KNOWLEDGE INFRASTRUCTURE, TECHNICAL PROBLEMSOLVING AND INDUSTRIAL PERFORMANCE: ELECTRONICS IN BRITAIN AND FRANCE
Geoff Mason *, Jean-Paul Beltramo ** and Jean-Jacques Paul **
* National Institute of Economic and Social Research, London ** Institut de Recherche sur l’Economie de l’Education, CNRS et Université de Bourgogne, Dijon.
Discussion Paper No. 189 November 2001
Acknowledgements: We are grateful to the UK Economic and Social Research Council for financial support for this project (Ref. No. L323252028). We would like to offer particular thanks to the companies, research institutes and other organisations which participated in the research and to Pari Patel for the use of SPRU patent data. We are also grateful for comments received on an earlier version of this paper presented at the DRUID Summer Conference on the Learning Economy, University of Aalborg, Denmark, June 2000. Responsibility for any errors is, of course, ours alone. Address for correspondence: Geoff Mason, NIESR, 2 Dean Trench Street, Smith Square, London SW1P 3HE E-mail:
[email protected]
ABSTRACT This paper reports on a detailed comparison of matched samples of electronics establishments in Britain and France to assess how cross-country differences in external research interactions and knowledge sourcing affect relative industrial performance. Recent improvements in British innovation and trade performance have been associated with an apparently more dynamic and market-driven pattern of research interactions and exchange of ideas with customers as well as a faster rate of new relationship-building with universities than was found in France. Knowledge transfer between organisations in Britain is further stimulated by the regular recruitment of experienced personnel who bring new personal networks of contacts with them. However, French establishments may in future benefit from their close relationships with public sector laboratories which support a relatively high level of investment in basic and strategic research.
1
1. Introduction
Continued disparities between countries in economic performance combined with the rapid pace of technological change have led to widespread interest in learning more about the extent to which such disparities may be attributable to differences in national-institutional structures. In this context particular attention is paid to intercountry differences in key components of ‘knowledge infrastructure’ – for example, universities, research institute networks and publicly-funded research and knowledge transfer programmes – which Smith (1997) suggests are best understood as the endproduct of cumulated investments in ‘social overhead capital’. 1 This research interest reflects growing awareness that, within any national economy, innovative performance now depends on the speed and efficiency with which knowledge is distributed and utilised as well as on national capabilities for generating new knowledge. In many industries the codification of an increasing proportion of the knowledge stock (Arora and Gambardella, 1994) together with the development of electronic networks for knowledge distribution has contributed to the development of ‘new modes of research’ which are more overtly based on collaborative techniques and the re-combination and re-use of existing stocks of knowledge (David and Foray, 1995). Furthermore, new technical advances have increasingly come to depend on the combination of previously distinct technologies, for example, the use of new materials in electronics. In this context, many authors consider that universities, publicly-funded research institutes and other elements of knowledge infrastructure can play an important role in promoting knowledge transfer between organisations and in assisting enterprises to identify and make use of knowledge generated elsewhere, often in fields which were previously beyond those enterprises’ internal competences.
This paper seeks to contribute to empirical investigations of how differences in national knowledge infrastructures affect firms’ external knowledge sourcing activities through a detailed comparison of matched samples of electronics
2
establishments in Britain and France. This comparison was expected to be of particular interest because, while there are several points of similarity between the British and French electronics industries (such as the historically important role of government spending on defence-related R&D), more recent developments appear to have contributed to very different operating conditions for electronics producers in each country.
Firstly, there have been very different rates of inward investment by foreign-owned electronics producers into Britain and France over the last two decades. For example, by 1996 the US foreign direct investment ‘position abroad’ in electric and electronic equipment manufacturing, a cumulative stock measure of capital that US-owned parent companies have provided to foreign affiliates, was some five times larger in Britain than in France. 2
Secondly, over the last 15 years, the British industry has experienced a much faster rate of ‘opening up’ of hitherto protected domestic markets for telecommunications and defence equipment than has occurred in France. For example, ever since the preparation of British Telecom (BT) for privatisation in 1984, BT has been free to diversify its capital equipment purchases between a wide range of foreign-owned and domestic suppliers (Morgan, 1989). By contrast, the leading French telecoms equipment suppliers have only recently been exposed to foreign competition in their home market (Brousseau, Petit and Phan, 1996 ; OECD, 1999).
Thirdly, marked differences have arisen between Britain and France in the role and functioning of ‘national laboratories’ which traditionally accounted for a large proportion of basic and strategic research in each country. In Britain formerly publicly-owned laboratories have been privatised and are now obliged to generate a large proportion of their annual income through short-term industrial consultancy 1
See Smith (1997: 92-96) for a discussion of the main reasons why knowledge infrastructure -- in common with physical infrastructure such as roads, harbours and electricity and telecommunication networks -- is more likely to be provided by the public sector than by private firms. 2 Estimated values of US foreign direct investment position abroad in electric and electronic equipment in 1996: US$3256 million (Britain); US$652 million (France). Source: Economics and Statistics Administration / Bureau of Economic Analysis, US Department of Commerce. This position is
3
activities. The same applies to the great majority of Research and Technology Organisations (RTOs) which were initially established with the primary objective of supporting private industry through knowledge and technology transfer activities (Rothwell and Dodgson, 1993). This contrasts with French public labs which are still largely financed by public funds and have thus been able to remain deeply involved in strategic research with time horizons of three years or more before any prospect of industrial commercialisation (Lawton Smith, 1997). 3
Fourthly, on a common measure of innovation output (patents granted in the US), the once very wide gap between the two industries now appears to have closed – a development that is surely ripe for investigation. During the 1970’s the French electronics industry enjoyed a substantial advantage over the British industry in patenting which peaked at a ratio of 2.4:1 in 1979 (Figure 1). 4 However, since then there has been a gradual decline in the French: British patent ratio to a point of nearparity in 1995 and 1996. This British catch-up in electronics patenting reflects, firstly, rapid growth in patenting by foreign-owned companies based in Britain; and secondly, a levelling off in the number of new patents granted each year to both foreign-owned and domestically-owned electronics companies in France (see Appendix Figures A1 and A2). 5
The relative decline in French patenting performance has occurred in spite of a rising average level of R&D expenditure (expressed as a proportion of output) in key sectors of electronics which is now considerably above the average level of R&D spending in the same sectors in Britain. For example, in radio, TV and communication equipment industries (including electronic components), the British R&D/output ratio fell precipitately between 1982-89 and in the last year for which
measured as the year end value of US parents’ equity (including retained earnings) in, and net outstanding loans to, their foreign affiliates. 3 ‘Public labs’ here refers to non-university research institutions which typically receive core funding from governments in excess of 50% of their financial resources. This category does not include the Centre National de Recherche Scientifique (CNRS) in France because of its assimilation into university research. 4 During this period heavy state subsidies for industrial R&D in France were coupled with the close involvement of government ministries in the restructuring and modernisation of the telecoms equipment sector (Morgan, 1989; Pavitt and Patel, 1989). 5 In 1996 foreign-owned companies accounted for 84% of total British electronics patents compared to only 37% in France.
