What is the optimal subsidy for residential solar?

How do we design clean energy subsidies to deliver greater benefits to society? Analytical answers to this question are scarce. Modeling should address both direct benefits from stimulating consumer adoption the year the subsidy is paid as well as indirect benefits from lowering future technology costs. We develop a benefit-cost analysis of residential solar subsidies in the US, disaggregated by state, accounting for technological progress, consumer adoption, and carbon and criteria emission reductions. We assess existing solar subsidies and also find the optimal subsidy schedule that maximizes net benefits starting in the year 2018. In the base case the national flexible optimal subsidy schedule begins at $585/kW and declines to zero in 14 years. The optimal subsidy starts higher and falls more quickly than the current federal tax credit, due to long-term benefits from early cost reductions and the need to reduce subsidies as a technology becomes cheaper. We also estimate state-by-state flexible subsidies which result in higher net benefits compared to a homogenous national subsidy ($2.8 billion versus $1.0 billion). Neglecting criteria pollution benefits, optimal subsidies accounting for technological progress and consumer behavior cost the government $45–49 per ton of carbon abated, much lower than the total abatement cost.