EFFORTS TO CHANGE BEHAVIOR WITH INFORMATION
Lack of information can be a serious internal barrier to action because it is not always obvious to an individual how to act effectively on his or her attitudes. This is especially the case for environmental protection, because the connections between behavior and its environmental effects can be impossible to discern from personal experience. Only expert analysts can tell which behaviors have the greatest effect on global warming or the extinction of species in distant tropical forests, so nonexperts cannot be expected to know what to do without some assistance. Even with a relatively simple problem, such as reducing energy use in the home, many people do not know which conservation actions are most effective, as we show in Chapter 10.
How much can be done to protect the environment by informing consumers? The best evidence comes from careful studies of deliberate interventions—studies that compare the behavior of people who have been informed with similar people who serve as a comparison group. In this section, we review several of these studies. We find first that simply providing straightforward information can make a difference, but mainly with easy, low-cost actions. We then look at other ways of providing information, methods based on principles of psychology and communication. These methods are much more successful, and illustrate what can be accomplished by information alone. We begin with studies involving simple, straightforward information, starting again with energy conservation examples.
Information, Plain and Simple
In the 1970s, in the early days of excitement of the modern environmental movement, researchers and governments began to put "conventional wisdom" into practice: They assumed that if concerned people were only told what to do, they would act to preserve the environment. This approach had very limited success, as the following examples illustrate.
Shortly after the Arab oil embargo of 1973 shook the faith of many Americans in the perpetual availability of fossil fuels, a number of U.S. gas and electric utility companies began preparing and distributing glossy informational brochures on how to save energy in the home. Some of these brochures targeted relatively simple, cost-free measures such as resetting thermostats on furnaces and air conditioners to use less energy in winter, and setting them even lower at night and when the home is unoccupied. The companies typically distributed the brochures by inserting them in the envelope along with the regular utility bill, a so-called bill stuffer. Note that there are few external barriers to making these simple changes, and that the American public in the late 1970s had a positive general attitude toward energy conservation. Thus, the main barrier to action seemed to be lack of information about which behaviors effectively save energy—the barrier that bill stuffers attempted to overcome.
Despite all this, the few reported studies of the effects of these bill stuffers on actual energy use yielded disappointing results. Thomas Heberlein (1975) conducted a small experiment just before the 1973 energy crisis in which he mailed a utility-produced brochure on electricity conservation to fifteen households in a Wisconsin apartment complex. His research team read electric meters throughout the complex for about twelve days before and after the brochures were received and found no change in electricity use by the control households and a small increase in use, though not a statistically significant one, among households that received the brochure. In a larger study, Samuel Craig and John McCann (1978) monitored the effect of a utility-produced pamphlet on how to cut electricity use by air conditioners. In early August of 1977, they sent pamphlets to about 800 apartments in New York City where the pattern of electricity bills indicated that air conditioners were in use. By 1977, energy was a national concern, so a strong effect might be expected. Nevertheless, a month later, the apartments that received the brochure along with a letter from Consolidated Edison, the local utility that produced the brochure, showed no change in their energy use compared to a control group that received no information.
The study also had a curious and more hopeful finding. Other apartments, randomly chosen to receive the same brochure along with a letter from the commissioner of the state's public utility regulatory commission, cut electricity use by 7 percent compared to the controls and the Con Edison group. Since air conditioners use only about 40 percent of household energy in the summer, the savings in terms of air-conditioning use was approximately 17 percent. This study shows that something_in addition to the_ information itself—something about the way information is __provided—can determine whether information works. In the next section, we return to the question- of what makes some information programs effective when so many others are not.
Some information programs, carried out both by gas and electric utilities and by government agencies, targeted much more difficult-to-take conservation actions, such as adding insulation to attics and walls or replacing energy-inefficient heating equipment. Such actions are often costly and many require major modifications to one's home. Put another way, these are actions for which there are major external barriers (limitations outside the individual). But there are also internal barriers, because people do not always know which actions are most important or how to take them.
Some of these information programs featured bill stuffer brochures, while others featured home "energy audits." As an example of the latter, consider a program started in 1977 by the Canadian government. The program, called ENER$AVE, offered all home owners a free computerized "energy audit." Participants filled out a questionnaire about their home, giving its age, size, form of construction, and other information. By return mail, each received a computer analysis with recommendations for home insulation, weather stripping, and other energy-saving actions, complete with estimates of the cost of each action, the energy and money that would be saved, and the "payback period"—the time it would take for the savings to repay the cost.
