This paper evaluates different policies for reducing carbon dioxide emissions and promoting the innovation and diffusion of renewable energy. It assesses the relative performance of policies based on their incentives for emissions reduction, efficiency, and other outcomes. The study considers how the nature of technological progress—whether through learning by doing or R&D investment, and the degree of knowledge spillovers—affects the desirability of different policies. Due to knowledge spillovers, optimal policy involves a portfolio of different instruments targeting emissions, learning, and R&D. In a numerical application to the U.S. electricity sector, the relative cost of individual policies in achieving reductions is roughly as follows: (1) emissions price, (2) emissions performance standard, (3) fossil power tax, (4) renewables market share requirement, (5) renewables production subsidy, and (6) R&D subsidy for renewables. Nonetheless, an optimal portfolio of policies achieves emission reductions at significantly lower cost than any single policy. The paper develops a unified framework to assess six different policy options for reducing greenhouse gas emissions and promoting the development and diffusion of renewable energy in the electricity sector. The model includes two subsectors: one emitting (fossil fuels) and one nonemitting (renewables), both assumed to be perfectly competitive and supplying an identical product, electricity. The model has two stages, each representing a specific number of years. Electricity generation, consumption, and emissions occur in both stages, while investment in knowledge takes place in the first stage and, through technological change, lowers the cost of renewables generation in the second. The paper evaluates the relative performance of different policies according to different potential goals: emissions reduction, renewable energy production, R&D, and economic surplus. It also assesses how the nature of technological progress—the degree of knowledge spillovers and the degree of innovation occurring through learning by doing or R&D-based innovation—affects the desirability of different policy measures. Given the presence of two market failures—an emissions externality and knowledge spillovers—no single policy can correct both simultaneously; each poses different trade-offs. However, some clear principles emerge. When the ultimate goal is to reduce emissions, policies that also create incentives for fossil-fueled generators to reduce emissions intensity and for consumers to conserve energy perform better than those that rely on incentives for renewable energy producers alone. The paper concludes that the nature of knowledge accumulation is far less important than the nature of the policy incentives. However, these forces can interact in important ways, since the presence of knowledge spillovers means that separate policy instruments are necessary to optimally correct the climate externality and the externalities for both learning and R&D. The paper also explores the possibility that policy measures can create “win–win” opportunities that both lower emissions and raise total surplus. Consideration of optimal policy combinations to address the knowledge externalities and environmental externality simultaneously is left for a subsequent paper.This paper evaluates different policies for reducing carbon dioxide emissions and promoting the innovation and diffusion of renewable energy. It assesses the relative performance of policies based on their incentives for emissions reduction, efficiency, and other outcomes. The study considers how the nature of technological progress—whether through learning by doing or R&D investment, and the degree of knowledge spillovers—affects the desirability of different policies. Due to knowledge spillovers, optimal policy involves a portfolio of different instruments targeting emissions, learning, and R&D. In a numerical application to the U.S. electricity sector, the relative cost of individual policies in achieving reductions is roughly as follows: (1) emissions price, (2) emissions performance standard, (3) fossil power tax, (4) renewables market share requirement, (5) renewables production subsidy, and (6) R&D subsidy for renewables. Nonetheless, an optimal portfolio of policies achieves emission reductions at significantly lower cost than any single policy. The paper develops a unified framework to assess six different policy options for reducing greenhouse gas emissions and promoting the development and diffusion of renewable energy in the electricity sector. The model includes two subsectors: one emitting (fossil fuels) and one nonemitting (renewables), both assumed to be perfectly competitive and supplying an identical product, electricity. The model has two stages, each representing a specific number of years. Electricity generation, consumption, and emissions occur in both stages, while investment in knowledge takes place in the first stage and, through technological change, lowers the cost of renewables generation in the second. The paper evaluates the relative performance of different policies according to different potential goals: emissions reduction, renewable energy production, R&D, and economic surplus. It also assesses how the nature of technological progress—the degree of knowledge spillovers and the degree of innovation occurring through learning by doing or R&D-based innovation—affects the desirability of different policy measures. Given the presence of two market failures—an emissions externality and knowledge spillovers—no single policy can correct both simultaneously; each poses different trade-offs. However, some clear principles emerge. When the ultimate goal is to reduce emissions, policies that also create incentives for fossil-fueled generators to reduce emissions intensity and for consumers to conserve energy perform better than those that rely on incentives for renewable energy producers alone. The paper concludes that the nature of knowledge accumulation is far less important than the nature of the policy incentives. However, these forces can interact in important ways, since the presence of knowledge spillovers means that separate policy instruments are necessary to optimally correct the climate externality and the externalities for both learning and R&D. The paper also explores the possibility that policy measures can create “win–win” opportunities that both lower emissions and raise total surplus. Consideration of optimal policy combinations to address the knowledge externalities and environmental externality simultaneously is left for a subsequent paper.