Association for Public Policy Analysis & Management

There are many policies and incentive programs to encourage the adoption of energy efficiency. The justifications for these policies are two folds. First, the negative externalities such as environmental pollution are not internalized into the electricity price. Second, various market inefficiencies induce suboptimal investment decisions. Novan and Smith (2018) provide empirical evidence that increasing block electricity rate can provide incentive for consumers to over-invest in energy efficiency because the private benefits accrued by consumers from energy efficiency (the energy bill savings) exceed the social benefits by 140% due to the fact that in most hours the marginal prices paid by consumers are higher than the marginal social cost of electricity supply. This implicit incentive to over-invest in energy efficiency suggests that policy makers might need to scale down energy-efficiency-promoting policies for increasing block rate customers. However, in order to help determine whether existing policies over-incentivize energy-efficiency investment, we need to quantify the social and private benefits of energy efficiency under other rate structures as well. In this study, we provide empirical evidence of hourly energy savings under Time-of-use (TOU) pricing. TOU pricing has higher marginal electricity prices during peak hours (when the marginal cost of supplying electricity is higher) and lower prices during off-peak hours. We compile a customer-level rich dataset of hourly electricity consumption for about 16,000 residential consumers from 2013-2016. We use a combination of matching, machine learning, and fixed effects panel regression approaches to estimate consumer energy consumption profiles from adopting energy efficiency while on TOU pricing. Then, using the estimated hourly energy savings, we calculate the private and social benefits from energy efficient AC. Private savings are calculated as the sum of hourly energy bill savings. Social savings are calculated as the sum of hourly private cost savings of providing electricity, avoided cost from pollutant and carbon emissions, and avoided investment in generating capacity. We obtain the following results. First, for TOU consumers, savings occur during both TOU peak-hours and non-peak hours. Second, price elasticities of consumers with energy efficiency are higher (more elastic) than those without energy efficiency. Third, the average annual private savings for TOU consumers from energy efficient AC is lower than the social savings by $50 per household (social benefits are about 20% higher than private benefits). Our results have important policy implications. We demonstrate that TOU pricing is unlikely to over-incentivize energy-efficiency investment and therefore other policy instruments are still needed to encourage energy efficiency adoption for TOU consumers. We also show that the price elasticities of consumers with and without energy-efficient AC are different. This indicates that policy makers should rely on empirical evidence to examine energy savings for different electricity pricing structures, instead of simulations assuming fixed price elasticities.