Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/119936 
Year of Publication: 
2011
Series/Report no.: 
51st Congress of the European Regional Science Association: "New Challenges for European Regions and Urban Areas in a Globalised World", 30 August - 3 September 2011, Barcelona, Spain
Publisher: 
European Regional Science Association (ERSA), Louvain-la-Neuve
Abstract: 
Innovation virtuously impacts on the degree of international growth, which in turn positively influences innovation activities and then firms�™ performance (Filipescu et al., 2009). Many authors have tried to identify and explain the relationship between these two phenomena at firm level. Only recently, few empirical studies investigate them at a more aggregate level (see e.g. Mariotti et al., 2008). Moreover the literature focuses only on one direction of causality, while scant attention has been paid to inspect empirically innovation and internationalization together (Kafouros et al., 2008; Filippetti et al., 2009; Frenz and Ietto-Gillies, 2007). This paper provides an empirical analysis of the mutual relationship of these two phenomena, taking into account various features of the regions themselves. The empirical study is conducted on data concerning 20 Italian regions covering the period 2000-2008. To better understand the complex relationship between internationalization and innovation, we refer to the Structural Equation Models (SEM). These are multivariate regression type models, in which response variables could in turn act as dependent and predictor within a system of equations, and all variables are assumed to influence one-another reciprocally, either directly or through other variables as intermediaries (Bollen, 1989; McAdam et al., 2010). Through the SEM the relationships are expressed by a set of parameters which explain the magnitude of the effect (direct or indirect) between independent (either observed or latent) and dependent variables. Indeed, internationalization and innovation could act as both dependent and predictor which measurement could be difficult then suggesting the use of latent variables, and where the system of indicators is complex enough to lead at a model specified through two-way relations intrinsically connected. Using SEM approach we are able to specify flexible models dealing with non-standard relations stylized along panel data structure, in which spatial and temporal dimensions do matter
Document Type: 
Conference Paper

Files in This Item:
File
Size





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.