Sources of resilience in the computer and software industries in France

Industry and Innovation, Aug 2001 by Nohara, Hiroatsu, Verdier, Eric

The history of the computer industry is usually divided into four main periods. The first was dominated by mainframe computers that were capable of undertaking large-scale computation tasks. The second saw the introduction of integrated circuits and the development of minicomputers. The third was characterized by the development of the personal computer, a phenomenon made possible by the invention of the microprocessor. We are now in the age of PC networks and the expansion of the "network of networks", the Internet.

Each new technological age has witnessed the emergence of new firms that have shaken the leading companies of the previous period without dislodging them from the markets in which they were previously positioned or preventing them from entering new markets (Malerba et al. 1998). The first of these leading companies was IBM. As an integrated producer of computer systems, "Big Blue" long played a dominant role and was instrumental in globalizing the computer industry. The next in line was DEC, with minicomputers, followed by Apple, Commodore and Compaq as PC specialists and, above all, the beneficiaries of the "vertical disintegration" of the industry, chief among them Intel and Microsoft. "Wintelism" (Borus and Zysman 1997) led to the decline of "proprietary systems" and placed designers of "operating systems" and microprocessors in a key position, to the detriment of computer manufacturers, "Big Blue" notably, but Compaq and Toshiba as well.

This shift in market power became even more pronounced with the diffusion of the Internet. The new beneficiaries were the producers of applications (SAP, Adobe), of interfaces (Netscape) and of languages (Sun) and "pure product definition companies", such as Cisco and 3COM. The result was a move away from proprietary systems towards open systems, which ensured compatibility between the standards of the various suppliers whose products and services provide the foundations on which the networks depend. Control of these standards and of the associated intellectual property rights are essential resources for those seeking to obtain competitive advantages in these new markets in the IT industry.

These characteristics, briefly summarized here, have already been described at length by many analysts of the changes that have affected technological paradigms in the IT industry (Mowery and Rosenberg 1998). However, this history of the industry is obviously to a large extent that of the American IT industry, since the pre-eminence of the American companies went unchallenged, with just a few exceptions. Japan was successful for a time in resisting IBM's domination by adopting the same type of mission-oriented policies as those developed in Europe, but even the Japanese finished by capitulating in the face of "Wintelism".

Moreover, the brief historical survey presented here runs the risk of making this account of IT history seem a "natural" one, the mere result of technological progress alone. In fact, this American pre-eminence induces us to examine the institutions and modes of industrial organization that fostered this particular form of path dependency ii l'americaine, in which success followed success in an uninterrupted sequence.

This acknowledgement, commonplace and trivial in itself, of the wide gap that exists between the European IT industry, particularly the French one, and its American counterpart leads us to combine this sectoral analysis with the national innovation systems approach (NIS) as developed, for example, by Freeman (1987). Freeman defines a NIS as a network of institutions operating in the public and private sectors whose activities and interactions introduce, modify and diffuse the new technologies. This approach stresses the specificity of the choices that shape the various national systems, in particular through public policies on education, academic research, legislation on patents and intellectual property and access to finance for emerging technologies.

The result is a certain dynamic irreversibility contained within "particular institutional infrastructures" (Dosi et al. 1988; Nelson 1993; Edquist 1997). Firms draw on the institutional resources of their countries of origin in order to construct their competitiveness and, more generally, to operate effectively in globalized markets. The interaction between firms and these institutions gives them access to more or less effective organizational and technological learning processes through which national industries acquire their particular configuration (Lundvall 1992).

From our point of view, the case of the French IT industry provides an emblematic case of the "embeddedness" of technological and industrial development in national innovation systems. This paper will attempt to explain why, at the various stages in the development of the IT industry, French hardware manufacturers were by and large incapable, despite certain flashes of inventiveness, not only of translating technological advances into industrial products but even of understanding the new opportunities those advances offered. We will focus in particular on the effects of the so-called "national champion" policy that was embodied in the Bull company and its various earlier incarnations. Over a long period, the French state maintained a policy that was commercially very protective and financially advantageous; it involved the payment of subsidies on such a scale that it became very difficult for new firms to enter the market. In the USA, on the other hand, the emergence of new competitors had long been strongly encouraged by access to venture capital; the success of this procedure was such that it became the cornerstone of current European technological policies.