Association for Public Policy Analysis & Management

Maintaining a reliable electricity supply with increasing shares of intermittent and decentralized renewable generation will require both upgraded infrastructure and greater flexibility within electric grids. Time of Use (TOU) rates are one of the many demand-side response measures that can facilitate this increased flexibility. Although pilot programs have established that households reduce peak electricity demand in response to TOU peak-time rates, it is not yet well understood which factors affect household acceptance of TOU outside pilots, nor whether these factors differ for low-income households. Recent research on potential TOU adopters has found perceptions of control, usefulness, and ease of use predict intentions to adopt TOU plans.

We build on these findings by investigating how enabling technologies (e.g., smart thermostats) influence TOU acceptance, and by focusing on the role of income as a predictor for pilot participants deciding to stay on TOU post-pilot. More specifically, this paper builds a model of residential customer TOU acceptance, defined as customer intent to stay on TOU rates after taking part in an incentivized pilot. We use a sample of several thousand residential electricity customers who took part in a large utility-sponsored TOU pilot in Southern California. Hourly electricity usage data and bill data will be used to model actual customer bill changes and peak demand times. Survey data on perceived control, household demographics, and smart thermostat ownership will also be used.

We hypothesize that among the factors which affect TOU acceptance will be: (1) customer perceptions of the extent to which TOU rates (a) are easy to understand, (b) facilitate management of electricity bills, and (c) allow control of comfort; and (2) actual customer ability to match demand to off-peak rates (thus actually facilitating bill management). The impact of perceived ability to manage bills on TOU acceptance is expected to be strongest among low- (vs. higher) income households, and negative perceptions may lead to either financial stress or to high drop-out rates from TOU programs. Smart thermostat ownership is expected to predict TOU acceptance among customers who perceive smart thermostats to provide additional control over cost and comfort while on TOU rates; if this is confirmed, financial incentives for smart thermostat purchase could increase TOU acceptance.

Results will reveal the extent to which TOU acceptance is influenced by perceptions of costs, comfort, technology ownership, and household income. If low-income groups are particularly influenced by perceived control over bill management, there is a strong case for future research to examine which messaging or technological strategies can be developed to focus specifically on increasing perceived ability to manage electricity bills within this group, potentially also increasing actual ability manage electricity bills on TOU. Findings will further inform utility program design and implementation. For instance, if resistance to TOU is based more on perceptions of control than actual achieved management of electricity bills, enhancing customer perceptions of control (e.g., through messaging) could improve TOU acceptance.