Rate structure effects and regression parameter instability across time-of-use electricity pricing experiments

Abstract

Using a pool of data from eleven time-of-use (TOU) pricing experiments, a consumer demand model is estimated, and simple but statistically exact tests are performed for rate structure parameter instability across these experiments. Rate structure effects considered in the analyses concern the length of time (during the demand cycle) that the respective prices are charged, intermediate pricing periods, demand versus energy charges, and voluntary versus mandatory participation. Pricing period length and energy charges are found to have significant effects but the significance of others rate structure effects are generally mixed. The hypotheses of parameter instability are accepted, in a number of cases, across experiments but rejected through time.