Subsidies for close substitutes: Aggregate demand for residential solar electricity

Subsidies promoting residential solar systems are intended to reduce carbon emissions by lowering demand for electricity from the grid. The ability of these subsidies to reduce grid demand hinges on how close, on aggregate, the two sources of electricity are to perfect substitutes. To test the efficacy of these policies, we form a tractable model of national residential electricity demand that identifies the aggregate substitutability between residential systems and electricity drawn from the grid. When estimated on the United States, we find that while the two are close to perfect substitutes, the degree to which substitutability is imperfect has material implications for policy. Subsidies inducing one $\mathrm{kWh}$ of residential solar electricity demand displace only 0.5 $\mathrm{kWh}$ of grid consumption. As an emissions reduction policy, subsidies had national abatement costs of $332 per MTCO${}_2$ in 2018.