Association for Public Policy Analysis & Management

Soft costs, such as customer acquisition, labor, and permitting, account for about 60% of installed solar photovoltaic (PV) prices in the U.S. Thus reducing soft costs is a key target in bringing PV costs down, thereby enabling a broader deployment of PV. Firm-level knowledge management practices, including maximizing learning-by-doing and capturing knowledge external to the firm, is a key element in understanding the variation in soft costs across firms. Accordingly, in this paper we cast a direct lens on learning mechanisms associated with managing solar soft costs. First, we examine how the geography of innovation – specifically the differing roles of local and non-local knowledge – affects soft costs. Second, we develop a local knowledge network to examine the network structure for insights into how soft costs knowledge flows through and, in turn, shapes those networks. Using this two-pronged approach, we develop an understanding of local and non-local soft costs knowledge flows and impact on installed PV prices.

Firms utilize multiple sources and processes in their technological learning, using both local and non-local sources, though researchers continue to argue about the importance of geography in technological learning. We use patent citation analysis to address two questions about firms’ technological learning process: 1) How do firm-level and local market characteristics influence firms’ choices on different learning sources (localized versus non-localized knowledge); and 2) What are the effects of different learning sources on the value of technological innovation? For our analysis, we develop a new database of PV balance-of-system patents in the U.S. between 2000 and 2014, and then track the location and impact of the backward and forward citations for each patent.

Next, we develop a local knowledge network to better understand the factors that influence local learning. The knowledge network in which a firm is embedded heavily influences firm-level knowledge, thus in order to develop a comprehensive understanding of how local knowledge flows impact solar soft costs, we explore the following question: who learns what (knowledge acquisition), from whom (knowledge production), and how (spillover mechanisms)? To address this question, we focus on the solar energy ecosystem in Austin, TX, mapping the network of actors and knowledge flows between those actors. Analysis of the emerging network offers insight into how 1) knowledge and networks evolve together, 2) information searching drives network development versus how existing network ties shape information availability, and 3) different types of knowledge flow through the network and the impact this has on cost reductions.

Together, these two analyses provide a better understanding of the impact of geography (local and non-local) on knowledge flows and the subsequent impact on technological learning and soft costs reductions. Overall, this paper sheds new light on policies and practices that are most likely to stimulate knowledge networks and spur local innovation ecosystems with the greatest potential to reduce soft costs, thereby enabling wider deployment of solar.