4
comparable data are readily available (1993) was only at a level of 5.3% compared to 12.3% in France.
6
Hence, in this period there appears to have been a considerable
divergence between input and output measures of electronics innovation performance in France relative to Britain. Furthermore, the pattern of relative performance in electronics implied by recent trends in patenting is broadly supported by indicators of trade performance as well. For example, in telecommunications equipment, UK exports rose from 4.2% of total OECD exports in 1985 to 7.6% in 1995; the equivalent proportions for French exports were 4.0% in 1985 and 5.6% in 1995.7
The paper is ordered as follows: Section 2 describes the methods and sample selection procedures used in our comparison of matched samples of electronics establishments in the two countries. Section 3 reports on British-French contrasts in within-enterprise knowledge flows of relevance to technical problem-solving. Sections 4 to 7 then compare the external knowledge sourcing activities of the two samples of establishments in some detail, focussing on their R&D linkages with other firms such as customers and suppliers and with universities and specialist research institutes and public laboratories. Section 8 concludes with a brief review of our main findings.
6
7
Source: OECD, DSTI (ANBERD and STAN industrial database). Source: OECD, Foreign Trade by Commodities, various issues.
5
Figure 1: Ratio of French to British patents granted in the United States in electronics and All Sectors, 1969-96 2.50
Ratio of French to British patents
2.00
1.50
1.00
0.50
Electronics
95 19
93 19
91 19
89 19
87 19
85 19
83 19
81 19
79 19
77 19
75 19
73 19
71 19
19
69
0.00
ALL SECTORS
Source: Data supplied to Science Policy Research Unit (SPRU) by the US Patent and Trademark Office Industrial classification (SPRU product areas): Electronics: 24 Telecommunications, 25 Semiconductors, 27 Calculators, computers and other office equipment, 28 Image and sound equipment, 29 Photography and photocopy.
6
2. Methodology and sample selection
The specific product area chosen for investigation was a fast-developing class of electronic components which have a wide range of potential applications in industries as diverse as telecommunications equipment, machinery manufacturing and defence systems.
8
In this sector the term ‘components’ typically refers to both physical
devices and ‘sub-systems’ which incorporate new software developments and is by no means confined to hardware. This type of product area was selected in order to focus attention on a ‘leading edge’ technology in which knowledge transfer between the science base and industry was a ‘live issue’ in both countries. In addition, the chosen product area had the attraction that in both countries ‘indigenous’ enterprises were well represented in the sector as well as third country multinationals. In total, visits were made to 26 British and 22 French establishments during 1996-97. As shown in Table 1, some 31 of these visits were to corporate establishments (either production sites, detached research centres within larger enterprises or smaller establishments which were principally engaged in R&D and prototyping activities at the time of our visits). The remaining visits were to private, non-profit or public research institutes, university departments or other organisations such as regional development agencies. Given the relatively small size of samples involved in labourintensive research of this kind, care was taken to ensure that the two samples of establishments were not only well-matched for product area but also comprised a mix of establishment employment-sizes which broadly reflected the population sizedistributions for the wider electronic component industries in each case (Table 2).
9
In both countries the corporate establishments comprised a mix of largely ‘domestic’ small and medium-sized enterprises and larger sites owned by multinational enterprises. Thus, both samples included establishments owned by large domestically-owned enterprises which are among the largest R&D spenders and producers of electronics patents in each country. At the same time the British sample 8
More precise details about the selected product area are withheld to ensure the confidentiality of participating companies and research organisations. 9 Estimated medians and quartiles of enterprise size distributions in the electronics components industries are as follows: Britain median 350 employees; lower quartile 85; upper quartile 1850; France 375, 60, 1600 (Sources: CSO, Annual Census of Production, 1993; SESSI, Enquete Annuelle d’Entreprise, 1995).
7
included a higher proportion of foreign-owned enterprises in order to reflect the greater importance of such enterprises in the wider British industry. The visits were geographically clustered in the following regions: Southern England, Scotland, Brittany, Ile de France and the Rhône-Alpes.
The establishments visited were initially identified through trade directories and listings of research organisations. Detailed information about employment and principal activities was sought by telephone before formal requests for visits were made. In Britain two thirds of corporate establishments which were formally approached for a visit agreed to participate. In France the response rate was just over 90%. The principal form of information-gathering during the visits was extensive semi-structured interviews with senior technical directors or managers. Nearly all visits included direct observation of production and/or research facilities. The visits lasted between half a day and a day and, if necessary, were followed up with further detailed enquiries by telephone or letter. The main characteristics of sample establishments are summarised in Table 2 together with code numbers used to refer to extracts from managerial interviews in subsequent sections.
8
Table 1: Number of establishments visited by type of establishment and employment size-group BRITAIN
FRANCE
Type of establishment: Corporate establishments: -- Production (a) -R&D, prototyping (b) -- Detached R&D centres (c)
10 4 3
8 5 1
Private non-profit research organisations/institutes
4
0
Public research institutes
0
4
Universities
2
1
Other organisations
3
3
TOTAL
26
22
Employment size-group (corporate establishments only): Under 100 100-499 500-999 1000-plus
6 6 2 3
5 8 0 1
TOTAL
17
14
Notes: (a) Production establishments including on-site research, design and development departments. (b) Establishments where R&D expenditures represented more than 50% of annual turnover. (c) Detached research centres within larger enterprises
9
Table 2: Main characteristics of samples of establishments
CODE
Establishment size-group (no. of employees) Principal activity (a)
BRITAIN B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 B11 B12 B13 B14 B15 B16 B17
1000-plus 1000-plus 500-999 500-999 100-499 100-499 100-499 100-499 Under 100 Under 100 Under 100 Under 100 1000-plus 500-999 100-499 Under 100 Under 100
Production Production Production Production Production Production Production Production R&D/prototyping Production Production R&D/prototyping Detached R&D centre Detached R&D centre Detached R&D centre R&D/prototyping R&D/prototyping
FRANCE F1 F2
1000-plus 100-499
Production Production
F3 F4 F5 F6 F7 F8 F9 F10 F11 F12 F13 F14
100-499 100-499 100-499 100-499 100-499 100-499 Under 100 Under 100 Under 100 100-499 Under 100 Under 100
Main end-user market (b)
Region (c)
Ownership and geographical orientation (d)
Telecoms Defence Telecoms Defence Telecoms/Defence Diverse industries Diverse industries Diverse industries Telecoms Telecoms Defence Diverse industries Telecoms Telecoms Defence Diverse industries Diverse industries
SE Scot SE Scot Scot SE Scot Scot SE SE SE Scot SE SE SE Scot Scot
FMNC DMNC FMNC DMNC DMNC Domestic FMNC FMNC Domestic Domestic Domestic Domestic FMNC FMNC DMNC Domestic Domestic
Ile de F R-A
DMNC DMNC
Ile de F Brit Other Other R-A R-A R-A R-A R-A Brit Ile de F Brit
DMNC FMNC DMNC DMNC DMNC Domestic DMNC Domestic Domestic FMNC Domestic Domestic
Defence Defence/ Diverse industries Production Telecoms Production Telecoms Production Telecoms Production Defence Production Telecoms R&D/prototyping Defence Production Diverse industries R&D/prototyping Diverse industries R&D/prototyping Diverse industries Detached R&D centre Telecoms R&D/prototyping Diverse industries R&D/prototyping Diverse industries
(a) See notes to Table 1 (b) Establishments are classified as serving particular end-user markets if those markets account for in excess of 50% of total sales. (c) SE = Southern England & East Anglia; Scot = Scotland; Ile de F = Ile de France; R-A = RhôneAlpes, Brit = Brittany; Other = Other French regions. (d) Codes refer to parent companies as follows: FMNC = Foreign-owned multi-national; DMNC = Domestically-owned multi-national; Domestic = Other domestically-owned enterprises.
