Few and far between: a scoping review of the mechanistic evidence in empirical research on household energy-saving interventions

Varying effect sizes and inconsistent findings signal the need for a better understanding of the causal pathways through which interventions in the field of household energy saving operate. Hence, we conducted a scoping review of the mechanistic claims related to the observed effects of behavioural interventions in the domain. We reviewed the available publication pool to record and categorise the evidence used to support these claims and then matched it to distinct intervention types. We also recorded the strategies used to test the mechanistic claims in the literature. From a starting pool of 1591 records, we found that 110 studies contained a mechanistic claim and reported primary outcomes from behavioural interventions on household energy saving. Most of the 110 studies (74.6%) contained a measure of at least one suggested mechanistic variable, while the remaining 25.4% contained a mention of at least one. When a mechanistic variable was measured, 80% of the analysis methods were correlational (simple correlation, mean comparisons and regressions). In less than 15% of the studies, the treatments were randomised on a mechanistic variable of interest. Moreover, we found evidence that the same kind of intervention has been related to different mechanisms ('mechanistic heterogeneity') and different kinds of intervention were reported to operate through the same mechanism ('mechanistic overlap'). We conclude that the empirical research in the field pays insufficient attention to the mechanisms behind intervention outcomes and that household energy consumption policies would benefit from more mechanistic evidence.