Peaking Interest: How Awareness Drives the Effectiveness of Time-of-Use Electricity Pricing
This paper examines a new machine-learning method to estimate heterogeneous electricity demand responses to time-varying prices in an experiment on Irish households.
Abstract
I apply and extend a new machine-learning method to estimate heterogeneous electricity demand responses to time-varying prices in an experiment on Irish households. The most important source of heterogeneity is consumer awareness, followed by information provision and baseline consumption. In-home electricity monitors doubled responses. Other household characteristics like demographics, appliance ownership, or house characteristics, were not predictive of heterogeneous effects. Surprisingly, households appeared to violate a central law of demand theory: while they responded to the existence of a price change, they were extremely insensitive to the magnitude of the price change. This suggests that “getting the prices right” is less important than getting consumers to pay attention in the first place.
