Science Tribune - Article - January 1997
The erosion of state capacity : Industrial research and public R&D policy in German information technology
Technical University Munich, Germany
E-mail : email@example.com
The author analyses the intervention of the German state in R&D activities in information technology. The development of state capacities, as analyzed in the case of R&D policy, seems to be typical of a general fundamental trend in modern society, namely, the structural de-coupling of social sub-systems such as the state and the economy. In terms of systems theory, it is as if an increase in the internal dynamics of social subsystems has led to a decrease in the state's capacity to intervene purposefully. This is not to say that the era of sovereignty and statehood is over, but rather that the structural preconditions for public policies have changed fundamentally.
Structure and dynamics of the German research system
The project we shall describe was carried out at the Max Planck Institute for the Study of Societies (a) and focussed on research in information technology (IT) (data processing, microelectronics, telecommunications and industrial automation). Three basic characteristics of the German research system were at its origin :
1. The German R&D system is highly differentiated. The most striking feature is the institutional differentiation of public research organizations (among them the Max-Planck Society and the Fraunhofer Society) outside the university system responsible for basic and applied research.
2. Industry accounts for the major share of R&D in Germany and this share has been steadily increasing (64% in 1990). Applied research and experimental development are virtually the exclusive province of industry.
3. The social subsystems involved - sciences and the economy - enjoy a high degree of functional autonomy from the state. This is regarded as a precondition for the German R&D system to function properly.
Consequently, German research policy has always been confronted with the challenging task of guiding and controlling R&D activities without destroying the autonomy of science and industry. The crucial questions that arise - and that we have addressed - are :
- Can the state intervene in the R&D activities of industry in a goal-oriented way?
- Can it promote the technological competitiveness of companies or avoid the risks inherent in new technologies ?
What is the state's room for manoeuver ?
Our approach did not consider "state capacities" as simply the sum of the state's legal competence and of its financial and human resources, but assumed that they are the result of three structural variables :
- the state's internal structure and its resources,
- the industry's internal organization and the companies' resources,
- the inter-organizational relationships between state and industry.
State capacities are thus the product of a complex configuration of public and private actors, each with their own interests, strategies and resources.
Of course, the success of public policies depends on the state's ability to provide and allocate resources in a strategic manner. Without adequate institutional capacities, appropriate instruments, and a degree of strategic coherence, governments would miss even the most favourable opportunities for intervention. But, however generous and well-administered government programmes may be, they will misfire if companies ignore them or use public money to replace internal funds that are already available. The success of industry-related public R&D policies therefore depends on the structural opportunities the companies offer and the restrictions they impose (competitive environment, strategies, internal organization, etc...). In addition, successful policies need close and stable relations between actors in the public and private domain so that governments know where political intervention is necessary and can take advantage of available opportunities.
With regard to neo-institutionalist state theory, this approach is broader than the neo-Weberian approach of, for example, Theda Skocpol et al (1). It is closer to Peter Katzenstein's concept of a policy network (2) and Peter Hall's institutionalist approach (3). In the realm of policy research, we combine the usual "top-down" approach with a "bottom-up" analysis as suggested by Paul Sabatier ten years ago (4).
Funding of German information technology
Our empirical analysis of German information technology (IT) has shown that industrial R&D activities and public R&D policies have been closely linked for a long time. Despite the public commitments to a market economy, every government - regardless of its ideological profile - has tried to promote industrial research in IT. Since the late 1960s, the state has launched many programmes intended to influence and stimulate company activity or, at times, overcome their reluctance. From the time of the first national programme in 1967 until 1989, the Federal Research Ministry spent about 10 billion DM in promoting IT. More than half of this money (approx. 5.5 - 6 billion DM) was given directly to the companies. In addition, German IT received a large share of the total funds distributed by European programmes (ESPRIT I and II, RACE I) in the second half of the 1980s.
How successful was state funding ?
Quite obviously, the German state was very active in promoting the IT industry in order to improve its international competitiveness. But were its interventions successful ? The main results of our project can be summarized in two hypotheses.
