With increased weather-dependent electricity production and electrification at the heart of the ongoing transition away from fossil fuels, peaks in electricity demand are set to increase and become increasingly difficult to meet, which threatens the functioning of the electric power systems that our society depends on. Time-varying electricity rates, which aim to incentivize electricity users to shift their electricity use in time, have been proposed as a key instrument in alleviating grid imbalances and bottlenecks. Previous research has found that users respond to such rates by shifting their electricity use in time, but there is great variability in the observed response between studies that remains unexplained. In other words, it is unclear how and why users respond to time-varying rates, and thus how these so called demand side response policies should be designed to provide the best results.

This thesis aims to improve our understanding of how and why (not) time-varying rates work by exploring how users respond to both price and non-price signals that aim to incentivize a shift of electricity use in time, and what motivates, discourages, enables and hinders them to respond. This is done through four separate studies that are carried out in contexts where users have been involuntarily subjected to interventions that aim to incentivize demand side response. Using several novel methods, research designs and understudied empirical contexts, the studies also illustrate how biases that are commonly observed in the literature can be avoided and how intervention effects that often remain overlooked can be captured.

The results suggest that users may hold different motives to respond to signals that aim to incentivize a shift of electricity use in time, including non-financial motives such as a care for the environment and a will to meet the expectations of others. The rhythms and schedules of people’s everyday lives are identified as the most important hindrances for people to engage in demand side response. Notably, there is no evident relationship between how much money users may save by responding to a signal and their actual response. Many users do hold expectations of saving money and claim to engage in demand side response as a result of those expectations, but the fact that users are rarely (if ever) informed of whether their expectations are met or not suggests that many users may actually be willing to engage in demand side response with little or no financial reward. However, there is a risk that users who expect to save money may refrain from or stop engaging in demand side response if their expectations are not met, which poses a potential threat to the long-run effectiveness of conventional price-based demand response programs. There is also a risk that users who primarily hold non-financial motives may be discouraged to engage in demand side response if monetary savings is the key selling point of demand response programs. Policymakers, professionals and researchers should explore these risks and alternative policies that address them, particularly policies that may be more appealing to users that hold non-financial motives to engage in demand side response. Doing so will be key to ensure that current and future demand side response policies are cost-efficient and effective, both today and tomorrow.