Introduction

Industrialization in the North created what has come to be called 'Fordism', referring to the assembly-line technology that played an important role in the success of the T-Ford automobile production at the beginning of this century. The Fordist type of production organization is usually coupled with 'Taylorism' in management. This implies a clear line of command, and a highly-developed division of labour and stratification within the factory, which places management and workers on different sides. This has been seen as a fundamental weakness of many Western economies by many critics, e.g. Best (1990). Furthermore, the Fordist paradigm has come under criticism in the wake of increasing difficulties in balancing an economic scale of assembly-line production with existing consumer demand.

Hence, elements of a post-Fordist industrial paradigm have emerged in the wake of economic troubles in the seventies and the eighties. The central feature of this discourse is the emphasis given to increasing flexibility. In developed countries this means automation, electronic information systems and robotization. In addition, neo-Fordism is based on semiautonomous groups of producers, often small firms. Their co-ordination depends on a centralized information system, which is often located within an established corporation. Through this form of organization, producers are able to respond efficiently to fluctuations in the volume and quality of sophisticated and differentiated consumer demand.

These developments have also suggested that industrialization does not necessarily have to mean an increasing share of mass production in large-scale enterprises. Indeed, the flexible specialization concept puts small industries at the centre of the industrial strategy debate. Piore and Sabel (1984) suggested that the deterioration in industrial performance in a number of Western countries results from the limits of the mass production model and saw flexible specialization as a future alternative. They emphasized in particular the decentralization of big factory chains and redeployment of productive forces in small units, which take advantage of flexible technologies. Dissolution of rigid mass production systems and introduction of more innovative ways of producing, using multi-purpose equipment and employing skilled workers, would enable crisis-ridden economies to react to continuous changes.

Schmitz (1989) explicitly discussed the applicability of these ideas to the Third World. He distinguished a small enterprise variant and a large firm variant of flexible specialization. In the first case, flexible specialization results from the clustering of small firms and a strong interfirm division of labour. The large firm variant exists when large firms decentralize and specialize internally or use specialized suppliers. The latter in particular has produced organizational innovations such as 'Just In Time' (JIT) inventory management techniques.

The small-firm variant of flexible specialization presumes that clusters of small producers can reach collective efficiency. Innovative behaviour is expected in such an environment and competition is tempered by co-operation. Suppliers of parts in the automobile industry can and do compete, for example, but co-operation between the 'assembling' firm and its suppliers, or between suppliers, in solving specific technological problems, also occurs.

Clustering of enterprises can enhance this co-operation and help the enterprises in surviving economic adversity by increasing their capacity to adapt to changes in the environment. In many cases small enterprise clusters are not only able to survive during hard times, but actually increase their share of total production, at the expense of mass producers (Piore and Sabel, 1984:12). This has inter alia been demonstrated by their ability to withstand and respond to the oil and debt crises.

The problem remains of how to measure flexibility, innovative mentality and collective efficiency. The last concept can easily be confused with localization, urbanization or other agglomeration economies. Innovativeness and flexibility at the level of enterprise clusters are not as easy to identify in the field. Several papers in this volume, however, suggest different approaches to this problem.

Enterprise environments and collective efficiency

A key element in the theory of flexible specialization is the realization that the individual enterprise cannot be understood in isolation from the specific environment in which it is operated. Both the structure and the efficiency of the enterprise depend on the products and services available from other private or public enterprises in the area, on the structure and qualifications of the labour force, and on the size and structure of the market. Agglomerations of differentiated interacting activities may achieve what Schmitz (1990) has called collective efficiency. The areas where they are clustered are often called Marshallian (industrial) districts after the English economist Marshall who wrote about such districts in the beginning of the twentieth century (see Pyke, Becattini and Sengenberger, 1990; and Pyke and Sengenberger, 1992).

If needed services or production inputs are not available on the market the enterprise will either have to produce them itself, accept an often much-reduced efficiency, or choose a technology which reduces its needs for external inputs. The choice of technology also depends on the structure of the labour market, and the choice of product depends on the size and structure of the market.

Flexibility in a production system is a response to instability and uncertainty in the market (Salais and Storper, 1992). Large-scale production requires a relatively large and stable market to be profitable. To secure the necessary market stability, large-scale producers are forced to opt out of the smallest and most unstable markets and leave them to niche producers, to subcontractors or un-serviced.

Enterprises venturing into these smaller and more unstable markets have, briefly,...