To handle surges in electricity demand during peak hours, utility companies often resort to meeting the excess demand by building additional capacity. The downside to this strategy is that a large proportion of the power generation capacity will become latently available, and thus not efficiently used. Another strategy is reducing peak hour electricity demand so that it meets the supply. Adjusting the demand to the supply is referred to as demand response (DR), and offers a significantly more environmentally friendly and sustainable approach to improving the efficiency of energy systems. In order to incentivise consumers to move their energy consumption to off-peak hours, or reducing it during peak hours, electric system operators and utility companies design DR programs that encourage consumers to change their energy usage patterns. One such program is the real-time pricing scheme, where electricity prices are adjusted dynamically, being higher during peak hours.

Real-time pricing schemes are designed around the principles of neoclassical economic theory, which assumes that consumers respond rationally to financial incentives. Studies investigating the elasticity of the own price of the demand for electricity in the residential sector have consistently shown small, but negative values, suggesting that financial motives do promote DR to a degree. However, research from environmental psychology suggests that non-financial motivations, such as environmental concern, also play a crucial role in shaping energy consumption behaviour.

This thesis investigates whether financial and environmental motivations influence how Norwegian households adjust their electricity usage during peak hours. Norway represents a unique case to study, as it is particularly well suited for DR programs. This was done using fixed-effect regression on a panel dataset consisting of the hourly electricity consumption of 1,136 households in Norway. Each regression model was characterised by different fixed effects, and each household was grouped by how they answered questions on what motivates them to partake in DR. In each model, groups corresponding to financial and environmental motivations were assigned a dummy variable which was interacted with indicators of peak hours, while all other households were used as a reference group. The resulting coefficients were analysed and discussed.

The baseline regression results show that neither households in the financially motivated
group nor the environmentally motivated group appear to reduce their electricity consumption during peak hours by more or less than the reference group. This suggests that households adjusted their electricity consumption during peak hours in a similar way, despite differences in stated motivation. These findings are consistent for all models, except when the model included hour-level fixed effects, which control for each household’s load structure. In that model, the coefficient on the interaction between the environmental group variable and the peak hour variable became significant at the 5% level, with a value of -0.0206. This effect corresponds to approximately double the electricity used by a single LED light bulb per hour. The robustness analysis supported these conclusions with a few minor exceptions.

Although the effect is significant, it is very small, making its practical relevance questionable. Furthermore, some limitations were opposed on this research by a relatively small sample size and the option for households to give multiple answers to motivational factors for partaking in DR. For these reason, this work can not provide definite answers to policy makers on whether DR programs are based on the correct assumptions, and whether focusing on the monetary incentives is most effective.