A Behavioural Rebound Effect: Results from a laboratory experiment

Significant attention has been paid to leveraging behavioural motivators (non-price interventions) to increase energy conservation (Allcott and Mullainathan, 2010). Technological change that improves energy efficiency is also important (The Global Commission on the Economy and Climate, 2014). While both focus on reducing energy use, these two strands of literature have yet to be joined to con-sider what behavioural effects might result from technology change. The direct rebound effect is the increase in consumption due an increase in energy efficiency and can be modelled as the rational response to a change in relative prices (Chan and Gillingham, 2015). This paper investigates whether there might also be a be-havioural rebound effect by looking at two potential sources. First, a behavioural rebound effect where pro-environmental behaviours are reduced after an improve-ment in energy efficiency. Second, moral licensing may increase the behavioural rebound effect if individuals who buy an energy efficient product subsequently give themselves psychological licence to reduce their pro-environmental behaviours even further. I develop a novel laboratory experiment to investigate these mechanisms, which can be cleanly isolated in the laboratory without the many confounds po-tentially present in the field, such as other motivations to reduce energy usage like saving money. Subjects must decide how to allocate their effort, in a real effort task, between earning money for themselves and reducing damages to a tree plant-ing charity. I find evidence for a behavioural rebound effect, which is estimated to be 32% in this laboratory setting. Moral licensing also occurs, increasing the size of the behavioural rebound effect, and it is strongest among subjects with a higher level of pro-environmental orientation of their attitudes and beliefs. The main driver of pro-environmental effort is shown to be beliefs about social norms. This paper extends the core model of the rebound eff(This abstract was borrowed from another version of this item.)