This dissertation develops a theoretical framework for the analysis of oligopolistic rivalry through cost-reducing research and development (R&D) expenditures in an environment where firms cannot fully appropriate the results of their R&D investment. This framework is developed in order to examine the incentives of firms to invest in R&D in oligopolistic markets, the implications for the structure and performance of industries subject to such spillovers and the nature of public policy in this context. Two lines of research are followed: one focuses on technological factors affecting incentives and performance, the other on the influence of strategic behaviour. With respect to technological factors, it is argued that the impact of research spillovers on market incentives and performance, and on the desirability of a public policy of R&D subsidies, depends crucially on the specific assumptions made about how a firm's production process is affected by its own R&D and that of its rivals. These assumptions are embodied in the knowledge production function and the associated cost function facing each firm. Issues of substitutability and complementarity between a firm's research and that of its rivals and between various components of a firm's own research are central in determining the impact of spillovers. The level and composition of R&D investment, production and profitability, concentration and monopoly power are all influenced by the impact of rivals' research on the marginal productivity of a firm's own research. Optimal subsidies are similarly influenced. This suggests that the effectiveness of such policies towards R&D investment depends on the nature of technology and on specific appropri-ability characteristics of industries. With respect to strategic behaviour, the dissertation extends previous research in the context of a multi-stage model of international competition in R&D to the case where R&D is subject to spillovers. This allows an examination of the importance of different behavioural assumptions (one-stage vs. multi-stage decision-making) on the impact of spillovers. The analysis also questions the results of strategic models obtained with full appropriability of R&D, whereby strategic behaviour results in higher equilibrium levels of R&D and production and lower profits. Strategic behaviour can result in lower R&D and higher profitability if research is difficult to appropriate. The conditions under which this occurs are explored fully. The characteristics of industrial policy in the context of strategic international competition are then explored. Models that assume fully appropriable R&D suggest the optimality of a policy of positive subsidies to the R&D expenditures of domestic firms. The existence of spillovers may reverse this result. The analysis thus casts doubt on the efficacy of a government policy of R&D subsidies in a strategy of "precommitment". This goes beyond the usual retaliatory arguments against such behaviour and points to the limits of such an interventionist approach in an international environment where R&D has some characteristics typical of public goods.

Using panel data analysis, this paper considers the impact of governance provisions on firm’sR&D investment decisions. The current paper also contributes to the literature on corporategovernance and innovation by introducing an interaction dimension which captures theinfluence that internationalization of U.S. firms may have on R&D investment decisions. Basedon a sample of 627 U.S. firms for the years 2008-2016 this paper’s results suggest thatgovernance provisions do not nurture or impede R&D investments, and therefore the resultscasts doubt on the existence of a relationship between governance provisions and R&Dintensity. The current paper’s findings also suggest that proposed effects of internationalizationdo not modify the relationship between entrenchment and R&D for U.S. firms.