Subsidies and costs in the California solar market: an empirical analysis

Abstract

Environmental concerns have prompted governments around the world to subsidize renewable energy markets. One of the major risks associated with subsidizing is that it may inflate costs. Thus, understanding the drivers of costs, and specifically how subsidies affect costs is crucial for evaluating and designing good subsidy policies. In this paper, we identify and estimate the cost drivers of solar photovoltaic systems in the California market using a semi-parametric regression model, and further quantify the cost-inflationary effect by simulation using machine learning techniques. We find evidence for significant cost inflationary effects of subsidies. The regression results suggest that a 1% increase in incentives per kW installed is associated with nearly 0.1% increase in costs per kW installed. Furthermore, simulations indicate that cut-off of subsidies in 2012 would have saved the California government US$1.15bn, while the extra costs imposed on end-customers would be only US$0.30bn. Our results suggest that a cut-off in 2012 would not have lead to a substantial jump in costs to end-customers at the cut-off point, and that costs would only be slightly higher for end-customers than with subsidies. The results indicate that an accelerated subsidy down-scaling may be desirable, with minimal adverse implications for end-customers.