Sunteți pe pagina 1din 21

John Cantwell

Simona Iammarino
The technological innovation of multinational corporations in the
French regions
In: Revue d'conomie industrielle. Vol. 109. 1er trimestre 2005. pp. 9-28.
Abstract
In a rapidly globalising economy, and particularly in the face of a process of economic integration such as that occurring in the
EU, regions forge an increasing number of linkages with other locations within and across national boundaries through the local
technological development efforts of multinational corporations (MNCs). By using patents granted to the largest industrial firms -
arranged by the region (NUTS 1) host to the research facility responsible - the paper explores the location of innovative activities
of MNCs in France, and the relationship between the profiles of technological specialisation of foreign-owned and indigenous
companies in the French regions.
Rsum
Dans le processus de globalisation de nos conomies, et particulirement face au processus d'intgration conomique l'uvre
en Europe, les rgions mettent en place de plus en plus de liens avec d'autres lieux dans et en dehors des frontires nationales
travers les efforts de dveloppement technologique des entreprises multinationales. En utilisant les brevets des grandes
entreprises industrielles - apprhendes par la rgion (NUTS1) abritant les capacits de recherche - l'article explore la
localisation des activits innovatrices des multinationales en France, et la relation entre les profils de spcialisation technologique
des entreprises trangres et des entreprises indignes dans les rgions franaises.
Citer ce document / Cite this document :
Cantwell John, Iammarino Simona. The technological innovation of multinational corporations in the French regions. In: Revue
d'conomie industrielle. Vol. 109. 1er trimestre 2005. pp. 9-28.
doi : 10.3406/rei.2005.3062
http://www.persee.fr/web/revues/home/prescript/article/rei_0154-3229_2005_num_109_1_3062
John
CANTWELL
Rutgers Business School and University of Reading
and Simona IAMMARINO
SPRU, University of Sussex and University of Rome "La Sapienza"
THE TECHNOLOGICAL INNOVATION
OF MULTINATIONAL CORPORATIONS
IN THE FRENCH REGIONS
Mots-cls : entreprises multinationales, innovation technologie, systmes rgionaux.
Key words : Multinational Corporations, Technological Innovation. Regional Systems.
I. - INTRODUCTION
The nexus between global and local processes has been investigated quite
extensively by the literature of the most recent years. One particular and cru
cial aspect of such a relationship lies in the creation and diffusion of innovat
ion, which, more than other economic processes, show rather complex pat
terns of distribution across space.
Indeed, as emphasised by Dicken, '"global' and 'local' are not fixed scales;
rather, they represent the extreme points of a dialectical continuum of complex
mutual interactions"
(Dicken, 1994, p. 103). As a consequence, neither the
orthodox approach - which traditionally considers both the (multinational)
firm and the local system as black boxes whose behaviours are determined by
exogenous factors; nor an entirely endogenous perspective - which tends to
explain structure and growth mechanisms as the result of purely internal
forces - seem appropriate to investigate the issue 'global versus local'. Rather,
structure and behaviour of the two "extreme points" need to be considered
within the context of their increasing interdependence, including both endoge
nous determinants and exogenous variables relevant to the analysis.
Following the line of our previous studies, this paper presents an analysis of
the location patterns of multinational firms' innovative activities in France.
REVUE D'CONOMIE INDUSTRIELLE n 109, Ie- trimestre 2005 9
The paper is divided into five sections. The following section summarises the
conceptual framework for the empirical study carried out here. After a short
description of the data used for the empirical investigation, section three ana
lyses the geographical distribution of innovative activities of large firms, both
indigenous and foreign-owned, across the French regions in the period 1969-
95. Section four documents the characteristics of the hierarchy of regional
research centres in France, by testing the relationship between the profiles of
technological specialisation in foreign-owned and indigenous French firms. As
we have suggested elsewhere, differences in the regional capacity of attraction
of high value added activities serve as a prerequisite to unveiling the potential
technological communication (or technological spillovers) that may be in exis
tence within these regions and, indeed, across EU national boundaries. Finally,
section five highlights the main implications that can be drawn, and our futu
re research agenda.
II. - MULTI NATIONAL CORPORATIONS, TECHNOLOGICAL
INNOVATION AND EUROPEAN REGIONS
It is widely accepted that the organisation of innovative activities can no lon
ger be illustrated simply on the basis of concepts such as the dichotomy of
market-hierarchies and the transaction costs mechanism as fundamental expla
nations of internalisation/externalisation of capabilities, functions and assets.
Indeed, the trend has been increasingly observed for multinational corpora
tions (MNCs) to establish internal (intra-firm) and external (inter-firm) net
works for innovation, which are characterised by different levels of territorial
and social embeddedness with reference to the location which hosts them.
Regions, even belonging to the same nation-State, show different characterist
ics that determine the degree of attractiveness and the amount of spillovers
that a region is able to draw. The choice of a particular location for a MNC to
invest in research and development activities is thus driven by several factors,
which can be summarised as "social capability" and "technological congruen
ce"
(Abramovitz, 1986; Fagerberg, Verspagen and von Tunzelman, 1994).
While the first refers to the overall ability of the region to engage in innovati
ve and organisation processes, the latter points to the distance of the region
from the technological frontier, or, in other words, its capacity to implement
the technical properties connected to the new knowledge.
Furthermore, it has been shown that MNC affiliates abroad have assumed a
predominant role in an increasing proportion of all the most advanced technol
ogies. The interpretation given is twofold: on the one hand, the ceaseless rel
evance of local innovation processes as reservoirs of different technical expert
ise in the globalisation era, and on the other, the outgrowth of an "organisa
tional capital" which allows the integration of several related technological
competencies across geographically dispersed units (Zander, 1997). Thus, the
development of cross-border corporate integration and intra-border inter-com
pany sectoral integration, as new forms of governance, makes it increasingly
1 0 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
important to examine where and how innovative activity by MNCs is interna
tionally dispersed and regionally concentrated (1).