I. The institutional hypothesis
The first hypothesis emphasises the institutional contingencies of successful interventions. The results of our project show that the organization of R&D in industry does offer opportunities for purposeful and selective intervention. Public programmes can influence, to some extent at least, the course and the scale of company research. However, these opportunities are restricted by institutional factors both within companies and the state structure. In brief, public interventions into industrial research are possible but limited. This conclusion runs counter to the latest variety of systems theory and to the conventional wisdom of neoliberal economic theory, which both maintain that external interventions into the economy are either impossible or inefficient.
Admittedly, Germany's R&D programmes to promote the IT industry have not been particularly successful, and their failure could well be taken as evidence for the ineffectuality of intervention into the economy and of repealing the laws of the market. If we take a closer look, however, we observe that very particular constellations of factors were responsible for the limited success of the programmes. For example, in data processing, the government chose a highly ambitious strategy in the 1970s and decided to make a frontal attack on IBM in the market in which IBM was strongest, the mainframe business. Such a strategy would have been very demanding and risky anywhere and, in Germany, it soon proved to be too much for both companies and state alike. To implement its strategy, the state needed to coordinate R&D programmes and public procurement effectively but, for institutional reasons and, in particular, because of the decentralized structure of the German state and its bureaucracy, the coordination came to nothing. As a consequence, the share of government contracts in data processing allocated to domestic IT companies was well below that in other countries with a national computer industry (France, US, Japan).
II. The hypothesis of declining state capacity
The second hypothesis argues that state capacity in R&D policy has been declining significantly since the early 1980s. In IT, there is plenty of empirical evidence to show that public intervention in industrial research has become much more difficult and, at times, well-nigh impossible. This is food for thought for those who want the state to engage in large-scale and expensive "high-technology wars". The erosion of state capacity in IT occurs at all levels and has been triggered off by several independent but interrelated factors.
Globalization of technologies, markets and companies
The most important factor is the globalization of technologies, markets and companies which has induced various changes both in the companies' internal structures and in their external relations. With regard to public R&D policy, three particularly important processes have been sparked off :
1. Decentralization of R&D. In the 1980s, companies began to reorganize their R&D activities. They wanted better exploitation of their capacities, speedier development of new products and greater responsiveness to customer needs. In practice, they cut back their central R&D laboratories and integrated R&D activities into their product divisions. This led to a decline in basic research by the big German IT companies (Siemens, Daimler, Bosch).
2. Internationalization of R&D. The major R&D activities of the big IT companies are no longer concentrated in one host country, if they ever were. A company like Siemens conducts about 20-25% of its R&D abroad.
3. Integration of R&D into global networks of inter-firm cooperations. The 1980s witnessed a tremendous increase in the number of joint ventures, strategic alliances and R&D cooperations. The result is a complex global network of inter-firm cooperation. In 1991/1992, Siemens was involved in about 1,400 cooperative projects and, among the company's major partners, there were only two German companies.
Because of all these developments, the target of public R&D policies has changed fundamentally. In contrast to the heyday of national technology policy, when the state had to deal with a small number of clearly distinct national companies, the target of public policies today is an everchanging, intricate network of "multi-domestic" corporations. This new constellation has blurred any distinction between "home" and "host" countries and complicated the definition, if any, of a "national" economic interest. Any "national technology policy to support specific forms of activities has become increasingly meaningless because there are so many leakages in both financial advantages and research results - to non-national companies" (5).
Changes in the internal structure of the state
Since the 1960s, when the first public programmes in IT were initiated, the internal complexity of the state has increased significantly and eroded state capacity. In Germany, despite the establishment of a national ministry responsible for science and research, we have a complex constellation of actors and institutions whose competences and resources are distributed over several institutional levels of decision-making and various territorial boundaries. Most importantly - and most interestingly from a political science perspective - there has been a Europeanization of R&D policies.