10
3. Internal versus external knowledge search and exchange
In electronics, as in many other industries, the pace of innovation has accelerated in recent years, not just because of rapid progress in science and technology but also because of increasing competitive pressures. These developments provide strong motivations for researchers to look outside, as well as within, their firms for relevant information on both the specialist needs of customers and the technical possibilities of meeting such needs.
Nonetheless, in both Britain and France it is commonplace for researchers to first look within their own enterprises for the ideas or knowhow needed to grapple with technical problems. Results from the European Commission’s Community Innovation Survey (CIS) during the 1990’s show that, when ‘innovating firms’ were asked to rank information sources for their ‘importance’ in enterprise innovation strategies, internal sources were highest-ranked in both countries. As might be expected, the importance attached to internal information sources was considerably greater in large enterprises than in smaller ones (Bosworth and Stoneman, 1996; Craggs and Jones, 1998; SESSI, 1995). Within large multi-site enterprises, the scale and effectiveness of internal knowledge search and exchange can vary according to managerial choices regarding the organisation of R&D activities. In both Britain and France we encountered a few large companies where our interviewees expressed concerns about organisational structures (such as product divisions being set up as separate profit centres) that inhibited communication and collaboration between divisions and between each division and central R&D facilities. It is beyond the scope of this paper to seek to generalise about inter-country differences in this respect. However, we did identify striking British-French differences in within-enterprise knowledge flows which arose from the larger proportion of foreign-owned (particularly US-owned) establishments in the British sample.
In cases where US-owned R&D facilities in Britain were part of global networks of R&D facilities [B1, B3, B13, B14], knowledge transfer within each company was
11
promoted by intensive, daily use of telephone and e-mail facilities and regular use of video-conferencing. Individual scientists and engineers also regularly crossed the Atlantic for face-to-face meetings and discussions, including Monday-Friday working in different locations. These interactions not only facilitated co-operation in problemsolving and minimisation of duplication but also helped to plug British-based researchers into the networks of contacts (with institutions as well as individuals) which had been built up by their counterparts in the US. Examples of the benefits of such trans-Atlantic interactions included one large US-owned production plant in Southern England which had developed new, low-cost assembly techniques in conjunction with engineers based in a Californian research centre [B3]; similarly, a small US-owned Scottish establishment was drawing on the experience of an American sister plant in its efforts to move to high-volume production of subassemblies involving its main components for which the design and development work had been carried out in Scotland [B8]. However, this favourable assessment of cross-border co-operation in R&D did not apply to foreign-owned ‘branch plants’ such as B7 which were primarily engaged in test and assembly activities, with only a very limited involvement in product design and development.
In the French sample the small proportion of foreign-owned establishments were also found to be playing a positive role in cross-border knowledge transfer within the enterprises concerned and in expanding the depth and range of work on product innovation. For example, in one foreign-owned production plant [F4], the French manufacturing engineers had built up extensive personal links with the parent company’s R&D centre in order to secure early involvement in the product design process which would help improve the ‘ease of manufacture’ – and thus lower the production costs -- of new products. However, the relatively low incidence of foreign ownership in the French sample meant that the main source of advantage to the French sample in within-enterprise knowledge search and exchange derived, not from trans-Atlantic networking as in Britain, but rather from high levels of internal mobility in the largest domestically-owned French enterprises.
Such internal mobility is still the prime characteristic of the career development of individual engineers and scientists in the French industry. For example, in French
12
production and R&D/prototyping establishments about 64% of externally-recruited graduate engineers and scientists engaged in R&D activity had been taken on direct from engineering school or university in the three years prior to visits, considerably higher than in Britain (41%).10 This reflected a common preference in France for confining external recruitment to relatively young engineers rather than looking for people with lengthy employment experience in other companies. In these ‘internal labour markets’ regular transfers of personnel between R&D laboratories, design and development departments and production engineering departments help to build up networks of informal relationships and promote the exchange of useful knowledge within each company.
However, although the efficiency of within-enterprise knowledge flows in France may be improved by inter-departmental mobility of engineers and scientists, the downside is a lesser infusion of new ideas and tacit knowledge which many Britishbased establishments gain by recruiting R&D staff with employment experience in other enterprises. The benefits of this kind of knowledge transfer were described by British managers as follows: ‘The main thing is that experienced people see problems from a different perspective…. They help us avoid a “Not Invented Here” attitude’ [B1]; ‘We try to think of what people can bring to us, if they have a particular kind of experience…. It’s a good way to absorb new technologies’ [B5].
10
This difference in recruiting practices is statistically significant at a 95% confidence level.
13
4. Research linkages with external organisations: overview
Most corporate establishments in the two samples had a wide range of informal and formal (contractual) research links with other organisations such as customer enterprises and university departments. Informal contacts typically involved the sharing of technical knowledge or advice about information sources in the expectation of future reciprocation or as a means of contributing to the resolution of technical problems of mutual interest. In some cases informal discussions and contacts between individual engineers and scientists in different organisations had led to the development of formal relationships and the level of informal interactions had subsequently been intensified in the course of implementing formal arrangements for collaboration. In other cases it was the formalised linkages which had come first, for example, companies coming together with other companies or university departments to apply for funding for joint research projects supported by national or European funding bodies. Subsequently, this had provided a basis for individual researchers to develop a range of informal modes of communication with counterparts in other organisations.
