Conference Agenda

Deployment subsidies for nascent technologies have moved to the forefront of climate action, industrial policy, and geoeconomics. To limit the substantial fiscal costs of such programs, policymakers routinely employ subsidy schedules that are nonlinear in the quantity agents deploy. However, nonlinearities can threaten a program's deployment goals because agents may respond to them at the intensive margin by reducing deployed quantity and at the extensive margin by ceasing participation. This paper characterizes optimal nonlinear subsidy schemes in which a principal trades off the benefits from deployment with its budgetary costs. Theoretically, it shows that nonlinearities are optimal only if adopters' participation responses are heterogeneous along the subsidy schedule. Yet, early-stage technologies typically exhibit limited heterogeneity in this dimension. As a consequence, the optimal subsidy rate is close to constant over a wide range of the schedule. In contrast to the celebrated Sadka–Seade “no-distortion-at-the-top” result, “no-distortion-at-the-bottom” emerges as an actionable policy prescription in this context. The insights generalize to screening problems broadly, and in particular to multidimensional screening à la Rochet and Choné (1998) and optimal nonlinear monopoly pricing. A quantitative analysis of the canonical German rooftop-solar program demonstrates the theoretical findings.

Rooftop solar photovoltaic (PV) adoption has expanded rapidly in Australia, resulting in one of the highest household adoption rates in the world. Yet the partisan politics and policy drivers of this growth, and its distributional welfare consequences, remain empirically underexplored. Using a balanced panel dataset of 137 Australian Local Government Areas (LGAs) over 2006-2023, this study examines how partisan governance and major policy incentives, specifically the Small-Scale Renewable Energy Scheme (SRES) and Feed-in Tariffs (FiTs), affect residential solar uptake, and how adoption in turn affects household energy expenditure. Employing a Panel-Corrected Standard Errors (PCSE) estimator, the paper finds that both instruments significantly stimulate adoption, with FiTs exhibiting a substantially stronger effect than upfront subsidies. Partisan alignment between federal and state Labor governments amplifies uptake above and beyond the individual-level effects of either government alone. While solar adoption reduces average household energy expenditure by up to 44% by 2023 relative to 2011 levels, income-stratified analysis reveals that higher-income LGAs capture significantly larger savings, raising important equity implications for the design of universal renewable energy incentive schemes.

Pay-as-you-go (PAYGo) solar electricity is a key technology for expanding basic energy access to remote, low-income populations, but it remains expensive relative to traditional energy sources. I study demand for PAYGo solar using a randomized experiment with 800 existing PAYGo customers in Kenya and Rwanda. The experiment randomly assigns incentives that lower the effective price of usage for consumers who meet monthly purchase thresholds. Although average demand is unresponsive, consumers with the highest pre-experimental demand increase their purchases by 6%–7% in response to the incentive. I estimate a lower bound on consumer surplus using these revealed-preference responses to bound welfare. Although I find large welfare gains for high-demand consumers, my bounds imply that expanding adoption of PAYGo solar by another 25% would require subsidies of 70%-80%.

This paper examines the design of corrective subsidies for positive externalities in a setting often observed in actual programs, where subsidy eligibility and amounts can be differentiated across market segments defined by recipients’ observable and hard-to-manipulate characteristics, such as income and location. It develops a synthesized framework to optimally target or differentiate subsidy rates among these segments by decomposing each segment’s optimal subsidy into three components aligned with the program’s objectives—Pigouvian (externality correction), Ramsey (cost-effectiveness), and redistribution. The relative significance of these three components depends on the shadow value of the subsidy budget, reflecting the trade-offs among the program’s objectives. I investigate how optimal subsidy targeting is influenced by factors such as the program budget, the magnitude of positive externalities, production costs, heterogeneity across segments in responsiveness to subsidies and in social welfare weights. Overall, this paper provides a theoretically grounded and empirically implementable framework for designing differentiated corrective subsidies, thereby complementing the existing positive literature on the efficiency and distributional impacts of corrective subsidies. By integrating Pigouvian, Ramsey, and distributional perspectives within a single social welfare maximization problem, the paper bridges the theory of optimal taxes and subsidies, applied welfare analysis, and real-world policy design.