Impact evaluation of the social electricity tariff program in Brazil: Evidence from a regression discontinuity design

This paper evaluates Brazil's Social Electricity Tariff Program and its ability to improve electricity affordability for low-income households using a sharp Regression Discontinuity Design (RDD). Using data from the 2017–2018 Family Budget Survey (IBGE, 2018), we compare households just below and just above the income eligibility cutoff to estimate the causal effect of eligibility on electricity consumption and expenditures. We find that eligible households consume on average 20 kWh less electricity per month than households just above the cutoff, consistent with under-consumption near the threshold. Eligibility also reduces monthly electricity spending by R$15.35 on average. While statistically significant, this level of relief appears too small to meaningfully ease budget constraints, potentially encouraging households to restrict electricity use to remain within more subsidized consumption brackets. The results suggest that the current benefit schedule may be insufficient to ensure adequate access to modern energy services—such as thermal comfort and information and entertainment—while maintaining payment affordability. We discuss policy adjustments to better align subsidy levels and target heterogeneous household needs, and highlight the importance of accounting for regional and seasonal variation in electricity demand.