Whilst the establishment in a foreign location facilitates the monitoring of
developments in different technological fields, it also enables the extraction of
local knowledge for MNC global networks (Cantwell, 1992, 1995; Chesnais,
1992; Granstrand and Sjolander, 1992; Dunning & Wymbs. 1997). In the case
of the former, the firm is likely to be active abroad in technologies where comp
lementarity between the strength of the host economy and its own expertise
exists. In the case of the latter, a firm locates its research facility abroad to
exploit the technological advantage of the host region either to reinforce its
own competence or to alleviate its weakness at home. On this basis, it was the
refore suggested that, behind location choices related to innovation, there are
the two typical types of agglomeration forces, which operate differently across
regions. The first can be termed "localisation economies", which are sector-
specific and tend to intensify intra-border sectoral integration through local
external networks between foreign-owned affiliates, indigenous firms and
local non-market institutions. The second are referred to as general external
economies, or "urbanisation economies", which attract all kinds of economic
activities and tend to strengthen cross-border intra-firm integration, allowing
the feedback of knowledge, expertise and information within networks of affi
liates.
Arising from this differentiation of agglomeration economies, it became pos
sible to distinguish between higher order and intermediate regional
centres (2). Such centres arise "as a consequence of the interaction and the
intensity of general external economies and localisation economies, which in
turn depend upon the characteristics of the regional system considered"
(Cantwell and lammarino, 1998, p. 387). Whilst in the case of an intermediat
e location, knowledge in specific technological fields is accessed and injected
into the multinational network, affiliates located in higher order centres can
enjoy a broader range of spillovers from the local environment. However,
whilst there is evidence that much of the technology developed abroad by large
firms lies in their core areas of strength (Patel and Vega, 1999) (3), MNC
research in foreign locations is also increasingly associated with a higher pro-
(1) For an in-depth discussion on the linkages between the globalisation of innovation and
regions - which highlights the importance of location in the globalisation process and the
refore the supremacy of a regional approach when analysing this phenomenon - see
Cantwell and lammarino (2003).
(2) The other extreme is that of lower order regions, i.e. technologically weak and backward
regions that have an inadequate innovative base in order to compete with other locations
and to be attractive for external flows of knowledge and technology.
(3) This suggests that adaptation and technical support to foreign manufacturing plants conti
nue to be major explanatory factors.
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 1 1
bability
of entry into new and more distantly related fields of technology. Such
knowledge-seeking activity is undertaken to help define the future directions
in the evolution of the corporations' sources of competitiveness (Pearce,
1999).
The relationship 'global versus local' implies a mounting competitive bid
ding in order to attract external sources of knowledge and technological comp
etence, by which to enhance the regional knowledge base (4) and tap into the
full growth potential of globalisation. The sufficient condition to take part to
the competitive bidding, especially within the European economic arena, is
clearly the dynamics (upgrading/diversification) of regional technological
advantage and the ability to provide opportunities in the most promising tech
nological fields. Such an ability is strictly related to the regional capacity to
engage in "institutional learning", i.e. to adjust the local institutional structure
in order to support, sustain and enhance the development of new technologies
and to adapt to the prevailing technological paradigm. Thus, the "institutional
comparative advantage" of the regional economy is often the underlying rea
son for a more dynamic technological performance and of the gradual shift of
the regional specialisation towards the fastest growing areas of innovative acti
vity. Obviously, not all regional cores are able to adjust their profiles of spe
cialisation to the highest technological opportunities: the cumulative and loca
tion-specific nature of technological process might eventually imply the rise
and the decline of technological poles within Europe.
III. - MNC RESEARCH LOCATION IN THE FRENCH REGIONS
As in our previous studies, the geographical unit of analysis used to explore
the French case is based upon sub-national entities that derive from normative
criteria, as classified by Eurostat in the Nomenclature of Territorial Units for
Statistics (NUTS), providing a single uniform arrangement of territorial sys
tems. The comparable NUTS level chosen here for France is NUTS 1. The
empirical investigation uses patents granted in the US by the US Patent and
Trademark Office (USPTO) to the world's largest industrial firms for inven
tions achieved in their French-located operations over the 1969-1995 period,
classified by the host region in which the research facility responsible is loca
ted. Each patent was classified into one of 56 technological sectors derived by
mapping from the primary classification of the USPTO and organising patents
into common technological groups (see Appendix for the resulting 56 sectors).
For further discussion of the data and the rgionalisation of the patent databa
se see Cantwell and Iammarino (1998, 2000, 2003).
(4) We use the word 'enhance' and not 'construct' because the existence of such a base is a
necessary, although not sufficient, prerequisite in order to have attraction capacity and
thus to be a regional centre of excellence, either as a higher order or as an intermediate
location.
1 2 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
A
few selected indicators by region - relative to the end of the period to
which our database on US patents refers - may provide a brief picture of the
relative size of the French regions. Ile de France appears to be the only region
well above the national average for most indicators: its economic relevance is
leading also when compared to the EU average - the index of GDP per capita
in the middle 1990s being above 175 with respect to the base (EU15=100).
More generally, the most economically advanced areas are certainly those
around the capital region, the centre and the centre-east of the country, which
show a GDP higher than the EU average.
Looking at innovation variables, according to the National Innovation
Survey the most innovative areas are identifiable within the southern and eas
tern belt of Ile de France (within the Bassin Parisien region) - where in fact
the greatest concentration of innovative firms is found by the Survey - whilst
the north-west and the south of the country appear to be relatively lagging
behind. Such a difference in the geographical spread of innovative activities is
mainly attributed to the regional industrial structure - i.e. the industrial sectors
most represented in the eastern part of France are those in which small and
medium enterprises (SMEs) show a relatively stronger propensity to
innovate - and to the effects of local policies for research and technological
development (Ministre de l'Industrie, 1994).
Even accounting for demographic and economic size, both in terms of R&D
expenditure (as a percentage of GDP) and of R&D personnel (as a percentage
of active population) le de France drives up the national figure: 3.3% against
a national average of 2.4% in the case of the former indicator, and 2.3%
against 1.5% with respect to the latter (5). It has to be remembered that, in
absolute terms, the highest R&D expenditure among all EU regions (at the
comparable geographical level) is indeed that of Ile de France: more than
11,400 million of euros in 1995 (the same figure, for example, for the South
East of the UK was approximately 6,000). Moreover, the French capital region
is at the top of the ranking among the EU regions which are both wealthy (per
capita GDP above the EU average) and innovative (R&D as a share of GDP
above the EU average) (Eurostat, 2000).
