中文摘要
量化农村电气化的福利效应常常受到非价格障碍的阻碍，这些障碍会在采用边际上抑制需求。我以肯尼亚和卢旺达800名已采用即用即付（PAYGo）太阳能的现有客户为对象，通过一项随机实验研究其采用后的需求。该实验随机分配激励，对达到月度购买门槛的消费者降低其使用的有效价格。尽管平均需求没有变化，但实验前需求最高的消费者在激励作用下购买量提高了6%–7%。我利用这些反应估计PAYGo太阳能消费者剩余的下界。我发现高需求消费者获得了较大收益，但低需求消费者的收益大幅下降。结合我的估计与文献中关于太阳能家庭系统环境外部性的证据，我计算出PAYGo太阳能补贴的公共资金边际价值在肯尼亚和卢旺达分别至多为1.7和2。
Abstract
Quantifying the welfare effects of rural electrification is often hindered by non-price barriers that depress demand at the adoption margin. I study demand for pay-as-you-go (PAYGo) solar using a randomized experiment with 800 existing PAYGo customers in Kenya and Rwanda post-adoption. The experiment randomly assigns incentives that lower the effective price of usage for consumers who meet monthly purchase thresholds. Although average demand is unchanged, consumers with the highest pre-experimental demand increase their purchases by 6–7% in response to the incentive. I use these responses to estimate a lower bound on consumer surplus from PAYGo solar. I find large gains for high-demand consumers, but benefits deteriorate substantially for low-demand consumers. Combining my estimates with evidence from the literature on the environmental externalities of solar home systems, I calculate that the marginal value of public funds for PAYGo solar subsidies is at most 1.7 in Kenya and 2 in Rwanda.