EL DEBER DE FORMACION DEL EQUIPO HUMANO LA FORMACION PROFESIONAL

Strange’s theory is a good starting point in understanding why states have decided to reduce the barriers both to trade and to investment by MNCs. Yet it rests on too simple a picture of technological change and its place in the production system. In particular, it rests on the unstated assumption that the Wxed cost of R&D are Wxed, not only in the Wnancial accounts of the corporation, but also in place, either geographically or organizationally. Strange does not address this point explicitly, but in order for her argument to make sense we must assume either that R&D always occurs in the same place (or at least in the same country) as production, or that R&D and production occur in the same company. If the fruits of R&D can travel across borders by being sold, licensed, imitated, or stolen, then how would a rise in R&D costs create pressure for increased international trade in goods, or for greater FDI? This brings us back to the theory of the MNC, and the problems of trading, transferring, and controlling productive knowledge.

Productive knowledge is often transferred between companies without trade in goods, and without foreign investment. Pharmaceutical companies, for instance, license the formulae for their drugs to various manufacturers around the world. The same applies to our auto manufacturing example: consider Fiat’s construction of turnkey plants and licensing of designs. The country that was then Yugoslavia built something of an auto industry in the 1970s. That was not because R&D costs were so low in the 1970s that the Yugoslav market, either on its own or even with the added scale aVorded by the export of the late, unlamented Yugo car, could sustain the costs of developing the next generation of automobile technology. It was because the Yugoslav state set up a company that licensed technology and product designs from Fiat: Fiat did the R&D, and Yugo bought the results. Or, to take a more successful example from the same period, Korea’s Hyundai licensed (and continues to license) technology from Japan’s Mitsubishi. It is hard to see how rising R&D costs would have made such deals so much more diYcult.

There are, however, plenty of cases in which productive knowledge has a high tacit component, and so is either diYcult or costly to transfer in this way. There are also cases in which a company controlling productive knowledge doesn’t want to license or sell it, preferring to keep knowledge in-house. Strange’s argument about R&D costs forcing down barriers to trade and barriers to FDI would still hold if either the feasibility of transferring knowledge, or the willingness of companies to transfer it, had declined in recent decades. And there is good reason to believe that changes in the organization of production have made simple, arm’s length knowledge transfer of productive knowledge between companies more diYcult, at least in the case of our automobile industry example.

Since the 1970s, production methods in the auto industry have changed dramatically, as have relationships between customers and suppliers along the supply chain of materials and parts. Many of the changes in question are associated with the production techniques pioneered by Toyota, starting in the 1950s, and with the sort of inter-Wrm relationships typical of Japanese manufacturing. The practices of Japanese manufactur-ers such as Toyota gave birth to such concepts as ‘‘lean production,’’ ‘‘just-in-time delivery,’’ and ‘‘total quality management,’’ which will be known to many readers.

When referring to these practices as a package, I will call them ‘‘Xexible mass produc-tion.’’ Both Xexible mass production, and the classical mass production it has replaced, will be examined in more detail in Chapters 9 and 10. For now, we need to consider brieXy how this change has aVected the development and transfer of productive knowledge.

Until the 1970s, the Detroit-style classical mass production was a standard to which industries around the world aspired. The Detroit system was Taylorist. Taylorism, or scientiWc management, involves giving each production worker very precise instructions about what to do and how to do it. These instructions are formulated by the company’s

24 THE GLOBAL ENVIRONMENT OF BUSINESS

managers and engineers after systematic study – hence the label ‘‘scientiWc.’’ Production workers themselves have little, if any, involvement in the study of methods or formu-lation of instructions.

The Taylorist separation of conception (by managers) and execution (by workers) was paralleled, under mass production, in the relationship between the lead company – be it Ford, Fiat, Renault, Nissan, or some other – and its suppliers of parts. The design and engineering departments of the lead company would design a car, including most of the parts. Some parts would then be made by the lead company, and others purchased from suppliers. Would-be suppliers would bid for the work of making parts on the basis of full technical speciWcations, spelled out on paper blueprints (today it would be CAD/

CAM Wles) provided by the lead company. Companies supplying parts needed some technical capabilities, of course – they needed to be able to make production decisions.