Turning to the indicator used in the present work, Table 1 reports the shares
of patents granted by the USPTO to large corporate inventors - both the lar
gest French firms and the largest French-located foreign-owned firms - attr
ibutable to research facilities based in the French regions (6). First of all, in
terms of absolute size of large firms'
patenting activity, France, showing
(5) It is worth mentioning that, in 1995, the capital region accounted for nearly 50% of total
national R&D expenditure (Eurostat, 1999).
(6) The dpartements d'outre-mer are not considered here, as they do not register any US
patents by large firms in the period considered. Thus, there are 8 regions included in Table
1, instead of the 9 NUTS1 French regions.
REVUE D'CONOMIE INDUSTRIELLE n 109, Ier trimestre 2005 1 3
28,106 patents in the period 1969-95 as a whole, represents less than one third
of the overall activity carried out in Germany (with 92,058 patents), lags well
behind the UK (with 35,219), but is far above Italy (with 7,040). The share of
foreign-owned firms in the overall total is 25.6%: this is consistent with other
studies on the economic role of foreign-owned affiliates, which in fact place
France in an intermediate position between the highly globalised character of
the research carried out in the UK and the endogenously-based strength of
German technological competence (OECD, 1999) (7).
It is interesting to report a few features which characterise corporate inno
vation in France, as they emerge again from the National Innovation
Survey (8). In general terms, foreign-owned affiliates show on average a stron
ger propensity towards product innovation than their nationally-owned count
erparts, which seem to be relatively stronger in process innovation. They dif
fer significantly also with regard to the sources of innovation: whilst foreign-
owned firms rely more on external sources, indigenous French firms build pr
imarily on in-house research (Dupont, 1994). To some extent, this might be
interpreted as a relatively more pronounced tendency by foreign-owned affi
liates to establish inter-firm networks for innovation as part of the overall stra
tegy of MNCs outside their country of origin. Moreover, foreign-owned firms
(7) France is ranked between the UK and Germany in terms of both the foreign affiliate share
of R&D expenditure and the share of production (turnover) in manufacturing (OECD,
1999).
(8) The Survey referred to the innovative activities undertaken in the period 1986-91. The
source of information is Dupont (1994).
Table 1 - Shares of US patents of both the largest French firms
and the France-located foreign-owned firms, attributable to research
in the French regions relative to France as a whole, 1969-95 (%)
REGIONS
Ile de France
Bassin Parisien
Nord-Pas-de-Calais
Est
Ouest
Sud-Ouest
Centre-Est
Mditerrane
Total France (absolute nos.)
French firms
58.3
8.4
1.3
3.6
2.3
4.7
17.4
4.0
20902
Foreign firms
58.2
14.0
1.0
7.0
2.2
1.9
6.9
8.7
7204
Total
58.3
9.8
1.2
4.5
2.3
4.0
14.7
5.2
28106
14 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
show on average a greater sensitivity to the association between technological
innovation and organisational innovation, which is a crucial aspect in the capac
ity of firms to source competencies in various foreign locations.
As emerges from Table 1, geographical agglomeration turns out to be as
outstanding in the French case as in other comparable economies: with refe
rence to the first regional core, more than 58% of the overall patenting activi
ty is concentrated in Ile de France, followed by the Centre-Est, with almost
15%, and Bassin Parisien, with slightly less than 10%. The latter region, howev
er, is the second most popular location for foreign research carried out in the
country (14%), whilst the Centre-Est ranks fifth in order of importance,
accounting only for 6.9% of foreign-owned France-based patenting, in spite of
having the second largest share of indigenous research (17.4%).
The high geographical polarisation of innovation generated in MNCs drove
our choice to restrict the analysis at a more detailed level to the three regions
mentioned above, since the absolute numbers of patents granted in the other
regional sites is too low for meaningful statistical analysis. It is necessary to
bear in mind, however, the differences in the degree of attractiveness of exter
nal resources that - even in comparison with the patterns of spatial distribution
of indigenous research - mark out these three regional systems, which per se
lend support, at least at first glance, to our hypothesis of the existence of a
regional hierarchy within national boundaries.
The sectoral distribution of patents demonstrates interregional differences to
an even greater extent. Table 2 reports the foreign shares by sector and region
(foreign-owned firms' percentage of total patents granted to large firms for
research located in the region) in the 27 years considered (9). The highest
contribution of foreign research to the regional total is recorded in Bassin
Passin (36.5%), which is much above that observed for le de France (25.6%,
identical to the national average); as expected, the Centre-Est is far less attrac
tive - with the foreign share accounting for just 12% of regionally -based
research.
It is interesting to note that, in all cases, the most relevant contributions of
foreign-owned research to the local totals are found in some fast-growing and
'general purpose' technological fields, which are likely to lie at the heart of
spillovers between indigenous and foreign-owned firms. This is particularly
the case of Information and Communication Technology (ICT) sectors - such
(9) Some of the technological sectors were dropped from the table on the grounds of the rela
tively small number of patents for corporate research in the country as a whole, the ou
tcome being that only 31 sectors are reported in Tables 2 and 3 and thus discussed in the
text. However, the total 56 refers to the total number of patents for all 56 technological
sectors. The key to the sectoral codes is given in the Appendix.
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 1 5
Table
2 - Foreign shares (foreign-owned firms' percentage
of total patents granted to large firms for local research in France),
by sector and region, 1969-95
SECTORS
3
5
7
9
10
11
12
13
14
16
17
20
23
28
29
31
33
34
35
36
37
38
39
40
41
42
49
50
51
53
56
Total Tech56
ILE DE FRANCE
3.6
13.9
22.1
17.5
4.8
40.5
35.3
16.0
25.0
12.8
21.0
28.9
20.4
7.3
29.4
12.8
35.6
38.4
17.9
29.9
14.0
20.5
17.7
30.0
35.4
13.0
10.5
27.3
2.2
30.9
16.1
25.6
REGIONS
BASSEV PARISIEN
10.8
44.5
38.3
10.9
5.3
11.5
50.0
25.6
49.4
30.5
36.2
52.0
28.6
33.3
37.6
15.8
50.0
50.0
25.0
60.0
65.9
56.6
66.0
77.9
75.6
51.4
45.2
4.8
32.8
45.8
9.8
36.5
CENTRE-EST
1.0
10.9
4.8
5.3
6.3
2.5
27.5
9.6
22.4
9.8
16.0
3.7
50.0
34.4
40.5
25.0
25.0
27.6
-
2.4
5.6
15.8
18.7
13.5
14.1
60.0
3.5
10.1
2.8
25.8
22.7
12.0
FRANCE
5.5
18.6
23.6
17.0
17.1
25.5
37.5
13.7
24.9
18.1
19.6
28.9
21.9
23.6
29.0
13.1
42.4
39.8
18.9
28.3
19.9
24.6
20.4
39.7
40.2
24.7
13.1
19.3
8.9
30.4
16.2
25.6
as Telecommunications (33, 34), Semiconductors (40) and Office equipment
and data processing systems (41) - but also of Other general industrial equip
ment (29) and Other instruments and controls (53). A high foreign share at
both national and regional levels is observable in Pharmaceuticals and bio
technology (12).
