Firma e território: três ensaios sobre inovação em ambientes periféricos

• Ano de defesa: 2007
• Autor(a) principal: Eduardo Goncalves
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Economia regional; Inovações tecnologicas; Indústrias; Argentina Indústrias Inovações tecnologicas; Brasil Indústrias Inovações tecnologicas
• Link de acesso: http://hdl.handle.net/1843/AMSA-725P3F

This thesis encompasses three articles which tackle three different dimensions of the relationship between the enterprise and the territory. The first one, A cross-section analysis of technological innovation for Brazil and Argentina, takes the national dimension of territory regarded as peripheral. Considered two out of the three most important Latin-American economies, they underwent a process of late industrialization that brought structural weakness and bottlenecks which comprise the background for the contemporary innovation behavior of established firms, either national or transnational. Despite the similarities of import substitution policies of both countries, their particular historical industrialization experiences explain the differences found in the results. They come from probit regressions with instrumental variables to control for endogeneity of the exports variable. In general, these results suggest an upgrade of the Argentinean and Brazilian firms in their patterns of technological innovation vis-à-vis the import-substitution period. The first improvement is related to pattern of knowledge acquisition, upgrading from the traditional purchase of capital goods, most of them imported, to the acquisition of disembodied knowledge by means of R&D purchases and more definitive forms of technological knowledge, such as patents, licensing and know-how. The second one is the role played by exports as an induction device to innovation, which in the specific case of Brazilian product innovating firms overcomes the traditional role played by imports. This is strong evidence that exports in these two countries are upgrading from their functional role as a device of import capacity creation to innovation inducement device. The second article, Technological innovation in large Brazilian firms, explores the innovating behavior of the large firm in Brazil, departing from findings in the first article related to role of a firms size to its innovation ability. The article brings two methodological improvements compared with the previous one. Firstly is the firms follow up in the 1998-2003 periods using panel data which take into account the non-observable factors influencing the innovation decision making of firms. Secondly, this is accomplished by conditional (fixed-effects) logistic regression. The results confirm the Schumpeterian hypothesis regarding the leading role of large firms to perform technological innovation. In addition, the regressions show that capital-goods acquisitions are far less important than disembodied forms knowledge to increase a firms propensity to innovate. In contrast with those of process, product innovations depends more on tacit forms of knowledge since fixed factors intrinsic to firms and time invariant, correlated to the error terms, are relevant to innovation. Finally, the third article, Organizational and territorial determinants on the innovation capability in Brazil, the region inside the Brazilian border is the territorial scale of analysis of the pivotal behavior of firms. From the methodological viewpoint, the article uses hierarchical regressions, providing an analytical device to estimate the impacts of firms features on innovation controlled by the characteristics of the territorial environment where it is located. The results show that the territory environment does influence the innovating behavior of the firm. In the Brazilian context, the propensity of a firm to innovate increases with: the geographical proximity to São Paulos micro region, the degree of local industrialization and the schooling level of local population. Local technological knowledge externalities are only relevant when they are measured by patents per capita.