Moreover, for reasons that will be explored in later chapters, relations between mass producers and their suppliers tended to be adversarial; when suppliers did have pro-duction problems, they worked them out on their own. They were not, however, expected to contribute to product design.

The hierarchical separation of conception and execution in the mass production process made both spatial and organizational separation of production activities rela-tively simple. It is easy to miss this point because both spatial and organizational separation are more prevalent in today’s production systems than they were under mass production – then it was common to integrate many stages of production in one building under one ownership, and today we have modular organization and ocean-spanning production networks. Even so, signiWcant spatial and organizational segmentation of production did occur under mass production, and when they did the requirements for interaction between the diVerent production units – the Xow of information, instructions, and products – were much less exacting than they are today. A mass production assembly plant, with its special purpose machines and detailed instructions, could be located almost anywhere, provided there was a supply of parts, and a small number of managers and engineers trained in the company’s methods. Most of the employees in such factories were semi-skilled, and required little special training either in their trades or in the company’s particular methods. Com-panies supplying parts needed the appropriate production capabilities, plus the blue-prints.

The Xexible mass production methods used in the auto industry today make much greater use of the knowledge of both production workers and suppliers. Within the lead company, they emphasize involvement of a skilled workforce in the continuous im-provement of the manufacturing processes they use. Production workers are expected to identify problems and opportunities for improvement, and to participate in developing solutions. This requires ongoing two-way communication and cooperation between production workers and the management/engineering teams. Because these methods

are considered important for both the maintenance of quality and the ongoing reduc-tion of costs, the lead companies expect their suppliers to follow similar practices.

Flexible mass production companies often expect a supplier to design the parts it will be supplying. This is sometimes called ‘‘black box design’’: if we imagine the lead company’s schematic sketch of the mechanical systems for a new car, certain parts and sub-assemblies might not be shown in detail, but represented by simple boxes. The lead company’s designers know what the part represented by the box needs to do, but they do not know how this will be achieved. Details – the contents of the proverbial black box – will be worked out later, in a collaborative process involving both the lead company and the supplier.

Black box design requires – and makes use of – distributed design capabilities. It is particularly useful in circumstances where technology is changing rapidly; under such circumstances, the distribution of design capabilities both makes it possible to keep product design up to date and ensures that suppliers have the technological capacity to adapt their production systems to the fast-changing technologies. Clearly, however, black box design demands more of the suppliers. Small suppliers that were capable of making parts from blueprints were not necessarily able to collaborate on the devel-opment of new parts; even if they were, the lead company may need just one fuel injection system for a particular model of car, so if suppliers are to be involved in design, the lead company will want just one supplier for this product. For these reasons, the transition to black box design has been accompanied by a move to fewer, and larger, Wrst-tier suppliers (Wrst-tier refers to the suppliers that deal directly with the lead company; the Wrst-tier suppliers, of course, have suppliers of their own, and so on).

The upshot of these changes in the way cars are designed and made is that knowledge about how to assemble a car or to make parts for it, has become more diYcult simply to buy on the market. This is not to say that such knowledge is no longer bought and sold – companies today buy and sell technology and design licenses more than they ever have.

With technologies and products changing as fast as they now do, however, making use of such licenses in order to build a car calls for greater technological capabilities than it did before. For that reason, small national car companies and stand-alone subsidiaries of multinationals both are obsolete. Small suppliers, without the capability to collab-orate on design and to supply an international production network, Wnd themselves in a similar situation. For a detailed discussion in the case of Argentina’s car parts manu-facturers, see Marcela Miozzo (2000); for a comparison of the cases of Argentina, Spain, Taiwan, and South Korea, see Mauro Guille´n (2003).

Thus, in the automobile industry, it has become more diYcult for companies serving small national markets to keep up to date on technology, design, and production methods. The reason is not, as Strange supposed, the rise in R&D costs (though that has occurred), but changes in the way products are designed, the ‘‘continuous improvement’’

26 THE GLOBAL ENVIRONMENT OF BUSINESS

approach to reWning production processes, and the consequent changes in relations between companies up and down the supply chain.