The German solar programme was a very prominent example of a deployment subsidy for a nascent green technology. To contain its substantial fiscal costs, the subsidy was nonlinear in the capacity of adopted solar panels. This paper finds that an optimal nonlinear schedule could further decrease costs; however, overall gains from using nonlinearities remain modest. A novel theoretical insight explains the quantitative finding: nonlinearities are cost-effective only if participation responses are heterogeneous, yet early-stage technologies typically exhibit low heterogeneity in this dimension. The analysis highlights that nonlinearities are not a panacea for reducing the fiscal burden of deployment subsidies.