Skip to content Skip to navigation

Energy Behavior Taxonomy

Energy Action & Behavior Attributes
A fundamental necessity to effective energy behavior change is to characterize desired behaviors. Does the behavior require much skill? Is it expensive? Must it be performed just once or several times a day? Will energy and economic savings be substantial or small?

Energy actions or behaviors, such as purchasing a refrigerator, turning off lights and cleaning a heating filter, are driven by various factors and can be described along different dimensions. Properly characterized and clustered behaviors can aid scholars, policy makers and energy industry professionals in prioritizing desired behaviors, targeting appropriate consumers and understanding relevant barriers.

This project developed a comprehensive set of 261 residential energy-efficiency behaviors and identified ten different attributes for classifying the behaviors. Next, the investigators developed a rating classification scheme and explored how the behaviors clustered together based on the ratings on these attributes. They also developed a framework for the systematic study of behavior attributes and how they cluster.

Among the findings:

  • Frequency of occurrence: Many behaviors (38 percent) in the set are performed only once every three or more years. These include such behaviors as major appliance purchases and home weatherization. Another 15 percent of behaviors occur every one to three years, such as small appliance purchases and appliance maintenance. The second largest set of behaviors (20 percent) occurs with very high frequency (multiple times a day), such as turning off lights and unplugging/turning off computers and entertainment devices.
  • Required skill level: Some 52 percent of behaviors in the set require very little or no skill, such as closing shades or drapes to keep heat in or out, turning off lights, or installing a CFL. Medium skills –e.g., reading instructions and/or having tools – are required for 21 percent of behaviors. Another 28 percent of behaviors demand significant skill such that an expert may be needed to perform the behavior, e.g., to install insulation or high efficiency windows.
  • Energy savings: Almost half of the behaviors (46 percent) are projected to save more than 750 kilowatt hours per year. About 20 percent of the behaviors have marginal energy savings of 1 to 25 kWh/year. Some 15 percent of the behaviors save 25-100 kWh/year, while 12 percent save 101-250 kWh/year, and 10 percent save 250-750 kWh/year.
  • Cost: More than half of behaviors cost no more than $20 – 43 percent cost under $5, and 11 percent cost $5 to $20. About 20 percent of the behaviors cost between $100 and $1,000, and 11 percent cost more than $1,000.
  • The behaviors were also characterized for household function, locus of decision, observability, home topography and appliance topography were also analyzed.
  • Researchers then organized the behaviors into five clusters: "Call an Expert" (73 behaviors, 28 percent of the sample); "Family Style" (66 behaviors, 25 percent); "Household Management" (49 behaviors, 19 percent); "Go Shopping" (47 behaviors, 18 percent); and "Behind the Scenes Work" (25 behaviors, 10 percent).

Publications and Presentations

Clustering household energy-saving behaviors by behavioral attribute (pdf)
Energy Policy92 pp 444-454
Roumpani, M., Flora, J.A., Boudet, H., Armel, C. (2016) 

Attributes of energy reduction behaviors
Behavior, Energy & Climate Change Conference
Roumpani, M., Flora, J.A., Boudet, H., Armel, C. (2011)

Future work

Researchers plan to extend their behavior change attribute framework and analytic techniques to transportation and food. In addition, they are examining survey ratings of behavior use, intentions to use, confidence in long term use, and barriers to use.