Information vs Automation and Implications for Dynamic Pricing

Abstract

Utility providers have an incentive to curb consumption during periods of peak demand because of the short-run costs of temporarily expanding output via older, more polluting capacity and the long-run costs of investing in new plants and/or transmission infrastructure. Demand reduction can possibly be accomplished by i) increasing price in these periods, ii) enabling consumers to better monitor their electricity usage using smart meters, and iii) providing home technologies which can automate price response. Using a field experiment randomized at the account level in two cities and with extremely disaggregate data on household electricity consumption, we estimate the effect of different price-schedules in combination with smartmeter and home automation technologies on electricity consumption. We leverage our disaggregate electricity consumption data in order to match households based on their pre-experiment consumption patterns. This allows us to estimate individual treatment effects that can be used to target specific households and greatly increasing the potential payoff of future interventions.