Researchers built an appliance recommendation website to help users learn about the electricity consumption and operating cost of both the refrigerator they own and new refrigerators. The latter also were shown with prices. Research results are based on the website’s users selecting refrigerator models to compare, not actual purchases, because consumers generally research online but make purchases in physical appliance stores.
Among the findings:
- Messaging with a strong energy focus was successful in nudging consumers toward more energy-efficient models.
- Messaging that stressed purchase prices over energy consumption and costs resulted in preferences for refrigerator models that consumed 594 kilowatt-hours per year on average. A strong emphasis on energy efficiency resulted in preferences for models that consumed 523 kWh/year.
- Pushing the rank of the most energy-efficient refrigerator model down the search results increased the average kWh/year of refrigerators saved to the list, indicating that rank may have a bigger effect than energy information.
- Energy information may have a very small effect or even none at all. Replying to search results with only photos, short descriptions, prices and kWh/year had similar results to the addition of models’ lifetime electricity costs, annual energy savings or both.
Publications and Presentations
Bunching with the Stars: How Firms Respond to Environmental Certification
Working paper, University of Maryland
Houde, S. (2013):
Selling Energy Efficiency Online with Simple Nudges
Manuscript in preparation
Houde S., Armel C., and McClure
How Consumers Respond to Product Certification: A Welfare Analysis of Energy Star
Working paper, Stanford University
Houde, S. (2012)
These results have a somewhat profound implication for behavioral interventions. An attempt to produce more energy-efficient purchasing behaviors by using an intervention derived from an analysis of what the underlying problem was—the first cost bias—failed to produce an effect. Resarchers attempted to address the tendency of people to make an appliance purchase decision based on the upfront cost of an appliance rather than the longer term energy savings. In contrast, where the intervention was driven simply by using some of the previously demonstrated most effective behavior change techniques, we saw a change in average kWh of between 10-20% depending on the measure used. Thus, for quick and effective results, it may make sense to first try the most effective proven behavior change techniques to date.
Future research should test additional manipulations to clarify how to best design online appliance recommendation websites to encourage efficient purchases, and which types of behavioral techniques are most effective.