forces, and thus by extension its role in the world economy, the region’s class structure, in determining the structure of the state, cannot be viewed as merely ‘transmitting’ the pressures of the world market and division of labour. Wallerstein cannot have it both ways: a labour control system as a class structure of exploitation which determines the character of the state, and a labour control system as reflecting the most suitable ‘productive’ technique for a given region in the world economy. His attempting to do so leads him to explicitly contradict his central contention that the state structure of a region is determined by that region’s place in the world division of labour. Thus at one point Wallerstein explains Japan’s and Russia’s unusual economic success, their ability to enter the world economy in semi-peripheral rather than peripheral status, as a result of their strong state structures.  Yet if so, does not the state determine the region’s economic role, rather than vice versa? On the other hand, France was by any reckoning an unusually strong state in the seventeenth century, yet it did not reside in the core.
Although Wallerstein appears to be on the right track in seeking to understand the relationship of the state to economic development—the development of the productive forces—in terms of the class structure of surplus extraction, he cannot come to a fruitful conclusion because, for him, class structure remains in the last analysis a mere adjunct to the developing productive forces, tied to the world division of labour. At the same time, Wallerstein’s understanding of state structure as economically determined via the world division of labour is closely bound up with his view of the function of the state system as primarily to enforce the transfer of surplus from periphery to core. The resulting quantitative conceptualization of states, in terms of their ‘strength’ or ‘weakness’, itself precludes any sensible analysis in terms of the structure of class.
If, in contrast, one attempts to view the state in relationship to class structure, understood as the social relationships by which an unpaid-for surplus is extracted by a ruling class from the direct producers, it is possible to get beyond largely unhelpful characterizations of states in terms of their quantitative strength or weakness, to a qualitative characterization in terms of their differing relationships to the development of the productive forces—a relationship precisely mediated by their differing relationships to the established systems of surplus extraction. In Marx’s words, ‘The specific economic form in which unpaid surplus labour is pumped out of direct producers, determines the relationship of rulers to ruled . . . It is always the direct relationship of the owners of the conditions of production to the direct producers—a relation always naturally corresponding to a definite stage in the development of the methods of labour and thereby its social productivity—which reveals the innermost secret, the hidden basis of the entire social structure, and with it the political form of the relation of sovereignty and dependence, in short the corresponding specific forms of the state.’ 
Weak States and Underdevelopment
From this point of view, the relationship of the state to economic development or underdevelopment in a given region—and ultimately to the world economy—follows the ‘logic’, as it were, of the surplus extraction relationship itself: of the needs of the ruling class, as conditioned by the character of their exploitative relationship with the direct producers. Thus, for example, Wallerstein is quite right to understand the development of the Polish state, in the first instance, in relationship to serfdom. But serfdom in Poland does not determine a state which is appropriately characterized as either strong or weak. Serfdom as a mode of production, as a system of surplus extraction, depended upon—was defined by—the forceful control exerted by the lords over the peasant tenants’ movements to ensure rent, especially through preventing the rise of a market in tenants or labourers. In the case of Poland, the instruments of force, law and administration which guaranteed the surplus extraction relationship of serfdom were held directly in the localities by the serf-lords themselves.
The initial result was a state which was largely decentralized in character. Yet, its relationship to the development of the productive forces, the economy internally or externally, cannot be usefully understood in terms of ‘weakness’. On the contrary, the stance of the state vis-`-vis the economy, both internal and external, ‘strongly’ served to enhance the class position of the serf-lord. Thus, on the one hand, the state’s opposition to the towns did not result from any direct opposition between merchants and nobles. It arose from the danger the towns posed to the lord-serf relationship, by providing places to which the serfs might flee. On the other hand, the policy of free trade pursued by the state did not represent any sort of surrender to the core states of the West. Indeed, the United Provinces (Holland), from whence came the bulk of imports into Poland, could hardly be characterized as strong. The point of a free-trade policy was to serve the interests of the serf-lords, directly through providing cheap industrial goods, and indirectly by undercutting local industry, which through offering an alternative for the peasants might have undermined the serf relationship. 
The state’s relationship to underdevelopment in Poland was not determined by its weakness vis-`-vis core states, but precisely by the degree of its strength in enforcing policies which tended to strengthen the serf mode of production. It was because the policies of the state enhanced the landlords’ ability to ‘maximize surplus’ within the given social relationships of serfdom (which themselves tended to undermine the basic productive forces in agriculture) that they simultaneously tended to minimize the chances for the economic development of Poland. In turn, the source of the state’s policy directions was obviously to be found in the overwhelmingly predominant position of the serf-lords, determined by their direct and immediate control over the direct producers and their surplus. The fact was that the Prussian state maintained through the seventeenth and early eighteenth centuries essentially similar policies, and a similar pattern of international economic relations, despite the fact that it was by most reckonings a ‘strong state’. This was precisely because it remained largely bound up within the same type of structure of serf-lord relations as obtained in Poland.  In short, the state’s relationship to development or underdevelopment in Eastern Europe, or other regions, is not most satisfactorily grasped in terms of its strength or weakness, because its contribution to the growth or stagnation of the productive forces is not primarily mediated by its relationship to ‘unequal exchange’ and a transfer of surplus (into or out of the region)—but rather by its interconnection with a region’s dominant class structures.
The ‘Primitive Accumulation of Capital’
In the final analysis, however, the whole discussion of unequal exchange leading to the transfer of surplus must be assigned a subordinate place in relationship to the question of the rise of development and underdevelopment. The argument that unequal exchange and the transfer of surplus are central is largely derived from the notion, widespread among Marxists, that a ‘primitive accumulation of capital’ was largely responsible for the uniquely successful development experienced by certain areas within the Western European core from the sixteenth century, as well as for the onset of underdevelopment in the periphery. But those who argue in this way miss the point. The notion of a ‘previous accumulation of capital’ was originally Adam Smith’s. Marx raised the notion in order to criticize it and transform it fundamentally, so as to get beyond the circular conceptions of economic development with which it was inevitably connected. As Marx posed the problem, ‘the accumulation of capital presupposes surplus-value; surplus-value presupposes capitalist production; capitalist production presupposes the availability of considerable masses of capital and labour-power in the hands of commodity producers. The whole movement, therefore, seems to turn around in a never-ending circle, which we can only get out of by assuming a primitive accumulation (the ‘previous accumulation’ of Adam Smith) which precedes capitalist accumulation; an accumulation which is not the result of the capitalist mode of production but its point of departure.’ 
Marx’s intent, therefore, was not merely to criticize the ‘just so story’ by which Smith himself explained so-called previous accumulation, but to reject Smith’s whole notion as fundamentally misconceived. No amount of accumulated money or wealth can explain the accumulation of capital, for this requires certain historically-developed social-productive relations. As Marx puts it, ‘There can therefore be nothing more ridiculous than to conceive this original formation of capital as if capital had stockpiled and created the objective conditions of production—necessaries, raw materials, instruments—and then offered them to the worker, who was bare of these possessions.’ (Marx’s emphasis).  At the same time, ‘In themselves, money and commodities are no more capital than the means of production and subsistence are. They need to be transformed into capital . . . So-called primitive accumulation, therefore, is nothing else than the historical process of divorcing the producer from the means of production.’ 
Those Marxists who, like Wallerstein, stress the significance of an original amassing of wealth in either money or natural forms often tend to beg the fundamental questions. In the first place, they do not say why such a build-up of wealth ‘from the outside’—from the periphery to the core—was necessary for further economic advance at the time of the origins of capitalism. Were there, for example, some sort of technological blocks requiring an immense concentration of capital to overcome: blockages which demanded even more resources than could be brought together from within the core?  Even more importantly, what allowed for, and ensured, that wealth brought into the core from the periphery would be used for productive rather than non-productive purposes? In particular, what determined that this would be used for the development of the productive forces, so as to increase the productivity of labour? And how was it connected with a continuing process of accumulation via innovation? Historically, the build-up of wealth, and its concentration in the hands of specific potential ‘investors’, has occurred time and again without discernable effect. It is only a system which is organized so that the accumulation of capital via innovation is enforced by the very structure of the social productive relations that can turn an accrual of potentially productive resources from outside to the service of economic development. In Wallerstein’s world system, no such enforcement mechanism is specified, precisely because the class-structured system of accumulation of capital based on free labour, where labour power is a commodity, is ruled out from the start. We are left to wonder why any wealth transferred from the core to the periphery did not result merely in the creation of cathedrals in the core and starvation in the periphery.
V The Class Structure of Economic Development and Underdevelopment
Neither development in the core nor underdevelopment in the periphery was determined by surplus transfer. Economic development was a qualitative process, which did not merely involve an accumulation of wealth in general, but was centrally focused on the development of the productivity of labour of the direct producers of the means of production and means of subsistence. This development of labour productivity, most significantly in agriculture, which occurred in parts of Western Europe in the early modern period, was dependent in turn upon the emergence of a social system which tended not only to equip the direct producers with capital and skill at the highest level of existing technique, but possessed the capacity to continue to do so on an increasing scale. In short, the uniquely successful development of capitalism in Western Europe was determined by a class system, a property system, a system of surplus extraction, in which the methods the extractors were obliged to use to increase their surplus corresponded to an unprecedented, though enormously imperfect, degree to the needs of development of the productive forces. Capitalism was therefore distinguished from pre-capitalist modes of production in requiring those who controlled production to continue to increase their ‘profits’ (surplus) largely by increasing what we have termed relative, not merely absolute, surplus labour. To account for capitalist economic development is, therefore, at least to explain the basis for this conjunction between the requirements for surplus extraction and the needs of the developing productive forces: on the one hand, its structure, or the reasons it held true; on the other hand, its origins, or how it came into being. It is a fundamental weakness of Wallerstein’s analysis that it never forces these questions to be directly posed.
East European Serfdom and Underdevelopment
Wallerstein, as we have seen, contends that the serfs in Poland and the other grain-exporting areas of north-eastern Europe were ‘working for [capitalist] landlords who paid them for cash crop production, a relation in which labor power is a commodity’ (rfd p. 400). It was thus, for Wallerstein, the rise of agricultural capitalism in East Europe based on ‘coerced cash crop labor’ (i.e., serfdom)—part and parcel of the incorporation of this region into the world market—which made possible the parallel transformation to a manufacturing capitalism based on wage labour in the West. In fact, however, it was precisely the non-capitalist organization of these East European economies which determined that they would not respond in a ‘capitalist’ manner to market opportunities; correlatively, that they would not be subjected to capitalist-type penalties for failure to do so; and consequently, that they would not be superseded by ‘more suitable’ forms of labour control/reward to labour, despite their inefficiency. Precisely because labour-power as a commodity did not obtain in these economies, they could not serve as a satisfactory basis for the development of capitalism in the West: for the emergence of a system characterized by a dynamic of accumulation and innovation.
