Why, Where and With Whom Do You Link? The Nature and Motivations of Linkages Within and Outside an Italian Local System

Abstract

Sacchetti S. Why, where and with whom do you link? The nature and motivations of linkages within and outside an Italian local system, Regional Studies. Building on the work of Richardson (1972) and on the need of firms to coordinate activities with others in search for related yet different competences, the paper proposes the results of a survey in the Italian Province of Reggio-Emilia (Emilia-Romagna, Italy). Concerns are related, in particular, with the geography of linkages, their nature, as well as with the profile of firms addressing specific types of objectives. Using categorical principal component analysis as well as correspondence analysis, the paper associates particular groups of motivations with firms' characteristics. The conclusions support the view according to which the increasing complexity and knowledge intensity of economic activities requires that inter-firm linkages are considered beyond spatial proximity and that knowledge-related motivations are associated with specific firms' profiles and patterns of specialization of activities.

Sacchetti S. Pourquoi, où et avec qui faut-il s'associer? La nature et les motivations des liens au sein et en dehors d'un système local italien, Regional Studies. A partir de Richardson (1972) et du besoin des entreprises de s'associer pour rechercher des compétences à la fois connexes mais distinctes, on présente les résultats d'une enquête menée dans la province italienne de Reggio Emilia (en Emilia Romagna, en Italie). On porte un intérêt particulier à la géographie des connexions, leur nature, aussi bien que le profil des entreprises qui abordent des objectifs particuliers. Employant une analyse en composantes principales catégorique aussi bien qu'une analyse par concordance, on associe des groupements particuliers de motivations aux caractéristiques des entreprises. Les conclusions viennent à l'appui de l'idée selon laquelle la complexité croissante des activités économiques à intensité de connaissance nécessite que les connexions interentreprises sont considérées au-delà de la proximité géographique et que les motivations qui se rapportent à la connaissance s'associent aux profils et à la spécialisation des entreprises spécifiques.

Organisation industrielle Connexions interentreprises Proximité géographique Economie basée sur la connaissance Spécificité des biens immobiliers Anlyse en composantes principales

Sacchetti S. Verknüpfungen – warum, wo und mit wem? Wesen und Motivation von Verknüpfungen innerhalb und außerhalb eines lokalen Systems in Italien, Regional Studies. Aufbauend auf Richardson (1972) und auf dem Bedürfnis von Firmen, auf der Suche nach verwandten, aber unterschiedlichen Kompetenzen ihre Aktivitäten mit anderen Firmen zu koordinieren, stellen wir die Ergebnisse einer Studie in der italienischen Provinz Reggio-Emilia (Emilia-Romagna) vor. Insbesondere befassen wir uns mit der Geografie von Verknüpfungen, ihrem Wesen und dem Profil von Firmen, die sich mit bestimmten Arten von Zielen befassen. Unter Einsatz einer kategorischen Hauptkomponentenanalyse sowie einer Korrespondenzanalyse verknüpfen wir bestimmte Gruppen von Motivationen mit den Merkmalen von Firmen. Unsere Schlussfolgerungen bekräftigen die These, dass Verknüpfungen zwischen Firmen aufgrund der zunehmenden Komplexität und Wissensintensität der Wirtschaftstätigkeiten auch jenseits der Ebene der räumlichen Nähe in Erwägung gezogen werden müssen und dass wissensbezogene Motivationen mit spezifischen Firmenprofilen und mit Spezialisierungsmustern bei den Tätigkeiten einhergehen.

Organisation der Industrie Verknüpfungen zwischen Firmen Geografische Nähe Wissensbasierte Wirtschaft Vermögensspezifizität Hauptkomponentenanalyse

Sacchetti S. ¿Por qué, dónde y con quién? La naturaleza y las motivaciones de los enlaces dentro y fuera de un sistema local italiano, Regional Studies. Basándonos en las teorías de Richardson (1972) y siendo necesario para las empresas coordinar actividades con otras empresas en busca de competencias relacionadas pero diferentes, proponemos los resultados de un estudio en la provincia italiana de Reggio-Emilia (Emilia-Romaña). Lo que nos preocupa especialmente es la geografía de los enlaces, su naturaleza y las características de las empresas que tienen en cuenta tipos específicos de objetivos. Analizado los componentes principales y categóricos y la correspondencia, asociamos grupos determinados de motivaciones con las características de las empresas. Nuestras conclusiones confirman que para obtener una mayor complejidad e intensidad de conocimientos de las actividades económicas es necesario que se tengan en cuenta los enlaces entre las empresas independientemente de la proximidad espacial y que las motivaciones relacionadas con los conocimientos se relacionen con las características de empresas concretas y los modelos de especialización de las actividades.

