Using Field Experiments to Understand Energy Efficiency

Paper Session

Saturday, Jan. 7, 2017 2:30 PM – 4:30 PM

Swissotel Chicago, Zurich F
Hosted By: American Economic Association
  • Chair: John A. List, University of Chicago

The Impact of Smart Technology on Consumer Welfare: Evidence From a Large Randomized Smart Grid Experiment

Matthew Harding
,
Duke University
Justin Ho
,
Columbia University
Carlos Lamarche
,
University of Kentucky

Abstract

This paper uses Big Data resulting from a large randomized controlled smart grid trial which exposed households to varying degrees of dynamic pricing in conjunction with different enabling technologies to evaluate the welfare impact of different dynamic rate schedules. Households were allocated to different treatment groups consisting of Time-of-Use rates and also Variable-Peak Pricing (whereby the price faced by consumers varied from day-to-day), while also randomized to technologies enabling different degrees of price/quantity salience and behavior automation. The paper measures the responsiveness of households to different price incentives at 15 minute intervals. Based on the experimental estimates we project the welfare cost to consumers resulting from increasingly real time price variation and contrast to the resulting cost savings to the utility. The paper measures behavior changes related both to demand response and longer term energy efficiency resulting from dynamic pricing.

Information Frictions, Inertia, and Selection on Elasticity: A Field Experiment on Electricity Tariff Choice

Koichiro Ito
,
University of Chicago
Takanori Ida
,
Kyoto University
Makoto Tanaka
,
National Graduate Institute for Policy Studies

Abstract

We develop a discrete-continuous choice model to characterize the link between plan choice, switching frictions, and subsequent continuous choice of service utilization. We then test the model predictions by using a randomized controlled trial in electricity tariff choice. We find that both information frictions and inertia prevent consumers from switching to a tariff that is privately and socially beneficial. While interventions to mitigate these frictions increased overall switching rates, they also incentivized relatively price-inelastic consumers to switch. We characterize this phenomenon by selection on elasticity and show how it affects the optimal rate design in the presence of switching frictions.

Smart Thermostats and Social Norms: Distributional Evidence From a Field Experiment

Alec Brandon
,
University of Chicago
John A. List
,
University of Chicago
Robert Metcalfe
,
University of Chicago
Michael Price
,
Georgia State University

Abstract

Over the past decade governments around the world have invested billions to support the development and installation of smart-grid technologies. We present evidence from a field experiment that considers a role for pro-social preferences in leveraging these investments via a smart thermostat with social norm framing of set point choices. To guide our investigation we develop a simple model that relates the household’s choice variable (temperature set point) to our observable (electricity use).
We then consider the testable predictions of a smart thermostat affecting one of two margins—pro-social motivations or adjustment costs—and find that higher order moments can parse the two models. In particular, a smart thermostat increasing prosocial motivations predicts a reduction in the mean and standard deviation of electricity use, while a reduction in adjustment costs predicts an increase in the standard deviation. We find that treatment caused an approximate 2.3 percent reduction in average electricity use and a corresponding 5.6 to 7.5 percent reduction in the standard deviation of
electricity use, suggesting an important role for social incentives in smart technology. Counterfactual simulations of the wholesale electricity market in California point to meaningful savings from adoption of such technologies at scale, with changes in variance driving more than one-third of the savings.

Are Consumers Poorly Informed About Fuel Economy? Evidence From a Randomized Trial

Hunt Allcott
,
New York University
Christopher Knittel
,
Massachusetts Institute of Technology

Abstract

Energy efficiency standards, including automotive fuel economy standards, have been justified on the basis of large private welfare gains to consumers, on top of any gains from externality reduction. Two frequently-proposed reasons for why consumers currently leave these private gains on the table are that they are imperfectly informed about the benefits of fuel economy or that they are informed but inattentive to these benefits. We test these by implementing two potentially-powerful informational interventions with two different consumer populations. First, we carry out extensive informational discussions with randomly-selected consumers from car dealerships in several different locations across the country. Second, we recruit vehicle shoppers from a nationwide survey panel and treat them with videos and other information about fuel economy. Under the assumption that our treated groups are attentive and fully informed to fuel economy, we use our treatment effects to bound the allocative distortions from inattention and imperfect information.
Discussant(s)
Dave Rapson
,
University of California-Davis
Michael Price
,
Georgia State University
Steven L. Puller
,
Texas A&M University
Catherine Wolfram
,
University of California-Berkeley
JEL Classifications
  • D0 - General
  • Q4 - Energy