Hence although informal interactions are hard to measure, the common overlapping of informal and contractually-based linkages means that some assessment of knowledge flows between sample enterprises and other organisations can be made by comparing the incidence of formal links. As shown in Table 3, the most striking inter-country differences was the fact that all but one of the
French sample
establishments had formal research links with public sector research institutes and laboratories of a kind which hardly exist any longer in Britain. The nearest equivalent in the British case was the much smaller proportion (one in five) of sample enterprises involved with private non-profit Research and Technology Organisations (RTOs).
By contrast, inter-country differences in research interactions along supply-chains were less apparent. ‘Research interactions/linkages’ are here defined as occurring only when establishments’ contacts with technical staff from customer or supplier
14
companies go beyond discussions about clients’ requirements and periodic progress reports on new product development to include two-way knowledge sharing and exchange. The unweighted measure of R&D interactions with customers shows that roughly three quarters of British establishments had collaborative links of this kind with customers compared with just under two thirds of their French counterparts. The incidence of formal technical alliances along the supply chain was also greater in Britain than in France. Against this, the employment-weighted measures of customer interactions in the French sample are slightly above the British level, reflecting the fact that in France – to a greater extent than in Britain – it was the larger companies (rather than SMEs) which were working closely with customers (Table 3, Rows 1 and 2). Given the small size of the samples concerned, no statistical significance can be attached to these British-French differences in measures of R&D involvement with customers, and the same applies to R&D links with suppliers (which were generally less intensive than interactions with customers in both countries). However, the incidence of ‘pre-competitive’ joint research projects and other forms of collaboration with enterprises outside the supply-chain was markedly higher in France than Britain (Table 3, Row 5).
Most of the pre-competitive research collaboration in the French sample was funded either through the government Fonds de recherche technologique (FRT) or through joint projects supported by public sector agencies for telecoms and defence procurement,
respectively,
CNET
(Centre
National
d’Etudes
Télécommunications)/France Télécom; and the DGA (Direction général pour l’armement) and DRET (Direction de la recherche et des études techniques du ministère de la défense). While France is well known to be ‘exceptional’ in the large share of public research funding directed to a small number of large firms (OECD, 1999:100), in this particular sample all the SMEs received some public funding as well, either from CNET or DGA or through the ANVAR programme which provides small and medium-sized enterprises with interest-free preferential loans. Thus all the French sample establishments received national government funding support for R&D compared to just under half the British sample establishments supported
15
through government schemes such as LINK or the Teaching Company Scheme (Table 3, Row 9). 11
In addition to differences in simple measures of the incidence of external knowledge sourcing activities in the two samples, our interviews and observations at establishment level highlighted several important and interesting inter-country differences in the nature and intensity of knowledge search and exchange which deserve detailed consideration. These findings are presented under four main headings: R&D interactions with customers and suppliers; research links with other firms (including competitors); relationships with universities; relationships with specialist research institutes and organisations.
11
The LINK programme supports collaborative pre-competitive research partnerships between private enterprises, universities and other science base institutions. The Teaching Company Scheme subsidises employment of new graduates and post-graduates on industrial company-based research projects.
16
Table 3: Incidence of research linkages with external organisations in national samples of corporate establishments Scores: Yes=1, No=0 for each establishment BRITAIN Employmentweighted average (a)
Unweighted Average Supply-chain interactions: Customers: - regular knowledge exchange (b) - full technical partnership (c) Suppliers: - regular knowledge exchange (b) - full technical partnership (c) Other external research linkages (d): Other firms (including competitors)
FRANCE EmploymentUnweighted weighted average average (a)
Tests for equality of respective sample means (Unweighted) (Weighted) (p-values)
0.76
0.73
0.64
0.80
0.4737
0.6657
0.53
0.59
0.36
0.73
0. 3541
0.4523
0.35
0.47
0.23
0.15
0.4865
0.1462
0.18
0.32
0.08
0.18
0.4442
0.1234
0.12
*
0.28
0.38
0.43
0.0924
0.4781
Public research institutes/ organisations (e) Private non-profit Research institutes/ organisations (f)
0.00
***
0.00 ***
0.92
0.98
0.0000
0.0000
0.18
0.30
0.00
0.00
0.1181
0.1119
Universities (g)
0.88
0.89 **
0.77
0.50
0.4276
0.0230
0.51 *** 0.00 0.28
1.00 0.14 0.21
1.00 0.15 0.06
0.0002 0.1143 0.8938
0.0026 0.6059 0.1419
n=17
n=13
n=13
External funding for R&D: National government Regional agencies European Commission No. of establishments
0.41 0.00 0.24 n=17
***
F-tests of equality of respective country averages: ***: British-French differences statistically significant at the 1% level or better **: 5% level *: 10% level Notes : (a) Establishments weighted for shares of total R&D employment in national samples. Detached research centres accounted for 30% of total R&D employment in the British sample as compared to 35% in the French case. [Calculations exclude one French production establishment (F1) for which data could not be obtained]. (b) Refers to contacts/meetings with customers/suppliers in course of product development process that often involve exchanges of technical knowledge and ideas which go beyond simply clarifying the details of customer/supply contract specifications. In addition to periodic technical discussions, customers/suppliers may also be involved with prototyping and/or testing. (c) Refers to close technical partnerships with customers/suppliers that involve regular scheduled meetings of key personnel plus multiple layers of informal communication to discuss details of product specification and problems encountered in course of projects; full involvement in prototyping and testing; projects may even have been initiated by customers/suppliers. Establishments classified to this category are also included under the heading ‘regular knowledge exchange’ with customers/suppliers in Row 2.
17
(d) Refers to formal (contractual) research linkages with firms outside the supply-chain, specialist research institutions and universities. (e) ‘Public’ here refers to non-university research institutions which typically receive core funding from governments in excess of 50% of their financial resources. (f) ‘Private, non-profit’ here refers to research institutions which typically receive some core funding from governments but not in excess of 50% of their financial resources, and are not classifiable to either the university sector or the business enterprise sector according to guidelines laid down in the OECD’s ‘Frascati Manual’ (OECD, 1994b). (g) Includes the Centre National de Recherche Scientifique (CNRS) in France.
