Association for Public Policy Analysis & Management

As global efforts mount to avoid the most catastrophic consequences of climate change, jurisdictions often turn to subsidy policies as a means to increase the diffusion of green technologies thereby reducing CO2e emissions. However, the nature of any increase in diffusion is unclear: it could be the result of either accelerated adoption decisions or of market expansion, i.e. additional adoptions. Additional adoptions would not have adopted in the counterfactual, while accelerated adoptions would adopt later in the absence of policy. While either explanation serves to improve the distal policy outcome of avoided emissions, understanding this additionality aspect in terms of the proximal policy outcome – accelerated diffusion – has important long-term consequences. Uncovering the nature of increased diffusion spurred by policy, acceleration or additionality, requires a complex understanding of the counterfactual.

In this paper, we use a generative model to explain the impact of the California Solar Initiative rebate program, the largest state-level solar PV rebate program in the U.S., to increase diffusion of residential solar photovoltaics (PV). Operationalized in an agent-based model, this generative approach separately captures financial and informational aspects of the individual adoption decisions that comprise an emergent diffusion phenomenon. At an individual level, we build a counterfactual scenario in which a subsidy policy was not implemented and thus distinguish acceleration impacts from additionality impacts. By establishing a counterfactual baseline for the proximal adoption outcome, we also create a new counterfactual for the distal avoided emissions outcome that impacts calculations of program benefits.