Innovation, Regions and Proximity: From Neo-Regionalism to Spatial Analysis

Abstract

Shearmur R. Innovation, regions and proximity: from neo-regionalism to spatial analysis, Regional Studies. There have recently been questions regarding how geographic proximity should be conceptualized in the study of regional innovation. This stems partly from different meanings of the term ‘innovation’ (incremental product and process innovation in this paper) and partly from the way space is usually conceptualized: regions are conceived as bounded territories with particular attributes (neo-regionalism). Drawing upon spatial analytical concepts, it is suggested in this paper that an alternative to neo-regionalism is to view space as a continuous field of opportunities, with accessibility to factors of innovation playing a key role for local innovativeness. An analysis of Québec innovation data corroborates this approach.

Shearmur R. Innovation, regions et proximité: du néorégionalisme à l'analyse spatiale, Regional Studies. La conceptualisation de la notion de proximité géographique dans les études sur l'innovation soulève actuellement des débats. Ces débats reposent en partie sur les différentes acceptions du terme innovation (dans cet article il s'agit d'innovations incrémentales de produits et de procédés), en partie sur la manière dont on appréhende l'espace: en général, celui-ci est conçu comme constitué de territoires délimités, chacun doté de caractéristiques particulières. En s'inspirant de notions d'analyse spatiale, cet article propose une manière alternative d'introduire l'espace dans les études d'innovation, c'est-à-dire de le considérer comme un champ continu de possibilités. Dans ce cadre, c'est l'accessibilité relative aux facteurs d'innovation qui serait la clé de l'innovation locale. Des résultats empiriques portant sur le Québec corroborent cette idée.

Systèmes d'innovation Analyse spatiale Néorégionalisme Québec Géographie Industrie

Shearmur R. Innovation, Regionen und Nähe: vom Neoregionalismus zur Raumanalyse, Regional Studies. In letzter Zeit wurde die Frage laut, wie geografische Nähe bei der Untersuchung regionaler Innovation konzeptualisiert werden sollte. Der Ursprung dieser Frage liegt zum Teil in den verschiedenen Bedeutungen des Begriffs der ‘Innovation’ begründet (in diesem Beitrag eine gesteigerte Produkt- und Prozessinnovation) und zum Teil in der Art, in der der Raum gewöhnlich konzeptualisiert wird: Regionen werden als abgegrenzte Gebiete mit bestimmten Merkmalen aufgefasst (Neoregionalismus). Ausgehend von räumlichen Analysekonzepten wird in diesem Beitrag als Alternative zum Neoregionalismus vorgeschlagen, den Raum als kontinuierliches Feld von Gelegenheiten zu betrachten, wobei die Verfügbarkeit von Innovationsfaktoren für die Innovationsfähigkeit vor Ort eine zentrale Rolle spielt. Bei einer Analyse der Innovationsdaten von Quebec bestätigt sich dieser Ansatz.

Innovationssysteme Raumanalyse Neoregionalismus Quebec Geografie Produktion

Shearmur R. Innovación, regiones y proximidad: del neorregionalismo al análisis espacial, Regional Studies. Recientemente se ha cuestionado cómo debería ser conceptualizada la proximidad geográfica en el estudio de la innovación regional. El origen de esta pregunta radica en parte en los diferentes significados del término ‘innovación’ (en este artículo, innovación incremental de productos y procesos) y en parte en cómo se suele conceptualizar el espacio: las regiones se conciben como territorios limitados con características particulares (neorregionalismo). Basándonos en conceptos analíticos espaciales, en este artículo sugerimos que una alternativa al neorregionalismo es considerar el espacio como un campo continuo de oportunidades en el que la capacidad de acceder a factores de innovación desempeña un papel fundamental para la capacidad innovadora a nivel local. Un análisis de los datos de innovación en Quebec corrobora este planteamiento.

Sistemas de innovación Análisis espacial Neorregionalismo Quebec Geografía Fabricación

Keywords:

• Innovation systems
• Spatial analysis
• Neo-regionalism
• Québec
• Geography
• Manufacturing

JEL classifications:

• O31
• R12

Acknowledgements

The author would like to thank Meric Gertler and David Wolfe for inviting him to participate in the Innovation Systems Research Network (ISRN), a forum within which these ideas were developed. The author would also like to thank David Doloreux, Réjean Landry, and Mark Freel with whom he had lively discussions about questions of geography and innovation. Finally, the Institut de la Statistique du Québec (ISQ), and its data access centre (CADRISQ), are thanked for facilitating access to the data. The usual disclaimers apply.

Notes

The connection between regional innovation and regional development is often assumed; however, it does not follow from the fact that innovation depends on regional attributes that regional development depends upon innovation. Evidence from some regions (for instance, aluminium-producing regions in Québec), shows that innovation can lead to regional decline in a context of industrial specialization (Polèse and Shearmur, 2002).

Thus, the literature on knowledge spillovers, principally associated with the analysis of patents or world-first products (Jaffe, 1989; Breschi and Lissoni, 2001; Ejermo and Grasjo, 2008; Feldman and Audretsch, 1997), is not the focus of the present paper. This is because the aim is not to understand how regional attributes feed into the global innovation system, but rather to understand which regional attributes seem to encourage the local adaptive behaviour of firms.

Storper (1995) argues that the region should be considered a ‘fundamental unit of social life’ (p. 191).

It should be pointed out that Storper (1995) believes that by providing a theoretical framework for this neo-regionalist approach, he can enable certain generalizations and comparisons to be made across regions.

This approach requires extensive data on the location of potential innovation inputs, and also requires detailed hypotheses as to the nature of these inputs (because each can be tested individually). Since many such inputs co-locate in metropolitan areas, and since it is not always clear which innovation process is occurring in which location (see above), the approach outlined by McCann (2007), whilst simpler and premised primarily on the frequency and importance of interactions, is also more robust to the extent that it need not specify the nature of the interactions.

The only area related to innovation studies in which a spatial analytic approach is relatively common is the study of knowledge spillovers, usually proxied by the study of patents and co-patenting (for example, Maggioni et al., 2007; and Greunz, 2004). Breschi and Lissoni (2001) argued that that is a very specific area of study. This body of work has not (yet?) fundamentally modified the geographic conceptualizations that prevail in innovation studies, partly because results relating to the distance over which knowledge is transmitted are inconclusive, and partly because patents are a very specific type of (major) innovation.

The organizations that participated in financing the manufacturing census were as follows: Institut de la Statistique du Québec, Ministère du Développement économique, de l'Innovation et de l'Exportation du Québec, Industrie Canada (région du Québec), Conseil National de Recherche du Canada (région du Québec), Ministère des finances du Québec, and Conseil de la Science et de la Technologie du Québec. The Innovation Studies Research Network and INRS research funds financed access to these data.

For the wording of innovation questions, see Appendix B(i).

For establishment-level product innovation, the regional effect enters the model with a 0.12 probability of being equal to zero. The cut-off for indicating statistical significance in the present exploratory analysis was 0.10.

For the definition of location quotients and specialization, see Appendix B(ii).