Many of the most effective energy-saving devices are hard to see. Moreover, the energy they save is invisible and difficult for most energy users to monitor with any accuracy. Other effects of energy savings, such as increased comfort or a small increase in disposable income, may also be difficult to perceive. As a result, people who successfully save energy do not necessarily know it and may even conclude that their efforts are useless. For example, consider a homeowner whose savings from adding insulation or a new furnace are masked by changes in weather or energy prices. If cooling or heating costs are higher after the investment than before, it is discouraging.

For these reasons, energy information programs should benefit from making energy savings more visible to those who achieve them. Such programs would gain credibility and the benefit of positive word-of-mouth advertising. When people learn the difference between effective and ineffective energy-saving actions they become better able to respond to emergency conditions that call for rapid changes in patterns of energy use.

Field experiments have explored various ways to provide energy users with an accurate account of their energy use. One method involves the installation of devices that monitor fuel or electricity use and provide digital readouts in cents-per-day, miles-per-gallon, or other useful units. In another method, people read gas or electric meters and provide householders with energy-use information on a daily or weekly basis, again using any of various simplifying measurements. Research has examined the effects of teaching people to read their own meters, sometimes combining this instruction with a commitment from the energy user to try to save a certain percentage of energy use. Still other research has examined variations on the presentation of information on utility bills, maintaining the regular monthly billing period, but providing information in a more useful form.

Many of these “feedback” procedures have proven effective in studies of residential energy use (Seligman, Becker, and Darley, 1981; Shippee, 1980; Winett and Neale, 1979). The degree of effectiveness depends on several factors. First, feedback must be credible—that is, related to behavior. For feedback about energy used for home heating and cooling, this usually means using weather-corrected units. But if the units being used are not understood, the information may be discredited (Winett and Neale, 1979).

Second, feedback is more effective when energy users have made a commitment to conserve energy (Seligman et al., 1981) or have set themselves quantitative goals for saving energy. In one study, participants in a feedback experiment who committed themselves to reducing energy consumption 20 percent used 12 percent less energy than people who were