This dissertation reflects on strategies for utilizing networks to catalyze a low-carbon energy transition. In particular, I focus on the role of networks for two policy strategies: (i) switching to alternative, low-carbon energy technologies, and (ii) reducing consumption of carbon-intensive, fossil fuel energy resources. Chapter 2 of this dissertation uses historical energy transitions to argue networks complicate prescriptive policy design. In particular, we illustrate the nature of the underlying technologies used to covert physical fuel inputs into energy services structures interactions between markets. The interconnections between markets, in turn, creates complex feedback loops that lead to simultaneous changes across the entire economy. We explore how these feedback loops impact energy efficiency policy in more detail in chapter 4. Chapter 3 investigates the role of networks in proliferating the diffusion of low-carbon energy technologies. We provide evidence that network formation is critical for the early success of emerging low-carbon innovations. When networks form at this early stage of the technology life cycle, they provide a platform for information exchange between early and future adopters, leading to lower search, transaction, and operational costs for future adopters. Lastly, in chapter 4, we develop a theoretical model that embeds industrial, energy efficiency improvements within a network setting to understand how interconnections between markets affects the outcomes of sector-specific low-carbon energy policy. Energy efficiency is often touted as a cure-all policy measure to reduce dependence on carbon intensive, fossil fuel resources. However, when markets are connected by the economy's production network, the outcome of energy efficiency policy is highly uncertain. Using a combination of theoretical and numerical analyses, we illustrate the structure of the economy's production network shapes the change in aggregate energy consumption following a sector-specific energy efficiency improvement.