Sources of Price Dispersion in U.S. Residential Solar Installations
Prices of solar PV have dropped dramatically, by half in just the past 6 years. But looking simply at prices paid today, there is considerable heterogeneity. For systems installed in 2014, the 10-90 percentile range for the observed $/W spans nearly a factor of 2. This apparent price dispersion raises policy-relevant questions, such as: why are consumers paying more than they need to? And would better-informed consumers increase the social benefits of solar PV? This paper analyzes price dispersion in U.S. residential PV installations between 2008 and 2014. Focusing on the most commonly used metric in previous studies of price dispersion, we use the quarterly coefficient of variation (CV) as our measure of price dispersion. We find higher levels of price dispersion in our data (0.22) than the average of 55 previous studies we reviewed (0.16). We also find that price dispersion has been persistent; it has remained above 0.15 since 2000 with no trend over that period. If anything, price dispersion has been increasing recently during the period for which we have complete data, 2008-14. Econometric analysis of the factors affecting price dispersion supports theories from the economic literature focusing on access to information and the costs and benefits of consumer search. Factors that increase the consumer payoffs of investing time in searching for information—system size and the value of solar—are associated with lower levels of price dispersion. Factors that reduce the costs of search—neighbors who have recently installed solar and having third-party quotes available—are also associated with less price dispersion. These results provide support for the importance of public efforts to enhance access to price information, e.g. by supporting price quote providers. The results also point to the particular need for information in nascent markets for PV in which access to the experience of neighbors is not available.