As expected - and in line with what was observed for the South East in the
UK, Baden Wrttemberg in Germany and Lombardia in Italy - the sectoral
16 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
distribution
of foreign-owned companies' shares of regional and national
research activity shows a very high degree of correspondence between le de
France and the country as a whole. However, contrary to what was observed
in the UK and Italian studies, which, for the regional centres other than the
most prominent one, showed more focused and locally specific sectoral pat
terns in the distribution of foreign participation in regional research, the other
two French regions indicate foreign contributions that are relatively wides
pread at a sectoral level. This is especially the case of Bassin Parisien, which
shows remarkably high foreign shares in all sectors classified as various cate
gories of electrical equipment (and not only in those identifiable as ICTs) -
such as Illumination devices (37), Electrical devices and systems (38) and
Other general electrical equipment (39) - and also in some chemical processes
(5 and 7). The Centre-Est shows foreign shares that are well above average in
a different set of mechanical fields, such as for example Mining equipment
(23) and Other specialised machinery (28), which have instead relatively lower
values in the other regional cores and in the country as a whole.
Looking at the evolution over time of, respectively, the French-owned and
foreign-owned firms' total number of patents, it emerges that both French and
foreign patenting have followed a rather similar and discontinuous path:
increasing in the earliest 1970s, with a sharp downturn in the latter half of the
decade - which fell to a trough in 1979 - and then increasing again up to the
beginning of the 1990s, since which time the trend has been declining (see also
Cantwell and Kotecha, 1997). The evolution of foreign shares between 1969
and 1995 shows a great deal of differentiation across regions. le de France and
the country as a whole have rather stable foreign shares, whose development
over time is almost coincident, becoming divergent only in the last years
observed. Bassin Parisien has been characterised by a very high degree of fluc
tuation: although the foreign contribution to total regional patenting has remai
ned the highest throughout almost the whole period, in 1995 it was more or
less at the same level as at the end of the 1960s. The Centre-Est has instead
followed a steadily increasing trend which, in the early 1990s, has enabled this
region to overtake the capital core in terms of foreign share of research activit
y.
In general terms, the above picture confirms the traditional representation of
the French innovative and economic system, which is known to be one of the
most geographically hierarchical in Europe. The centralisation of actors and
activities - firms, capital, infrastructures, innovation, lobbies - within the Paris
region has been recognised as one of the key structural factor of the French
economic geography, and also the gradual process of decentralisation underta
ken during the 1970s and the 1980s was organised in hierarchical terms (some
times viewed rather as a 'metropolitanisation' process). Indeed, contrary espe
cially to the Italian examples of agglomeration - such as the famous industrial
districts - French local systems are much more 'urban systems', i.e. growth
poles represented by large industrial cities (a typical example is, Grenoble in
the Centre-Est), with weak connections with the rest of the territory, highly
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 1 J
integrated within the national economy and showing direct relationships with
the State (Le Gales et al., 1999). Moreover, as also the comparison within the
EU area clearly highlights, since the 1980s le de France has captured the bulk
of economic growth in France in relation to the internationalisation and multi-
nationalisation processes that have involved the national economy (see also
Mucchielli, 1998; Crozet et al, 2004).
IV. - TECHNOLOGICAL SPECIALISATION AND THE REGIONAL
HIERARCHY IN FRANCE
4.1 Comparative regional technological advantages
As highlighted elsewhere, one of the main drawbacks of using absolute numb
ers of patents is the difficulty associated with then making comparisons be
tween the activity of heterogeneous areas of technological endeavour. Since the
propensity to patent is higher in certain fields of activity (for example, phar-
maceuticals), this poses potential problems when undertaking comparative
analyses. This can be circumvented, however, by employing the Revealed
Technological Advantage (RTA) index, a technique first applied by Soete
(1987) and subsequently developed by Cantwell (1989, 1993). It is a proxy for
technological specialisation and is calculated in the following way:
where: P^ = number of patents granted in region i in technology j
Pwj = number of world patents granted in technology j
The RTA for a given region in a specified technology will vary around unity.
An index greater than one indicates a relative advantage (or specialisation) in
this technology, whereas an index less than one points to a relative disadvan
tage (10).
Table 3 reports the RTA values by region and France as a whole (11), and
sector. The overall picture seems to indicate some peculiarities, also with res
pect to our previous EU studies. First of all, although the technological advan
tages of both the nationally-owned and foreign-owned firms are - as expec
ted - more widely dispersed at the sectoral level in the country as a whole (the
cross-sectoral variance in the RTA index of France is the lowest), the sectoral
(10) Note that this is a proxy for relative (as opposed to absolute) advantage. A small region
could demonstrate a high RTA in a particular sector but this could actually be associated
with a low patent count in absolute terms.
(11) In this case the index obviously considers 2, where i = 1,...., 9 (the 9 NUTS 1 regions).
1 8 REVUE D'CONOMIE INDUSTRIELLE n 109, Ie' trimestre 2005
concentration is higher for indigenous firms than for foreign affiliates in all
cases (12). The highest value is found in the capital region, where local firms
show the strongest technological competence in Bleaching and dyeing (10)
- which however does not attract at all external resources, as appears in terms
of foreign shares - and Special radio systems (35) - which instead is also the
highest comparative advantage of foreign research located in the region.
