Do low-income electricity subsidies change peak consumption behavior?

Increases in electricity consumption during peak hours place additional strain on the electric power system, which can be partially mitigated if foreseen years in advance. Smart meter data, with hourly resolution or better,... [ view full abstract ]

Increases in electricity consumption during peak hours place additional strain on the electric power system, which can be partially mitigated if foreseen years in advance. Smart meter data, with hourly resolution or better, allow improved characterization of the effects of various programs and other interventions on residential and system load shape. We characterize the load shape impacts of the California Alternate Rates for Energy (CARE) program, which provides 3.2 million households an average electric bill subsidy of 33%. We use hourly electricity consumption data from roughly 30,000 randomly selected households from Pacific Gas and Electric service territory to estimate the hourly effect of enrollment in the CARE program on household electricity consumption using a fixed-effects regression model. We find that the CARE program is associated with an average increase in electricity consumption of 13% [11%, 16%]. The increase is relatively constant throughout the day, with no two hours statistically distinguishable from each other. We find suggestive evidence that the largest increase in electricity consumption in all three regions occurs between 7pm and 10pm, generally after summer peak demand. The overall increase is smallest in the cooler Coast, largest in the warmer Inland Hills, and in the middle in the hot Central Valley. These estimates of regional differences in the effect of the CARE program can help policy makers and utilities understand the energy effects of changes to low-income electricity subsidies.