CRIC Project - Patent Analysis of Trajectories of Innovation: The Chemical Sector
(Project completed)

• The Problems Addressed
• Empirical Work and Data Sources
• Key Results and Outcomes
• Significance of Results and Outcomes
• Key Publications

The Problems Addressed

This research fits into the CRIC research agenda by addressing the micro-foundations of innovation systems and innovation policy. Innovation 'systems' have been conceived in the literature in a variety of ways. Here we focused on the fact that systems can emerge and develop around particular technologies and leave a very tangible record in the patent data. This provides a methodological opportunity for mapping systemic behaviour. The scope of the research was to deepen our understanding of the 'dynamic processes' by which the competitive and innovative landscape underpinning the chemical industry have changed. A specific focus was to analyse the innovation processes shaping the trajectories of technological development.

Empirical Work and Data Sources

We chose to look at chemical and related technologies and the chemical and allied products industry for several reasons. The chemical industry is interesting from a dynamic process case-study perspective given the major changes taking place in both its own behaviour and in its innovative and competitive environment. Moreover, chemicals is one of the few technological sectors where Europe still enjoys a substantial advantage in international competition and innovativeness. This cannot be said of many other sectors. Evidence revealed (see key findings below) how a number of sub-sectors of the industry have played an important role in several technological (and economic) upswings in the twentieth century.

Patents are considered a more important strategic weapon by firms in chemicals and related areas than in other sectors, so that trends in chemical patenting may be considered to represent trends in inventiveness and innovativeness quite closely. Hence, the empirical work was conducted using panel data analysis of patent data and associated patent statistics. Patents are particularly appropriate for this kind of study of the structural dimensions of change and development as they are recorded at a detailed level of disaggregation as well as over long historical periods. However, for some parts of the study we used a combination of both quantitative indicators (i.e. all individual and corporate patents granted in the US since 1890) and qualitative case study accounts, as we have found the two approaches provide complementary information. Where appropriate the chemical industry and chemical technologies are also analysed in relation to the wider external environment including other sectors within electrical/electronics, mechanical and transport.

Moreover, until now our ability to interpret the dynamic processes of technological development and corporate innovation has been severely hampered by the lack of suitable indicators. This is especially true for historical studies. Patent data have been used extensively in recent studies, but almost invariably these data have been constrained to US patents data for the relatively recent past (from 1969 or later, after which time the US patent Office has made partially available its raw data on corporate affiliations of patents granted). Other than in a very rudimentary way, the long-term issues have not been clarified. This study has been a part of improving on that situation by permitting detailed insight into the corporate structure of patenting over a hundred years period (1890-1990); using accumulated patent stocks calculated from individual and corporate patenting of leading American and European companies belonging to corporate groups in major international industries, classified at a very detailed level of disaggregation.

Key Results and Outcomes

The analysis addressed three broad themes.

Theme 1: The evolution of technological paradigms

Increased inter-relatedness:

The first study examined the evolution of technological opportunities historically. By analysing the complexities behind changing technological opportunities, technological 'regimes' were traced empirically.

The results suggest the existence of two major technological regimes during the 20^th century. The first regime extended from the opening of the twentieth century until the 1940s. It was characterised by intra-sector technological diversification and the formation of a structure of specialised engineering and science-based fields. In essence this regime was one in which firm patenting behaviour shows a sharpening of the technical boundaries between the 'grand sectors' of manufacturing (science-based Chemicals and Electrical/electronics technologies on the one hand, and the engineering-based Mechanical and Transport technologies and Non-industrial technologies on the other hand). By contrast the second regime, in the second half of the century, shows more interaction between these fields. It is suggestive of an historical shift towards more integrated technological systems through the fusion of diverse and formerly separate branches of technology.

The new paradigm governing the evolution paths of trajectories of technological opportunities builds to a greater extent on inter-group complementary and interrelatedness rather than more isolated individual channels of development. Hence, this evidence revealed how the evolution of technological opportunities has become increasingly interrelated, broad-ranging and complex, in which trajectories previously following isolated channels of development are bought together.