Besides the latter sector, in both le de France and the country as a whole the
technological specialisation of foreign-owned affiliates overlaps with that of
local firms also in the 'general purpose' technological areas of Miscellaneous
metal products (14), Other general industrial equipment (29) and
Telecommunications (33), and in Pharmaceuticals and biotechnology (12). In
the latter case, it should be remembered that France imposes a local research
requirement on companies that sell to its health authorities, thus the matching
between foreign research and the local one may be partially due to this regu
latory constraint. From Table 3 it is rather evident that the sectoral concentrat
ion of indigenous research in the capital region drives the distribution of forei
gn research to a much greater extent than in the case of France, where indige
nous firms (and, only in some cases, foreign affiliates) are advantaged also in
Other organic compounds (11), Metallurgical processes (13), and in some
'general purpose' technologies (or GPTs) such as Other specialised machinery
(28) and electrical systems (37 and 38). Another relevant point to notice is that,
notwithstanding the fact that le de France records the highest concentration of
innovative efforts of indigenous large firms, the cross-sectoral variance of
foreign-owned technological activity is the lowest amongst the three regions.
Contrary to what was observed in the UK and in Italy, the other two French
regional centres display, even at a first glance, a relatively lower overlapping
between nationally-owned and foreign-owned firms' innovative activities.
Particularly in the case of Bassin Parisien - where comparative advantages of
both categories of firms are found in some chemical processes (5 and 7), in
some specialised equipment (16 and 17), in general and specialised industrial
equipment (28 and 29) and in Rubber and plastic products (49) - the differen
ce is striking between local and foreign technological profiles in some groups
of technologies. In fact, while indigenous large firms are highly specialised in
the group of other organic chemicals (9 and 11) - where foreign research not
only is completely despecialised, but also in terms of the contribution to the
regional total its share is negligible - foreign -owned affiliates show a very
strong specialisation in various categories of electrical equipment (37, 38 and
39), Semiconductors (40) and Internal combustion engines (42), in which indi-
(12) This is consistent also with studies on MNC locational choices of production activities in
the French Departments (Crozet et al., 2004). It is also worth mentioning the high comp
arative advantage of French-owned firms in Aircraft (44), shown by le de France, Bassin
Parisien and France as a whole: in this sector, a traditional point of strength of the count
ry, no foreign patents are recorded at all.
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 1 9
Table
3 - RTA index (French and Foreign firms)
relative to the world, by sector and region, 1969-95
SECTORS
3
5
7
9
10
11
12
13
14
16
17
20
23
28
29
31
33
34
35
36
37
38
39
40
41
42
49
50
51
53
56
Total Tech56
Standard deviation
ILE DE FRANCE
French firms
1.723
0.906
0.705
0.367
5.147
0.709
1.588
0.883
1.273
0.891
1.042
0.436
1.277
1.375
1.590
2.303
1.673
0.976
4.844
0.830
1.533
1.212
0.942
0.724
0.846
1.128
0.657
0.641
1.250
1.108
1.427
1.000
1.059
Foreign firms
0.185
0.427
0.582
0.227
0.752
1.403
2.521
0.491
1.237
0.381
0.804
0.515
0.954
0.316
1.929
0.988
2.691
1.771
3.071
1.029
0.724
0.911
0.589
0.903
1.346
0.491
0.224
0.700
0.081
1.439
0.797
1.000
0.747
REGIONS
BASSIN PARISIEN
French firms Foreign firms
3.647
1.309
1.416
1.891
2.490
1.693
0.809
0.751
0.884
1.529
1.302
0.758
0.617
1.376
1.003
1.022
0.491
0.496
0.604
0.151
0.604
0.624
0.420
0.605
0.107
0.682
1.945
2.434
1.585
0.591
2.137
1.000
0.788
0.768
1.824
1.526
0.402
0.240
0.382
1.404
0.448
1.496
1.164
1.285
1.426
0.429
1.195
1.049
0.333
0.853
0.861
0.349
0.392
2.028
1.411
1.417
3.700
0.572
1.251
2.786
0.214
1.342
0.867
0.401
1.000
0.774
CENTREEST
French firms
2.523
1.701
1.642
2.253
0.997
2.202
0.420
1.590
0.552
1.065
1.195
0.790
0.297
0.695
0.383
0.092
0.213
0.385
0.000
0.495
0.658
1.519
0.672
0.443
0.284
0.255
5.095
1.060
1.282
0.489
0.472
1.000
1.008
Foreign firms
0.194
1.519
0.601
0.920
0.487
0.413
1.166
1.238
1.166
0.846
1.666
0.222
2.171
2.663
1.912
0.225
0.519
1.074
0.000
0.088
0.283
2.092
1.131
0.507
0.341
2.801
1.368
0.869
0.272
1.248
1.016
1.000
0.751
FRANCE
French firms Fore
2.450
1.108
0.919
0.924
3.489
1.139
1.269
1.039
1.198
1.025
1.248
0.580
1.091
1.168
1.294
1.494
1.238
0.841
2.978
0.703
1.104
1.112
0.929
0.623
0.636
0.794
1.602
0.885
1.163
0.920
1.380
1.000
0.640
ign firms
0.417
0.735
0.825
0.548
2.085
1.130
2.205
0.479
1.150
0.656
0.885
0.684
0.885
1.046
1.531
0.652
2.647
1.615
2.007
0.806
0.793
1.053
0.692
1.191
1.238
0.756
0.698
0.613
0.329
1.168
0.773
1.000
0.559
genous competence turns out to be rather weak (also in terms of shares of total
regional patenting, much lower than the equivalent foreign shares).
In the Centre-Est, where two of the most important urban centres of the
country, namely Lyon and Grenoble, are located, the sectors in which both
foreign-owned and national firms register a comparative advantage are Rubber
and plastic products (49), Electrical devices and systems (38), Metal working
equipment (17) and Chemical processes (5). As in Bassin Parisien, specialisa
tion profiles differ substantially in chemical technologies, which represent a
point of strength in the local pattern of specialisation whereas foreign-owned
patenting is rather marginal and despecialised; on the other hand, the latter is
remarkably strong - both in terms of RTA values and foreign shares - in some
specialised and general machinery (23, 28 and 29), in which fields indigenous
firms' competence in the region turns out to be quite feeble, in spite of the rela
tive strength shown at the national level.
20 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
4.2.
The regional hierarchy
In order to further investigate our hypothesis of a geographical hierarchy of
regional centres within the national borders, a simple regression analysis was
carried out.
