This study evaluates the distribution of electricity subsidies to residential customers in Addis Ababa, Ethiopia in 2016 that results from the current increasing block tariff (IBT) structure. Customer billing data supplied by the electricity utility was matched with socioeconomic information collected from a survey of 987 households, and used with a utility-specific estimate of the costs of electricity service to estimate household-specific subsidies. The analysis differentiates between primary customers, who own an electricity account, and nonprimary customers, who pay a primary account holder, and presents the first detailed analysis of the incidence of subsidies provided to households with private versus shared electricity connections. Results show that households in the poorest quintile received 7% of the total subsidy, while households in the wealthiest quintile received 37%. The majority of households with shared connections were in poorer quintiles, and 80% in the poorest quintile were nonprimary customers. The regressive outcomes of the IBT structure are explained by two of its attributes: (1) the volumetric rates in all tariff blocks are substantially below the total average cost of delivery of the service; and (2) there is only a moderate association between electricity consumption and household wealth.