Behavioral Interventions and Water Conservation
Paper Session
Saturday, Jan. 7, 2017 3:15 PM – 5:15 PM
Swissotel Chicago, Zurich G
- Chair: Anant Sudarshan, University of Chicago
The Effect of Social and Consumption Analytics on Residential Water Demand
Abstract
Water utilities in arid regions, such as California and Texas, face challenges in meeting residential demand under conditions of supply uncertainty, which is expected to be amplified through increased variation in precipitation under climate change. In response, utilities and technology companies are innovating, often together, on ways to share information to customers with the aim of improving water-use efficiency and to encourage conservation during dry periods. We partner with one company, DropCountr, to investigate how information provided in home water-‐use reports shared via a web-based platform affects household consumption behavior. Similar to Opower’s residential energy-use reporting, the home water-‐ use reports we consider provide consumption analytics, social comparison, and conservation information to residential users, primarily through digital communications. We estimate the effect of household enrollment in a pilot program using DropCountr’s web-based home water-use reports for a city in Northern California. The data used in this analysis includes two years of historical consumption, along with approximately ten months of data under the pilot program. Households receive digital communications informing them about several features of their usage, such as: individual household patterns in water consumption; comparison to other users in the service areas; and, information on rebate and conservation opportunities. We find that the pilot program has a statistically and economically significant conservation effect on water consumption at the household level for those customers who enrolled in the service. We estimate this effect to be a 6% reduction in average monthly consumption for the average enrolled household. We contrast these results with preliminary analysis for a large city in Texas that has also adopted DropCountr’s web-based platform for providing home water-‐use reports.Becker Versus the Behavioralist: Using Targeted Messages to Promote Compliance With Outdoor Water Restrictions
Abstract
During the summer of 2007, we partnered with the Truckee Meadows Water Authority to implement a natural field experiment designed to promote compliance with outdoor water restrictions that limit outdoor water use to two assigned days per week. The experiment was embedded in a daily monitoring project of more than 4,800 residential water consumers that provided daily use data over an eight week period. Households were randomly assigned to either a control group or one of three treatment conditions; (i) a group that received a reminder of the watering restrictions and associated schedule, (ii) a group that received information on the restrictions and a normative appeal to adhere to the schedule given prevailing drought conditions, and (iii) a group that received information on the restrictions along with a message noting that TWMA had observed unusual patterns of usage in their neighborhood and an appeal to adhere to the schedule. Consumption for households in the experiment was observed during a four-week pre-intervention period and for four weeks after the letters were mailed. Results of the experiment show that the letters generated an approximate 6.5 to 15.2% reduction in noncompliance and a significant reduction in daily use on unassigned days. Moreover, we observe heterogeneity in the effects of the various letters on use on assigned days. While the schedule and monitoring letters lead to increased use on assigned days, the normative appeal leads to a significant reduction in use of assigned days. To explore the long-run impacts of treatment, we monitor a random subset of the households during the summer 2008. Interestingly, both the normative appeal and monitoring letters have lasting impact although the effects are less pronounced than those observed during the summer 2007.Enduring Effects of Changes in Billing Frequency: Evidence From Urban Water Use
Abstract
Little is known about the way consumers update their behavior in intermittent choices settings. I exploit a natural experiment in which residential water customers were switched from bi-monthly to monthly billing to assess the long-term responses to increases in billing frequency. I find that customers increase water consumption, but the estimated average treatment effect weakens over time and its profile differs based on historical consumption patterns and preferences for outdoor water use. Inattentive consumers respond to the treatment by a smaller magnitude and for a shorter duration. Notably little heterogeneity is found among wealth groups in the dynamic response to monthly, which suggests that more frequent billing may not aid in smoothing expenditures.Discussant(s)
Anant Sudarshan
, University of Chicago
Katrina Jessoe
, University of California-Davis
Gabe Lade
, Iowa State University
Steven Sexton
, Duke University
JEL Classifications
- Q5 - Environmental Economics