Developments complement and extend rather than substitute:

By exploring the process by which technological evolution has increasingly been channelled into more complex technological systems, we have also shown that knowledge embodied in old technological fields within old systems is integrated in newer systems. This suggests that knowledge embodied in old paradigms is generally not destroyed but complemented and extended in new ones. Hence, new innovation systems tend to be offshoots of a creative incremental technological development process in a variety of areas.

Theme 2: The S-shaped image of the technological growth cycle and the diversity of technology dynamics

The S-shaped image of the technological development patterns:

Since the work of the business cycle theorists in the 1930s, no attempts have been made to study empirically the long term evolution paths of individual technologies using long time series. Hence, the second theme of the study considered the cyclical nature of technological development, in which special attention was paid to the presence and confirmation of the Sigmoid-shaped (or S-shaped) image of the technological growth cycle.

Examples of S-shaped growth paths in technological activity within chemicals:

The diversity of technology dynamics:

The nature of the S-shaped technological growth curve was conceptualised and extrapolated in depth using the many sub-sectors of the chemical sector as the units of analysis. Each technological growth cycle was characterised in relation to the parameters and properties which guide its evolutionary path. Special attention was given to characterising the common phases of cyclical development, and to identifying the duration of each technological growth cycle as well as the timing of 'take-offs' in innovative activity. The study confirmed the very diverse nature of technology dynamics in terms of cycle duration, the timing of technological takeoff, technological opportunity, and the level of accumulated socio-economic capability or historical impact.

Three historical technological upswings periods:

The study then explored how these technological cycle takeoffs appeared to be clustered within certain historical epochs. They were subsequently analysed in relation to the industrial sectors with which they co-evolved.

Note: It is relevant to mention that, as this research is based on a patent data classification scheme, we only measure activity in a certain technological area rather than a specific innovation. Thus, when trajectory takeoffs re-appear within new innovation systems over time, they vary significantly in nature and content as well as form, function and impact.

Overall, the study demonstrated two time-bands (normally associated with upswing periods) in which chemical technological systems took off since the opening of the twentieth century, and we would forecast a third.

The first (around 1930) comprised a system based on coal as a feed-stock, chemical process technology and technology related to products such as rubber and dyes. The complex of interrelated industrial sectors participating in this evolution was a relatively narrow range of sectors most associated with the competencies in the component technologies of the system, due to strong technology-industry overlap. The second technological upswing period (in the 1960s) launched a technology system based on petrochemical feed-stocks and synthetic materials with broad industrial applications. It was a much more complex system with extended technological boundaries both in terms of the total number of technological component technologies making up the system and in terms of the different overall scope of those component technologies. Finally, we forecast an emerging system which moves us into the life sciences and the exploitation of bio-technology, with a large scope for the pharmaceutical industrial sector and potentially for the food sector.

Theme 3. Industrial Dynamics

Innovation processes were then examined in relation to Industrial Dynamics which proved to be an appropriate way to shed light on the distributed nature by which trajectories evolve. In the Industrial Dynamics approach long term competitiveness is analysed in relation to the firm's dynamic supply capabilities (e.g. technological capabilities, adjustment capabilities, local networks), or the co-operative arrangements among firms and other institutions. Hence, Industrial Dynamics is compatible with the resource-based theories of the firm.

Hence, this part of the study on trajectories of innovation applied a capability perspective to the dynamics of industries. The changing composition of an industry system is derived from the changing distribution of capabilities of firms within industrial sectors, participating in the generation and exploitation of a common knowledge base. In this context, the co-evolution between the wave-like development patterns of clusters of innovation trajectories (identified in Theme 2) and the corporate capabilities within industrial structures were identified. This 'competence bloc' view on systemic change illuminated the systemic aspects of innovation and competition at the microeconomic level. It highlighted how industry systems evolve within technological structures, and vice versa.

The study illustrated how industrial sectors are coming together in adopting and developing technological trajectories within broader technological systems. This is in turn also reflected in the industrial sectors' broadening their technological competencies. This indicated an historical trend towards a more complex industrial base, and a more distributed future development of technologies. That is, distributed innovation processes within industrial systems supporting the development of technological trajectories, as they evolve through several decades, were empirically derived in the research.

