Green tariffs increase emissions – moral licensing leads to increased energy consumption

V Conclusion The findings from our quasi-experimental econometric analysis of household electricity consumption based on metered data from a large provider of green electricity in Germany suggest that switching to a green tariff leads to a non-monetary renewable rebound effect of around 7.7%. Further, our results imply that this renewable rebound effect is persistent over at least four years, possibly reflecting moral licencing effects that induce households to permanently change their habitual behaviours and/or to acquire additional electricity-consuming technologies. For our data, these findings appear to be driven by households in the first switching period (2004/2005). For later periods, we mostly fail to find statistically significant results. Because we have no information on household characteristics or individual attitudes, our data unfortunately did not allow us to account for potential heterogeneity in findings by socio-economic groups. Such analyses must be left for future research. Further, because of missing data on electricity tariffs of the previous provider, we had to use proxies for a large share of customers in our data set. Thus, we cannot robustly estimate the potential (negative) monetary rebound for switching from a grey tariff to a possibly more costly green tariff. Yet, our findings have several implications for policymakers and electricity providers. In particular, to limit potential negative effects of the non-monetary renewable rebound, policy makers could provide additional information. More specifically, such information may vividly describe the mechanisms which have been found to result in non-monetary renewable rebound effects in response to tariff switching. In addition, such information could explain the negative environmental effects such as higher greenhouse gas emissions pertaining to the renewable rebound effect. Such information could be targeted at individuals likely to subscribe to green tariffs, i.e. females, households/individuals with large incomes, high education levels, low electricity consumption, and ecological orientation (e.g. Sommer ; Ziegler ). Regarding policy analysis and design, ex-ante policy assessments carried out to define energy and climate policy targets should take renewable rebound effects into account. Failure to do so may lead policy makers to systematically underestimate the costs of achieving these targets. Our findings further suggest that these assessments should be based on difference-in-difference estimations rather than before-after analyses. Because the latter do not take into account developments which would have happened without the treatment – in our case moving into a larger dwelling, for example – they may lead to an overestimation of renewable rebound effects. Last but not least, utilities’ procurement of green electricity needs to account for a non-transitory increase in electricity consumption for new and old customers switching to a green tariff.