The cross-sectional regression analysis, used to ascertain the relationship
between foreign-owned and indigenous firms' research activities, was carried
out firstly for each of the three regions (the relevant subscript i = 1,2,3) and
for France as a whole for the overall period 1969-95. In the regional cases, the
adjusted version of the RTA index was used, for the purpose of overcoming the
potentially skewed distribution owing to a smaller number of patents at the
sub-national level (unlike for the country as a whole) (13). The regression was
run across all 56 technological sectors (subscript j) for the following equation:
adjRTAFORjj = a + adjRTAFREjj + e^ [1]
where i = 1 for le de France, i = 2 for Bassin Parisien and i = 3 for Centre-
Est. For France as a whole we have: RTAFORj = a + RTAFREj + Ej
The period was then subdivided into 1969-82 and 1983-95, in order to test
the relationship between the technological specialisation of foreign-owned
subsidiaries and that of indigenous firms at the two different geographical
levels - regional and national - over time. In this latter case we adopted a
Granger notion of sequential causality, and we ran the regression across 47
technological sectors (14) :
adjRTAFOR]jt = a + adjRTAFREijt_, + eijt [2]
and at the national level: RTAFORt = a + RTAFREjt.] + eJt, where t refers
to the period 1983-95 and t-1 to 1969-82.
Table 4 reports the statistics of the regional regressions for equation 1. First
of all, contrary to the results obtained in the case of the South East in the UK,
Lombardia in Italy and all six higher order German regions, in the main tech
nological core of the country, le de France, the aggregate patenting activity of
foreign firms located in the region is dependent upon the technological spe-
(13) The adjusted RTA is given by: adjRTA^ = (RTAy - l)/( RTAy + 1), ranging from -] to +1:
values between 0 and 1 (between 0 and -1) indicate a comparative advantage (disadvant
age) of region (/) in sector (/') relative to the world.
(14) The reason for having a smaller number of sectors in the lagged cross-section model is
that, when we subdivided the period 1969-95, we dropped all the technological sectors
with an overall number of patents less than 600 in the world total in both 1969-82 and
1983-95. The purpose was to avoid the inclusion of sectors with a relatively low propens
ity to patent at the world level.
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 2 1
Table 4 - Results of the regional regressions for 1969-95 (equation 1)
le de France (i=l)
adjRTAFRE i}
INPT
LM Diagnostic Statistics A
Serial Correlation X2(l)
Normality X2(2)
Heteroscedasticity X2(l)
Bassin Parisien (i=2)
adjRTAFRE 2j
INPT
LM Diagnostic Statistics A
Serial Correlation X2(l)
Normality X2(2)
Heteroscedasticity 5C"(1)
Centre -est (i=3)
adjRTAFRE 3j
INPT
LM Diagnostic Statistics A
Sena 1 Correlation X2(l)
Normality X~(2)
Heteroscedasticity X2(l)
Coefficient
0.22839
-0.18912
0.33063[0.565]
2.1949[0.334]
0.66196[0.416]
Coefficient
-0.045365
-0.21065
0.6642[0.415]
3.6451[0.162]
2.4285[0.119]
Coefficient
0.13920
-0.33500
0.4068[0.524]
4.0318[0.133]
4.1528[0.042]
Standard Error
0.11200
0.049363
Standard Error
0.13856
0.072102
Standard Error
0.16532
0.090740
T-Ratio [Prob]
2.0392[0.046]**
-3.8312[0.000]***
T-Ratio
[Prob]
-0.3274[0.745] -2.9215[0.005]***
T-Ratio
[Prob]
0.8420[0.403] -3.6919[0.001]***
No. of observation: 56
*** significant at 1%
** significant at 5%
* significant at 10%
A the LM test statistics reported are asymptotically distributed as a X random variable;
where X2 (DF=l)0.05 critical value is 3.84 and X2 (DF=2)0.05 critical value is 5.99
cialisation of indigenous firms (the coefficient is significant at 5%). Therefore,
as a first approximation, this would suggest that le de France is configurable
as an intermediate region, rather than as a region at the top of the geographic
al hierarchy within its own country, as it displays the pattern of technological
overlapping between foreign subsidiaries and local firms that was found to be
typical of intermediate regional centres of excellence. This is also true when
looking at the regression over time.
22 REVUE D'CONOMIE INDUSTRIELLE n 109, Ier trimestre 2005
Bassin Parisien and France as a whole show instead a pattern which is repre
sentative of higher order locations at the top of the hierarchy: the specialisa
tion of foreign-owned firms does not depend on the technological advantage
of French firms, suggesting that there might be other factors - more general
kinds of spillovers and locational advantages - bringing foreign firms to locat
e their research facilities into these areas. The case of Centre-Est is again less
clear-cut: the results obtained seem to indicate a relative degree of overlapping
in the technological profiles of foreign-owned patenting in the second period
(t) and French-owned patenting in the first time period (t-1) (the coefficient for
Centre-Est is significant at 10%), whilst the correlation is not significant for
the overall period.
Yet, these results are puzzling only at a first glance. First of all, as also
expected, France emerges as a country at the top of the European ranking, in
line with its relative position, as already discussed, somehow half-way bet
ween Germany and the UK, and much above Italy in terms of the relative size
of its innovation potential. Indeed, the absence of any significant relationship
between foreign innovative activity and that of indigenous firms was found as
in the case of the UK and Germany, contrary to the overall Italian model, in
which the specialisation of the foreign and Italian large firms matched for the
overall period and even more through time.
As far as le de France is concerned, there are clear factors underpinning the
peculiarity of the results obtained. As also mentioned above, from our previous
findings it turns out that foreign-owned research in higher order regions is not
attracted by any particular local strength, but rather by inter-industry agglo
meration externalities, sourcing especially 'general purpose' technologies. In
the case of le de France the regional comparative advantages lie, in the main,
precisely in the areas of the leading GPTs. This is consistent with what has
been shown by Cantwell and Iammarino (2001) - carrying out a more in-depth
inspection of change, stability and strengthening of technological comparative
advantages across European regions over time - in which the French capital
region records one of the strongest concentrations in terms of number of sec
tors showing a consolidation of technological specialisation; such a process
seems indeed to have occurred particularly in GPTs and in core technological
systems. This suggests that foreign-owned firms from a wider range of indust
ries than locally are attracted into the region, but their technological focus is
then similar to the local structure of comparative advantages - i.e. overlapping
in ICT [telecommunications and radio systems (33 and 35), a traditional
French strength], metal working and general machinery (14 and 29), and gener
al instruments (53), all of which are leading GPTs.
