Econométrie de la Propriété industrielle

Grand Amphithéâtre Lumière - Lyon
14 & 15 Mai 1998

Résumés des interventions / Abstracts

Compétition et comparaison internationales

W. Meeusen, G. Rayp
Université d'Anvers - RUCA
Middelheimlaan 1
B-2020 ANVERS (Belgique)

The determinants of the international competitiveness of the manufacturing industry of 11 OECD countries are examined over the period 1970-1995. International competitiveness is measured as the share of exports of each country in the 12-country total. Determinants include relative unit labour cost as an indicator of gross profit margins, patent shares as indicators of product quality, trademark shares as proxies for relative marketing efforts, and relative capacity use as a proxy for relative capacity to deliver.

The study can be seen as an extension of previous work by Soete (1981), Fagerberg (1988), Greenhalgh a.o. (1990 and 1994) and Amendola a.o. (1993). Our contribution consists of two parts the inclusion within a technology gap interpretation of competitiveness of the commercial dimension of innovation (measured by trademark shares), and the explicit consideration of two alternative theoretical frameworks in which the technology gap explanation of competitiveness might be valid. We are therefore able to indicate the sensitivity of the results to the theoretical viewpoint one takes.

The determinants of competitiveness are first examined within the more conventional international trade equilibrium framework. Non-stationarity of the variables however indicates rather slow, if any, adjustment to the projected trade equilibrium. This implies that, analytically, an error correction model (ECM) is the most appropriate. As the stylised facts of the 11 OECD countries considered seem to suggest that the ‘Kaldor paradox’ (positive correlation between competitiveness and relative unit labour costs) has some relevance, the Johansen-Juselius ECM approach (which allows for endogenity of the variables) is opted for.

Yet, for most countries only one cointegration vector can be identified (for one country cointegration is even rejected as such) and the space of the four included variables remains largely unbounded. Moreover for half of the countries examined the cointegrating vector was at odds with the necessary restrictions on the coefficients consistent with a ‘competitiveness equation’. Hence, cointegration of the variables seems rather weak, which raises some questions concerning the validity of an equilibrium viewpoint on international trade. We re-estimated therefore the competitiveness relationships assuming continuously shifting trade positions and permanent disequilibrium conditions in which strong hysteresis effects are present.

Despite their different underlying assumptions, both approaches strongly support the ‘technology gap’ interpretation of international trade flows. Cost indicators have relatively weak explanatory power. The S&T indicators perform better. Traditional S&T variables, such as the part of R&D outlays in GDP, or the investment rate in manufacturing show less clear-cut results, but the patent share variable of the contrary gives good results. Surprisingly, also the trademark shares yield results which are statistically significant for most countries.

The different R&D indicators for explaining international competitiveness are discussed in function of their a priori merits, and of the empirical results obtained. Particular attention is devoted to trademark shares.


B. Clarysse, B. Sloan
European Commission DG XII
AS-4 square de Meeûs, 8
1000 BRUSSELS (Belgique)

There has been a growing perception in recent years that Europe’s S&T system is in a rather paradoxical situation in that, on the one hand, the educational and scientific research base is acknowledged to be of high quality, while, on the other hand, it seems to be failing to translate this very healthy scientific base into strong technological and economic performance.

This paper attempts to shed light on this so-called "European Paradox" by exploring the linkages between indicators of scientific and technological output (publications and patents), and between indicators of technological investment and indicators of economic performance, while relating all these to the underlying R&D effort. The performance of the EU, and of the individual member states, is compared with those of the US and Japan. The aim is not to create a complex model of the innovation process, but rather to contribute to the debate on this topic by raising certain questions and gaining some helpful insights.

In particular the paper seeks to explore a number of questions: Is this paradox specific to Europe or do we find it in the US and Japan? Is the core problem the mismatch between science and technology or is it related to inefficient commercilaization? Can the paradox within Europe be generalized across all member states and industries?

After an examination of the linkage between the EU’s scientific and technological outputs in relation to the corresponding research efforts, the analysis moves closer to the marketplace, to look at Europe’s performance in terms of hi-tech trade and its relationship with technology investment. What emerges is a rather variegated picture, with significant differences between member states and industries, and a contrast between its rather moderate technology output and its improving trade performance.


