Variations on the Normative Feedback Model for Energy Efficienct Behavior in the Context of Military Family Housing

Abstract

In discussions on energy efficiency, the topic often involves the development of a new technology. But models that target human behavior change can also generate significant energy savings, often with less expense. One such model that has been employed to millions of households across the United States is the normative feedback model. This model integrates the most salient research from the field of social norms into the utility billing process. Residents receive a home energy report that compares their energy consumption to the average energy consumption of their neighbors. Resulting behavior changes have led to energy savings of between 1.4% and 3.1% (Alcott, 2011). This research tests three variations of the normative feedback model, with the aim of determining its boundary conditions and improving it. Each variation aligns with one of the three research objectives: 1) to determine whether normative feedback generates significant energy savings when applied to households that are not billed for utility usage, 2) to determine whether increasing the proximity of comparison increases the energy savings generated by normative feedback, and 3) to determine energy savings associated with the implementation of a normative-based utility billing system. The first two variations were tested through an experiment conducted at Joint Base Andrews in Maryland, where residents are not currently billed for their utility usage. Residents received normative feedback via home energy reports for three consecutive months. Results were analyzed through both a differences-in-differences analysis and a multiple regression analysis, and an overall energy savings of 3.8% was identified. Thus, the normative feedback model can generate energy savings even in the absence of a billing system and could therefore be employed in the two-thirds of military family housing that are not yet billed for utilities, with resulting savings of approximately $19.3 million annually. In the home energy report experiment, residents were compared at three different levels of proximity: neighborhood, street, and next-door neighbor. The analysis identified an energy savings of 3.8% at the neighborhood level, 4.9% at the street level, and 2.8% at the next-door neighbor level. These results indicate a "sweet spot" in setting the proximity level of comparisons. By increasing the proximity from the neighborhood to the street level, energy savings increased, which is consistent with expectations based on a previous study in a different context (Goldstein, Cialdini, & Griskevicius, 2008). But increasing the proximity further to the next-door neighbor level actually reduced the energy savings. Therefore, to maximize effectiveness, future applications of normative feedback for energy efficiency should make comparisons at the street level of proximity. And finally, this research investigated the use of normative feedback as the basis for a utility billing system. Such a system, as implemented at Fort Belvoir in Virginia, establishes a monthly baseline equal to the average energy consumption for that month. Residents make payments for their consumption over the baseline and receive payments for their consumption under the baseline. A multiple regression analysis found that the implementation of this billing system into a community not previously billed for utility usage generated an energy savings of 14.1%. This result takes an important first step towards the development of a billing program optimization model for the military's transition to utility billing.