Template-Type: ReDIF-Article 1.0 Author-Name: Javier López Otero Author-Name: Gustavo Antonio Contreras Cabrera Author-Name: Rosa Jordá Borrell Title: Aplicación de un modelo de regresión a análisis de los conductos de transferencia tecnológica del Cluster aeroespacial andaluz Abstract: Resumen:Esta investigación tiene por objeto aplicar un modelo de regresión para analizar el impacto de las externalidades de conocimiento sobre la innovación de una muestra de firmas aeroespaciales andaluzas en tres ámbitos geográficos: local, nacional, e internacional. El estudio distingue entre externalidades obtenidas a través de conductos voluntarios y no-voluntarios. Para ello se realizó un acopio de información sobre las firmas del clúster aeronáutico andaluz mediante una encuesta propia. Los resultados evidenciaron que las externalidades internacionales no tenían significación y que las nacionales destacaban sobre las locales por albergar empresas más intensivas en conocimiento.Abstract:This article aims to: 1. apply a regression model to analyze the impact of externalities of knowledge about the innovative activity of aerospace companies located in Andalusia. 2. Addressing this impact from the perspective of three geographical areas (local, national, international). And, 3. Study, on the one hand, externalities captured by voluntary technology transfer and, on the other, those that deal with involuntary channels, determining their spatial boundaries. So, this work aims to analyze the characteristics of knowledge externalities generated in the aviation industry in a region such as Andalusia, as the pieces of research previously conducted on this subject have focused mainly on advanced countries or regions. To that end, 134 companies in the Andalusian aeronautical cluster have been surveyed and our own questionnaire has been made. This questionnaire includes about 160 questions divided into ten sections that have yielded more than 180 variables. Moreover, following the guidelines of analysis Griliches, Jaffe, Feldman, Cohen, Kepler, etc., studied in the theory section, a model to analyze the externalities of knowledge generated in the Andalusian aeronautical industry has b een developed . The proposed model would look like: 〖Log (R〗_i)= α+〖(β〗_0*Log(R+D_(Loc ) )+β_1*Log(R+D_nac)+β_2*Log(R+D_int)*R+D_company+〖(β〗_( 3)*〖HR〗_Loc+β_4*〖HR〗_nac+β_5*〖HR〗_int)*〖HR〗_company+e_i Ri is an indicator of technological results, R & D values the investment in research and development carried out by public and private external agents. This variable has been broken down into three geographical areas: local, national or international. HR (Human Resources) lies in the proportion devoted to R & D of total workers in each public and private employees outside agency, and ei represents the model residuals. In this research the variable technological results was used as the dependent variable and was obtained from the field work done. This is the summation of all the technological achievements that have produced each of the companies in the sample. On the other hand, each of the connections (between a company in the sample and public and private external agents) can be categorized into five channels as detailed in Figure 1, and are grouped into two groups: voluntary, and non-voluntary channels. In the first group, the impact of R & D from external agents on firms through the sale of technology, and through technological cooperation agreements, is analyzed. In the second case, the impact of R & D from external agents on the firms in the sample, through channels such as hiring employees, informal relationship between employees and observing other agents, has been studied. RESULTS The multiple linear regression model applied to the results obtained by voluntary channels, has produced a 0.82 r2 parameter, an FIV index between 1-3, heteroskedasticity tests have been negative and the model coefficients are shown in Table 3. These results verified that only three of the six variables introduced are significant, particularly the variable R & D at local and national levels, and variable staff qualifications at national level. In the case of variable R & D dominates national R & D on the local one. Thus, the coefficients obtained from the R & D variables allow to reject the research hypothesis 2. The R & D performed by external agents nationwide impacts more on the technological results of the companies in the sample than do the external agents established at local or international level. The contribution of R & D of international companies has been discarded as not being significant. Thus, the results have not seen a relationship between the generation of technological results and international companies even if it is in these areas where most technological knowledge is generated. The variable qualifications of human resources within the external agents is only significant at national level, while local and international firms are not representative, so that it can be concluded that the research hypothesis 3 has not been fulfilled. The model applied to the results obtained through non-voluntary channels yielded a 0.66 r2 parameter, a next to 1 VIF index, heterocedasticity tests were negative and the model