As Wallerstein points out, chronic food shortages had been endemic to mediaeval Europe under the régime of serfdom. Indeed, the general tendency of the European serf-lords to resort to labour-squeezing approaches to surplus maximization had given rise to long-term diminishing returns in agriculture—leading to a ‘general crisis’ of productivity, of production, and ultimately of population throughout most of Europe at various points during the fourteenth and early fifteenth centuries. Yet, Wallerstein notwithstanding, the serf-bound economy of Eastern Europe which responded to a growing West European market for food in the sixteenth and seventeenth centuries could not solve this chronic difficulty during the early modern period. The fact is that the serf-economy could at very best respond sluggishly to demand, even the very intense pressures communicated from Western Europe during the early modern period. In the short term, output was simply never adjusted to meet changing prices; rather, prices fluctuated as a result of varying supplies due to very uneven harvests.  In the long term, in the face of prices which continued to rise precipitately throughout the sixteenth century and into the latter part of the seventeenth, Polish national marketable grain surpluses actually declined as a result of declining productivity. At the same time, total national output increased only weakly during the second half of the sixteenth century; it actually ceased to grow from the end of that century, at the precise moment that prices were making their most rapid ascent; and it may have begun to decline during the first half of the seventeenth century. 
It is true that Polish exports rose more impressively than output in this period, although still nowhere near as much as grain prices.  Yet this was made possible only through drastically cutting into peasant subsistence—a method which obviously had definite limits, and was bound in the long run to rebound against the lords themselves. It was feasible to increase ‘absolute surplus labour’ by increasing the time spent by peasants working on the lord’s estate and by cutting down the size of the peasants’ plots. Yet each of these methods tended to undermine the peasants’ ability to produce their means of subsistence and means of production. Since the peasants were responsible for supplying both the labour power and the means of production to cultivate not only their own plots but the lords’, surplus-squeezing methods meant in the long run a decline of productivity on the lords’ own estates. Since transport costs always limited the land area which could potentially be subject to cultivation for purposes of export onto the world market, the growth of Polish grain output had to slow down. 
The irony of this situation is that throughout this period, serf peasant plots maintained significantly higher productivity than did the lords’ demesnes. Indeed, so long as they reached a certain size, they could and did market a larger product per acre than did the lords’ demesnes. Peasant production, despite the limitation of serfdom, remained more productive than did that of the lords. In this light, the labour of the ‘coerced cash crop’ system could hardly be considered the ‘most suitable’ method of producing for the world market—for underwriting development elsewhere—as Wallerstein contends. Its weakness as a system of production for the world market is indicated by the fact that, despite the orientation of the entire economy to exports, it could send out at best 5 per cent to 7 per cent of its total grain product.  Of course, this is to be expected, for as one author succinctly concludes, ‘the main reason for the emergence of this system [of demesne serfdom] was not that it produced more food, but that it enabled the ruling class to increase its revenue.’  Indeed, the inferior ability of the ‘coerced cash crop’ system to develop production became especially evident over the long run. Thus the areas around Danzig and in Silesia, where serfdom existed only in an attenuated form if at all, were the last regions to experience the general trend to declining productivity and general crisis, if they did so at all. 
The wars which totally disrupted the sputtering Polish economy from the middle of the seventeenth century did not allow for the foregoing dynamic of retrogression to fully work itself out. But the eighteenth-century picture tells the story. Long after the recovery from military destruction, Poland’s serf economy had been unable to overcome its inability to develop the forces of production. As a result, levels of productivity had sunk to new lows. Meanwhile, the rise of a capitalist system of agriculture in the West, to which I shall shortly refer, meant the emergence of severe competition on the world markets for grain. Polish grain exports in the later eighteenth century had shrunk to less than half of what they had been in the sixteenth century.  Yet despite Wallerstein’s claim that the capitalist character of the serf or coerced cash crop method of labour control meant that it would, like all other capitalist methods of labour, ‘expand or diminish in function of its profitability’, serfdom remained firmly entrenched in Poland—in spite of the obviously precipitate decline in returns, absolutely and relatively to others. The drop in profitability did not determine the replacement of the coerced cash crop method by more suitable methods of labour control which could more effectively compete on the world market. Serfdom remained as incapable as ever of developing more efficient means of production, but showed no signs of being replaced by competitors.
Serfdom and the World Market
But this is hardly surprising, once we cease to look at serfdom as a method of labour control chosen by capitalists—who from this viewpoint could easily choose another method if and when it became more competitive and more profitable than serfdom—and see it as expressing a historically evolved system of class relations. Since the serf-lords had direct (non-market) access to their own means of subsistence (serf-peasant output from their demesnes), they did not have to buy on the market necessities for reproduction; their ability to survive, to reproduce, was independent of their ability to ‘hold their place on the market’. Indeed, from the point of view of the Polish economy, the goods imported from the West (almost exclusively by the Polish lords) were in the fullest sense luxuries. The Polish producers might not be able to hold their own in the world market, and consequently might be less able to buy luxury imports, but this would not ‘put them out of business’. As a result, no entrepreneurs, either from inside or outside the system, no matter how great the superiority of the productive methods they could potentially put into play, could replace the serf lords through competition. Far from being capitalist, Poland remained impervious to capitalist development. Its serf class structure ensured that existing means of production—land, labour power and means of production—remained locked away from potential capital accumulation. Precisely because free wage labour had failed to emerge, neither labour power nor means of production had emerged as commodities; as a result, these use values simply could not constitute a field for capitalist investment and development, because they were not, could not be, combined as exchange values under the sway of capital for production at the socially necessary rate. They were already merged by means of a class structure which fused the direct producers with the means of production and subsistence.
In sum, the growing connection of Poland with the world market—the growing impact of trade—did, in accord with Marx’s generalization, ‘facilitate the production of surplus destined for exchange in order to increase the enjoyments, or wealth of the producers (here meant are owners of the products)’. On the other hand, as Marx also theorized, growing production for exchange was ‘incapable by itself of promoting and explaining the transition from one mode to another’. It was this two-sided ‘dynamic’, under the pressure of trade, which constituted at the broadest level the ‘structure of underdevelopment’ in early modern Poland and elsewhere: the growth of surplus extraction in response to the market, without the transformation of the mode of production which was required to make possible the development of the productive forces so as to increase the productivity of labour. This determines, at the most general level, the intensified use—indeed ‘using up’—of labour power, as well as of natural resources, but without an offsetting acceleration of the social productive forces which could make for a correspondingly (or more greatly) increased productiveness of labour power and increasingly effective use of natural resources. Specifically then, the class structure of serfdom in Poland determined underdevelopment by stifling the productivity of the direct peasant producers, thereby undercutting the home market for means of production and means of subsistence, and at the same time determining that what market there was would be largely in luxuries. To the degree, therefore, that Poland was ‘incorporated’ within the world market, its economy was increasingly strangled; to the degree to which trade (later) declined, the Polish economy stagnated. However, neither of these trends was determined by the rise of trade and the world market, but fundamentally by a class structure of serfdom which precluded the emergence of an ‘internal’ dynamic of development, while ensuring that any commercially induced dynamic from ‘outside’ would ultimately lead to retrogression.
Correlatively, the fact that Poland was increasingly connected up to the world market under the serf economy rendered it decreasingly able to contribute to the development of the ‘capitalist’ world economy. Since Poland’s economy could not develop its output via the growth of the productive forces in response to Western demand, the Western European economy’s industrial development was limited, to the extent it depended upon Polish food exports. On the other hand, since the Western economies were producing primarily luxury products for Poland’s use, they were not contributing goods which could feed back into the productive process in Poland so as to better equip Poland’s agricultural producers to respond to Western needs. This paradoxical ‘dependence’ of the Western European economies on serf-bound Eastern Europe, and correlative ‘independence’ of the latter from the former, was expressed in the long-term decay of the terms of trade of Western manufacturing commodities against Polish agricultural commodities. This decay reflected most of all the relatively far greater inflexibility of Polish agricultural supply in comparison to that of Western supply of manufacturers (as well as such other commodities as wine, sugar and spices), but also the comparatively far lesser significance of the demand for imports from the West (primarily luxuries) by the Polish economy than the demand for Polish agricultural output (necessities) by the Western economy.  Because core manufacturing production for the periphery was still largely in luxuries, while peripheral raw material production for the core in necessities was based on pre-capitalist social-productive relations, the economic relationship between early modern eastern and western Europe was quite analogous to that between country and town in feudal mediaeval Europe. The traditional character of this interdependence was hardly symptomatic of the rise of a capitalist world economy.
It is a central contention of Wallerstein’s that the ‘combination’ of free labour in the core countries with coerced labour in the periphery is ‘the essence of capitalism’ (mws, p. 127)—and Wallerstein demands that we examine the ‘relations of production of the whole system’, i.e. the European world economy, rather than merely the relations of production of each local economy. But surely to the extent that the early modern ‘European world economy’ actually met Wallerstein’s specification—to the extent, that is, that it was defined by the interconnected systems of production based on coerced cash crop labour in the periphery and based on free labour in the core—it remained fundamentally ‘pre-capitalist:’ a sort of renewed feudalism, with a somewhat wider scope. The lack of a real breakthrough was indeed reflected in the inability of the ‘modern world economy’ to provide the material underpinnings for continuing economic-industrial growth in most of Europe  through the early modern period decisively better than had the serf-based economy of Europe during the mediaeval period. Specifically, much as the mediaeval economy was stricken by general crisis of production leading to demographic crisis in the fourteenth century, so the early modern economy was struck again by general productivity crisis leading to demographic difficulties in the seventeenth century.  Like its predecessor, the ‘general crisis of the seventeenth century’ was a crisis of actual scarcity, typical of pre-capitalist modes of production and reflecting directly their inability to develop the productive forces. It was not a crisis indicative of the dominance of capitalism, exemplified by ‘poverty amidst plenty’ and manifesting precisely development of the productive forces in the framework of capitalist social-productive relations.
French Peasant Freeholding and Retarded Development
If, then, it is necessary to recognize that the serf-based agriculture of Eastern Europe could not prevent the ‘general crisis of the seventeenth century’, it also must be emphasized that the economy which emerged throughout most of Western Europe following the fall of serfdom was itself similarly unable to break out of the age-old cycle of declining agricultural productivity, despite the substantial penetration of commerce. This is because, through much of Western Europe, notably much of France, serfdom had been succeeded not by capitalism, but by an economy dominated by what were essentially peasant freeholders. Petty property in land set definite limits to the development of production, especially in basic food products—which posed the fundamental barrier to long-term economic development. At the same time, like serf property, peasant property was not subject in any simple and direct way to be easily superseded by other potentially more productive class organizations, even under market pressure. In result, the potential for development via the town/country, manufacturing/agricultural division of labour, based on the growth of labour productivity especially in the countryside, was severely restricted.