18
5. R&D interactions along the supply chain In Britain several of the smaller sample establishments
reported very close
relationships with their main customers which involved multiple layers of informal communication between technical staff on both sides as well as regular scheduled meetings of key personnel to discuss changes to product specifications and ways to tackle technical problems which had arisen in the course of product development projects. These customers typically also had a full involvement in prototyping and testing and, indeed, in some cases projects had been largely initiated by customers in the first place. For example, one small Scottish start-up firm [B16] with strong links to universitybased researchers described ‘a constant interchange of ideas’ with a large foreignowned multinational company which had signed up to make use of the new technology they were developing. Although the start-up firm was responsible for pushing forward the core innovations at the heart of the project, the customer firm was described as a valuable source of knowhow on device design. Another small establishment in Scotland [B8] had long been accustomed to having ‘engineer to engineer contact’ with its customers but had tended to regard such involvement as largely concerned with ‘getting the specification right and trying to educate the customer’ (for example, by suggesting design changes which would reduce manufacturing costs). However, the nature of their relationships with customers had recently changed with entry into a new collaborative relationship with a US-owned multinational client to develop a new range of products in conjunction with the larger company’s R&D centre in England. The fluid nature of R&D interactions along the supply-chain in Britain was also exemplified by SMEs in southern England. A manager at B6 described a process whereby prospective customers would approach several small firms like themselves to discuss their product requirements and might even commission prototypes from more than one supplier. Relationships were only likely to deepen if engineers from both the customer and supplier firm worked well together while maintaining a mutually beneficial balance between knowledge sharing and retention: ‘We learn a lot from customers and vice versa…. but we don’t want to give too much knowhow
19
away’.
Another small firm [B11] regularly participated in technical meetings
involving not just its own customers but its customers’ customers as well and the information exchanged at these meetings often led to the development of new relationships with subcontractors for various specialist services. In this kind of changing interactive environment rival companies are well positioned to monitor each other’s progress through customer and supplier contacts as well as through a general ‘grapevine’ in the industry based on personal contacts and a steady flow of individual engineers and scientists moving from jobs in one company to another. Although as noted above, each company is aware of disadvantages of giving away ‘too much’ information, they also have experience of the positive advantages of two-way knowledge sharing which provides incentives to stay active in new relationship building. The recruitment of experienced personnel who bring new personal networks of contacts with them is another factor which helps to maintain the momentum of relationship building between companies. However, as might be expected, the proportionate importance of newly-formed relationships with customers and suppliers – with their attendant risks and uncertainties – was much less for larger British-based companies than SMEs. This was particularly true of large defence equipment suppliers such as B2 and B4 which tended to work within the context of long-established relationships with customers and suppliers. In the French sample the most intensive collaborative relationships between firms all seemed to hinge around a small number of large domestically-owned companies, particularly the leading customers and suppliers in defence and telecoms equipment sectors. Not only did these large firms have long-established relationships with each other but SMEs also looked to them as sources of both research funding and expertise. Examples of research interactions which had been fostered in this context included a newly-established small defence equipment supplier [F8] whose initial product development had been heavily subsidised by the state armaments authority (DGA). Its R&D projects had been largely motivated by two large customer enterprises which had specific applications in mind and the small firm concerned had been able to draw on the knowledge and expertise of technical staff at these large companies’ own research centres throughout the projects. In another case, a small
20
firm [F14] offering design and development services to telecoms equipment producers had benefited from access to specialist equipment belonging to France Télécom (its main customer) and from regular technical discussions with other customers in the same region where a majority of R&D personnel in each organisation had over many years built up personal contacts through interactions with France Télécom’s national research centre (CNET). With the aid of government procurement agencies and state-owned enterprises, therefore, these patterns of knowledge search and exchange among supply-chain partners were if anything more structured and well-entrenched in France than in Britain. This reflects a long heritage of French industrial planning based on the supply-chain, with state agencies initially involved in constructing networks of subcontractors rather than leaving such decisions up to prime contractors (Storper, 1993). However, there was less evidence of French companies engaging in the more risky type of relationship-building found in Britain where (as described above) many of the interactions between sample establishments and their customers were based on relatively new relationships which both sides had initially entered into in a relatively speculative manner (often involving partners from other regions or other countries). For example, two of the smaller companies in the French sample which had been set up in order to commercialise scientific discoveries in public research institutes had maintained exceptionally close working relationships with those research institutes but their relationships with customers had not developed beyond discussions of those customers’ specific
requirements [F9, F10]. In other cases where supply-chain
interactions did involve two-way knowledge-sharing this tended to be confined to well-established relationships with large clients. One medium-sized French-owned company [F7] had gradually reduced its dependence on defence-related markets in recent years but had not developed close collaborative research relationships with civilian clients to match those it had earlier built up with large defence equipment producers. At another medium-sized company [F5] attempting to make a similar transition away from defence and aerospace markets, a manager said that, in the event of technical problems, they were still most likely to turn to their long standing ‘grands clients’ for assistance and advice.
21
6. Research links with other firms (including competitors)
Only two of the 17 British sample establishments had formal research links with firms outside the supply-chain compared with five of the 13 French establishments. In both British cases this involved pre-competitive collaboration with European rivals in the context of European Commission (EC) programmes while most of the French partnerships between rival firms had been fostered by national government agencies rather than the EC.
Some examples of research collaboration between direct competitors in the French sample had no parallel among British sample establishments. In one case [F6] two French companies had both tendered for a government defence agency (DGA) contract and the agency had helped to bring the two rivals together to carry out the design and development work. Among other things this partnership had benefited from a new emphasis on early involvement of manufacturing specialists in the design process, an approach which both partners had hitherto tended to neglect. Another French company [F7] had participated for more than eight years in joint research with rival firms in the context of a university-based Groupement d’Intérêt Economique (GIE).12 Although this project was seen as taking too long to deliver commercially useful results, F7 was prepared to continue investing in it for far longer than would normally be contemplated by equivalent British firms.
12
This ‘Economic Interest Group’ is one of several research partnerships in France which have been established on the basis of 1967 legislation which allows for formal research collaboration between enterprises.
22
7. Research links with universities
In this rapidly-changing area of electronics, large proportions of British and French establishments had formal research links of one kind or another with university-based researchers (Table 3) and all of them at least engaged in periodic informal contacts. In both countries several managers conveyed a strong sense of urgency in gaining access to academic knowledge sources so as to plug gaps which had been identified in in-house expertise. For example, a small French firm [F9] catering to niche markets in telecoms and medical equipment areas was heavily dependent on the physics and chemistry knowledge of CNRS and other university-based laboratories for the development of new materials. In Scotland a large defence supplier [B4] provided core funding to a university-based research centre and also regularly sought help from universities with ‘panic problems’, particularly in relatively unfamiliar areas such as chemistry and metallurgy. However, a closer investigation of university linkages brings out some inter-country differences as well as similarities.
As Table 4 shows, just under half the French establishments purchased technical services from universities whereas four in five of the British establishments either employed university-based technical consultants or hired specialist equipment from universities or purchased both these types of service; this difference is statistically significant at a 97% level of confidence (Table 4, Row 2). Many French firms had no need of university contacts of this kind because of their continuing close involvement with public research laboratories which tended to preclude the need for using university-based consultants and equipment.