H. Grupp, G. Muent
Fraunhofer Institute ISI
Breslauer Street 48
D-76139 Karlsruhe (Allemagne)

Recent developments in trade theory focus on the effects of innovation and knowledge accumulation. In this contribution we ask whetner pattems of specialisation in technology. and trade are converging or diverging in industrialised economies since the early 1960s. On the basis of a particular model by Grossman and Helpman and alternative approches of evolutionary theory it can be shown that continuous innovation and intra-sectoral knowledge transfer lead to either convergent or divergent patterns of specialisation, depending on the degree of openness.

An empiral annalysis with trade and patent data hints at converging trends within Europe whereas patterns among the trad blocs USA. Japan and europe seern to be diverging. The descriptive analysis is supplemented by and econometrical gap model which is based on a fifth order autoregressive time serries analysis.


R. Paci, S. Usai
University of Cagliari
Dept. of Economic and Social Sciences
Via Sant'Ignazio, 78
09123 CAGLIARI (Italie)

Recent studies have emphasised the concept of National Innovation Systems [Nelson R. (1993) and Freeman C. (1995)] while neglecting the presence of huge regional differences in technological capabilities within each country. Indeed, at the regional level threre is an even more important role for technological specialisation due to agglomeration economies [Adhur (1989) and Krugman (1991)]. This paper, following some recent contributions [Caniels (1997) and Verspagen (1997)], explores the spatial distribution of patenting activity across the whole regions of the European.

Union. The analysis is based on a original databank on regional patents statistics starting from the data collected by the European Patent office (EPO), we bave assigned each patent. to its region of origin through the postal code of trie inventor's residence. More precisely, our series refer to patent applications, classiflical by the inventor's rcgion for the 12 countries of the European Union over the eighties (1980, 1985 and 1990 - for a total of 53,270 patents). As for the geographical split we have considered 109 national and sub-national units selected in order to ensure a certain degree of economie homogeneity and administrative functionality. Another interesting feature of our database is that it allows us to illustrate the sectoral technological Specialisation of each rcgion given that we have converted the original IPC data (which is unsuitable for economie comparison) to the NACE classification at the three digit level.

The main results worth highlighting are as follows : first, the technological activity in the EU appears to be highly concentrated, although concentration tends to décline over the period. This results from the huge differences between southern and northern Europe. Moreover, as expected, the degree of disparities in the productivity distribution appears much lower with respect to innovative activity both at the European and at the country level. As a result, the correlation coefficients between the regional distribution of innovative activity and labour productivity turns out to be positive but net very hi gh (around 0.43).

More Surprisingly there appears a weak negative correlation between technological concentration and aggregate productivity, that is the European regions which enjoy a more homogeneous distribution of their tcchnological capability across different industrial sectors appear to be also characterised by a higher productivity level. This outcome suggests the presence of positive externalities and spillovers in technology and production that favours those regions that succed in covering a broader range of technological activities.


S. Negassi
CESSEFI - Paris I / SESSI - Ministère de l'Industrie

F. Favre
SESSI - Ministère de l'Industrie

E. Pfister

L’objectif de ce travail est d’étudier la composante stratégique de la formation d’alliances dans le domaine de la R&D et en particulier la spécificité des alliances regroupant des firmes appartenant à plusieurs nations différentes. Il apparaît alors intéressant de comparer les formes coopératives à d’autres formes d’internationalisation de la R&D. Ce versant spécifiquement internationale sera approfondi au sein des deux axes suivants :

1. Il convient de replacer la coopération internationale des firmes en R & D dans le cadre plus générale de leur stratégie d’internationalisation et donc d’essayer de déterminer les raisons pour lesquelles les firmes s’engagent dans des formes d’internationalisation relevant de la coopération plutôt que d’autres formes. Il semble fondamental de réintégrer la coopération stratégique dans les modèles de concurrence oligopolistique internationale.

2. Nous nous proposons ensuite d’évaluer l’influence de la coopération des firmes sur leur performance en termes de l’innovation avec une attention toute particulière sur l’origine géographique des partenaires de cette coopération, et les transferts de technologies.

Notre analyse se situe au niveau individuel des firmes. Nous disposons à cet effet de bases de données à la fois nouvelles et extrêmement riches. Elle se compose de données SESSI et EUROSTAT L’OST et la Commission Européenne sont aussi nos fournisseurs d’informations statistiques.

L’analyse micro-économétrique résulte des travaux initiés par Hildreth et Houck, développés par Swamy et généralisés par Balestra-Negassi.