Julie Chao, Lawrence Berkeley National Laboratory
If your utility company were to send you a letter challenging you to use less energy than your neighbors, would you respond? How would you decide whether to buy a refrigerator that is slightly more expensive but consumes significantly less energy? What would it take to get you to drive your car less and use transit more?
At Lawrence Berkeley National Laboratory (Berkeley Lab) scientists and engineers are pursuing a broad spectrum of technological solutions for using less energy. But a handful of researchers are also looking at human behavior and decision making as a key factor in reducing energy consumption. The payoff could be huge.
“It has been estimated that changes in household behavior and personal transportation could save five to nine quads of energy per year in the United States,” says Berkeley Lab scientist Rick Diamond. “The total U.S. energy used in 2010 was 40 quads for buildings, so it’s big.”
Berkeley Lab researchers — including economists, ecologists and building scientists — have been studying behavior for decades, but the field has gotten more attention in recent years as global climate change has become an increasingly urgent problem.
However, just as quitting smoking can be difficult, changing behaviors around energy consumption is also not easy. Utilities, universities and private companies are hiring not only engineers to study and implement behavior-based solutions but also sociologists and even psychologists.
“People always think of behavior change as fast, easy and cheap. It’s neither fast, easy nor cheap — that’s a myth,” says Diamond. “Too often people think you get behavior change by education, such as putting up a poster that says, ‘turn off lights.’ But education is not the same as behavior change. You can give people information. Whether they choose to act on it depends on their motivation, their ability to take action and the cultural context for that behavior.”
One type of behavior-based program that more utilities are trying in recent years involves using nonstandard, psychological ways to persuade people to reduce their energy use, such as setting a goal for themselves or competing against their neighbors to use less energy. The programs — some of which were established in part based on Berkeley Lab research for California policymakers — can target one-time behaviors (such as changing thermostat settings), habitual behaviors (turning off lights) or purchasing behaviors.
“These types of programs are all pretty new,” says Berkeley Lab economist Annika Todd. “We don’t really know what people are doing, how they’re responding, and how long any behavior changes will last. So we wrote a report basically saying how you can measure savings from these kinds of programs.”
Todd’s study for the State and Local Energy Efficiency Action Network, which looked at dozens of behavior-based programs, recommends randomized controlled trials — with a “treatment” group and a “control” group of households — as the best way to ensure the validity of energy savings data. Some companies that had done this found electricity and gas savings in the range of 1 to 2 percent.
“That’s actually pretty good, especially from a cost-effectiveness standpoint,” Todd says. “It gets as good as or higher in cost effectiveness than other programs. The cost is just to do the mailings. Once the program gets rolled out, from a pilot to full scale, it will probably be even cheaper.”
Next she will be studying the question of persistence, or how long behavior changes last. “If they’re buying refrigerators the savings could last 10 years. If they’re turning off lights, it might last only a few days,” she says. “We don’t really know.”
Institutional behavior is another area that Diamond is studying. In a project for the Department of Energy with two other national labs — Oak Ridge and Pacific Northwest National Laboratories — Diamond is using a social science framework to develop case studies and training modules for federal agencies to meet sustainability objectives.
In one case study of the U.S. Navy’s San Diego area, which achieved savings of 73 million gallons of water and a 37 percent reduction in energy usage, Diamond’s team found that technology alone was not enough. “It was not just a question of installing technology; they worked with people, got their input, told them what was going on,” he says. “Everything was done with a behavioral perspective and they were able to achieve large savings.”
The team is now stepping up in a more direct role of helping federal agencies achieve their energy goals. “We will go to federal agencies and say, ‘this is what your peers have done, this is why we think it’s successful,’” says Diamond.
Human behavior is a crucial aspect of technology in many instances. “If people can’t understand how to use something they won’t use it,” says Berkeley Lab researcher Ed Vine, who has been studying behavior for 30 years. “So we may have great technologies, but nobody’s using them.”
Programmable thermostats are a classic example: meant as an energy-saving device, some are too complicated for people to understand. Now Berkeley Lab has developed the first method to quantify and rank the usability of thermostat interfaces. The resulting score allows manufacturers, regulators and others to rate the usability of controls and other products with complex interfaces. The methodology has been adopted by Energy Star for its next thermostat specification. Future work will focus on measuring the usability of lighting controls in commercial buildings, vehicle instrumentation and heat pump water heaters.
Another behavior-related research project at Berkeley Lab looked at the potential of energy conservation to help California reach its goal to reduce emissions of heat-trapping gases to 80 percent below 1990 levels by 2050. Berkeley Lab researchers Max Wei and Jeff Greenblatt focused on daily habitual actions — such as turning off lights, recycling, and personal transportation — that are sometimes associated with lifestyle changes.
“Our approach is to treat these actions as market adoption items,” Wei says. “What we find by looking at non-energy behaviors, such as seat belt usage, smoking reduction, and poultry consumption is that many of them follow a market adoption ‘S’ curve. It starts with a low minority of people doing the action, then a curve upward as more people adopt the action, typically to some saturation level. Many changes take decades.”
Wei and his colleagues selected about a dozen behavior actions and estimated a starting adoption rate and final adoption rate in 40 years based on a set of attributes and barriers, such as how visible the action is to others, the economic cost and cultural barriers. The list of actions includes healthier diets, line drying clothes, taking public transit, “ecodriving” (slower speeds, fewer hard stops and starts to reduce gas consumption) and reducing solid waste.
The researchers found two sets of actions to have the largest potential impact: reducing vehicle miles (such as by driving less, using public transit, or telecommuting more) and reducing solid waste (by recycling more and sending less waste to landfill). Together these can cut greenhouse gas emissions by 10 to 15 percent by 2050. “California’s electricity production is relatively clean, so the potential savings can be even greater in other parts of the country,” Wei says.
Diamond and Vine helped launch the Behavior, Energy & Climate Change conference (BECC) five years ago, which has also helped galvanize interest in the area. One year the organizers decided to conduct an experiment on their own conference-goers. Instead of asking people to check a box if they wanted the vegetarian option for lunch, they asked people to check a box if they wanted meat. As a result, people taking vegetarian meals jumped to 80 percent, compared to less than 20 percent the previous year.
“You’re not telling people what to do, you’re framing choice architecture. You’re changing how people make decisions,” Diamond says. “There’s a long history of this in marketing and advertising. Maybe we can do it more effectively in the energy area.”