As with research interactions along the supply-chain, other differences between the two samples in enterprise-university linkages which emerged from interviews related to the degree of stability in such relationships – something which is not captured in simple measures of the incidence of interactions. The qualitative evidence on this score suggests that relationships between many French companies and universities, once formed, are rarely subject to change and re-formation. One large domesticallyowned multinational [F3] had long relied on certain universities and CNRS labs
23
(based in universities) for scientific advice, tests and analysis and it was increasing its involvement in medium term research programmes in conjunction with these same universities in order to meet the requirements of various government and EU funding programmes. It was taken for granted that they ‘had to avoid certain [other] universities because of their close relationships with competitors’. At another Frenchowned multinational [F7] it was even said that they preferred to work with university and CNRS research teams that they already knew (described as ‘équipes privilégiées’) rather than with unknown teams, even if the unknown teams were identified as specialists in the problems needing to be resolved.
Table 4: Research interactions with universities Scores: Yes=1, No=0 for each establishment BRITAIN
Unweighted average Informal contacts with universities:
FRANCE Employmentweighted Average (a)
EmploymentUnweighted weighted Average average (a)
Tests for equality of respective sample means (Unweighted) (Weighted) (p-values)
1.00
1.00
0.86
0.93
0.1143
0.2570
0.65
0.78 **
0.38
0.31
0.1641
0.0203
Purchase of university services: - technical expertise and use of specialist equipment - use of equipment only - no purchase of university technical services TOTAL
0.18
0.05
0.08
0.03
0.4442
0.9030
0.18 ** 1.00
0.17 ** 1.00
0.54 1.00
0.65 1.00
0.0380
0.0126
Formal research contracts: - PhD students - other research contracts
0.59 0.53
0.72 ** 0.66
0.38 0.46
0.28 0.32
0.2849 0.7240
0.0385 0.1232
No. of establishments
n=17
n=13
n=13
n=13
F-tests of equality of employment-weighted averages: ***: British-French differences statistically significant at the 1% level or better **: 5% level *: 10% level (a) See Table 3, Note a.
This tendency towards stability in enterprise-university research links in the French sample was reinforced by some SMEs [F5, F14] whose university contacts were confined to their immediate region and the fact that firms often relied on individual
24
staff members for personal contacts with universities. Since many French firms typically confine new recruitment to newly-qualified graduates or to people with small amounts of work experience – and, in particular, the large companies refrain from recruiting from each other -- this means that very few French engineers and scientists have been able to expand their personal contacts to cover a range of different universities by working with a series of different employers.
In Britain there were also several examples of longstanding relationships between particular companies and universities which resembled those found in France and there were similar concerns about confidentiality. However, these factors did not prevent the majority of British sample enterprises from actively pursuing opportunities to develop new collaborations which could prove advantageous. Indeed, the relatively high mobility of experienced engineers and scientists in Britain contributed to substantial fluidity and change in research relationships with universities. For example, at the large telecoms equipment producer B1, the basic approach was described as ‘defining unmet research needs and looking for the best university to carry them out’. These choices were often guided by suggestions from individual researchers who had developed extensive networks of university contacts prior to joining B1. Another foreign-owned multinational [B3] had initially taken proximity to certain universities into account in choosing its British R&D locations but was still ready to develop new relationships with other British universities in order to gain access to specific expertise and equipment.
To some extent the faster rate of new relationship building in British enterpriseuniversity links reflected the considerable financial pressures on universities to generate contract income and thus to take a pro-active approach to identifying prospective industrial research partners. This pro-activity on the part of British universities partly accounted also for the higher level of industrial support for PhD students in the British sample (Table 4, Row 6). 13
13
In some cases these financial
British-French differences in industrial support for PhD research were statistically significant at the 96% level. This gap occurred even though the proportions of sample establishments receiving government subsidies for such involvement were much the same. About 30% of British sample establishments were involved with PhD training through the CASE programme whereby doctoral students are jointly funded by government research council and industrial partners. In France a similar
25
pressures had led to increased tensions between universities and sample establishments on IPR (Intellectual Property Rights) issues and one medium-sized establishment [B10] had elected to expand its own research activity rather than enter into formal collaboration with the university with which it had been negotiating. However, in a majority of other sample establishments, contacts initiated by universities had generally contributed to the breadth of enterprise-university research interactions.
proportion of sample establishments were involved with the CIFRE scheme through which doctoral students are taken on as employees by sponsoring companies for three-year periods.
26
7. Enterprise interactions with specialist research institutes
As noted above, all but one of the French establishments had close working relationships with publicly-owned research centres and laboratories (Table 3) -- a type of institution for which there are few equivalents in Britain following the privatisation of most former public sector labs and sharp cuts in core funding for the Research and Technology Organisations (RTOs). All four of the private, non-profit research organisations in Britain which participated in this study reported that, in order to survive as independent organisations in the market-place for research services, they needed to increasingly engage in short-term industrial problem-solving on a consultancy basis. However, in two cases the organisations concerned had been able to maintain some involvement in collaborative research projects with universities through participation in the government-funded Postgraduate Training Partnership (PTP) scheme.14 At one of these labs, managers said that, without such university linkages, their staff would suffer ‘complete loss of the knowledge base’ needed to stay up to date in its specialist areas of electronics design and software.
In the British sample only three establishments had regularly used the services of private non-profit research organisations such as the RTOs, typically engaging them as testing centres or in one case seeking help to resolve specific problems in the assembly of components which fell within a particular RTOs area of expertise. The main reasons for other British establishments not making more use of RTO services were described as, firstly, their relatively high fees and, secondly, a perception that research organisations of this kind were out of touch with ‘leading edge’ developments in electronics. 15
14
The PTP programme provides support for partnerships between selected RTOs and universities designed to promote involvement of PhD students in industrially relevant research projects 15 Similar reservations have been expressed by German electronics firms about Fraunhofer institutes which are also required to generate a large part of their income through private contract work. Although the Fraunhofer institutes perform their role of intermediating between the academic science base and industry with notable success in mature industries such as mechanical engineering, there seems to be much less scope for such a role in fast-changing R&D-intensive areas of electronics where both SMEs and larger firms are well capable of making direct contacts with university-based researchers (Mason and Wagner, 1999).
27
In France public sector laboratories and technical agencies have been given incentives since the early 1980’s to establish subsidiary enterprises and develop commercial partnerships with other firms but this has been aimed more at stimulating the exploitation of innovations than at transforming them into private sector operations (Chesnais, 1993). Hence leading research institutes involved with the electronics industry such as CNET – the France Télécom research centre -- and the Laboratoire d’Electronique, de Technologie et d’Instrumentation (LETI) in Grenoble were still largely financed by public funds at the time of our visits.
16
This had
enabled French public labs to build on their close personal and physical links with universities and the CNRS, maintain their own strategic research capacity
and
remain closely in touch with new technological developments in electronics.