To further corroborate this interpretation, an additional inspection was car
ried out with reference to the industry of output of large firms located in the
French regions. In the case of le de France, this exercise confirmed that fore
ign-owned firms are from a different range of industries as compared to French
firms, as shown by a negative correlation coefficient between the two RTA dis-
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 23
tributions when constructed by industry of output. Therefore, it is possible to
conclude that this is clearly a higher order region, insofar as the similarity of
the indigenous and foreign technological profiles can be viewed as a coinci
dental consequence of the regional pattern of technological specialisation
within the firms of different industries.
With regard to the other two regions, as seen above Bassin Parisien does not
create problems in its identification as a higher order centre. The Centre-Est,
instead, should be definitely categorised as an intermediate regional location.
The main argument, supported quite clearly looking at patents at the firm level,
is that foreign-owned firms are here attracted to conduct local research in sec
tors in which there is indeed a strength owing to local firms but, given that the
foreign entrants are often classified in a related rather than the identical indust
ry, they locate allied research also into what (for them) is the primary techno
logical field (which is not necessarily a local specialisation, although it can be
related to the area of local strength). In other words, in the case of Centre-Est
there is indeed a specialised attraction of the kind which is expected in tech
nologically intermediate regions, but this is obfuscated by the fact that it relies
in part on technological overlaps between firms of related but distinct indust
ries. However, at the aggregate level, only in this region was a positive corre
lation found between the overall distribution of indigenous and foreign-owned
firms across industries, giving further support to the idea that the exchange of
competence and knowledge in this location is much more focused than in the
other two regional cores and mainly of an intra-industry kind, and at the very
least occurs between quite closely related industries.
V. - IMPLICATIONS AND CONCLUSIONS
The French case has shown that the regional hierarchy is far from being a
clear-cut and rigid classification of local contexts within and across EU natio
nal borders. A categorisation without an identification of the types of areas
would have implied an oversimplification of the complex interactions between
the global and the local dimensions of the generation of technological innovat
ion. As already highlighted, distinctiveness and specificity are the key-words
in analysing national and regional systems of innovation. The results obtained
for France and its regional cores fit with our existing framework, providing
further insights on the nature of the regional hierarchy on the basis of the rel
evance of GPTs and core technological systems in higher order centres, and on
the technological overlapping of large firms operating in sometimes different
but allied industries in intermediate regions. We have concluded that in both
le de France and Bassin Parisien - as in France as a whole - the exchange of
knowledge is essentially inter-industry, while in the intermediate region of the
Centre-Est it has much more of an intra-industry character. It has been shown
that foreign-owned affiliates tend to seek out all-round centres of excellence
either for the diversification of their capabilities and/or for the development of
GPTs, in which case their distribution by industry of output is likely to be rela
tively dispersed given the rather horizontal nature of such 'general purpose'
24 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
technologies. On the other hand, specialised technical expertise may attract
external resources also from related industries - a mechanism through which
intermediate centres might enjoy technological spillovers over a wider secto
ral spread, possibly evolving into higher order regions.
However, the above results need to be interpreted also in the light of the
peculiar character of the French innovation system, which distinguishes it
from other EU countries to a rather large extent and provides a sounder fr
amework for the explanation of underlying regional and national differences.
These characteristics can be summarised as follows (Chesnais, 1993):
the pervasive State involvement in the generation not only of general scient
ific and technical knowledge, but also of technology per se, in the form of
patentable and immediately usable new products or processes;
as already highlighted, one of the most distinctive features of the French eco
nomy is the relatively low level of territorial concentration, which differentiates it
especially from the Italian case. The general economic centralisation within the
capital region has always been a key structural feature of France, and even decent
ralisation has been strictly organised by the central State;
the existence of vertically structured and strongly compartmentalised tech
nological sub-systems, such as electrical power, telecommunications, electro
nics, aerospace, transport systems, all involving a strong alliance between the
State and public/private firms belonging to the oligopoly of the French indust
ry. As also mentioned above, the role of the State in technological innovation
is comparatively stronger than that in other EU countries, insofar as it is hea
vily present also in sectors in which government interference is usually quite
low, such as in the chemical-pharmaceutical group;
large and very large firms have been the almost exclusive partners in such
public-private technological alliances. Unlike in the case of Germany and,
especially, Italy, in France the active role played by SMEs in the overall inno
vation system is rather limited.
The reasons for these peculiarities are rather complex and deeply rooted in
historical processes. However, as pointed out by Chesnais (1993), two elements
at least need to be mentioned: the inherent historical weakness of French capi
talism, at the heart of the need for government support, and the role played by
the "Grandes coles" in shaping the strong linkages between the State apparat
us and the private sector. These elements underlie also the kind of hierarchical
segmentation of the geographical space which is specific to the French case. In
fact, Paris has been defined as the "thinking head of the national productive sys
tem" (Beckouche, 1991). The Paris region, and its complex concentration of
large innovative French corporations, have benefited extensively of the vast pro
jects of the Etat high tech aimed at building up a national capacity in leading
GPTs. Thus, in the case of le de France, it seems possible to argue that the
regional system "was massively structured by a particular technocratic combi-
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 25
nation
of public and private actors", which is now in a process of adaptation and
restructuring to global competition (Le Gales et al, 1999).
Finally, as the cases of le de France and Centre-Est have shown, there is the
need to further improve our understanding of some aspects of the effects of
innovation and globalisation on firms and regions - i.e. technological spillo
vers - by examining more in detail the patterns of MNC technological versus
production specialisation in each region. Differences between the two specia
lisation profiles may be indicative of technological diversification by industry,
and hence potential technological overlaps between industries, with important
implications for the evolution of the regional innovation system as a whole.
REFERENCES
ABRAMOVITZ, M. (1986), "Catching up, forging ahead, and falling behind", Journal of
Economic History, 46, 385-406.
BECKOUCHE, P. (1991), "French high-tech space: a double cleavage", in G. BENKO and
M. DUNFORD (eds.), "Industrial change and regional development: the transformation of
new industrial spaces", London and New-York, Belhaven Press.