coefficients are shown in Table 4. The model results show that only two variables have significance, these are: The R & D of national companies and R & D of international companies. The coefficients obtained show that there is a broader geographical scope for externalities than the one achieved by the voluntary channels. Thus, the R & D conducted by distant external agents (national and international) has a greater effect on the technological results of the Andalusian companies than the R & D performed by local external agents does. In the two areas that have been significant the international one has received the highest coefficient. Therefore, in view of these results hypotheses 2 and 3 in non-voluntary channels should be rejected. The explanation for this is: 1. the large proportion of reverse engineering as the most quoted channel among the non-voluntary ones; and 2. it has always been made about products and services of national and international external agents. CONCLUSIONS Analyses of the externalities of knowledge allow us to conclude the following: - Any investment in R & D carried out in Spain, especially in the Madrid cluster, impacts in a positive way in the Andalusian cluster. - If the Andalusian aerospace cluster directed its production to more knowledge-intensive segments, the investment in local R & D would impact in a more positive way in the Andalusian business cluster. - Bearing in mind that the limit of tacit knowledge is the national, Andalusian companies, to benefit from the externalities of knowledge generated in countries that invest heavily in R & D, would have to establish industrial plants in these countries. This possibility would optimize the acquisition of tacit knowledge through non-voluntary channels. - If there were more companies producing knowledge in Andalusia, as FIC, or Tier1, both foreign and domestic, the externalities of knowledge might be less controlled and there would be a greater proportion of channels of collaboration and non-voluntary. Comparing these results with other studies like the one conducted by Niosi and Zhegu (2005) on the Montreal aerospace cluster, it proves the Montreal firms have a greater sensitivity to international knowledge externalities. This can be explained by the different position of the Andalusian cluster to the Montreal cluster in the overall aircraft production chain. In the first case, small firms predominate, little internationalized, with low-level skills and depend on the demand for one or two FIC. Instead, the cluster Canadian firms located several large FIC and Tier1, with multiple competencies, and working in collaboration with other firms TIER1 and FIC worldwide. Thus, the Andalusian little international presence limits its sensitivity to foreign R & D, while the Montreal cluster, being globalized, is more sensitive to international R & D. Moreover, the Andalusian cluster model results have similarities with those obtained by Cooke and Ehret (2009) in their analysis of the aerospace cluster of Wales. According to both authors, the existence of technological knowledge in the companies and employees of Wales, was key to the allocation of work packages to FIC firms in that cluster. Skills and competencies in manufacturing carbon fiber of the Spanish aerospace firms have favored the acquisition of important contracts with Airbus Spain and Airbus Military, which has resulted in a transfer of technological knowledge from the latter to the companies of the Andalusian cluster. In sum, most of the previous studies have analyzed knowledge externalities indirectly Autant-Bernard and Massard (2000); Pakes and Griliches (1984), or through some of the analyzed channels Arundel and Geuna (2001) and Ehret Cooke (2009). However, in this research externalities of knowledge have been studied through a wide range of possibilities, and also have been grouped in voluntary and involuntary channels. This leads to the conclusion that the most decisive voluntary knowledge externalities in the Andalusian aerospace industry are the trading of technology. And it also allows to carry out an assessment of the knowledge lost / gained by the Andalusian cluster companies, and the ones related to it, through reverse engineering, informal communication of employees, or hiring a worker. Also, previous work on externalities such as the ones by Autant-Bernard and Massard (2000); Trajtenberg and Jaffe (2002), have studied them through traditional sources such as patents. However, the uniqueness of the aerospace industry, to the effect that there is a close link with the State defense, lead to many technological results not being published in any Intellectual Property Office, and technology transfer being sought by other means, fundamentally business relationships. Classification-JEL: R1 Keywords: Industria Aeroespacial, Regresión multinomial, Conocimiento Tácito, Externalidades de Conocimiento, Transferencia de conocimiento, Aerospace Industry, Knowledge transfer, Multinomial regression, Tacit Knowledge, Knowledge Externalities Pages: 189-220 Volume: 2 Year: 2015 File-URL: http://www.revistaestudiosregionales.com/documentos/articulos/pdf-articulo-2474.pdf File-Format: Application/pdf Handle: RePEc:rer:articu:v:2:y:2015:p:189-220