Market segmentation and time-of-use rate plans: a consumer choice analysis
Date
2019
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
Modern electric utilities face a complex set of expectations – they must provide electricity to all customers within their service territory and meet consumer electricity demand while also being regulated to make sure the plans and prices they impose on their customers reflect what would be competitive plans and prices. As residential consumer demand for energy continues to increase, meeting this demand during peak times becomes increasingly difficult and costly. Increased strain on the shared grid during high use periods can cause interruption of service, increased costs, and increase the reliance on non-renewable sources of energy. Time-of-use (TOU) electricity service plans, which have a higher price per kilowatt hour during the highest use time periods, can be implemented by utilities to try to incentivize customers to shift their usage to times of the day that have lower overall demand for energy. Setting a price that is high enough to get customers to change their behaviors to use energy at different times of the day, but not too high as to reduce customer utility requires having intimate knowledge of consumers in a service territory as well as their behaviors and preferences. Discrete choice experiments can help determine consumer preferences overall, but it is very likely that different types of consumers will have different preferences for plan types as well as plan rates. This study aims to see if market segmentation can be used to help accurately identify which customers are likely to choose to switch to a TOU rate plan by utilizing Esri Tapestry Market Segmentation data and a 2014 discrete choice experiment where respondents were asked to choose between a status quo plan and two time-of-use rate plans. In this study, a nested logit model is estimated and an out-of-sample test is performed to test the hypothesis that the model results accurately predict the actual respondent choice better than a nested model that is estimated by grouping respondents by demographic group. Using the entire sample, the estimated nested logit model is used to simulate adoption rates for a set of two different time-of-use rate plans across different LifeMode summary groups. The modeled results show the difference in willingness to adopt a TOU plan between the groups. It is found that a LifeMode group that is comprised of young, low-income, diverse individuals is most likely to switch from the status quo plan to a TOU plan. Conversely, a group of retirement age individuals who do not routinely utilize the most up-to-date technology is least likely to switch to a TOU plan.