In late 1980, a group of Canadian professors of business administration (McDougall, Clayton, and Ritchie, 1983) surveyed a sample of homeowners, most of whom had completed the ENER$AVE survey about two years before. They asked whether the household had undertaken any of six energy-saving actions that were sometimes recommended by the ENER$AVE program: adding insulation in attic, walls, basement, or over unheated areas, installing weather stripping and caulking, or installing storm windows. If the household had taken any of these actions, the respondent was asked whether the action occurred within the past two years. The researchers assumed that if ENER$AVE was effective, the households that had participated would have taken more of these actions in the last two years than the comparison households. After excluding' actions that people reported they had done more than two years before, and which would therefore probably not have been recommended by the ENER$AVE audit, they found that households that had not participated in ENER$AVE reported having taken 45 percent of the energy-saving actions over the previous two years. The households who participated said they had taken 46 percent of the actions.
This is not much of a difference, and is too small to be statistically reliable. Of course, the study is not definitive. The ENER$AVE participants may have been more likely to have forgotten what changes they made in their homes (although there is no particular reason to expect this), and it is possible that the people who participated in the program—or their homes—differed from nonparticipants in some important respect that the study did not measure. However, this study finds about the same thing as studies of other computerized home-energy audits that used different research methods and asked different questions. This sort of information program appears to have little overall influence on how people use energy at home.
Why did ENER$AVE's computerized energy audits have so little effect? One possible answer that occurred to many conservation advocates was that the audits did not offer good enough information. When a homeowner says there is insulation in the attic, the computer cannot tell how much. Neither can it tell how well caulking or weather stripping has been applied. But if the energy audit is done personally, by a trained energy analyst, the computer can get better information. Moreover, the analyst can explain the recommendations and answer the homeowner's questions.
Following this logic, in the late 1970s U.S. gas and electric utility companies began offering customers free or low-cost on-the-spot energy audits. Soon afterward, the federal Residential Conservation Service program required the states to see that these audits were available to households at a minimal cost. Were these programs effective? Table 4-3 reports the results of two early evaluations.
These two programs appear to have been partially effective. They increased the frequency of a few energy-saving actions, but had no effect on most of them. More specifically, the energy audits increased the frequency of relatively low-cost behaviors (caulking, weather stripping, and modifying water heaters), but not expensive ones (insulating walls, ceilings, and floors). Apparently, the energy audits removed the information barrier to action, but not the external barriers that prevent householders from taking expensive energy-saving actions. Consequently, the only behaviors that changed were the ones for which information was the only significant barrier. The conclusion is hopeful in that it shows that detailed, accurate information can make a difference. But it is also discouraging in a larger sense. Success was only partial, and it required a significant investment of money and the time of trained personnel in interacting one-on-one with householders. Moreover, this effort failed to change the behaviors that have the greatest energy-saving potential, because these were precisely the ones with major external barriers (see Chapter I0).
Better Ways to Provide Information
We have seen that simply providing people with straightforward information has weak effects on only a limited set of behaviors. This section shows that behavior-change programs can be much more successful if they pay attention to the way they provide the information. The successful programs we describe in this section found ways to deliver information that caught people's attention and made the information credible.
Feedback. One approach to making information more effective is to tie it directly to people's behavior. Beginning in the 1970s, psychologists began experimenting with a method that, instead of telling people what to do to save energy, offered higher quality information about how much they were already using. The experiments provided regular, usually daily, feedback on how much energy a household was using and on what that rate of energy use would cost by the end of a month. Some studies used simple technology, for example, students reading electric or gas meters every day and leaving a note on the front door. Other studies used electronic monitoring devices, installed in a prominent place in the home such as on a wall near the kitchen sink, with the information made available automatically. Such devices are capable of providing feedback by the hour, minute, or second, but most of the early devices were not so advanced. Feedback systems provide information much more easily than reading a utility meter, and in a form that is personalized and easy to understand.