Thus, on the one hand, the small size of the peasant holding set definite limits to the advance of production. The key advances connected with the agricultural revolution in basic grain production required, as we shall see in a moment, large-scale consolidated holdings. The fact that peasant property in France tended to be divided at inheritance only exacerbated the problem. At the same time, although the freeholding peasantry paid only a fixed rent (usually rendered nominal after centuries of inflation), they were increasingly subject to heavy exactions in the form of taxation by the absolutist state. The fact that state taxation took such a large share of the agricultural surplus—surplus which was not ploughed back into agricultural production, but was spent unproductively, largely on the financing of war—meant in effect a surplus-extracting structure which was in some ways as retrogressive as serfdom. 
On the other hand, although the peasant food producer might sell part of his output on the market, the reproduction of his means of production and the maintenance of his plot did not depend on it. In the first instance, his plot provided the means of subsistence to survive as a peasant holder. In other words the peasant did not have to produce at a competitive level to survive, because he did not have to hold his place on the market; because he did not have to sell in order to reproduce. More efficient producers could not undermine and replace the peasant holding through competition. Because the direct producer was not separated but attached to the means of production and subsistence, great masses of use values in the form of labour power, land and means of production were thus kept outside the field of capital investment—in spite of the opportunity for profit offered by a growing market, signalled by increasing food prices, through the early modern period. Potential land accumulators with the means of production to improve were to an important extent stymied, or at least slowed down.
This is not to say that peasant proprietorship was invulnerable, but merely to emphasize that the ‘rise of trade’ could promote a process leading to its undermining and supersession only with the greatest difficulty, and over a long period of time.  It is true that with the development of the market, the appearance of new commodities might ‘create new needs’, inducing the peasant to attempt to increase output. Yet there was a significant counter-tendency: the peasants’ predilection to diversify their own production—rather than specialize—in order to produce as many as possible of their necessities on their own plot, precisely to avoid market dependence. Within such a structure, they might significantly increase output for market purchases merely by extending and intensifying their labour, thus applying to themselves the methods of extracting ‘absolute surplus labour’.  At the same time, the processes by which new consumer goods might actually become ‘necessities’, may tend to be long and drawn out. Indeed, it is only when peasants come to have recourse to the market for improved tools, that they become especially vulnerable to competition. As Rosa Luxemburg thus concluded, the peasants’ control over agricultural production, ‘would not immediately and, under European conditions of ownership, only with great difficulty, submit to capitalist domination’. 
To undermine peasant property, it was necessary to deprive the peasantry of their ability to produce their own subsistence, and the mechanisms by which this seems to have been accomplished best in early modern France were generally ‘extra-economic’:  on the one hand, the subdivisions of holdings at inheritance which might push the size of the plot below the minimum necessary for subsistence, leaving the peasant vulnerable to be squeezed out through debts; on the other hand, the increase in taxation which might directly force the peasantry below subsistence. Yet neither of these processes led easily or directly to the supersession of peasant production by capitalist property relations and the rise of new methods of production. On the one hand, in the face of a sea of petty proprietors, it was difficult for potential accumulators who might succeed in taking over peasant plots to actually bring them together physically, so as to form large consolidated holdings suitable for advancing the productive forces. On the other hand, since the general system of peasant proprietorship created a permanent demand for land to be used for subsistence purposes, land values tended to be pushed up; and there was every incentive for the non-peasant owner either to resell land or to lease it, profiting from ‘squeezing’ rather than improvement.  Indeed, even where large plots could be constructed, there was relatively little agricultural progress, as exploiters (be they owners or tenants) tended to employ labour-intensive techniques, so as to take advantage of the mass of cheap labour available from the massively dominant subsistence (or sub-subsistence) peasant sector. Certainly, despite the high prices and development of trade which obtained in the early modern period, significantly widespread agricultural improvements did not occur in time to prevent the general crisis of agricultural productivity, demography and industrial production which gripped France during the seventeenth century.
England and the Rise of Agrarian Capitalism
In light of the barriers presented to the development of the productive forces by both serfdom and entrenched petty proprietorship, it is especially striking that the only areas of Europe which appear to have in the main escaped the general crisis of the seventeenth century had also, during the early modern period, experienced the rise of capitalist class relations of production in agriculture. Thus we have, classically in England, the rise of that ‘three-tiered’ relation of landlord/capitalist tenant/free wage labourer, around which Marx developed much of his theory of capitalist development in Capital. On the one hand, this capitalist agrarian structure made possible, to an unprecedented extent, the accumulation of capital especially through innovation in agriculture. On the other hand, the same structure made such productive investment ‘necessary’, at least in tendency.
So, in the first place, the landlords had been able to gain control over large consolidated blocks of land. This was a result not only of the decline of serfdom, but of the general short-circuiting of the emergence of small peasant proprietorship in the land—a process to which I shall return below. Large farms appear to have made possible the introduction of new techniques—notably up-and-down husbandry and various systems of irrigation—which transformed agricultural production. Those techniques appear to have been far more adaptable to large-scale production requiring large holdings, than to peasant agriculture. 
At the same time, and perhaps equally important, the organization of production on the basis of the emergence of labour power as a commodity, in particular via the separation of the direct producers from the land, tended to impel a process of capital accumulation in agriculture founded in the capitalist relations themselves—i.e. an agrarian economy based on free tenants and free wage labourers. In particular, the tenant farmers who took over as capitalists in the countryside—as leaseholders—had to specialize and to introduce new methods in order to be able to sell their products at the market price, a requirement for survival. For if the tenants could not produce at the market price, either they would see their profits decline or they would be unable to pay the rent.
In short—in marked contrast with the peasant proprietor—if the capitalist tenant had higher than average costs, the result of his using backward methods of production, he would be subject to a two-sided pressure. On the one hand, if he tried to pay the going rent, his profit rate would decline further below the average, his funds for accumulation would therefore fall off, his potential for further investment would be cut back, and his position on the market would be even further undermined. On the other hand, if he tried to pay a lower rent, he would be subject to eviction by the landlord, who could then seek a new tenant more able to make the necessary improvements to compete on the market.
In fact, the pressures on the capitalist tenant to make the average rate of profit, so as to be able to pay the going rent, put the landlord often in the position where it was to his own interest to take over, in part, the role of capitalist investor—providing especially the investments on farm infrastructure, such as enclosure, farm buildings, etc. (and thus receiving himself an increment over and above the rent in the form of profit from the tenant). In this way, there emerged something like a symbiotic relationship between capitalists and landlords, which was in fact reflected in various rental or leasehoalding arrangements. These in particular secured the tenants’ capital investments, protecting them from confiscation by the landlord via increasing rents. 
Given this context, the ‘world market’ is unquestionably quite significant. For it was clearly the European demand for English cloth exports which exerted the original pressure for the development of English cloth manufacturing, especially from the later fifteenth century; this, in turn, created demand for agricultural products that induced the English landlords and their capitalist tenants to consolidate holdings and to improve. The indispensable contribution of the world market was thus, typically, to provide the concentrated and continuous demand necessary to induce a transformation toward the application of fixed capital and cooperative labour (especially in agriculture), in the presence of already favourable social-productive or class relations, marked by the dissolution both of serfdom and entrenched peasant property. The point is that industrial export booms had been a common occurrence in mediaeval Europe, and continued to take place elsewhere in Europe throughout the early modern period. What distinguished the English development from those in other places was the continuity of industrial growth throughout the period, in the face of stagnating, even declining overseas markets—and in this respect the agrarian structure was pivotal.
Thus one witnesses in England, by the latter part of the seventeenth century, the emergence of a highly flexible system of agriculture which not only could respond to market pressures through specialization and improving the output, but which in certain respects had to do so—given the capitalist tenant’s need to sell on the market in order to reproduce and hold on to his lease. The result was a system which supported a continuously increasing agricultural productivity which, as a result, provided for a symbiotic relationship between agriculture and industry; each provided a market and ever cheaper means of production and consumption for the other. On the continent, by contrast, agriculture and industry were set in conflict with one another by the prevailing class systems of production based on serfdom or peasant proprietorship. There, market pressure in favour of agriculture simply drove up food prices, because the class structures of production set up barriers to increasing output via accumulation and innovation. Higher prices determined that a greater share of the society’s total productive powers would go to immediate subsistence, and thus correlatively undermined the society’s ability to support manufacturing.
The English situation was just the opposite. With developing agricultural productivity, England was able to support a much increased section of the population off the land by the end of the seventeenth century, and especially in manufacturing. At the same time, agricultural demand made possible the emergence of a growing home market, not only for industrial goods and products for general consumption, but also for agricultural means of production. The latter, in turn, tended to further increase agricultural productivity. It was the strength of the home market which provided the indispensable basis for the development of the English economy through the period of the general crisis of the seventeenth century, when elsewhere industry was contracting. So while much of European export industry declined during the seventeenth century as a result of its dependence upon pre-capitalist agricultural ‘hinterlands’—a dependence which had only become more complex, but had not fundamentally changed—English industry continued to grow because of its construction on increasingly capitalist agricultural foundations. It was the same developing home market which provided, by the end of the eighteenth century, an indispensable foundation for the industrial revolution. 
VI. The Origins of Capitalism
In sum, the development of agrarian capitalism in England can be said to have resulted from the emergence of a historically-evolved correspondence between, on the one hand, the methods which the ruling class of landlords and tenants were required to use in order best to increase their surplus and, on the other, the requirements of the development of the productive forces. The methods applied by the ruling class—in particular the drive towards accumulation via innovation in agriculture—are thus incomprehensible simply as their own choice. These were given, as it were, by the class structure; by the system of surplus extraction relations with the direct producers in which the ruling class found itself. This system set up certain limits and thus foreclosed certain options for obtaining surplus; at the same time, it opened up certain promising possibilities. Thus, on the one hand, the ruling class could not increase its surplus, through simply forcefully squeezing the peasantry to thereby increase absolute surplus labour. This limitation had developed because the landlords had been unable to prevent the dissolution of serfdom due to peasant resistance in the later mediaeval period. The peasantry was no longer unfree, so that it could at will move, buy land, or lease it. On the other hand, the ruling class was well positioned to increase relative surplus labour dramatically, by bringing in new techniques on large consolidated farms. Indeed, as we have seen, the societal organization of production on the basis of individual producing units, with access only via the market to the means of subsistence and means of production, more or less forced the capitalists to accumulate via innovation—in order to be able to sell, to hold their place on the market, and thus to reproduce. This possibility (and pressure) for accumulation had emerged (following the dissolution of serfdom) as a result of the landlords’ success in gaining control of the land, specifically by preventing the peasants from doing so.
From this viewpoint, the origins of capitalist economic development, as it first occurred in England, are to be found in the specific historical processes by which, on the one hand, serfdom was dissolved (thus precluding forceful squeezing as the normal form of surplus extraction) and, on the other, peasant property was short-circuited or undermined (thus opening the way for the accumulation of land, labour and the means of production). Clearly, this two-sided development is inexplicable as the result of ruling-class policy or ruling-class intention, but was the outcome of processes of class formation, rooted in class conflict. Peasant resistance had broken serfdom in Western Europe, in spite of landlord attempts to maintain it. But in Eastern Europe, the landlords prevailed and prevented this outcome. Correlatively, the application of landlord power had forclosed the emergence of widespread peasant proprietorship following the downfall of serfdom in England. But elsewhere on the continent the peasantry succeeded in gaining the land. It is these contrasting outcomes of processes of class conflict—dependent in turn on contrasting evolutions of class society and disparate balances of class forces at different points in time—which are at the heart of the original transition from feudalism to capitalism, and which require to be understood if the onset of capitalist economic development is to be fully comprehended. 
