Abstract:

We study how climate policies shaped solar and wind deployment in 49 OECD+ countries from 1990 to 2023. Combining capacity data with policy stringency from the OECD Climate Actions and Policies Measurement Framework, we estimate event-study difference-in-differences models for diffusion speed (the annual growth rate of log installed capacity) around policy onsets and strengthenings, and embed these responses in an S-curve framework to map growth-rate changes into counterfactual capacity paths. Three findings stand out. First, policy design and timing matter more than simple presence: positive feed in tariff (price and duration) reforms and renewable expansion planning reliably accelerate deployment, while carbon pricing, emissions trading systems, and renewable portfolio standards do not show robust short-run effects; coal exit measures yield delayed gains, mainly for solar. Second, policy-induced increases in growth rates are transient but cumulate into level differences: relative to a no-policy diffusion baseline, the policy bundle roughly doubles solar capacity and increases wind capacity by about 30 percent, with feed-in tariffs and renewable expansion planning accounting for most of this boost. Significant cross-country heterogeneity exists in total policy-induced boosts, along with a moderate correlation between solar and wind outcomes. Third, effectiveness depends on the stage of diffusion and on technology: well-designed deployment support introduced at low penetration delivers much larger proportional gains than the same instruments implemented later, and solar is more policy-sensitive than wind. The results imply that policy portfolios aimed at rapid decarbonisation should prioritise early, credible deployment support tailored to technology and system constraints, rather than economy-wide pricing instruments.

Citation:

Tankwa, B., Ravigné, E. & Farmer, J.D. (2026), 'Climate policy reforms and the acceleration of solar and wind diffusion', INET Oxford Working Paper Series, No. 2026-01
Download Document (pdf, 14.426 MB)