Association for Public Policy Analysis & Management

Large-scale deployment of distributed solar photovoltaic (PV) technologies is widely considered to be an important piece in addressing the environmental impacts of the electricity sector, while increasing resilience of the electricity system and energy access. While solar costs have come down dramatically over the last decade, costs need to come down even more to realize the full promise of solar. There is general consensus that the next big tranche of cost reductions in solar will come not from hardware (cells, module, etc.), but rather from “soft costs”, including permitting, financing, overhead, marketing, and other non-hardware costs associated with installing solar systems. Previous literature on solar costs has focused on how market structure, installer scale, demographics, ownership, policy, and system components influence the solar PV installation price (Nemet et al., 2016; Gillingham et al., 2014). However, the unexplained variation is still large (>70%) after controlling for these factors, which suggests the need for further work to identify the potential for cost reductions of solar PV installations. To address this research opportunity, this paper focuses on the role of technological innovation, knowledge spillover, and upstream-downstream collaboration to reduce costs of solar PV installations.

Specifically, this paper aims to address the following three related questions: 1) what are the characteristics of inverter manufacturers-installers network; 2) whether and how installers’ technology innovation and inverter manufacturers-installers network – especially knowledge spillover from upstream inverter manufacturers to downstream installers – can contribute to cost reductions of solar PV installations; 3) whether and how the innovation effect and network effect depend on local market and policy conditions. To address these questions, we develop a new database of solar PV balance-of-system (BOS) patents in the distributed PV market from the USPTO database between 2000 and 2014 by using Boolean keywords searching. We then matched the patent database with the dataset of small-scale PV installations in the U.S., allowing us to build regression models with installation price as the dependent variable and innovation and spillover variables, plus other controls, as explanatory variables.

This paper sheds lights on the drivers of cost reduction of solar PV installations, with a focus on the role of technology innovation and upstream-downstream network connections. Our preliminary network analysis indicates that large installers have more arm’s-length ties with inverter manufacturers compared to small installers, but they have more embedded ties with innovative inverter manufacturers. Our preliminary regression results show that only the total capacity of collaboration between installers and innovative inverter manufacturers significantly reduces the solar PV installation prices. By including both technological innovation and installers’ experience into the model, this paper contributes to separating the effect of productivity improvement (i.e., by repeated working on technologically similar installations) and the effect of technology innovation (i.e., changes of working process or installation methods). Moreover, studying the geographic embeddedness of the translation from technological innovation and knowledge spillover to cost reductions helps identify the importance of local conditions, such as local market structure and local knowledge base, to facilitate the application of innovative knowledge.