In contrast, on the basis of Wallerstein’s method in The Modern World System, it becomes almost impossible actually to pose the problem of transition in a clear and consistent way. This is because Wallerstein persists in seeing the emergence of different, but in his view equally capitalist, systems of labour control as the product of ruling-class choices, under the incentives and sanctions of the world market. The first problem of such an approach to the origins of capitalism—the transition from feudalism to capitalism—is that it is wholly ‘unrealistic’: ruling classes were not simply free to choose the manner by which they could exploit the direct producers; not free to choose the so-called systems of labour control/reward to labour. Yet if one looks further, and accepts for a moment Wallerstein’s premise of ruling-class freedom of choice vis-`-vis systems of labour control/rewards, it can be seen that the whole problem of the origins of capitalism, conceived as a transition from some previous class-structured system of production, must actually be dissolved. For if, with the rise of trade and the process of incorporation of a region within the world division of labour, the ruling class can be said to choose the most suitable method of labour control in order to maximize its surplus, it has to be admitted that such a ruling class, whatever its relations to the exploited before the rise of the market, were already capitalists in potential. The previous mode of exploiter-exploited relations becomes irrelevant. For once the exploiters have access to the world market, they are free to invoke whatever system of labour control they please, and will do so subject to the demands of the world market itself and their ability to compete on it most effectively. In other words, the rise of trade determines the emergence of capitalists and capitalism; ‘transition’ becomes the result not the source of capitalism.
Yet if one is to reason consistently in this manner, one is immediately beset with very serious logical and historical problems. This is because it is still necessary to set out the processes through which those exploiters-become-capitalists actually implemented their choices, so as to transform their positions from pre-capitalist to capitalist exploiters. These questions become extremely pressing if one begins at any specific point in history before Wallerstein’s rise of the world market; in particular, if one takes as a starting-point the system of feudalism based on the surplus extraction relations of serfdom which were so widespread in mediaeval western Europe, and then tries to explain the passage to capitalism on the basis of Wallerstein’s ruling-class choices. In fact, it is impossible to account for this transition in these terms—and in attempting to do so Wallerstein is forced into a series of very questionable historical and logical arguments.
The Problem of the Starting-point
Thus if Wallerstein ‘begins’ with a society where serfdom was predominant, and if he wants to explain the transition to capitalism in Europe, he is faced right off with an enormous problem. For if one begins with serfdom, and the ultimate source of the rise of capitalism (the rise of free labour in the West/coerced crop labour in the East) is to be found in the ruling-class choice of methods of labour control/rewards to labour, one must argue that, at some point, some ruling class actually chose to abolish serfdom and perhaps replace it with wage labour. Wallerstein is unwilling to make this assertion, for it never happened.
On the other hand, if Wallerstein wishes to begin his explanation (of the transition to capitalism) following the fall of serfdom, he is faced with a double problem: 1. How was serfdom in fact overcome? 2. What system came into being following the fall of serfdom, which was still not yet capitalist? As to the first question, Wallerstein argues that the trans-European demographic crisis of the fourteenth and fifteenth century led in both Eastern and Western Europe to the effective end of serfdom, by the period 1450–1500. As Wallerstein states, ‘In the 14th and 15th centuries, the social structure of Poland was not markedly different from that of France or England. The demographic contraction had led there, as in western Europe, to increased right of the tiller of the soil vis-`-vis the landowner (the decline of feudalism)’ (tpn, p. 96). This argument is unacceptable for two reasons.
In the first place, it rests on a form of demographic determinism, which will not stand up. Wallerstein’s argument is that the drop in the land/labour ratio, caused by the late medieval demographic decline, made labour more scarce relative to land, and that therefore the serfs were in a better bargaining position with their masters: they now could bargain for an end to the limitations on mobility and on access to land which defined serfdom—in short, bargain for their freedom. Yet this argument assumes what it is necessary to prove: that the serfs could in fact bargain in this way. To assume that they could is to assume that de facto the serfs had already gained their freedom; that there had thus emerged a free market in tenants and labourers. But it was the essence of serfdom that the unfreedom it entailed prevented just such a market and precisely this sort of bargaining. In other words, the argument skips over the key link in the chain of explanation: the way the serfs were (or were not) able in fact to break the lords’ control so they could (or could not) actually bargain with them.
Secondly, the demographic crisis does not, in historical fact, appear to have issued in the decline of serfdom throughout Europe. Serfdom did disappear in the West. But it actually rose in the East. The different ‘outcomes’ confirm the illogic of Wallerstein’s demographic determinism.  Indeed, most authorities on Eastern Europe, including the ones Wallerstein relies on most heavily,  agree that serfdom had been firmly established in Poland before the big impact of the world market—so that it cannot be explained as a response to the world market.
Yet even suppose Wallerstein were correct that serfdom had ceased to exist in both East and West Europe by 1450–1500, what had taken its place? Wallerstein never specifically answers this question. However, he argues that from this point, as the world market began to develop, each region’s ruling class selected their different ‘most suitable’ methods of labour control. English agrarian capitalists turned to free wage labour. The East European ones turned to serfdom. By implication, the English landlords had the choice to use serfdom, but did not take it. On the other hand, the Eastern lords could have chosen free wage labour, but passed up this option. How unrealistic are these notions in historical terms should be obvious.
In the first place, the possibilities of English landlords re-enserfing English peasants did not exist. Indeed, in this period (1450–1500) they were having difficulty, at least in certain areas, in retaining property in the land, as peasants were pushing to fix rents and fines. Had the peasants been successful, especially under the long-term inflationary trend, this would have meant the de facto extinguishing of rent; the dissolution of the landlords’ ability to extract a surplus.  In the East, on the other hand, the landlords did of course generally succeed in enserfing the peasantry. But this can hardly be understood as the outcome of a simple choice. All over Europe throughout the later middle ages, in England, France and many other places, the ruling class of landlords had attempted to do precisely the same thing, to keep the peasants enserfed. But they had failed in the effort. Moreover, even in parts of grain-exporting Eastern Europe—notably the region around Danzig, as well as lower Silesia and lower Saxony—serfdom was not made to prevail. In this light, the general success of the Eastern European landlords in imposing serfdom cannot be understood merely in terms of their intentions, but obviously demands a much fuller analysis of the conditions making for their ability to control the peasants.  Finally, the case of France demonstrates, in itself, the absurdity of Wallerstein’s argument when taken to its logical conclusion. Here, as we have seen, a great part of the peasantry not only was able to get out from under serfdom, but to gain essential proprietorship of a large part of the land. Are we to conclude that this was the outcome of the choice of the French landlords, who voluntarily gave up their control over the peasantry and their property rights to create the most suitable structure of labour control for the world market?
The point is that whatever the apparent historical plausibility of his account, were we to follow Wallerstein in conceiving the rise of what he calls methods of labour control/reward to labour as a function of ruling-class choice, under the pressures and limits of the world market, the whole question of the mode of production—of exploiter/exploited relations—which preceded capitalism becomes irrelevant to its emergence, and the question of transition simply ceases to exist as a problem. This is because the exploiting class, whatever its relation to the exploited in the pre-capitalist era, merely required the onset of trade to take on the role and mode of operation of capitalists—and adopt the now appropriate method of labour control/reward to labour. In other words, we are back to the Smithian world, in which the rise of capitalism is the rise of the world market—but with one difference, which is more apparent than real.
For the Adam Smith of The Wealth of Nations, Book 1, the world before capitalism is composed of potential individual profit maximizers, ready to expand production, as a result of their own egoistic motivations, on the basis of the most appropriate specialization, as soon as they are involved in the reciprocal links of exchange which emerge with the historically determined expansion of trade. In contrast, for Wallerstein, the world before capitalism is composed of individual exploiters in various (somewhat vague) relations to the exploited,  but ready to specialize in the method of exploitation most suitable for their production on the world market. This, indeed, is the European world he thinks must have existed in the later fifteenth century. It is, in short, essentially Smith’s world of atomistic egos; but this time they can specialize not only in the most productive technique of production, but the most productive technique of exploitation. It is no wonder that Wallerstein refers to the ‘world-system’ of the sixteenth century as a ‘one-class’ system, for it is only the capitalists (themselves created by the world market) and their motivations which in his account really matter for the historical development of capitalism. Hence, in effect, world history before capitalism, in so far as it is a history of class society, plays no part in determining either the fact of the emergence of capitalism or the character of capitalist economic development. The rise of the world division of labour, based on the commercial expansion of Europe, both gives us the origins of capitalism apart from any transition through class conflict, and the form of its economic development apart from any class structure of capital accumulation.
VII. The ‘Development of Underdevelopment’
The onset of a capitalist dynamic of development was thus, in its first appearance, made possible as an unintended consequence of class conflicts—conflicts in which the peasantry freed themselves from the extra-economic controls of the ruling class, while the latter secured ownership of the land. The resulting overall class structure of production and reproduction made possible an unprecedented degree of correspondence between the needs of surplus extraction and the continuing development of the productive forces through accumulation and innovation, especially in agriculture, by way of the application of fixed capital on the basis of increasingly co-operative labour. The original emergence of capitalist development is, therefore, incomprehensible as a phenomenon of ‘money’, ‘trade’, ‘the production of commodities’ or of ‘merchant capital’. The very significance of these forms depends on the class structure of production with which they are associated. They perform indispensable functions in production and reproduction under capitalist social-productive relations. On the other hand, by themselves, by their ‘self-development’ (the widening of commodity production alone) they cannot bring about the emergence of capitalist social-productive relations and a pattern of economic development in response to the demands of profitability on the market.
To see the action of money or trade or of merchant capitalists as being behind the original emergence of capitalism is, therefore, circular: for it is to account for the origins of capitalism by the action of capitalists functioning in a capitalist manner. It is for analogous reasons that it is necessary to reject the idea that the mere extension of the world market via the action of merchants to stimulate increasing commodity production in new areas determines a pattern of underdevelopment, as in Eastern Europe. For this is merely to turn the Smithian argument on its head: to contend that the demands of production for profit on the market determine the rise of class relations and productive forces which enforce, not the development of capitalist production (as with Paul Sweezy), but the rise of economic backwardness (as with André Gunder Frank and Immanuel Wallerstein). The method of Book 1 of The Wealth of Nations cannot be used to explain the poverty of nations.
