Behavioural mechanisms of household electricity consumption

A central goal of the Energy Strategy 2050 is helping households use resources more efficiently. Incomplete information and limited attention are two key obstacles to making resource-efficient choices. In the context of electricity use, it has been hypothesised that the rollout of smart meters would improve information and sharply reduce energy use. However, evidence indicates that smart metering in countries with moderate baseline consumption, such as Switzerland, Germany or Austria, reduces energy use by only about 3 percent.

The aim of this project was to test whether behaviour-specific feedback can help in increasing these moderate conservation effects, and whether feedback combined with behaviour-specific incentives to conserve electricity can further reinforce conservation efforts. For this purpose a large-scale field experiment was conducted, using data from more than 1000 private households. All experimental groups receive a smart phone app that provides them with feedback on their electricity use. Five experimental groups with differing feedback and incentives were formed: (i) a group receiving aggregate information on their electricity use; (ii) a group additionally receiving behaviour-specific feedback for five categories of electricity use; (iii) a group that in addition to (ii) receives incentives to improve their behavioural efficiency on their least efficient behaviour; (iv) a group that in addition to (ii) receives information on how they have improved their least efficient behaviour compared to other participating households; (v) a group that in addition to (iv) receives incentives for improvements relative to other participating households. There was also a matched (non-experimental) control group of households that had smart meters installed by the electricity company for unrelated purposes.

No significant conservation effects were found when providing aggregate information on electricity use alone. However, the results are not significantly different from other studies which find small (2% - 3%) conservation effects for comparable households. By contrast, there are significant and quantitatively important increases in energy efficiency found for all groups providing behaviour-specific feedback: households reduce electricity consumption by 6% to 10% compared to aggregate feedback. Thus, adding behaviour-specific feedback roughly triples the conservation effects compared to typical estimates for aggregate smart metering information. Consistent with previous research, one also finds that households with high baseline use save significantly more energy in response to behaviour-specific feedback.

There are no significant differences in conservation effects across the four groups receiving behaviour-specific feedback. Thus, adding financial incentives to encourage electricity conservation neither increases nor decreases conservation efforts compared to the group receiving only disaggregated feedback. These results suggest that attention is a key mechanism to sustain conservation effects.

The results also show pronounced time-of-day patterns. Conservation effects from disaggregated feedback are largest during the hours when households are at home and have the highest electricity use. They are essentially zero for all other hours. During peak hours, behaviour-specific feedback reduces electricity use by 10% to 20%.

A strong pattern of habit formation wasn’t found. By contrast, it was shown that, when households receive fewer messages from the smart phone app, treatment effects become weaker. This again suggests that attention is a key mechanism to help households conserve energy.

Implications for research

The results underscore the key role of attention and information in realising conservation effects. They show that individuals are quite strongly motivated to conserve energy when informational conditions are such that electricity use becomes sufficiently visible. Incentives to conserve do not affect conservation efforts significantly.

Implications for practice

The results show that the rollout of smart meters can have important side benefits in terms of electricity conservation. However, the results also indicate that providing aggregate feedback on electricity use is not good enough. Utility companies must provide a breakdown of electricity use into the different components in order to achieve significant resource conservation.

The role of social information, incentives, and habits in household electricity consumption