Significance of Results and Outcomes

The work in theme 1 supports the view that increasingly inter-related technology changes evolve in an incremental, accumulative and path-dependent fashion in which development complement and extent rather than substitute. The results of theme 2 have implications for our understanding of the evolutionary paths of individual technologies, and the way we might understand technological systems, and evolving waves of innovation, change and development. In particular the results within theme 1 and 2 indicated that although patterns of technological development within systems change in nature, function and impact historically, the knowledge embodied in old systems (e.g. based on the 1930s interwar technology) are successfully 'transferred' into another (e.g. based on the 1960s postwar technology).

The research in theme 3 illustrated how the knowledge-bases upon which firms draw, exploit and develop have become increasingly complex, and how this co-evolves with the changing business opportunities governed by evolving technological regimes. The research within this theme also illustrates the limitations of the standard industrial classification scheme, if firms are to be grouped in relation to their closest 'dynamic' competitors, and why a system perspective is more appropriate.

The overall research project on the analysis of trajectories of innovation shows a historical shift towards more integrated technological systems through the fusion of diverse and formerly separate branches of technology. This is combined with an historical trend towards a more complex industrial base, and a more distributed future development of technologies in which industrial sectors are coming together in adopting and developing technological trajectories within broader technological systems. This is in turn also reflected in the industrial sectors' broadening their technological competencies.

Overall, the study reinforces and adds to the CRIC view of how the emergence and shaping of innovative activity may be distributed across a wider matrix of structures, inter-relationships, constraints, and tradeoffs throughout the economic system.

Key Publications

Andersen, B. (1998): "The Evolution of Technological Trajectories 1890-1990", in Structural Change and Economic Dynamics, Vol.9.no1. Pp.5-34

Andersen B (1999): "The Hunt for S-shaped Growth Paths in Technological Innovation: A Patent Study, in Journal of Evolutionary Economics. 9, Pp.487-526

Andersen, B. (1999): "The Complexity of Technology Dynamics: Mapping Stylised Facts in Post-Schumpeterian Approaches with Evidence from Patenting in Chemicals 1890~1990", in Francisco Louca (ed.): Perspectives on Complexity in Economics, UECE Press (ISEG), Pp.111-142

Andersen B (2000): Technological Change and The Evolution of Corporate Innovation: The Structure of Patenting 1890-1990. Edward Elgar: Cheltenham

Andersen B and Lundvall B-Å (forthcoming): "Industrial Dynamics", in Michie. J. (ed), Reader's Guide to the Social Sciences, Vol.I and II, London: Fitzroy Dearnorn Publishers

Andersen, B, Metcalfe J.S., Tether, B.S. (2000): "Distributed Innovation Systems and Instituted Economic Processes", in Metcalfe, J.S. and Miles, I (eds.) Innovation Systems in the Service Economy: Measurement and Case Study Analysis, Dordrecht: Kluwer, Pp.15-42

Andersen, B and Walsh V. (2000): "Co-evolution within chemical technology systems: A competence bloc approach", Industry and Innovation. Vol.7 No.1. Pp. 77-115

[Back to projects]

Top

CRIC has combined with PREST to form the Manchester Institute of Innovation Research (MIoIR).

New book: Trust in Food, A Comparative and Institutional Analysis by Unni Kjaernes, Mark Harvey & Alan Warde.

CRIC Final Report to ESRC:"Main Report" and "CRIC Performance Indicators 1997-2006".

'Instituted Or Embedded? Legal, Fiscal and Economic Institutionalisation of Markets' by Mark Harvey

'Beyond Efficiency and Market Shares: Competition within the Finnish Games Industry' by Mirva Peltoniemi

'Accounting for Economic Evolution: Fitness and the Population Method' by Stan Metcalfe

'Innovation and Final Consumption: Social Practices, Instituted Modes of Provision and Intermediation' by Andrew McMeekin & Dale Southerton

'Alfred Marshall’s Mecca: Reconciling the Theories of Value and Development' by Stan Metcalfe