Frank and his Critics
From this perspective, it is impossible to accept Frank’s view, adopted by Wallerstein, that the capitalist ‘development of underdevelopment’ in the regions colonized by Europeans from the sixteenth century—especially the Caribbean, South America and Africa, as well as the southern part of North America—is comprehensible as a direct result of the incorporation of these regions within the world market, their ‘subordination’ to the system of capital accumulation on a world scale. Frank originally explained this rise of underdevelopment largely in terms of the transfer of surplus from periphery to core, and the export-dependent role assigned to the periphery in the world division of labour.  These mechanisms clearly capture important aspects of the functioning reality of underdevelopment. But they explain little, for, as the more searching critics of Frank’s earlier formulations pointed out, they themselves need to be explained. In particular, it was stated, they needed to be rooted in the class and productive structures of the periphery. 
However, in more recent work, Frank has attempted to respond to his critics specifically by integrating an analysis of internal class structure into his theory of underdevelopment. He argues that ‘underdevelopment is the result of exploitation of the colonial and class structure based on ultraexploitation; development was achieved where this structure of underdevelopment was not established because it was impossible to establish. All other factors are secondary or derive from the basic question of the type of exploitation.’  By this reasoning, it was the relations of exploitation which came to dominate Latin American and Caribbean production for export, especially slavery and other sorts of enforced-labour systems, which determined underdevelopment. Thus, ‘The colonial and class structure is the product of the introduction into Latin America of an ultraexploitative export economy, dependent on the metropolis, which restricted the internal market and created the economic interests of the lumpen bourgeoisie (producers and exporters of raw materials). These interests in turn generated a policy of under- or lumpen development for the economy as a whole.’  Perhaps we could paraphrase Frank’s argument in the following terms: on the one hand, growing production for the market stimulated by world demand determined increasing pressure to extract greater surplus; on the other hand, the establishment of class systems of production based on the direct use of force determined that this increasing output would be achieved through the extension of absolute, rather than relative, surplus labour—with familiar results.
It is important to notice that certain conclusions of Frank’s earlier analysis are not logically bound up with such a model, although Frank does not make this clear. In the first place, the rise of underdevelopment, by this reasoning, does not require a commercial system enforcing unequal exchange (whereby the core gets more than it gives up in business transactions with the periphery) to explain it. Indeed, the exchange of raw materials from the East European periphery for manufactured goods from the West European core in the early modern period may have brought about a transfer of surplus from the core to the periphery. This was possibly also the case in the West Indies in the seventeenth and eighteenth centuries, where terms of trade favourable to the colonies seem to have helped to bring about a huge build-up of wealth in the hands of West Indies planters, even at the expense of their English trading partners.  Yet these areas ‘underdeveloped’ in prototypical fashion.
This is not to deny that there was a long-term transfer of surplus away from the periphery. It is to root this in a different dynamic. Thus it was the characteristic feature of the forced labour systems that their difficulties in developing the productivity of labour through the application of fixed capital were more than ‘counterbalanced’ by their success in reducing the costs of labour through reducing the subsistence of the work force.  As a result, the market for capital goods was limited and the market for consumer goods was similarly cut down, while there arose significant demand for luxuries by the ruling class. Meanwhile, precisely the fact of forced labour in agriculture, either in pure form (slavery) or in correlation with peasant possession of subsistence plots, undermined the economies’ ability to develop a free wage labour force for industry. In this context, Frank’s comment that ‘Because of commerce and foreign capital, the economic and political interests of the mining, agricultural and commercial bourgeoisie were never directed toward internal economic development’  could be misleading. It was not the specific national character of the capital or the commercial connection with the metropolis which determined a flow of potential investment funds ‘out of the system’. It was the class-structured character of the profit opportunities which determined that: 1. there would be relatively little investment even in the home industry (the mines and plantation); 2. what industrial production there was for the home market would be carried on largely in the metropolis because it could be more profitably organized there, leading to a flow of investment funds from periphery to core; 3. there would be considerable expenditure on luxury production which would not increase the productive capacity of the system.
In other words, the development of underdevelopment was rooted in the class structure of production based on the extension of absolute surplus labour, which determined a sharp disjuncture between the requirements for the development of the productive forces (productivity of labour) and the structure of profitability of the economy as a whole. On the one hand, this class structure determined a general antagonism between the demands of profit-making and the development of the productive forces in the fields subject to world market demand, by discouraging the advance of fixed capital and undermining the development of skill, since production was based on forced labour (while low payments to labour power encouraged the adoption of labour-using techniques). On the other hand, it determined a generalized lack of profitability for the remainder of the economy, precisely because this was generally compelled to support export production through ‘contributing’ cheap or free labour power and means of subsistence (by way of forced levies), without receiving any investment to raise labour productivity. Thus the ‘subsidizing’ of the ‘export sector’ was generally accomplished on the basis of the intensification of various forms of peasant production; and this, in turn, posed powerful barriers to development throughout the economy, through making difficult the application of fixed capital and the rise of co-operative labour, as well as, more generally, the full emergence of labour power as a commodity.
In the second place, it cannot be deduced from Frank’s revised account of the class structure of underdevelopment that what determined the colonies’ backwardness was their role in the world system; their production of raw materials for export. During the early modern period, grain for example was produced for export in many different areas of Europe under different class structures of production, with very different resulting patterns of economic development or underdevelopment. So Frank’s comment that ‘ultra-underdevelopment . . . was characteristic of an export economy’  must be carefully qualified. It was not the fact of production for export which determined export dependence; it was the class structure through which export production was carried out (based on ultra-exploitation/methods of absolute surplus labour) which determined that increasing export production would lead to underdevelopment rather than development. Otherwise it would not be possible, for example, to account for the impressive development of the grain-exporting economy of the Middle Atlantic colonies in the colonial period.
Market, Profit, Class
Indeed, it does not seem that Frank’s more recent formulations have fully broken from the neo-Smithian presuppositions which informed his earlier work. For in the last analysis, like Wallerstein, Frank continues to regard class as a phenomenon of the market; of the abstract needs of capital or production. Although he now does focus on class, Frank continues to treat it as a derivative phenomenon, arising directly from the needs of profit maximization. ‘The relations of production and the class structure . . . developed in response to the predatory needs of the overseas and the Latin American metropolis.’  As with Wallerstein, the demands of the market, of profit, determine the class structure, subject only to the limitations of geography and demography—as if the significance of these factors was not, in turn, to a great extent socio-historically determined, and as if the potential for profit did not itself depend on the class structure.
Specifically, in the American colonial case, accordiang to Frank, an ultra-exploitative class structure was imposed, subject to the possibility of producing certain demanded products (minerals, staples) and the availability of native labour. If geographical conditions permitted, says Frank, minerals would be mined or agricultural products produced through ultra-exploitative class systems. If the native mode of production could be harnessed to this task, so much the better; otherwise the best adapted (most suitable) forced-labour system would be imposed upon the indigenous population, or slaves would be imported.  In any case, if there was potential for profitable production for export for the world market, there was no way to prevent underdevelopment. Frank could not be more explicit about this. For the only way to avoid ultraexploitation and underdevelopment was to be useless to the world market. ‘The greater the wealth available for exploitation, the poorer and more underdeveloped the region today; and the poorer the region was a colony, the richer and more developed it is today.’  It was only the ‘neglect’ of certain areas, consequent upon their lack of economic potential, which allowed them to develop.
Thus, in particular, Frank accounts for the contrasting facts of the development of the northern parts of North America, and the corresponding underdevelopment of the southern parts of North America, the Caribbean and most of South America, precisely by stating that what made possible the development in the north was the absence of mines and raw materials.  The northern colonies were not underdeveloped, because they were ignored by the capitalist ultra-exploiters and furthermore allowed to share with England some of the advantages of Empire. On the other hand, the gold and silver found in Mexico and South America, along with the suitability of these regions for the production of agricultural raw material demanded in the metropolis, determined that they would be victims of class systems of ultra-exploitation.  It is these points which need to be closely examined.
The Case of Caribbean Sugar
The rise of Caribbean sugar production, especially on Barbados, appears to fit Frank’s theory perhaps best of all. Here there arrived an original settler population from England, which set itself up as ‘yeoman farmers’. Yet with the growing demand for sugar, and its introduction on the island around 1640, the class structure was rapidly transformed. Plantations owned by a small number of Europeans, and operated on the basis of a huge influx of African slaves, replaced the small farms which had previously dominated the island within the space of a couple of decades. Apparently, the class structure was immediately determined by the needs of the market and of capital accumulation. 
Still, a few questions need to be asked. First, how was the ‘so-called primitive accumulation of capital’ accomplished? In other words, how did the separation of the population of small farmers from the land actually take place? Had these producers actually owned the land? And had they, furthermore, produced their own subsistence needs (in addition to the tobacco they had sent out for export)? If so, how were they got rid of so easily?  Was the availability of land in North America, where most of them emigrated, a relevant factor in inducing them to sell out? Such questions should not be assumed away, for the supersession of peasant populations (if that, indeed, was what they were)  is not always such an easy proposition—the automatic consequence of economic necessities. 
But that is not really the major problem, which was rather the establishment of a class system of production apparently in accord with the demands of capitalism. Having got rid of the small farmers, how was it possible for the merchants and planters to establish the plantation system for the production of sugar? The obvious answer: by buying slaves. Yet this only pushes the question back a step. Why were slaves available to be used? Before they could be bought, the slaves had to be ‘produced’; more precisely, they had to appear on the market ‘as commodities’. But this poses large questions, namely of the formation of class systems of ‘production’ and appropriation of slaves in Africa (or elsewhere). The point here is not to enter into the debate concerning the degree to which the formation of such a structure marked the emergence of a new mode of production, or merely the adaptation and intensification of an already existing one. It is to argue that its existence should in no way be assumed; that the needs of capitalism, or capitalists, are not in themselves enough to explain it. This is especially because class formation, or the intensification of exploitation, is generally an outcome of class conflict, and this outcome itself needs to be accounted for. 
The Case of Colonial Virginia
The relevance of this question is clarified by the very great difficulty, if not impossibility, of enslaving the European settlers themselves in the colonial context. In Virginia, for example, the demand for tobacco from England and Europe set in train a demand by planters and merchants for increased output for export, and a consequently increasing pressure on the direct producers to increase their output. In this case, the direct producers for the planters and merchants were for the most part indentured servants, subject to work for their masters for a specified number of years before gaining their freedom. In this situation, the way to ensure and increase output was for the planters to intensify their servants’ labour, extend their terms of service, and close off their access to land by engrossing it themselves. These processes were indeed set in motion. Yet actually to accomplish them required increasing class exploitation and oppression and, in return, class conflict. From the 1660s, the Virginia colony was wracked by class conflict, by a succession of conspiracies and revolts, set off by the resistance of servants and ex-servants to the oppression of the planters, and culminating in 1676 in Bacon’s rebellion—the greatest social conflict in the pre-revolutionary history of North America.  In fact, the planters were in the long run unsuccessful in either seriously depressing the condition of European servants or preventing them from getting land. The existence of a massive class of small tobacco farmers is a characteristic feature of Virginia’s eighteenth-century social and political structure.  Had the planters, therefore, depended upon the labour of the European colonists, it might have been impossible to construct plantations—due to the results of class struggles in the South. Of course, as it turned out, plantations did, in the long run, come to dominate Southern society—but this was on the basis of slavery. Had it not been for the outcome of processes of class formation and class conflict in Africa, the development of Southern society, indeed society throughout the Western hemisphere, might have been very different. Capitalism, itself, cannot account for it.
