Modelling the world-system I: Commodity Chains

Ever since his first exposition of the world -systems perspective in 1974, Wallerstein has insisted that the defining characteristic of a world - system is a unitary division of labour across the world - system. Furthermore he has emphasised that this division of labour is constituted to produce necessities; products essential to the continued existence of the world-system.  Exactly how this unitary division of labour functioned was not spelled out until Wallerstein and his colleague Hopkins began their work on commodity chains at Binghamton in the late 1970s. The following quote from Wallerstein summarises their ultimate conclusions very well: "A world-economy is constituted by a cross - cutting network of interlinked productive processes which we may call 'commodity chains,' such that for any production process in the chain there are a number of 'backward and forward linkages,' on which the particular process (and the persons involved in it) are dependant."

 

What Wallerstein is arguing here is that commodities are the outcome of complex process of manufacture, and that said manufacture is conducted in diverse geographical locations within the world-system. A final commodity - one that is marketed to end consumers - is itself composed of a number of other commodities and may well have been part processed, manufactured or assembled before taking its final form. To illustrate this let us take two examples from the early days of the world - system and the present.

 

Let us look at a nice, simple commodity, fundamental to life in the mid sixteenth century - bread. At this point in time the emerging economic core of the world -system was the Netherlands. Rye bread is the bread of choice for good Dutch burghers and was generally bought at neighbourhood bakeries. The rye flour from the bread was milled from rye grown in Poland; the salt used to flavour the bread came from the Basque country; the copper baking trays and iron stove would have been made in Germany, most likely in Franconian towns like Nuremburg and Augsburg and the metals were mined in the Alps, the Ore Mountains of Saxony or in Bohemia. The wood from which the oven was fired and out of which the bakery was constructed came from Norway. Specifically Dutch in this mix was the yeast and the water.

 

Turning to our own times, what product could be more universal than a pair of Levi jeans? It transpires that a pair of jeans is the outcome of around 43 different process (including growing cotton and mining copper and zinc) conducted across 4 continents).

 

What Wallerstein is therefore emphasising is that commodities, even in the early years of the world -system, were and are produced not in one locality but on a system-wide basis. Hence the emphasis on the unitary division of labour. We can thus (theoretically at least) trace the contours of the world -system by tracing out the lengths of the commodity chains. Commodity chains thus constitute the geographical bounds of a world -system.

 

More significantly, commodity chains actually structure the component parts of the world -system - the famous core - periphery dichotomy. Let us read Wallerstein's own words again:  "As the commodity chains have become longer and more complex, and involved more and more machinery, there has been a constant pressure of the strong against the weak. The pressure has concentrated more and more of the processes in the chains that are easiest to 'monopolise' in a few areas - 'core' processes in 'core' areas - and more and more of the processes that require less skilled  and more extensive manpower that is easiest to keep at a low income level in other areas - 'peripheral' processes in 'peripheral' areas."

 

Now this is a very revealing quotation. Most summaries of Wallerstein's world - systems model take the existence of core and periphery as axiomatic and start by stating that a world - system is divided into core and periphery. Well clearly that is correct, BUT the two poles are in themselves the product of way the commodity chains function. Cores and peripheries are produced because the core and peripheral processes are concentrated in those areas. It is the clustering of production processes and their relative complexity and automation that produce core and periphery. The commodity chains are thus not only geographically, but structurally constitutive of the world -system.

 

In my reading this has not been a point that Wallerstein has emphasised, and perhaps he himself has not fully drawn the implications of his argument. (Although given the first 1974 volume of The Modern World-System does dwell at length on the development of the grain supply to north western Europe from Poland - Lithuania he could argue that it is there the be teased out). But the commodity chain explication does strongly imply - to my reading - that we need to see the establishment of commodity chains ad both logically and historically prior to the formation of the modern world -system, and possibly all others too. (although Wallerstein is always admirably cautious in retrospectively extending categories of analysis developed for the capitalist world market).

