Abstract: This paper examines the U.S. federal EV subsidy’s income cap, introduced under the 2022 Inflation Reduction Act, and its environmental and redistributive effects. First, using variation in exposure to the income-cap across ZIP-codes, I estimate that a 22-percentage-point drop in subsidy eligibility reduces EV registrations by 21 percent, implying a price elasticity of 2.9 for the marginal group. I develop an optimal-tax framework for income-capped subsidies and derive cap-conditional sufficient statistics for the jointly optimal per-unit subsidy and income cap. The optimal subsidy rate is larger when higher marginal external benefits coincide with greater price responsiveness, and when subsidy take-up is inversely related to income; the cap itself determines which households enter these moments. Calibrating a homothetic CES demand block, I trace an equity–environment frontier: as inequality aversion rises, the optimal cap tightens and the optimal subsidy falls; total environmental benefits decline, yet net fiscal efficiency can increase because inframarginal transfers are concentrated among high-income, high-adoption bins. The framework generalizes to other income-capped clean-energy programs and yields implementable rules for jointly choosing generosity and eligibility.
Abstract: This paper examines the intertemporal behavioral response of households to the phase-out of the U.S. federal electric vehicle (EV) tax credit. I exploit the 2019 credit reduction for Tesla and General Motors - triggered when each manufacturer reached the 200,000-EV sales threshold - to study how consumers adjust the timing of EV purchases in anticipation of scheduled incentive changes. Using high-frequency vehicle registration data from New York State between 2014 and 2021 and a two-way fixed effects event-study design, I document large and sharp intertemporal substitution for Tesla. Monthly Tesla sales rose by roughly 3,000 units per model in the two months preceding the credit cut relative to the phase-out month, over twice the model-level average monthly sales prior to the phase-out, indicating substantial purchase acceleration rather than net demand expansion. In contrast, I find limited evidence of retiming for General Motors. The results highlight that not accounting for intertemporal substitution leads to overestimation of the welfare and adoption effects of EV subsidies. I discuss price adjustments and salience as potential mechanisms and outline how these findings inform the design of dynamic clean-energy incentives and the evaluation of the 2022 Inflation Reduction Act’s reforms.