CANTWELL, J.A. (1989), "Technological Innovation and Multinational Corporations", Oxford
Basil Blackwell.
CANTWELL, J.A. (1992), "The internationalisation of technological activity and its implica
tions for competitiveness", in O. GRANSTRAND, L. HKANSON and S. SJLANDER
(eds.), "Technology Management and Internationa] Business: Internationalisation of R&D
and Technology", Chichester, John Wiley.
CANTWELL, J.A. (1991), "The theory of technological competence and its application to inter
national production" in D. McFETRIDGE (ed.), "Foreign Investment, Technology and
Economic Growth", Calgary, University of Calgary Press.
CANTWELL, J.A. (1995), "The globalisation of technology: what remains of the product cycle
model?", Cambridge Journal of Economics, 19, 1, 155-174.
CANTWELL, J.A. and KOTECHA, U. (1997), "The internationalisation of Technological
Activity: the French Evidence in a Comparative Setting", in J. HOWELLS and J. MICHIE
(eds.), "Technology, Innovation and Competitiveness", Aldershot, Edward Elgar.
CANTWELL, J.A. and IAMMARINO, S. (1998), "MNCs, Technological Innovation and
Regional Systems in the EU: Some Evidence in the Italian Case", International Journal of
the Economics of Business, 5, 3.
CANTWELL J.A. and IAMMARINO, S. (2000), "Multinational Corporations and the Location
of Technological Innovation in the UK Regions", Regional Studies, 34, 3, 317-322.
CANTWELL J.A. and IAMMARINO, S., (2001), "EU Regions and Multinational
Corporations: Change, Stability and Strengthening of Technological Comparative
Advantages", University of Reading Working Paper.
CANTWELL J.A. and IAMMARINO, S. (2003), "Multinational Corporations and European
Regional System of Innovation", London and New-York, Routledge.
CHESNAIS, F. (1992), "National Systems of Innovation, foreign direct investment and the ope
rations of multinational enterprises" in B.A. LUNDVALL (ed.), "National Systems of
Innovation", London, Pinter.
CHESNAIS, F. (1993), "The French National System of Innovation", in R.R. NELSON (ed.)
"National Innovation Systems. A Comparative Analysis", Oxford and New-York, Oxford
University Press.
26 REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005
CROZET,
M., MAYER, T. and MUCCHIELLI, J.-L. (2004), "How do firms agglomerate? A
study of FDI in France", Regional Science and Urban Economics, 34, 27-54.
DICKEN, P. (1994), "The Roepke Lecture In Economic Geography: Global-Local Tensions:
Firms and States in the Global Space Economy", Economic Geography, 70, 2, 101-128.
DUNNING, J.H. and WYMBS, C. (1999) "The Geographical Sourcing of Technology Based
Assets by Multinational Enterprises", in D. ARCHIBUGI, J. HOWELLS and J. MICHIE
(eds), "Innovation Policy in a Global Economy", Cambridge: Cambridge University Press.
DUPONT, M.J. (1994), "Les filiales trangres en France: des atouts matres pour innover", in
ministre de l'Industrie etc.
FAGERBERG, J., VERSPAGEN, B. and von TUNZELMAN, N. (eds.) (1994), "The dynamics
of technology, trade and growth", Aldershot, Edward Elgar.
GRANSTRAND, O. and SJLANDER, S. (1992), "Internationalisation and diversification of
multi-technology corporations", in O. GRANSTRAND, L. HKANSON and
S. SJLANDER (eds.), "Technology Management and International Business:
Internationalisation of R&D and Technology", Chichester, John Wiley.
Le GALES, P., ANNIELLO, V. and TIRMARCHE, O. (1999), "The Governance of local eco
nomies in France: after the Satet and national champions, emerging structure of local
governance?", Mimeo.
Ministre de l'Industrie, des Postes et Tlcommunications et du Commerce extrieur (1994),
"Les chiffres cls. L'innovation technologique dans l'industrie", Paris.
MUCCHIELLI, J.-L. (d.) (1998), "Multinational Location Strategy", Greenwich, JAI Press.
OECD (1999), "Measuring Globalisation. The Role of Multinationals in OECD Economies",
Paris.
PATEL, P. and VEGA, M. (1999) "Patterns of internationalisation of corporate technology:
location versus home country advantages", Research Policy, 28, 2/3: 145-155.
PEARCE, R.D. (1999), "Decentralised R&D and strategic competitiveness: globalised
approaches to generation and use of technology in multinational enterprises (MNEs)"
Research Policy, 28, 2-3,157-178.
SOETE, L. (1987), "The Impact of Technological Innovation on International Trade Patterns:
the Evidence Reconsidered", Research Policy, 16, 101-130.
ZANDER, I. (1997), "Technological diversification in the multinational corporation - historical
evolution and future prospects", Research Policy, 26, 209-227.
See Appendix next page
REVUE D'CONOMIE INDUSTRIELLE n 109, 1er trimestre 2005 27
Appendix
- The 56 technological sectors
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
Food and tobacco products
Distillation processes
Inorganic chemicals
Agricultural chemicals
Chemical processes
Photographic chemistry
Cleaning agents and other compositions
Disinfecting and preserving
Synthetic resins and fibres
Bleaching and dyeing
Other organic compounds
Pharmaceuticals and biotechnology
Metallurgical processes
Miscellaneous metal products
Food, drink and tobacco equipment
Chemical and allied equipment
Metal working equipment
Paper making apparatus
Building material processing equipment
Assembly and material handling equipment
Agricultural equipment
Other construction and excavating equipment
Mining equipment
Electrical lamp manufacturing
Textile and clothing machinery
Printing and publishing machinery
Woodworking tools and machinery
Other specialised machinery
Other general industrial equipment
Mechanical calculators and typewriters
Power plants
Nuclear reactors
Telecommunications
Other electrical communication systems
Special radio systems
Image and sound equipment
Illumination devices
Electrical devices and systems
Other general electrical equipment
Semiconductors
Office equipment and data processing systems
Internal combustion engines
Motor vehicles
Aircraft
Ships and marine propulsion
Railways and railway equipment
Other transport equipment
Textiles, clothing and leather
Rubber and plastic products
Non-metallic mineral products
Coal and petroleum products
Photographic equipment
Other instruments and controls
Wood products
Explosive compositions and charges
Other manufacturing and non-mdustnal