In consequence, the proportion of French sample establishments with formal R&D linkages with specialist labs and research centres was even higher than that involved with universities. In addition to playing an intermediary role between enterprises and the academic science base, the French public labs were effectively part of that science base but with a clearer remit than universities to promote the commercialisation of new research results. Indeed, three of the smaller French sample firms had been started up as off-shoots of LETI or similar public labs and had continued to work in close collaboration with lab-based researchers on further product development.
These relationships have often proved fruitful in terms of innovation and technical problem-solving. At one LETI start-up firm the managers described their relationship with their founding laboratory as a ‘beautiful example of co-operation with an upstream lab which feeds us with basic research…. we are in permanent dialogue [échange en permanence] with them’ [F8]. However, some of the start-ups have found it difficult to achieve independence in commercial markets. For example, F10’s managers realised soon after its formation that it would not be able to survive by relying solely on the LETI-developed technology with which it had started. They embarked on an extensive period of new product development which was jointly
16
Since 1st March 2000 CNET no longer exists and many of its former researchers have been absorbed into France Télécom’s R&D department. Other former CNET researchers have dispersed elsewhere, either within France Télécom or other enterprises or into CNRS research units or into retirement.
28
funded by ANVAR, prospective clients and its own sources. This project has now apparently been described by ANVAR as ‘une réussite technique mais un échec commercial’ (‘a technical success but a commercial failure’). After some years F10 remains a very small firm which has cut back heavily on new R&D and survives by serving niche markets which it has identified in the aerospace sector.
In the French telecommunications equipment sector, CNET’s direct contribution to enterprise R&D projects in 1996-97 was greatly reduced compared to previous decades but it continued to work with a variety of industry- and university-based partners. Here too there were mixed reports in terms of commercial outcomes. For example, at F3 some new product lines had been successfully developed on the back of earlier basic and strategic research carried out at CNET. More recently (in 200001) the company has adapted to the closure of CNET by entering into new arrangements for long-term R&D, including a laboratory associated with CNRS teams which include former CNET researchers.
But at a medium-sized telecoms supplier [F5] the technical director expressed some nostalgia for the past division of research labour with CNET whereby ‘they did the fundamental research and we did the applied research’. At the time of our visits the company continued to work closely with CNET and regional universities on new product development but projects were much more likely to run into problems of ‘discordance between the technical objectives and price’. One recently-completed project was described in the following terms: ‘The product is ready, the problem is commercial’. In this case the strengths of the knowledge infrastructure in their region provided only partial compensation for disadvantages such as its relatively late start in adjusting to a more competitive marketplace.
29
8. Summary and conclusions
Previous comparisons of the British and French electronics industries have tended to emphasise their similarities in terms of industrial structure and government policy contexts. For example, in both countries defence-related spending on research and development has traditionally accounted for a relatively high proportion of government R&D compared to other European countries, and continues to do so in spite of recent cuts in defence spending. 17 Indeed, in a well-known taxonomy, Ergas (1987) classified both countries as ‘mission-oriented.... (with) big science deployed to meet big problems’ as compared with other more ‘diffusion-oriented’ countries such as Germany.
The present study has highlighted several similarities between the two industries arising from their common background in defence-related research and the predominance of former ‘national champions’ in product development supply-chains involving SMEs, universities and specialist research organisations. But it has also identified several important differences
in the workings of national knowledge
infrastructures in Britain and France which appear to have important implications for the technical problem-solving capacity and innovative performance of electronics enterprises in the two countries.
The key differences in external knowledge sourcing patterns between the two countries are, firstly, the involvement of nearly all the French establishments with public sector research labs of a kind which hardly exist any longer in Britain; secondly, an apparently more dynamic and market-driven pattern of research interactions and exchange of ideas with customers in Britain than in France; and thirdly, a faster rate of new relationship building in enterprise-university R&D interactions in Britain than in France.
17
In 1996 defence budget R&D as a proportion of total Government Budget Appropriations or Outlays for R&D (GBAORD) was just under 38% in Britain and 30% in France compared to 10% in Germany and 5% (1995) in Italy (OECD, Main Science and Technology Indicators, 1998/1, Table 64).
30
These contrasts strongly reflect inter-country differences in public funding of medium- and long-term research and in state involvement in industrial organisation and restructuring. The public laboratories involved with the French electronics industry have, until recently at least, helped to maintain relatively high levels of investment in basic and strategic research and organisations such as LETI have also been given incentives to establish subsidiary enterprises and develop commercial partnerships with other firms. However, examples were found in France of new product development programmes being more oriented to technical achievement than to commercial success. We conjecture that this may partly reflect the relatively slow opening-up of French markets for telecoms and defence equipment to foreign competition. In addition, the long heritage of French state involvement in supplychain planning appears to have contributed to French enterprises’ research relationships with supply-chain partners being more entrenched and less susceptible to change and development than in Britain.
In Britain the momentum of new relationship-building along the supply-chain is maintained by (among other things) intense commercial pressures on enterprises to solve technical problems quickly and speed up the commercialisation of new products. Knowledge sharing between organisations is further stimulated by the regular recruitment of experienced personnel who bring new personal networks of contacts with them. The more fluid and rapidly-changing nature of enterpriseuniversity R&D interactions in Britain strongly reflects the budget constraints enforced on British universities which have led them to adopt a highly pro-active approach to research collaboration with industry.
Taken together these British-French differences in external research interactions and knowledge sourcing – along with the very high level of investment by US-owned enterprises in the UK -- may help to explain why recent trends in innovation performance have favoured the British electronics industry relative to its French counterpart – even in spite of a much higher rate of investment in R&D in France. One hypothesis that emerges from our findings is that the greater ‘openness’ to new ideas and knowledge of British R&D networks may be particularly advantageous in a
31
fast-changing high-tech industry such as electronics.
18
Further research using
predominantly quantitative techniques is needed to test this hypothesis while taking due account of the range of other factors besides high-level skills mobility and knowledge transfer which affect relative performance. Appropriate weight also needs to be given to the relative importance of foreign-owned enterprises in plugging British-based researchers into global – and especially trans-Atlantic – networks of knowledge search and exchange as compared to the lower level of foreign investment in the French industry. It is also possible that the UK industry benefited directly from US federal investments in basic research.
In conclusion, it is important to note that recent trends in relative performance may not persist far into the future. The relatively weak level of investment in basic and strategic research in electronics in the UK itself continues to be a potential Achilles heel which could lead to long-term deficiencies in knowledge generation and the ability of enterprises to absorb relevant knowledge produced elsewhere. Conversely, the French electronics industry is in the early stages of adaptation to deregulated markets for telecoms and defence electronics products and could now conceivably make rapid improvements in competitiveness.
