Distributional effects of residential energy tax credits: A machine learning approach

The rollout of the Inflation Reduction Act in 2022 introduced new tax credits to accelerate adoption of renewable technologies. However, studies into how existing credits are distributed across income strata are sparse. We used a machine-learning based approach to analyze how residential energy tax credits are distributed across income brackets. We hypothesized that the level of tax incentives claimed would increase as the level of income increased. We created a logistic regression model to compare the log-odds of uptake and a linear regression model to compare the relative amounts of tax credits claimed across income brackets. A comparison of these models showed how uptake of credits differs across income brackets, with a non-linear relationship between income and usage of energy tax credits. While we observed a general trend of increasing tax credit utilization with higher income, the increase was not uniform throughout. Our results suggested that existing incentives may not effectively reach middle-income households, potentially due to eligibility limitations for lower-income programs and insufficient financial capacity to afford sustainable technologies. Further research is crucial to understand the specific barriers faced by this income group, which could include limited access to information about programs, the high upfront cost of technology adoption, or a lack of targeted support. Addressing the challenges faced by middle-income households, who may not qualify for low-income assistance but still face financial barriers to adopting sustainable technologies, is vital to maximizing the impact of such incentives and ensuring broader access to the benefits of a sustainable future for all income levels.