Energy-Efficiency Investments in Public Housing: What Determines the Rebound Effect?

In the coming decades, many countries need to improve the energy efficiency of their building stock, to realize the climate and renewable energy goals. In the Netherlands, this involves more than 5 million dwellings and many billions euros in costs. Around half of these costs will be carried by public housing providers who are required by regulation to radically improve the insulation and heating in their extensive dwelling stock. Existing research suggests that the energy savings from energy retrofitting in homes strongly depend on residents’ behavioural responses. Savings predicted by engineering models are often not realized due to the rebound effect – consumers tend to increase their energy consumption after retrofitting (e.g. Fowlie et al., 2018, Davis et al., 2014, Alcott and Greenstone, 2017, Aydin et al., 2017). Much less is known however about the determinants of the rebound on a household level: how does it differ by type of technology, type of household and dwelling. Our study aims to fill this knowledge gap, specifically for the public housing sector. We exploit unique dwelling-level data on some 2 thousand energy retrofitting investments performed by Dutch public housing providers between 2015 and 2018. Detailed longitudinal information on the energy efficiency measures per dwelling is merged to the restricted access data about the socio-economic characteristics, dwelling characteristics and energy consumption of the resident households. We identify econometrically the effect of retrofitting on energy savings with a propensity score matching methodology, by comparing the changes in the energy use in retrofitted dwellings and in similar non-retrofitted houses. Then, the rebound effect is derived and its determinants are examined. We extend the literature by providing new microdata-based evidence for a European country, comparing the behavioural effects of various energy-saving home technologies and looking specifically at the public housing sector. This paper has practical implications. Detailed insight in the size and determinants of household behavioural responses to retrofitting measures is crucial to optimally shape energy transition (e.g. Ossokina et al., 2020). Our paper will give public housing providers tools to anticipate on behavioural responses of tenants by customizing energy retrofitting measures and selecting tailored investments for specic household and dwelling groups.