Foreign subsidiaries, innovation and spillovers in the manufacturing industry in Colombia

Resumen

This PhD thesis contributes to the analysis of technology internationalization, through foreign direct investment (FDI), and its effects on the competitive and innovation capacities of host countries, paying special attention to developing contexts. In particular, the objective has been to study the effects of FDI on the host technological capabilities through a better understanding of technological strategies of foreign subsidiaries in the Colombian manufacturing industry. The first issue addressed in this PhD dissertation was the differences in innovation performance of foreign subsidiaries compared to their domestic counterparts. The focus of the research was to assess the differences in innovation input and innovation output of subsidiaries compared to their domestic counterparts and the explanatory factors of these divergences. This quantitative analysis, using cross section data of Colombian manufacturing firms, yielded the following findings: • Foreign subsidiaries show a superior technological performance than local firms in this country. This allows affirming that the former can be considered as a potential source of knowledge able to generate positive spillover effects. • Foreign firms use comparatively greater internal and external knowledge inputs to innovate. At the internal level, activities that require intermediate or basic technological capacities predominate, while those knowledge flows within the multinational groups are important as the external side; also, the flow of knowledge taking place with domestic organizations (clients and suppliers, and, to a lesser degree, universities and research centers). • One result to highlight is that in contrast to prior evidence about the relationship between internationalization and innovation in more developed countries, subsidiaries of multinational firms in Colombia have a similar innovation performance to national firms connected to international markets, i.e. the technology gap of foreign and exporting domestic firms is not so large. The second analysis pursued was developed thorough three interrelated objectives. The first one was to assess the effect of internal, external and dual sources of knowledge on the innovation performance of foreign subsidiaries. I have argued that internal and external networks are crucial determinants of the development of distinctive competences in foreign subsidiaries. The second was to study the relationship among these knowledge sources to explain subsidiary´s technological capacities. I postulated that focus only on local or internal context is limited because it is likely that there are interdependencies or trade-off mechanisms between these types of networks, which in turn it may have important implications for the innovation performance of subsidiaries. The third objective was to analyze the role of absorptive capacities in networking capabilities. I proposed that accumulation of firms’ absorptive capacities defines a self-reinforcing process with networking activities because own innovation capabilities allow to subsidiaries to exploit external and corporate opportunities to create valuable knowledge. By using a panel data approach, a first finding was that foreign subsidiaries connected to external and intra-corporate networks have a better innovation performance, both in terms of innovation input as well as the generation of technological innovations. However, the possibility to find subsidiaries that carry out R&D and that are more research intensive, depends much more on their connection to the local innovation system, and specifically, connection with client, suppliers, competitors and R&D organizations with the purpose to innovate. Another finding was that there are significant complementarities between internal and corporate knowledge linkages, suggesting that subsidiaries sometimes combine corporate capabilities with external ones for innovation development. This is opposed the view that the external and corporate embeddedness are merely in conflict with one another or subject to a trade-off between them. Also, absorptive capacities become a key aspect to increase the likelihood of making use of both internal and external linkages. The third part of the thesis aimed to analyze the effect of the technological heterogeneity of foreign subsidiaries in the generation of intra-industry knowledge spillovers beneficial to domestic owned firms, using firm-level panel data for Colombian manufacturing firms. The analysis was developed in two basic stages. In the first one, an analysis cluster was performed with the aim to identify types of foreign subsidiaries according to their technological responsibilities, i.e., whether they can be classed as creating or exploiting units. In the second stage, I examined the relationship between the productivity of domestic firms and the foreign presence for two types of model specification: (i) the conventional model of spillovers effects in which FDI is simply treated as a homogeneous block and (ii) the subsidiary centered model in which was distinguished the differential effect of types of foreign investment. Many spillovers models assume that foreign subsidiaries are homogenous. The empirical results confirm our main proposition that competence-creating subsidiaries generate greater positive productivity effects on domestic manufacturing firms, in the same sector, than do units identified as competence exploiting, i.e. creative responsibilities of subsidiaries have a broad explanatory power to understand the possibility that FDI generates positive externality effects. In fact, subsidiaries oriented mostly to technologically exploitative activities do not generate knowledge spillover effects. In contrast, the estimation of the conventional model of spillover effects, where foreign investment is treated as a homogenous block in terms of technological capabilities, shows that the empirical analysis does not yield statistically significant results; revealing the limitations of considering subsidiaries as a homogeneous group with passive technological behavior, for research and public policy purposes. Finally, in the fourth part of the research was analyzed the conditional factors that could explain the generation (or absence) of productivity spillovers from foreign investment across regions (provinces) in Colombia, using firm level panel data for manufacturing industries. Also, regional disparities are included in the analysis of the Colombian manufacturing industry. The results permit to demonstrate why technologically active clusters generated over the basis of foreign subsidiaries are not likely in this country. The difference between absorptive capacities at both regional and firm level, as well as the role of industrial specialization, the level of embeddedness and the international connections are key elements to understand spillovers at industry and spatial levels.