Optimising eco-feedback design

Eco-feedback has become one of the most popular behavioural interventions for promoting household water and energy conservation. Since its inception, it has been adopted by various companies and governments around the world as one path to addressing climate change. Due to its ubiquity, eco-feedback interventions have been designed in various ways, potentially leading to heterogeneity in its treatment effects. This thesis investigates the different components of eco-feedback interventions, and how these can moderate its treatment effects Through four field experiments, I study the moderating effects of duration, frequency, medium, and to an extent, content of eco-feedback interventions. I find that 1) eco-feedback is effective at reducing household water/energy consumption across various contexts, achieving between 1-2% reduction in consumption, 2) the effects of the treatment attenuates over time once the treatment has ceased, 3) the medium by which the feedback is delivered is critical to its effectiveness, 4) delivering feedback for both water and energy at the same time may have a negative effects, and 5) the treatment effects are heterogeneous, mostly based on a household’s baseline consumption. Insights from this thesis should help inform the design of future eco-feedback interventions to better maximise its effects in the most cost effective manner.