Nor did the profitable incorporation of a region into the capitalist world market as a raw material exporter necessarily determine the rise of a class system of ultra-exploitation and consequent development of underdevelopment. In the Middle Atlantic colonies, especially Pennsylvania, as elsewhere, colonization had progressed on the basis of the establishment of small-holding farmers. As elsewhere, the demands of production stimulated increasing demand for surplus for export, in this case grain and flour. Yet, here the small farmers were not surpassed (as they were in the West Indies) through the establishment of an ultra-exploitative class structure.  It can hardly be said, on the other hand, that the small-farming system was most suitable for merchants or planter capitalists interested in maximizing gains on the world market. New techniques were at best very slowly introduced. Indeed, despite production for export, these farmers can hardly be said to have been specialized. On their relatively large farms, they made sure of their subsistence and reproduction by producing a full range of necessities. It was a surplus over the requirements for reproduction which went to market.  Clearly, a greater output would have been possible with specialization. Had it been possible to reduce these farmers to tenancy, this specialization might have been accomplished. As it was, however, their very control over the land made them largely invulnerable to takeover through competition. On the other hand, they appear to have possessed the class power to prevent any direct attempts at expropriation.
At the same time, despite the relatively limited advances made by the Pennsylvania farmers, they were able to become a significant factor on the world grain market, apparently as a result of their large holdings and their freedom and flexibility in exploiting them. But in this case incorporation into the world market (by means of a peasant-dominated class structure) did not determine underdevelopment. Pennsylvania farmers apparently marketed 40–50 per cent of their output, some in the colonies, a good part overseas.  With the returns, they were able to provide a growing home market. At the same time, their productive capacity allowed a growing section of the population to move into non-agricultural production off the land. A network of small towns grew up in close relationship with the rural producers. Most significantly the development and prosperity of Philadelphia, one of the great colonial cities, was made possible by its symbiotic relationship with its hinterland.  A slow, but significant dynamic of economic development was set in motion.
Frank’s original formulations aimed to destroy the suffocating orthodoxies of Marxist evolutionary stage theory upon which the Communist Parties’ political strategies of ‘popular front’ and ‘bourgeois democratic revolution’ had been predicated.  Frank rightly stressed that the expansion of capitalism through trade and investment did not automatically bring with it the capitalist economic development that the Marx of the Manifesto had predicted. In the course of the growth of the world market, Chinese Walls to the advance of the productive forces might be erected as well as battered down. When such ‘development of underdevelopment’ occurred, Frank pointed out, the ‘national bourgeoisie’ acquired an interest not in revolution for development, but in supporting precisely the class system of production and surplus extraction which fettered economic advance. In particular, the merchants of the periphery backed the established order, for they depended for their profits on the mining and plantation enterprises controlled by the ‘reactionaries’, as well as the industrial production of the imperialists in the metropolis. But even the industrial capitalists of the periphery offered no challenge to the established structure—partly as a consequence of their involvement in luxury production serving the upper classes—while they merged with the ‘neo-feudalists’ through family connections and state office. As Frank asserted, to expect under these circumstances that capitalist penetration would develop the country was, by and large, wishful thinking. To count on the bourgeoisie for a significant role in an anti-feudal, anti-imperialist revolution was to encourage a dangerous utopia.
Yet, the failure of Frank and the whole tradition of which he is a part—including Sweezy and Wallerstein among others—to transcend the economic determinist framework of their adversaries, rather than merely turn it upside down, opens the way in turn for the adoption of similarly ill-founded political perspectives. Where the old orthodoxy claimed that the bourgeoisie must oppose the neo-feudalists, Frank said the neo-feudalists were capitalists. Where the old orthodoxy saw development as depending on bourgeois penetration, Frank argued that capitalist development in the core depended upon the development of underdevelopment in the periphery. At every point, therefore, Frank—and his co-thinkers such as Wallerstein—followed their adversaries in locating the sources of both development and underdevelopment in an abstract process of capitalist expansion; and like them, failed to specify the particular, historically developed class structures through which these processes actually worked themselves out and through which their fundamental character was actually determined. As a result, they failed to focus centrally on the productivity of labour as the essence and key to economic development. They did not state the degree to which the latter was, in turn, centrally bound up with historically specific class structures of production and surplus extraction, themselves the product of determinations beyond the market. Hence, they did not see the degree to which patterns of development or underdevelopment for an entire epoch might hinge upon the outcome of specific processes of class formation, of class struggle. The consequence is that Frank’s analysis can be used to support political conclusions he would certainly himself oppose.
Thus so long as incorporation into the world market/world division of labour is seen automatically to breed underdevelopment, the logical antidote to capitalist underdevelopment is not socialism, but autarky. So long as capitalism develops merely through squeezing dry the ‘third world’, the primary opponents must be core versus periphery, the cities versus the countryside—not the international proletariat, in alliance with the oppressed people of all countries, versus the bourgeoisie. In fact, the danger here is double-edged: on the one hand, a new opening to the ‘national bourgeoisie’; on the other hand, a false strategy for anti-capitalist revolution.
True, bourgeois revolutions are not on the agenda. International capitalists, local capitalists and neo-feudalists alike have remained, by and large, interested in and supportive of the class structures of underdevelopment. Nevertheless, these structures have kept significant masses of use value in the form of labour power and natural resources from the field of capital accumulation. Until recently, of course, the class interests behind ‘industrialization via import substitution’ have not, as a rule, been strong enough to force the class structural shifts that would open the way to profitable investment in development. However, with contracting profit opportunities in the advanced industrial countries and the consequent drive for new markets and cheap labour power, potentially available in the underdeveloped world, such interests may now receive significant strength from unexpected quarters. Should a dynamic of ‘development’ be set in motion as a consequence—and that is far from certain—it could hardly be expected to bring much improvement to the working population of the underdeveloped areas, for its very raison d’être would be low wages and a politically repressed labour force. But this would in no way rule out its being accomplished under a banner of anti-dependency, national development and anti-imperialism.
Most directly, of course, the notion of the ‘development of underdevelopment’ opens the way to third-worldist ideology. From the conclusion that development occurred only in the absence of links with accumulating capitalism in the metropolis, it can be only a short step to the strategy of semi-autarkic socialist development. Then the utopia of socialism in one country replaces that of the bourgeois revolution—one moreover, which is buttressed by the assertion that the revolution against capitalism can come only from the periphery, since the proletariat of the core has been largely bought off as a consequence of the transfer of surplus from the periphery to the core. Such a perspective must tend to minimize the degree to which any significant national development of the productive forces depends today upon a close connection with the international division of labour (although such economic advance is not, of course, determined by such a connection). It must, consequently, tend to overlook the pressures to external political compromise and internal political degeneration bound up with that involvement in—and dependence upon—the capitalist world market which is necessary for development. Such pressures are indeed present from the start, due to the requirement to extract surpluses for development, in the absence of advanced means of production, through the methods of increasing absolute surplus labour.
On the other hand, this perspective must also minimize the extent to which capitalism’s post-war success in developing the productive forces specific to the metropolis provided the material basis for (though it did not determine) the decline of radical working-class movements and consciousness in the post-war period. It must consequently minimize the potentialities opened up by the current economic impasse of capitalism for working-class political action in the advanced industrial countries. Most crucially, perhaps, this perspective must tend to play down the degree to which the concrete inter-relationships, however tenuous and partial, recently forged by the rising revolutionary movements of the working class and oppressed peoples in Portugal and Southern Africa may be taken to mark a break—to foreshadow the rebirth of international solidarity. The necessary interdependence between the revolutionary movements at the ‘weakest link’ and in the metropolitan heartlands of capitalism was a central postulate in the strategic thinking of Lenin, Trotsky and the other leading revolutionaries in the last great period of international socialist revolution. With regard to this basic proposition, nothing has changed to this day.
I wish to thank Alice Amsden, Johanna Brenner, Temma Kaplan, Barbara Laslett, Richard Smith and Jon Wiener for reading this manuscript and offering criticisms and suggestions. I am also grateful to Theda Skocpol for sending me, in advance of publication, her review essay on Immanuel Wallerstein’s The Modern World System, which was very helpful to me, especially on problems concerning the early modern European states.
 See, for example: ‘England has to fulfil a double mission in India: one destructive, the other regenerating—the annihilation of old Asiatic society, and the laying of the material foundation of Western society in Asia.’ In ‘The Future Results of British Rule in India’, in Karl Marx, Surveys From Exile, London 1973, p. 320.
 See, for example: ‘The obstacles presented by the internal solidity and organization of pre-capitalistic national modes of production to the corrosive influence of commerce are strikingly illustrated in the intercourse of the English with India and China . . . English commerce exerted a revolutionary influence on these communities and tore them apart, only in so far as the low prices of their goods served to destroy the spinning and weaving industries, which were an ancient integrating element of this unity of industrial and agricultural production. And even so, this work of dissolution proceeds very gradually. And still more slowly in China, where it is not reinforced by direct political power.’ Capital in three volumes, New York 1967, III, pp. 333–4.
 Capitalism and Underdevelopment in Latin America, New York 1969, p. 9.
 Ibid. pp. 14–15.
 It should be made clear that Frank, in more recent writings, has attempted to modify and deepen his analysis of underdevelopment through taking greater account of ‘internal class structure’. However, his retention of the theoretical approach of his earlier works has prevented him from fully accomplishing his aims. See below, pp. 83–91.
 New York 1974 (mws). In the following discussion, I treat this book together with a series of closely related articles by Wallerstein which further clarify and amplify his themes. These include: ‘The Rise and Future Demise of the World Capitalist System: Concepts for Comparative Analysis’, Comparative Studies in Society and History, XVI (January 1974), pp. 387–415 (rfd). ‘From Feudalism to Capitalism: Transition or Transitions?’, Social Forces, LV (December 1976), pp. 273–81 (ffc). ‘Three Paths of National Development in Sixteenth Century Europe’, Studies in Comparative International Development, VII (Summer 1972), pp. 95–101 (tpn). ‘Dependence in an Interdependent World: The Limited Possibilities of Transformation Within the Capitalist World Economy’, African Studies Review, XVII (April 1974), pp. 1–27. (diw). Henceforth, when quoting from Wallerstein’s works, I will indicate the source through using the indicated abbreviations, with page numbers, placed in parentheses in the text.
 Modern World System, pp. 16–20. See also ‘Rise and Future Demise’, pp. 390–92.
 Obviously, this is not to deny that the methods of absolute surplus labour are used, indeed extensively and systematically used, under capitalism, for of course they are.
 Capital, I, Penguin/nlr edition, London 1976, p. 645.