Organización de la industria Enlaces entre empresas Proximidad geográfica Economía basada en el conocimiento La especificidad de bienes Análisis de componentes principales

Keywords:

• Organization of industry
• Inter-firm linkages
• Geographical proximity
• Knowledge-based economy
• Asset specificity
• Principal component analysis

JEL classifications:

• C14
• D2
• L2
• R12

Acknowledgements

The author would like to thank Rosa Arboretti, Lisa De Propris, Francesco Galassi, Francesco Sacchetti, Luigi Salmaso, Roger Sugden, Ermanno Tortia, and Marcela Valania for comments and discussion. Thanks also to Antares' fellows, in particular to Moreno Balzani, Alessia Bernardi, Simona Boari, Annalisa Campana, Lorenzo Ciapetti, Davide Consolini, and Lucio Poma. Data analysis was based on Antares' databases. The authors are grateful to the Editor and to two anonymous referees for their comments and suggestions. The usual disclaimers apply.

Notes

Boschma (2005) compares the nature of different forms of proximity, their interrelations, and their impact on learning and innovation.

The survey was designed jointly with Antares, Faculty of Political Science, University of Bologna-Forli, Italy.

Originally, the list provided by Assindustria was meant to address all manufacturing and service firms with more than ten employees. However, some of the respondents turned out to be of a smaller size. This was because initial data listed by business associations were not updated. Besides, the present analysis focuses on manufacturing and, therefore, the authors consider a sub-population of 1670 firms.

The rate of response to the postal questionnaire was affected by the fact that the survey was undertaken by Assindustria, the business association of large firms. Firms that were affiliated had a higher motivation to answer with respect to firms which were not (namely, smaller firms).

Ex-post, a stratification based on firm size (in terms of the number of employees) shows that firms with fewer than 50 employees are underestimated in the sample (proportions are 85% in the province and 65% in the sample), whilst firms with more than 50 employees are over represented: firms with 50–100 employees are 8.20% in the province and 13.07% in the sample; firms with 101–200 employees are 4.25% in the province and 12.42% in the sample; firms with 201–500 employees are 2.04% in the province and 6.54% in the sample; and firms with more than 500 employees are 0.48 in the province and 2.61 in the sample. Data on the number of manufacturing firms in the province are based on ISTAT (2001).

Size categories in terms of employees are: 4–29, 30–50, 51–100, 101–200, 201–500, and more than 500.

The data also show that the informality of relations is not (although weakly) compatible with firms linked to a parent company (the correlation index = –0.106). Throughout the analysis, only significant correlations (Cohen's Kappa) at the 0.01 level will be mentioned. Unless differently specified, the Cohen's Kappa measure of association (exact) is used, which also gives strong results for nominal dichotomic variables.

The Goodman–Krusal index = 0.067 and is significant at the 0.05 level.

For each of the objectives, the share of firms using international partnership over the share of firms relying on national partnerships was calculated.

In correspondence analysis, the graphical output is the most important. The angle existing between two categories (points) passing through the centroid of the plot (the point where the two axes cross) indicates: no correlation when the inclination is 90°; a positive correlation if the angle is less then 90° (the intensity of the correlation increases the smaller the angle is); and a negative correlation if the angle is larger then 90° (the intensity of the correlation increases as the angle becomes closer to 180°).

As the present data on network objectives are nominal, the authors have proceeded with CatPCA at first and then, using transformed numerical variables, applied factor analysis.

Loadings inferior to 0.3 have been excluded. It is usual to regard component loadings as high if they are greater than 0.6 (positive or negative) and moderately high if they are above 0.3 (Kline, 1994, p. 6). Given in parentheses are the lowest values of those loadings that recur twice.

The components identified by factor analysis were controlled using the reliability measure Cronbach's alpha. Of the five components, values between 0.7 and 0.8 were only obtained for the first three. Generally, alpha > 0.7 are considered to be acceptable. See Appendix 1 for the variability explained by components.

‘Load’ in factor analysis expresses the correlation between a variable and a component.

The Goodman–Krusal correlation index was used. It gives strong results when relating ordinal and nominal dichotomic variables.