Energy labels that make cents: A randomised controlled trial to test the effect of appliance energy rating labels

Last registered on February 28, 2017

View Trial History

Pre-Trial

Trial Information

General Information

Title

Energy labels that make cents: A randomised controlled trial to test the effect of appliance energy rating labels

RCT ID

AEARCTR-0002052

Initial registration date

February 28, 2017

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published

February 28, 2017, 11:53 AM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Country

Australia

Region

Primary Investigator

Name

Michael Hiscox

Affiliation

Harvard University

Contact Primary Investigator

Other Primary Investigator(s)

Additional Trial Information

Status

In development

Start date

2017-02-28

End date

2017-04-19

Keywords

Environment & Energy

Additional Keywords

Energy efficiency, star rating, labels, behavioural economics, loss aversion, consumer, appliances

JEL code(s)

C93 (Field Experiments), D03 (Behavioral Microeconomics), D12 (Consumer Economics)

Secondary IDs

Abstract

Reducing household energy consumption is a key component to boosting Australia’s energy productivity and the Australian Government employs a range of strategies to reduce household energy use and emissions. The Australian Energy Rating Label provides consumers with information about the energy-efficiency of appliances in a standardised format (energy consumption calculated by kWh per year). Evidence suggests that consumer understanding and interpretation of the label could be substantially enhanced. The study aims to examine ways to improve consumer understanding of energy efficiency ratings and the alignment of consumer preferences with their choice in household appliances. In particular, the study aims to test whether a new energy rating label, designed using insights from behavioural economics, might improve customer engagement with higher energy-efficient household appliances.

External Link(s)

Registration Citation

Citation

Hiscox, Michael. 2017. "Energy labels that make cents: A randomised controlled trial to test the effect of appliance energy rating labels." AEA RCT Registry. February 28. https://doi.org/10.1257/rct.2052-1.0

Former Citation

Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Experimental Details

Interventions

Intervention(s)

The trial is designed with three experimental arms into which customers visiting the online retailer’s website during the trial period will be randomly assigned. Individuals in the control group will view the website in its current form, with no energy rating labels associated with products. Individuals in the first treatment group will see the same products with the standard energy rating label that is currently mandated for use on those appliances when they are sold in physical (bricks-and-mortar) retail stores in Australia. Individuals assigned to the second treatment group will see the products with the newly designed energy rating label.

Intervention Start Date

2017-02-28

Intervention End Date

2017-03-31

Primary Outcomes

Primary Outcomes (end points)

The primary outcome is customer engagement with the product, which will be measured by the rate at which customers click to view detailed information about specific test products from the product category (filter) pages.
Secondary outcomes will include: purchasing intent (measured by the rates at which customers click to add test products to their shopping cart), purchases (rates at which customers click to complete payment for test products), and interest in the energy rating label (clicks to view more information about the label).

Primary Outcomes (explanation)

Secondary Outcomes

Secondary Outcomes (end points)

Secondary Outcomes (explanation)

Experimental Design

The study will be conducted in partnership with an Australian Government agency and an online retailer. It will use an individually randomised controlled trial with three experimental arms to determine the most effective way to support consumers to purchase products with lower predicted energy consumption over the life cycle of the product (kWh) using energy rating labels. Individuals who visit the online retailer’s website during the trial period will be randomised into one of three experimental arms. Aggregate (non-identified) data on behaviours of individuals in each experimental group will be analysed to assess the effects of different products labels.

Experimental Design Details

Individuals in the control group will view website pages in their current form, with no energy rating labels associated with products. Individuals in the first treatment group will view the current energy rating label positioned on the product category (filter) pages and on the product page for each test product. Individuals in the second treatment group will view the new behavioural economics-informed energy rating label positioned on the filter and product pages for test products. Four product categories will be included in the trial: televisions, fridges, washers and dryers. Within each of the four product categories, labels will be assigned to eight of the most popular products.

Randomization Method

Randomisation will be done by a computer algorithm.

Randomization Unit

At the individual level.

Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters

N/A

Sample size: planned number of observations

Approximately 48,000 non-unique observations are expected over a 30 day period. The observations are non-unique as individuals can log back into the website several times. We expect to have 13,000 unique observations over a 30-day window.

Sample size (or number of clusters) by treatment arms

We expect to have 13,000 unique observations over a 30-day window. This means approximately 4,333 individuals will be randomised into each of the experimental groups.

Minimum detectable effect size for main outcomes (accounting for sample design and clustering)

A total of approximately 48,000 non-unique visitors to the filter pages over approximately 30 days will provide at least 80% power for two-sided tests (at the 5% significance level) to detect substantial ≥5% differences in conversion rates between the experimental groups.

Supporting Documents and Materials

IRB