Social comparison and electricity consumption: evidence from a randomised experiment on Swedish households

Sammanfattning: This thesis examines whether social comparison is a good tool to lower households' electricity consumption during peak hours on the electricity grid. A randomized control experiment was conducted with 196 households in Umea , Sweden. All households were sent an email with information about electricity consumption during peak hours as well as electricity conservation tips. The emails to the treated households also consisted of a comparison of their average electricity consumption during peak hours relative to the equivalent of similar households. Our results indicate that social comparison can be used to lower electricity consumption during peak hours by 7 percentage points for households consuming more than the average prior to the treatment. An even larger effect of 8 percentage points can be found when only examining villas. Furthermore, we can observe a boomerang effect where treated villas benchmarked below average increased their average peak consumption with 9 percentage points.