This thesis studies innovation networks, R&D spillovers, spatial spillovers, productivity growth and associated adjustments in Taiwan high-technology industries. A dynamic production modeling is built up to estimate R&D and its spillover effects among and within high-technology industries to study the dynamic effects of inter-, intra-industry R&D and spatial spillovers and exogenous technical changes on output growth for Taiwan high-technology industries. The thesis involves analysis of R&D spillover effects among and within high-technology industries to study the dynamic perspectives of the innovation diffusion. As leading high-technology and network expand the horizons of economic agents¡¦ production possibilities and decisions, the spillovers between productive entities are likely to be of increasing importance in the knowledge-based economy. Understanding these productive inter-dependencies, and their potential to motivate various types of spillovers require modeling and measuring their existence and impacts. We provide a conceptual and empirical framework for measuring and evaluating various types of spillover mechanism, which allows us both to quantify the cost-effects and evaluating the contribution to productive performance. We explore the temporal, spatial, and industrial spillovers using a dynamic cost function model that explicitly parameterizes the spillover weights and econometrically estimates them. We extend the dynamic external spillover model framework, described in Tsai and Chen (2001), which assumed that each firm derives an optimal plan so that the expected present value of current and future costs stream is minimized. First, we study the spillover effects in the dimension of industry considerations and geography. Second, to identify all kinds of spillover sources, to assess the spillover processes, and to evaluate the contributions of such inter-dependencies in productive performance we incorporate the concept of spillover ratio into the industry dynamic model. The temporal, intra-industry, inter-industry, and spatial R&D spillovers are distinguished and, in doing so, provide a richer account of innovation, learning and the means which are encouraged by geographical proximity.