 

We are not down with commodity chains quite yet either. Obviously for such a geographically diverse process of production to function, commodities and part commodities must ne transported between locations. Again Wallerstein has some interesting points to make: "These production processes usually require physical transportation of commodities between them, and frequently the transfer of 'rights' to commodities in a chain are made by autonomous organizations, in which case we talk of the existence of 'commerce.'  Commerce is frequent, but far from universal, as the mode of linkage, and in no way essential to the functioning of a commodity chain, except at the very end when the final consumable product is sold to the final consumer. Both the great merchant companies of the seventeenth and eighteenth centuries and the contemporary multinational corporation have been structures that eliminated much (though seldom all) of the commerce in the interstices of given commodity chains."

 

The above quote is extremely significant. One of the consistent criticisms of Wallerstein from orthodox Marxists is that his model of the world market is geared towards exchange and commerce and slights production; that it is to quote the first such critique by Robert Brenner in 1975, "neo Smithian". Sadly this critique persists to the point of being recycled by Heike Gerstenberger in 2005. But what is clear is that Wallerstein is actually talking about production and subordinated to exchange and / or commerce to production. In his model exchange is not about "trade" but about the changing title to commodities or part commodities during the process of production. In some case, the transfer of title is between discrete economic actors, located in divergent political entities and so can be (mis)represented as trade; in other cases, it occurs as internal transfer within horizontally integrated economic entities. "Trade" for Wallerstein remains merchant trade, and occurs between world -systems, and invariably involves precocities. This is the arbitrage business of merchants and is quite separate to the activities conducted within a world - system which remain oriented to production.

 

Wallerstein also asserts that early modern merchant companies were similar in operation to contemporary multinationals in that they often conducted a form of vertical integration of commodity production. This is an assertion that raises the ire of orthodox Marxist critics, who insist that chartered companies are pure representatives of merchant capital. But is this always the case? There is powerful empirical evidence to support Wallerstein's view. The VOC, for example, can be seen as a merchant capitalist enterprise engaged in the exchange of precocities between two world - systems - silver from the European world - system and spices from the China centred east Asian world -system. But the VOV very quickly transformed itself into a carrier within that East Asian system, shipping silver to India where it was exchanged for textiles, which were shipped to Japan for copper and gold, shipped in turn to Manilla to exchange fro Chinese manufactures which were in turn exchanged for spices at Batavia and Malacca. In this case the VOV functioned as an element in the transfer of necessities within the East Asian system, subordinated to production of commodities. Even more significantly, the VOC itself became the producer of the spices sought in Europe. Thus the inhabitants of Ambon were exterminated in a war in the 1630s and the island resettled with slaves required to grow and supply cloves to the VOC. Here we see the VOC practising vertical integration, producing, transporting and storing a commodity that was turning into a necessity in Europe. And what about the Dutch West India Company, pioneering the sugar plantation complex in Brazil? Here is another case of a merchant company operating as a vertically integrated producer and supplier of a commodity, which as Wallerstein points out, became a necessity for European population growth and manufacture, providing as it did a quick burst of energy to workers engaged in production in the core. And as a final example, the British Royal Africa Company, engaged in the supply of slaves from West Africa to the Caribbean. Slaves after all are also a commodity - labour power - and were a necessity to the production of that new nutritional necessity of the European world -system, sugar. The acquisition and transport of slaves across the Atlantic is thus more than merely merchant capital - it constitutes part of the production of the commodity labour power. The production component of the Africa Company lay precisely in delivering a quantity of slaves, healthy enough and rendered compliant by the brutality of the voyage, to the plantations of Jamaica.

 

my argument here is that the orthodox Marxist insistence on this distinction of merchant capital from productive capital is very hard to delineate in the actual historical record, and depends upon a great degree of abstraction to work. The insistence on maintaining this distinction seems to me more idealist and even moralistic, than anything else. The evidence I suggest, lies with Wallerstein.

 

The starting point then for modelling the world - system, lies in identifying and mapping (as far as possible) the commodity chains at the relevant point in time. Identifying the relative complexity of the production processes involved will allow us to determine the core and peripheral regions within the system. Political entities that contain a predominance of core regions within their borders can be considered as core countries or states, and the same applies with regard to peripheral countries. (As Wallerstein stresses, these are labels of convenience and not accurate; no country is entirely given over to core processes, and peripheral processes can also be found within political borders. Indeed the German world - systemist H -H Nolte, has developed this conception further in his work on "inner peripheries". Core and peripheral countries are shorthand; such concepts are actually not found in Wallersteinian world - systems analysis.)