 The Transition from Feudalism to Capitalism, (expanded nlb edition), London 1976. The critique of Sweezy in this exchange by Maurice Dobb, as well as (implicitly) in Dobb’s Studies in the Development of Capitalism (Cambridge and New York 1963) is, of course, of fundamental importance—as is Ernesto Laclau’s critique of Frank, in ‘Feudalism and Capitalism in Latin America’, nlr 67 (republished in his Politics and Ideology in Marxist Theory, nlb, London 1977). I hope my great debt to both these writers will be apparent throughout this essay.
 Ed. Edwin Canaan, New York 1937. It should be emphasized that the following discussion of Smith relates exclusively to his ‘model’ in Book I, the part generally taken up by contemporary economists. It does not take into account Smith’s very rich socio-historical analyses found elsewhere in The Wealth of Nations.
 See Marx’s characterization of Smith as ‘the quintessential political economist of the period of manufacture’, due to ‘the stress he lays on . . . division of labour’, but ‘the subordinate part he assigned to machinery’ Capital, I, p. 468.
 Capital, I, p. 453 (emphasis added).
 Wealth of Nations, pp. 13, 17–21.
 For Smith’s interesting comments along these lines, see Wealth of Nations, pp. 385–7.
 Capital, III, pp. 326–7.
 Transition from Feudalism to Capitalism, pp. 44–5. Sweezy’s position is ambivalent and self-contradictory; he is aware of the difficulties with his line of argument—e.g. the implications of the ‘East European case’—yet does not adequately deal with them in reaching his conclusions.
 Thus, Sweezy states, ‘Dobb often seems . . . to assume that only the villein stood to gain from the abolition of serfdom. He tends to forget that “the enfranchisement of the peasants was in reality the enfranchisement of the landowner, who, having henceforth to deal with freemen who were not attached to his land, could dispose of the latter by means of simple revocable contracts, whose brief duration enabled him to modify them in accordance with the increasing rent from the land” ’. Transition from Feudalism to Capitalism, p. 45, note 15 (Sweezy is quoting from Pirenne).
 Dobb, Studies in the Development of Capitalism, pp. 38–42; Eleanor Searle, Lordship and Community: Battle Abbey and its Banlieu, 1066–1538, Toronto 1974, pp. 147, 174–5, 183–94, 267–327. B. H. Slicher Van Bath, The Agrarian History of Western Europe, A.D. 500–1850, London 1966, pp. 178–9.
 There was, in fact, another alternative, commonly resorted to during the medieval period: i.e. military conquest.
 See Eric Kerridge, The Agricultural Revolution, London 1967, pp. 181–221. Similarly, with the new irrigation systems, ‘the floating of the water meadows’, ibid. pp. 251–67.
 Thus if, by virtue of his position, the lord had been unable to apply force and, specifically, to tie the serf to the land, the serf might have had access to open lands, or at least to the plots of other lords. In forcefully controlling the serf, therefore, the lord made sure a rent would be paid, while precluding the possible cut in rent which might have resulted had the serf been free to bargain over rent with other lords.
 Eric Kerridge, Agrarian Problems of the Sixteenth Century and After, London 1969, pp. 121–6; W. G. Hoskins, ‘The Leicestershire Farmer in the Seventeenth Century’, Agricultural History, XXV (1961); G. E. Mingay, ‘The Size of Farms in the Eighteenth Century’, Economic History Review, 2nd series, XIV (1962); Searle, Lordship and Community, pp. 147, 174–5, 183–94, 267–329; R. Brenner, ‘Agrarian Class Structure and Economic Development in Pre-Industrial Europe’, Past and Present, No. 70 (February 1976).
 R. H. Hilton, The Decline of Serfdom in Medieval England, London 1969, pp. 29–31; also Hilton’s ‘Freedom and Villeinage in England’, Past and Present, No. 31 (July 1965), p. 31.
 Transition from Feudalism to Capitalism, p. 42.
 Ibid. p. 43.
 Ibid. p. 40.
 Cf. Perry Anderson, Passages from Antiquity to Feudalism, London 1974.
 Witold Kula, An Economic Theory of the Feudal System, London 1976, pp. 62–6ff.
 See, e.g. M. M. Postan, ‘Medieval Agrarian Society in its Prime: England’, in Cambridge Economic History of Europe, I, 2nd edn., ed. M. M. Postan, Cambridge 1966; Brenner, ‘Agrarian Class Structure’, pp. 47–51.
 See Brenner, ‘Agrarian Class Structure’, pp. 54–6, for a fuller critique of the notion that towns were responsible for the dissolution of feudalism/serfdom.
 Grundrisse, London 1973, p. 505.
 Ibid. p. 508.
 Transition from Feudalism to Capitalism, pp. 41.
 Ibid. p. 35.
 Grundrisse, pp. 509–10.
 Ibid. p. 505.
 Without labour power as a commodity there is not, then, commodity production in the full sense of the term. The use of the term ‘production of commodities’ in reference to pre-capitalist production for exchange can therefore be misleading, for it can convey the impression of a conceptual and historical continuity between pre-capitalist and capitalist ‘commodity production’; it is the discontinuity which must be emphasized.
 Capital, III, p. 328.
 ‘Production based on exchange value and the community based on the exchange of these exchange values . . . and labour as general condition of wealth, all presuppose and produce the separation of labour from its objective conditions.’ Grundrisse, p. 509.
 This is not precisely accurate. For there is a significant range of social-productive forms where the direct producer does not appear as a free wage labourer, but where exchange value can nonetheless be said to predominate. That is because, in these cases, the direct producer’s property in the means of production does not allow him direct access to the means of subsistence and reproduction. He is, therefore, required to sell on the market to reproduce and survive, thus forced to sell at ‘the socially necessary rate’ or go out of business. Examples of such ‘transitional forms’ would be the peasant producer of industrial crops, without landed property enough to provide him means of subsistence (especially food), as well as the independent urban artisan (with no guild protection). Other examples would be systems of free tenantry (without wage labour), where the tenants hold on terminable money lease from the landlord, as well as ‘putting out’ systems, where the producers are dependent on merchant suppliers for raw materials. What determines that all these forms are ‘transitional’ is that they allow for a more or less direct transition to formally capitalist class relations and co-operative labour under the pressures of competition on the market. What may tend to prevent this transition, even under market pressures, is on the one hand the ability and willingness of the direct producers to accept extreme increases in the rate of absolute surplus labour in order to continue to produce at ‘the socially necessary rate’: in other words, they ‘exploit themselves’—with the goal of holding onto ‘their property’—in order to be able to be competitive with productive units using more advanced techniques (co-operative labour combined with fixed capital). On the other hand, there is the reluctance of ‘capitalists’ (merchants, usurers, landowners) to make the change to fixed capital through the introduction of labour-saving techniques, because their ownership of fixed capital puts them at the mercy of the ups and downs of the market (see, for example, the transition from ‘putting out’ to the factory system). Marx, of course, referred to these forms as ‘simple commodity production’. From our point of view, however, it must be emphasized that they qualify as such not only, or even primarily, because they produce for the market, but because they do not have property in (or non-market access to) the means of subsistence. Thus, the ‘classical’ free peasant proprietors (along with many other sorts of agriculturalists) would not fit the category because they retained property in/direct access to the means of subsistence (even if they produced a great deal for the market).
 Grundrisse, p. 502.
 See Wallerstein, quoted above, p. 31.
 Capital, I, p. 450.
 See B. H. Slicher Van Bath, ‘The Yield of Different Crops in Relation to Seed, c. 1810–1820’, Acta Historica Neerlandica, II (1967). See also below, footnotes 57–60.
 See Unequal Exchange, New York 1972.
 ‘Rise and Future Demise’, p. 408.
 Capital, III, p. 791.
 On the policies of the Polish state and underdevelopment, see M. Malowist, Croissance et Régression en Europe XIVe–XVIIe Siècles, Paris 1972.
 F. L. Carsten, ‘La noblesse de Brandebourg et de Prusse de XVIe au XVIII Siècle’, in R. Mousnier, (ed)., Problèmes de Stratification Sociale, Paris 1968.
 Capital, I, p. 873.
 Grundrisse, pp. 508–9.
 Capital, I, pp. 874–5.
 It is, for example, today widely accepted that fixed capital requirements in manufacturing, even through the first stages of the industrial revolution, were relatively small.
 Kula, Economic Theory of the Feudal System, pp. 108–11; A. Mączak, ‘Export of Grain and the Problem of Distribution of National Income in the Years 1550–1650’, Acta Poloniae Historica, XVIII (1968), pp. 78–9.
 J. Topolski, ‘Le Commerce des denrées agricoles et croissance économique de la zone baltique aux XVIe et XVIIe siècles’, Annales ESC, March–April 1974, pp. 433–4; J. Topolski, ‘La régression économique en Pologne de XVIe au XVIIe siècle’, Acta Poloniae Historica, VII (1962), pp. 32–7; A. Wyczanski, ‘Le niveau de récolte des céréales en Pologne du XVIe au XVIIIe siècle’, Third International Conference of Economic History, Paris 1968, pp. 585–90.
 ‘The reduction of the supplies of grain to the port of Danzig in relation to the demand is absolutely certain’ [from the beginning of the 17th century]. Topolski, ‘Le commerce des denrées agricoles’, p. 433.
 Topolski, ‘Le commerce des denrées agricoles’; Topolski, ‘La régression économique’; Mączak, ‘Export of Grain’.
 L. Zytkowicz, ‘An Investigation into Agricultural Production in Masovia in the First Half of the 17th Century’, Acta Poloniae Historica, XVIII (1968), pp. 117–18; Mączak, ‘Export of Grain’, pp. 77, 95–6; Topolski, ‘Le commerce des denrées agricoles’, p. 431; Topolski, ‘Régression économique’, p. 42.
 Zytkowicz, ‘An Investigation’, p. 118.
 Topolski, ‘Régression économique’, pp. 47–8.
 Wyczanski, ‘Le niveau de la récolte’, pp. 586–9; A. H. John, ‘English Agricultural Improvement and Grain Exports’, 1600–1765’, in D. C. Coleman and A. H. John (ed)., Trade, Government and Economy in Pre-Industrial England, London 1976, p. 56.
 Kula, Economic Theory of the Feudal System, pp. 119–20ff. It should be noted that less than three quarters of the grain imported in Amsterdam was re-exported, mostly to southern Europe in exchange for salt, spices and wine—so that a very significant part of the European ‘modern world system’ simply involved exchanges of raw materials between ‘peripheries’. See K. Glamann, European Trade 1500–1700, London 1971, pp. 39–45.
 There were pivotally important exceptions; but, as we shall see, these had a different basis of development.
 Cf. E. J. Hobsbawm, ‘The General Crisis of the Seventeenth Century’, in T. Aston, (ed.), Crisis in Europe 1560–1660, London 1965.
 See Brenner, ‘Agrarian Class Structure’, pp. 45–6, 72–5, and references cited there. Cf. Marx’s comments on the barriers to the development of the productive forces built into peasant ‘proprietorship of land parcels’ in Capital, III, pp. 804ff.
