Goals, Incentives, and Information in Water Conservation

How to cite

Byrne, David; Goette, Lorenz; Berg, Anthony van den; Jones, Alana. 2022. "Goals, Incentives, and Information in Water Conservation." AEA RCT Registry. AEARCTR-0009028. https://doi.org/10.1257/rct.9028-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

David Byrne* , Lorenz Goette , Anthony van den Berg , Alana Jones

* Primary investigator

Registry ID
AEARCTR-0009028
DOI
https://doi.org/10.1257/rct.9028-1.1
First registered
Last updated
Status
Completed
Trial dates
2022-02-02 – 2022-05-29
Countries
Australia (Victoria)
JEL codes
C93, D61, D83, D91, L95, Q25, Q55
Registry record
http://www.socialscienceregistry.org/trials/9028

Abstract

Ensuring sustainable water usage is an ever-pressing concern for Australia and internationally for counties like Singapore. By partnering with a water utility, this project aims to combine large scale field experiments, machine learning, and data from smart devices that provide real-time feedback. This new knowledge will underpin the empirical development of new behavioural economic models (habit, beliefs formation, and reference-dependent decision making with rational inattention). These models will be able to be immediately deployed, informing policies designed to promote water conservation, encourage efficient water network investment, and help households anticipate and reduce water costs, particularly among disadvantaged groups.

Intervention

The experiment examines how goals, information, and incentives provided through GreenBe’s Water Savers mobile app influence water usage. We experimentally manipulate the design of weekly Water Savers Challenges, specifically with goal difficulty and rewards for achieving goals, on the mobile app to gain such behavioral insights.

Experimental design

2 x 2 experimental design that considers combinations of {high incentives for meeting goals, low incentives for meeting goals} x {easy water usage reduction goals, hard water usage reduction goals}

Design details

Randomization method
Randomization was done in the office by a computer
Randomization unit
Household
Clusters
1000 households
Observations
5,376,000 half-hourly water usage levels at the individual level (1000 households x 16 weeks x 7 days/week x 48 intervals/day)
Arms
200 households control, 200 households in one of 4 different experimental conditions that have 2 x 2 combinations of {high incentives, low incentives} x {easy goals, hard goals}
Primary outcomes
Household water usage at the individual level measured within 30-minute intervals from smart water meters.
Secondary outcomes
Post-trial survey outcomes on water-saving actions households took to conserve water to meet their goals.

Cited by

No citation data available for this registration yet.

Keywords