This paper explores the factors that drive firms' decisions to integrate or disintegrate through a detailed, historical analysis of vertical integration in the disk drive industry. The disk drive was founded and dominated from 1956 to 1985 by firms such as IBM, Control Data, Storage Technology, and Century Data, which were extensively vertically integrated into the design and production of most of the components used in their drives. In the 1980s, these firms were driven from their positions of industry leadership by a group of non-integrated entrant firms such as Seagate, Conner, and Quantum, which designed drives, purchased standard components, and assembled them. Yet even though a progressively larger share of industry units has been produced by non-integrated firms, each of these entrant firms has moved aggressively toward greater integration since 1989. They have, in essence, chosen to swim upstream.
The paper finds two primary reasons for moves to greater or less vertical integration in the industry. The first is the degree of modularity and component standardization in product designs. If components and the way they interact within the design become standardized, the paper predicts that disintegration will occur. Product designers-assemblers that can "mix and match" components procured from competing suppliers have advantages in product development cycle time, flexibility in adapting new technologies, and efficient system design. Technology change may increase or decrease the degree of modularity in design, however, and as a result, an industry might go through cycles of integration, disintegration, and reintegration as technologies emerge, mature, and are supplanted by new technologies.
The second driver of vertical integration is scale economies. Although in the industry's early years there were extensive economies of scale that enabled the largest disk drive manufacturers to earn extraordinary profits, by the 1990s unit volumes had become so great and fixed costs had been so tightly squeezed that the six major drive makers each have similar manufacturing cost positions. As product differentiation has become more difficult, margins in the assembly business have been driven to minimal levels. Scale economics are more attractive in component manufacturing, so assemblers such as Seagate and Conner have integrated even to standardized, modularly interchangeable components such as disks.
There is some evidence that competitive forces over time may flatten industry scale curves so that the attractiveness of being in component manufacturing is likely to diminish over time. Similarly, design economics generally militate toward modularity and component standardization. One might expect, therefore, that the industry will go through multiple cycles of integration, disintegration, and reintegration. The framework developed in this paper to explain changes in the extent of vertical integration in disk drives seems, in fact, to have considerably power in explaining patterns of dis-integration and re-integration in a diverse set of other industries.
TOM
31 pages
| Back to 1995-1996 Working Papers | Copyright © President and Fellows of Harvard College