In the 1980s, the European Community (EC) became a major player in R&D policy. Based on new legal competences granted by the Single European Act, the EC established a very comprehensive framework of programmes aimed at improving the competitiveness of European industry. The sector that gained the most attention was IT; it was allotted 40% of the EC research budget. The EC programmes (ESPRIT, RACE etc.) targetted the big IT companies in the member states which now receive more than half of their public R&D subsidies from the EC. Admittedly, these subsidies are only a small share of their overall R&D budget (nowadays only 2% of Siemens' research is financed by public money) but they can be highly significant. For example, between 20 and 30% of the European IT companies long-term generic research has been conducted within the various EC programmes.
An additional feature of the Europeanization of R&D policy is EUREKA, an intergovernmental organization established in 1985 to promote cooperation among European companies in high technology. Although EUREKA has no budget of its own and depends on national programmes and funds, it, nevertheless, also adds to the institutional complexity of R&D policy in Europe.
Clearly, R&D policy in Europe has become entangled in a complex, variable geometry of supranational, intergovernmental and national (and sometimes even regional) actors and institutions with their distinctive political interests and institutional logic. This has not resulted in any increase in state capacity - if it makes any sense at all to use the term "state" for such a kind of political order - but just the opposite. Because of the plurality of actors, institutions and programmes, the internal coherence of public policy has been weak, programmes frequently overlap and goals conflict. Microelectronics is an excellent example. In the 1980s, the EC launched ambitious programmes aimed at stimulating R&D cooperations between European companies in order to improve their competitiveness in the face of US and Japanese rivals. At about the same time, however, the German government promoted the cooperation between German and US companies (Siemens and IBM), and regional governments such as North Rhine-Westphalia subsidized the settlement of Japanese companies !! By and large, the companies ignored these confusing lines of political battle and, ultimately, major IT companies, regardless of nationality (Siemens, IBM, and Toshiba), joined forces to develop the next generations of advanced microchips.
Changes in government-industry relations
Globalization of the economy and Europeanization of the state have also transformed relations between the state and companies. The clientelist pattern of interest intermediation, typical of national technology policy in Europe in the 1960s and 1970s, has disappeared and has been replaced by loosely coupled "issue networks" (6). Wherever R&D policy issues are at stake, we now encounter a myriad of round tables, working groups and committees in which the companies are represented. The large number of such bodies do not, however, reflect the companies' influence in policy-making but, rather, their growing difficulty in defending their interests within the complicated political decision-making processes.
This new pattern of government-industry relations offers both advantages and disadvantages. On the one hand, the state's autonomy from industry has increased, on the other, industry's growing independence makes obtaining binding commitments from companies more difficult. And if the companies' cooperation is required for a R&D programme to be successful, this implies that, despite its growing autonomy from industry, the state has, in fact, diminished capacity.
It will be one of the most challenging tasks for the social sciences in the years to come to analyze the chances of success of public policies in this new economic and political context.
(a) The project (1988-1994) was undertaken jointly by Edgar Grande and Jurgen Hausler at the Max Planck Institute. The results were published in a co-authored monograph "Industrieforschung und Forschungspolitik" (1994) in the Institute's publication series and in a Habilitationsschrift by E.G. "Vom Nationalstaat zur europaischen Politikverflechtung" (1994). The project laid the foundations for ongoing research on the Europeanization of R & D policy and on the political dynamics of multilevel systems of joint decision-making.
1. Skocpol T, Evans P, Ruschemeyer D (eds). Bringing the state back. Cambridge University Press, Cambridge, 1985.
2. Katzenstein P. (ed) Between power and plenty. University of Wisconsin Press, Madison, WI, 1978.
3. Hall P. Governing the economy. Polity Press, Cambridge, 1986.
4. Sabatier P. Top-down and bottom-up approaches to implementation research. J Public Policy, 6, 21-48, 1986.
5. OECD. Strategic industries in a global economy. OECD, Paris, 1991, p. 100.
6. Heclo H. Issue networks and the executive establishment. In : The New American Political System. Anthony Krieg (ed). American Enterprise Institute, Washington DC, 1978, pp 87-124.