Association for Public Policy Analysis & Management

The impact of everyday energy consumption activities is often invisible to consumers. Information strategies that aim at correcting this information asymmetry are increasingly common. Energy saving tips (e.g., Schultz et al., 2007; Slavin et al., 1981) and monetary incentives (e.g., Bittle et al., 1979; Wilhite and Ling, 1995) are regarded as significant ways to reduce energy consumption by changing customers’ day to day behavior. However, previous studies rely on survey to capture customers’ behavioral change, which show inconsistent result on whether information nudges customers’ behavior in consuming energy (Delmas et al., 2013). In this paper, we link energy related information to objective energy consumption measures for the city of Tallahassee in Florida. City of Tallahassee is served by a publicly owned utility, which actively uses a messaging package to communicate with its customers in an effort to shape their energy consumption behavior. We collected monthly utility bill inserts for 60 months during the 2011-2015 time period. The messages have been manually classified to represent energy saving tips and monetary incentives. We then draw a random sample of 3,000 households out of 110,000 residential customers. Relevant data are collected from various sources including City of Tallahassee Utility, Leon County property appraiser, Leon County voter registration record, census bureau, and national climate data center. In the analysis, we performed a panel data analysis of household level energy consumption using types of message, housing characteristics, weather, and social-economic status as explanatory variables. Our study shows both energy saving tips and monetary incentives serve as significant factors for energy conservation behavior. We then test if different messages have varied impacts based on social economic status (SES), including political ideology, house value, race and education. Our preliminary results show that customers respond to messages in different way, based on their SES characteristics. This study contributes to existing literature by expanding the analysis of information nudge to objective, fine granularity data, which helps to resolve the inconsistent findings from previous studies. To practitioners, our finding shed light on how to effectively nudge customers’ energy conservation behavior by strategically targeting on their concern with regard to the characteristics of customers.