Solar panel adoption among Mexican small and medium-sized commercial and service businesses

Abstract

We analyze the determinants of adoption of distributed solar photovoltaic systems, focusing on small and medium-sized commercial and service firms. We use monthly billing data that are perfectly matched with data from a novel survey that gathers information on electricity consumption, stock of electric equipment, and a rich set of firm characteristics in the Metropolitan Area of Aguascalientes, Mexico. Using an econometric model, we find evidence that a set of explanatory variables such as business characteristics, the economic sector, ownership status, stock and usage of equipment and appliances, presence of other solar technologies, and views about the use of renewable energy are important determinants of the probability of adoption of solar panel systems. Furthermore, using machine learning methods to identify the best predictors of solar adoption, we indirectly validate the theory-driven empirical model by assessing a large set of explanatory variables and selecting a subset of these variables. In addition, we investigate relevant cases where a priori solar panel adoption seems to be cost-effective but structural adoption barriers and adoption gaps might coexist for certain groups of electricity users. We also calculate the social cost savings and the avoided CO2 emissions. Finally, based on our results, we provide several policy implications and recommendations.