Flexible Specialization and Systems-house Forms of High-technology Industrial Production

Available published statistics do not allow us to distinguish between flexibly specialized producers and systems houses, as such, within the high-technology industrial ensemble of Southern California. However, most of the small producers are presumably flexible specialists of one variety or another, and many of the largest electronics, aircraft, and space equipment manufacturers can certainly be seen as being systems houses. The question before us is, how do these two segments of the local economy behave and interact, and how do they contribute (or not contribute) to the perpetuation and growth of industrial districts?

Flexible Specialization

In conformity with the powerful line of argument first laid out by Piore and Sabel (1984), many scholars have sought to ground analysis of current technological-institutional structures of production in the

idea of "flexible specialization." The idea is generally considered to apply with special force to firms that produce small batches of output in constantly changing product and process configurations, and it stands in direct opposition to the notion of mass production, i.e., a form of industry characterized by standardized outputs made in extremely long runs. Flexibly specialized firms are thus able to engage in high levels of product differentiation, customization, and semicustomization, and they often turn necessity into a virtue by transforming their endemic instability into an opportunity for constant technological upgrading and design innovation.

Because flexible specialization usually entails a breakdown of internal economies of scale and scope in the production process, it tends to be associated with small or medium-sized and vertically disintegrated units of production. These units then typically become caught up in dense linkage networks through which they are able to tap into significant external economies of scale and scope. With the resurgence of flexible specialization in modern capitalism, therefore, has come a pervasive tendency to reagglomerate economic activity. Sabel (1989) has argued, too, in favor of the view that a convergence between large firms and industrial districts may be occurring, in which the former increasingly externalize many kinds of production activities and are subsequently reorganized within wider transactional networks. We have observed above that in the high-technology industrial districts of Southern California a definite downsizing of production units has been systematically occurring. And this observation is consistent with the wider findings of analysts such as Acs and Audretsch (1990), Loveman and Sengenberger (1990), and Birch (1987), who have shown that significant decreases have occurred in establishment and firm sizes (in manufacturing and services) in the major capitalist economies since the 1970s. This tendency stands in marked contrast to an earlier period over much of this century in which rising concentration seems to have been the rule (Prais 1976).

Systems-house Production

Despite the latter remarks, large production units remain—and will no doubt continue to remain—an extremely important element of the economic landscape of North America, both inside and outside of industrial districts. In the high-technology industrial districts of Southern California today, large production units flourish in particular

abundance in the guise of systems houses. Specifically, the term "systems house" is used here to designate any industrial establishment that (a) has massive internal economies of scope, very often flowing from a design-intensive or R&D-intensive production process, that (b) therefore has a variegated internal structure and employs large numbers of workers, and (c) manufactures complex outputs in small batches, very often (but not necessarily) over a lengthy production period. At any one time, a given systems house may be engaged in several different production programs run largely in isolation from one another, but with certain managerial, purchasing, technical, etc., functions in common. Output specifications may change radically from one program to the next. Sometimes it may take months or even years to complete one production cycle (as in the case, for example, of the Space Platform currently under construction by McDonnell Douglas Astronautics, or the Space Shuttle produced by Rockwell Space Systems Division).

Systems-house producers are, as it were, latent flexible specialists that have been unable to escape from the force of internal economies of scope. That is, in the absence of the various synergies holding their many and differentiated internal parts together, they would presumably fragment into networks of smaller, more specialized producers with greater ease of entry into and exit from different product markets. Moreover, just as flexibly specialized firms were present in the period of fordist mass production (though rarely at the leading edges), so too were systems houses in the form, for example, of major motion picture studios, big publishing houses (but not printing works), shipyards, and so on; and if they are perhaps most commonly found in technology-intensive sectors today, they are by no means restricted to high-technology industry but are also found in a wide variety of other manufacturing and service sectors.

Flexible Specialization Systems Houses, and Postfordism

Together, the flexibly specialized and systems-house segments of Southern California's high-technology economy constitute a peculiar kind of postfordist industrial system. Both of them represent opposite poles of a continuum of technological and organizational possibilities, and at the same time, both stand in marked opposition to mass production. A suggested tripartite ordering of these forms of manufacturing activity is laid out in figure 3.4 where fundamental archetypes are

Figure 3.4
Types of production units arranged by establishment size and
batch size.

arrayed according to characteristic establishment size and batch size. These two dimensions of variation of course stand in as proxies for other variables such as levels of product standardization, routinization of production processes, length of production cycle, and so on. To be sure, any given production unit may not in reality conform precisely to one of the three archetypes, and composite forms can be identified, as for example, in the case of specialized car manufacturers that combine elements of all three and would hence tend to be located somewhere in the interior of the triangle shown in figure 3.4.

Notwithstanding their evident differences from flexibly specialized producers, systems houses are assuredly an authentic element of the postfordist flexible economy. They make much use of flexible equipment in the form of computerized and robotized machines, they focus on the production of small batches of varied products that may go through many design changes as they are in process, and they drive forward large segments of the rest of the flexible economy, for even though they may be marked by intense internal economies of scope, they invariably consume a wide range of diverse, variable, and often unpredictable inputs (whereas in mass-production industries the trend is toward standardization and streamlining of inputs as far as possible). Systems houses, in brief, are typically interconnected via intricate transactional networks with large numbers of flexibly specialized

producers. For the same reason, they are very often found in dense flexible-production agglomerations. In the high-technology systems houses of Southern California, workers are frequently unionized, with the International Association of Machinists and the United Auto Workers the dominant organizations. Workers employed in smaller units of production in the region, by contrast, are rarely unionized. Paradoxically—but as a reflection of the genesis of Southern California's high-technology industrial systems houses in the period from the 1930s to the 1950s—union contracts tend to be rather "fordist" in nature. These contracts typically lay much emphasis on detailed job descriptions, seniority in layoff and recall procedures, and a strict line of demarcation between management and labor. In many establishments, attempts are now being made to reorganize the old labor-relations system and above all to set up workers' teams with rotating job responsibilities, though the latter experiments are still in a state of flux. In a number of recent interviews with representatives of management, it was also found that high-technology systems houses in Southern California rarely put just-in-time methods of input-output organization into practice, probably because of the variability and unpredictability of the production schedules that they face.