 For this general perspective, see Rosa Luxemburg, The Accumulation of Capital, New York 1968, esp. chapter 27 (‘The struggle against natural economy’) and chapter 29 (‘The struggle against peasant economy’).
 See A. V. Chayanov, The Theory of Peasant Economy, Homewood, Ill. 1966.
 Luxemburg, Accumulation of Capital, pp. 395–6.
 See Luxemburg’s comment: ‘A natural economy thus confronts the requirements of capitalism at every turn with rigid barriers. Capitalism must therefore always and everywhere fight a battle of annihilation against every historical form of natural economy that it encounters, whether this is slave economy, feudalism, primitive communism, or patriarchal peasant economy. The principal methods in this struggle are political force (revolution, war), oppressive taxation by the state, and cheap goods . . . capital is faced with difficulties because vast tracts of the globe’s surface are in possession of social organizations that have no desire for commodity exchange or cannot, because of the entire social structure and the forms of ownership, offer for sale the productive forces in which capital is primarily interested . . . If capital were here to rely on the process of slow internal disintegration, it might take centuries. To wait until the most important means of production could be alienated by trading in consequence of this process were tantamount to renouncing the productive forces of those territories altogether.’ Accumulation of Capital, pp. 369–70.
 Marx, Capital, III, pp. 810–11.
 Brenner, ‘Agrarian Class Structure’, pp. 61–4; Kerridge, Agrarian Problems, pp. 121–6; Mingay, ‘Size of Farms’; E. L. Jones, ‘Editor’s Introduction’, to Agriculture and Economic Growth In England 1660–1815, London 1967.
 Kerridge, Agrarian Problems, p. 46; E. L. Jones. ‘Agriculture and Economic Growth in England, 1660–1750: Agricultural Change’, Journal of Economic History, XXV (1965).
 See Jones, ‘Editor’s Introduction’; Jones, ‘Agriculture and Economic Growth in England’; A. H. John, ‘Agricultural Productivity and Economic Growth in England, 1700–1750’, Journal of Economic History, XXV (1965); A. H. John, ‘Aspects of English Economic Growth in the First Half of the Eighteenth Century’, Economica, new series, XVIII (1961); D. E. C. Eversley, ‘The Home Market and Economic Growth in England 1750–1780’, in E. L. Jones and G. E. Mingay (ed.), Land, Labour, and Population in the Industrial Revolution, London 1967.
 I have offered certain suggestions concerning these processes of class conflict in my article, ‘Agrarian Class Structure’, pp. 47–75.
 Wallerstein himself, at certain points, admits this, thus directly contradicting what he says elsewhere. Thus he writes, ‘The recession of the fourteenth century, however, led to opposite consequences in western and eastern Europe. In the west, as we have seen, it led to a crisis of the feudal system. In the east, it led to a “manorial reaction” ’ (mws, p. 95).
 Wallerstein’s main authorities on Poland, Jerzy Topolski and Marian Malowist, both state very clearly that serfdom had been established in Eastern Europe before the onset of the world market and that the world market cannot explain the rise of serfdom. Topolski writes, ‘The genesis and development of the seigneurial demesnes based on the corvée were linked to important processes which had preceded the era of the expansion of the export of agricultural supplies.’ ‘Le commerce des denrées’, p. 432. Malowist argues, ‘I believe, however, that the foreign demand for agricultural products from the East and Centre of Europe cannot be considered as the decisive cause of the profound changes which were produced in the rural economy of Poland and its neighbours in the fifteenth and sixteenth centuries, that is, the increase of the seigneurial demesne and the introduction of the system of corvée.’ ‘Le commerce de la Baltique et le problème des Luttes sociales en Pologne aux XVe et XVI siècles’. La Pologne au Xe congrès international des sciences historiques ` Rome, Warsaw 1955, p. 129 and passim.
 See Brenner, ‘Agrarian Class Structure’, pp. 61–3; C. Dyer, ‘A Re-distribution of incomes in fifteenth-century England?’, Past and Present No. 39 (April 1968); B. J. Harris, ‘Landlords and Tenants in England in the Later Middle Ages: The Buckinghamshire Estates’, Past and Present No. 43 (May 1969), pp. 146–60; S. T. Bindoff, Ket’s Rebellion (Historical Association Pamphlet), London 1949, pp. 7–9.
 See Brenner, ‘Agrarian Class Structure’, pp. 51–2; Hilton, Decline of Serfdom, pp. 36–59; Malowist, ‘Le commerce de la Baltique’, p. 129; Topolski, ‘La régression économique’, p. 41.
 One of the symptomatic problems of Wallerstein’s entire enterprise is his inability to provide a consistent conceptualization of European feudalism. For him, it is neither, apparently, a ‘world empire’ nor a ‘world economy’—its relationship to the development of capitalism is never made clear. At the same time, one may ask if, for Wallerstein, feudalism is actually a class society, for he consistently characterizes it in terms of the ‘so-called reciprocal nexus we identify with feudalism, the exchange of protection for labor services’ (emphasis added) ffc, pp. 278–9.
 See above, pp. 27–9.
 See Frank’s résumé of the arguments of his leading critics in Lumpenbourgeoisie: Lumpendevelopment. Dependence, Class, and Politics in Latin America, New York 1972, pp. 1–9. See also Laclau, ‘Feudalism and Capitalism in Latin America’.
 Lumpenbourgeoisie: Lumpendevelopment, p. 19.
 Ibid., p.14.
 Robert Sheridan, Sugar and Slavery. An Economic History of the British West Indies 1623–1775, Baltimore 1973, pp. 467–71; also chapter 1.
 See, e.g. L. C. Gray, History of Agriculture in the Southern United States to 1860, 2 vols., Gloucester Mass. 1958, I, pp. 474–5 and, in general, chapter 20.
 Lumpenbourgeoisie: Lumpendevelopment, p. 23. (emphasis added).
 Ibid. p. 22.
 Ibid., p. 23.
 Frank, ‘Development and Underdevelopment in the New World: Smith and Marx vs. the Weberians’, Theory and Society, Winter 1975, pp. 441–2, 462–3, note 39 and passim.
 Lumpenbourgeoisie: Lumpendevelopment, p. 19.
 ‘Then why did the British fail to devote the same “attention” to their New England and Middle Atlantic Colonies [as to those in the Caribbean] . . . Because these regions lacked all the conditions necessary to attract that kind of attention and impose a manner of monopolizing and extracting surplus through low wages and unequal exchange, and to develop a mode of production that would develop underdevelopment as existed elsewhere in the New World’, ‘Smith and Marx vs. the Weberians’, p. 442.
 ‘Smith and Marx vs. the Weberians’, pp. 442, 451–6; Lumpenbourgeoisie: Lumpendevelopment, chapter 1.
 ‘Smith and Marx vs. the Weberians’, pp. 446–7.
 It is possible, for example, that a population of yeoman farmers was never very well-entrenched to begin with. If this was the case—and I am not contending that it necessarily was—the ‘elimination’ of the mass of small farmers would be no mystery, since the majority of them must have been only tenants, with no claims to the land beyond their terminable leases. It is interesting in this respect that, according to contemporary sources, their were 764 proprietors in Barbados in 1638 before the ‘sugar revolution’ of the 1640s, and 754 proprietors in Barbados in 1667 after it was to a large extent completed. Apparently, also, land was mal-distributed from the beginning. (On the other hand, it is difficult to know how to evaluate the reliability of the contemporary estimates, while these, moreover, fail to take into account the tumultuous developments of the intervening years.) See Richard S. Dunn, Sugar and Slaves. The Rise of the Planter Class in the English West Indies 1624–1773, Chapel Hill 1972, pp. 50–51.
 It is possible that the small farmers of the West Indies were not peasants, in that they did not produce their means of subsistence on their own plots, but depended upon the market for these. In this case, their reproduction would have depended on sales of their products (in order to purchase). As ‘simple commodity producers’ in the strict sense, they would have been subject to elimination directly through economic competition. If this was the case, they were ‘transitional’ from the start, and we can understand their superçession by the great sugar producers who could make a larger profit on their plantations than could the small tobacco-producing farmers. For these distinctions, and the terminology, see above footnote 43. See, in this respect, Marx’s comment on the American colonial economy: ‘In these colonies, and especially in those which produced only merchandise such as tobacco, cotton, sugar, etc. and not the usual foodstuffs . . . right from the start, the colonists did not seek subsistence, but set up a business’. Theories of Surplus Value, Moscow 1968, II, p. 239.
 See above, pp. 73–5.
 See Walter Rodney, ‘African Slavery and other Forms of Social Oppression on the Upper Guinea Coast in the Context of the Atlantic Slave Trade’, Journal of African History, VIII (1966), p. 434; A. G. Hopkins, An Economic History of West Africa, New York 1973, pp. 104, 106. Both of these authors naturally see the development and/or intensification of slavery as responsive to the world market, but they do not adequately explain the specific character of the processes of class formation and class conflict which made this response possible.
 Edmund S. Morgan, American Slavery-American Freedom. The Ordeal of Colonial Virginia, New York 1975, pp. 215–70 and passim; T. H. Breen, ‘A Changing Labor Force and Race Relations in Virginia 1660–1710’, Journal of Social History, Fall 1973, pp. 3–25; Theodore Allen, ‘ “They Would Have Destroyed Me”: Slavery and the Origins of Racism.’ Radical America, May–June 1975, pp. 41–64.
 See, e.g., Aubrey Land, ‘Economic Behavior in a Planting Society’, Journal of Southern History, November 1967, pp. 473–5 and passim; Aubrey Land, ‘The Tobacco Staple and the Planter’s Problems: Technology, Labor, and Crops’, Agricultural History, January 1969, pp. 69–81.
 James T. Lemon, The Best Poor Man’s Country. A Geographical Study of Early Southeastern Pennsylvania, Baltimore 1972, pp. 42–70, 98–117.
 Ibid. pp. 150–83.
 For example, about one-third of the total wheat produced in the area was exported. See Lemon, Best Poor Man’s Country, pp. 180–83; Lemon, ‘Household Consumption in Eighteenth-Century America and its Relationship to Production and Trade: the Situation Among Farmers in Southeastern Pennsylvania’, Agricultural History, January 1967, pp. 68–70; D. Klingaman, ‘Food Surpluses and Deficits in the American Colonies, 1768–1772’, Journal of Economic History, September 1971, pp. 557–8, 562; James F. Shepherd, ‘Commodity Exports from the British North American Colonies to Overseas Areas, 1768–1772: Magnitude and Patterns of Trade’, Explorations in Economic History, Fall 1970. pp. 5–76.
 Lemon, Best Poor Man’s Country, pp. 118–49; 180–83, 223–7; Lemon, ‘Household Consumption’, pp. 59–67.
 See André Gunder Frank, ‘Not Feudalism—Capitalism’, Monthly Review, December 1963, pp. 468–78 and passim.