This dissertation uses a structural approach from sociology to explain technological changes as a cumulative process shaped by both technological networks and business networks. The thesis contributes to the field of technology studies by applying network analysis and teasing out mechanisms at different levels in the process of technological development. Analyses are done through two empirical studies and an agent-based model. The empirical objectives of the thesis include: establishing the importance of network effects by examining how an organization's connections to other innovative firms drive their development of new technologies, and establishing whether these structural effects persist at an aggregate level in differentiating regional performance in innovation. This thesis performs a systematic analysis of the effects of network structures on innovation by implementing innovative methodologies to study the process of interest and develops a novel measure. The goal is to put the development of technological change into a structural perspective and test the idea that interaction through complex networks of managers and inventors affects the speed and likelihood of innovation. The main argument of this thesis is that multiple networks among innovative actors facilitate the process of technological change. Using the United States patent data 1975--1999 as the underlying data on technological development and innovation, I show how networks of inventors and of corporate boards affect the rate of innovation at the business organization level and at the region level. I find that the interactions between these two kinds of networks among innovative actors foster the process of technological change. In particular, interactions among inventor networks and director networks produce a positive feedback dynamic that promotes technological change. This dissertation also uses agent based modeling to perform a simulation experiment devoted to pinpointing the relationship between structure and innovation. It defines an "innovation" as a search and learning process that results in a successful product. I introduce a two-phase model of collective innovation that accounts for the dynamism and deliberateness of the innovation process. I explore how interaction structures among a group of actors affect the speed at which the group can collectively innovate. In addition, I discuss the robustness of the effects of the network structure under different mechanisms of learning and with variations in actors' learning capabilities.