Implications for UK science and technology policy
Although a two-country comparative study of this kind cannot serve as the basis for detailed policy recommendations, our findings do have some implications for the broad direction of British science and technology policy:
1. Some of the apparent strengths of the British electronics industry -- industrial involvement in PhD training and the openness/diversity of R&D networks involving 18
The literature on networking displays widespread recognition of a tension between the need for R&D networks to be well ‘embedded’ in the social and institutional contexts within which they operate and the need for the constituent members of networks to be ‘open’ to new sources of knowledge and ideas which lie outside those networks (Granovetter, 1983, 1985; Cooke and Wills, 1999). This study’s findings suggest that, in a rapidly changing and developing industry, a high degree of embeddedness in networks may actually restrict the knowledge search and exchange activities of member organisations and their R&D staff.
32
firms and universities -- derive in part from government-funded programmes such as CASE, PTP and LINK which appear to have yielded many positive benefits. There is a case for further expansion and strengthening of such programmes, with efforts made to reduce the time delays and bureaucratic procedures which inhibit some firms from participating in them. Particular emphasis should be placed on using such programmes to increase the involvement of foreign-owned electronics enterprises in British-based R&D activity.
2. Science and technology policy needs to keep the advantages of ‘regional clustering’ in perspective. In fast-changing industries such as electronics, it appears to be a strength of the British industry that so many researchers and enterprises are engaged in knowledge search and networking on an international scale. Regional concentrations of enterprises, universities and specialist research organisations provide many benefits for R&D collaboration and for knowledge exchange and transfer in general. However, in fastchanging industries such as electronics, clusters will only thrive and develop if they function as key nodes in global innovation networks. The criteria for government assistance to potential new foreign investors need to reflect these dual objectives.
3. An ‘impatient’ budget-constrained approach to R&D investments has the positive effect of encouraging universities and research institutes to seek out industrial funding in a pro-active way. However, some restoration of basic and strategic research activity in Britain may now be required in order to maintain and build on the British electronics industry’s improved innovation performance. In order to achieve this objective, a range of policies should be considered, including an increase in core funding for selected university departments and research organisations. 4. In this context the development of the new Faraday Partnerships is welcome since they promise to serve many of the functions which this study has attributed to Postgraduate Training Partnerships, namely, to help shore up basic and strategic research while supporting industry-based postgraduate research training and encouraging networking between firms, RTO’s and universities. The key questions here concern the scale of this programme and the level of new resources to be committed to it.
33
References Arora, A. and Gambardella, A. (1994), The changing technology of technological knowledge: general and abstract knowledge and the division of innovative labour, Research Policy, 23: 523-532. Bosworth, D. and Stoneman, P. (1996), Information and technology transfer in Europe: national, European and global aspects, Paper presented at the EC International Conference on Innovation Measurement and Policies, Luxembourg, May. Brousseau E., Petit P. and Phan D. (1996), (eds.) Mutations des Télécommunications, des Industries et des Marchés, ENSPTT / Economica. Chesnais, F. (1993), The French National System of Innovation, in R. Nelson (ed.), National Innovation Systems: A Comparative Analysis, New York/Oxford: Oxford University Press. Cooke, P. and Wills, D. (1999), Small firms, social capital and the enhancement of business performance through innovation programmes, Small Business Economics, 1: 1-16. Craggs, A. and Jones, P. (1998), UK results from the Community Innovation Survey, Economic Trends, 539: 53-59. David, P. and Foray, D. (1995), Accessing and expanding the science and technology knowledge base, STI Review, No. 16, Paris: OECD. Ergas, H. (1987), The importance of technology policy, in P. Dasgupta and P. Stoneman (eds.), Economic Policy and Technological Performance, Cambridge: Cambridge University Press. Granovetter, M. (1983), The strength of weak ties, American Journal of Sociology¸ 87(1), 1360-1380. Granovetter, M. (1985), Economic action and social structure: the problem of embeddedness, American Journal of Sociology, 91(3): 481-510. Guillaume, H. (1998), Rapport de Mission sur la Technologie et l’Innovation, Rapport au premier ministre, Paris. Lawton Smith, H. (1997), Adjusting the roles of national laboratories: some comparisons between UK, French and Belgian institutions, R&D Management, 27(4): 319-331. Mason, G. and Wagner, K. (1999), ‘Knowledge transfer and innovation in Britain and Germany: “Intermediate institution” models of knowledge transfer under strain?’, Industry and Innovation, 6 (1), 85-109. Morgan, K. (1989), Telecom strategies in Britain and France: the scope and limits of neoliberalism and dirigisme, in M. Sharp and P. Holmes (eds.), Strategies for New Technology: Case Studies from Britain and France, London: Philip Allan. OECD (1994b), Proposed standard practice for surveys of research and experimental development - Frascati Manual, Paris. OECD (1999), OECD Economic Surveys 1998-99 : France, Paris. Pavitt, K. and Patel. P. (1989), L’accumulation technologique en France : ce que les statistiques de brevets tendent à montrer, Revue d’Economie Industrielle, 51 : 10-32. Rothwell, R. and Dodgson, M. (1993), The contribution which research and technology organisations make to innovation and competitiveness in UK industry, Association of Independent Research and Technology Organisations (AIRTO) Paper 93/1. Schmoch, U. (1999), Interaction of universities and industrial enterprises in Germany and the United States – a comparison, Industry and Innovation, 6(1): 51-68. SESSI (1995), L’Innovation Technologique dans l’Industrie: L’innovation en Solitaire est un Mythe, Service des Statistiques Industrielles (SESSI), www. evariste.org/sessi46.htm. Smith, K. (1997), Economic infrastructure and innovations systems, in C. Edquist (ed.), Systems of Innovation: Technologies, Institutions and Organizations, Pinter: London. Storper, M. (1993), Regional ‘worlds’ of production: learning and innovation in the technology districts of France, Italy and the USA, Regional Studies, 27(5): 433-455.
34
APPENDIX A: Figure A1: Patents granted in the US to foreign-owned and domestically-owned firms in the British electronics industry, 1969-96 400
Number of patents
300
200
100
95 19
93 19
91 19
89 19
87
85
19
83
81
79
77
Domestically-owned
19
19
19
19
73
71
75
19
19
19
19
19
69
0
Foreign-owned firms
Source and industrial classification: See notes to Figure 1 in main text.
Figure A2: Patents granted in the US to foreign-owned and domestically-owned firms in the French electronics industry, 1969-96
400
Number of patents
300
200
100
Domestically-owned firms
95 19
93 19
91 19
89 19
87 19
85 19
83 19
81 19
79 19
77 19
75 19
73 19
71 19
19
69
0
Foreign-owned firms
Source and industrial classification: See notes to Figure 1 in main text.
35