The theory of feedback is a simple application-of operant-Teirning_theory from_psychology (Skinner, 1938). 1f people are motivated to save energy, or to lower their energy bills, they will repeat whatever behaviors produce that reward. But it is difficult for people to tell which behaviors work because energy savings are not directly visible, and money savings are only realized once a month when the utility bill arrives—much too infrequently to help them learn what they have done to lower the bill. Feedback devices let people teach themselves how to save energy. In terms of learning theory, feedback acts as a signal of a reinforce—financial savings—that is slow in coming. Feedback provides much more specific and valid information th-in—a—general brochure or even an expert's energy audit because it is directly related to the householders' actual behavior and because it tells what people actually have saved, not only .aneslimatt_____ of what they might expect to save.
The effect of energy---use feedback depends on several factors. To change everyday behavior, it needs to be sufficiently frequent, and it is probably titost_effccfive if it is available immediately before and after people—have done something to try to save energy (Seligman et al., 1981; Shippee, 1980). It must be related to behavior in understandable ways. For example, feedback about energy used for home heating and cooling should be corrected for variations in weather (Winett and Neale, 1979). Otherwise, the large, weather-related changes in the need for heating or cooling can hide the effects of people's actions. It should also use units of measurement the householders can easily understand, such as dollars saved. And feedback is more effective when it concerns an energy source that is a large portion of the household budget (Winkler and Winett, 1982). That is, information works better when people have a strong financial motive to learn from it.
Overall, feedback experiments demonstrate under controlled conditions that real households during the late 1970s cut their energy use by around 10 percent immediately after feedback started and that the savings continued for at least several months, with feedback still being provided. The immediate savings indicate that the change was accomplished by altering behavior rather than by installing energy-saving equipment such as more fuel-efficient furnaces or appliances.
Although frequent feedback works, its effect is of limited magnitude and staying power. Because it operates mainly by getting people to use less, rather than by encouraging people to install equipment that can give the same comfort for less energy, the energy savings from feedback will sooner or later be perceived as sacrifices. (An argument has been made that annual or semiannual feedback may encourage people to install energy-efficient equipment, whose benefits can be seen most easily if they are averaged over a long period of time [Layne et al., 1988].) And feedback only works if the participants are strongly motivated. If the experiments were repeated in the mid-1990s, when there is no talk of a national energy crisis and when energy prices are no longer such a large portion of most people's incomes, feedback might be much less effective than it was in the late 1970s.
Modeling. One can also make information more effective by using a presentation that combines concepts from behavioral psychology and communication research. Richard Winett and his colleagues (1982) demonstrated a program that effectively reduced people's energy use without having them invest in new equipment or sacrificing comfort. The program featured twenty-minute videotapes of a young couple, much like most of the people in the Virginia apartment and townhouse complex where the experiment was conducted, demonstrating ways to save energy. For example, the tape on saving energy in the summer showed how to use fans and natural ventilation in the evening to save on air-conditioning, how to dress in lightweight clothing, how to shift the time and place of eating and cooking, and so forth. The script was carefully designed to present energy saving as a positive action. It used the visually compelling medium of television to demonstrate the desired behavior, and it employed the behavioral technique of modeling: the demonstrations were by people the audience could readily identify with and imitate. Participants in both the experimental and control groups in the study by Richard Winett and his colleagues (1982) also attended a forty-five-minute meeting in which they were instructed on the proper use of window fans, the insulating value of different items of clothing, and how to use a hydro thermograph installed in their homes to monitor temperature and humidity. Some of the participants were also given daily energy-use feedback for thirty days.
Compared with the group that only attended the meeting, the group that saw the videotape used 10 percent less household electricity immediately, and 19 percent less three weeks later. The savings for air-conditioning, which was the target of the program but is only a fraction of household energy use, were obviously much larger. Participants who also received feedback saved even more. The savings were accomplished with little or no change in indoor temperature, and the participants in the different groups reported the same levels of comfort. A companion experiment in the winter produced similar results. People saved more than 25 percent of the electricity used for heating. They did this mainly by lowering indoor temperatures, but because they were instructed in how to make the change slowly and to adapt with warmer clothing, they reported a level of comfort equal to that of the comparison group.
Winett's experiment demonstrates energy savings of over 20 percent from a carefully constructed information program. It is reasonable to ask, though, whether this sort of intensive effort, with meetings, feedback, and a specially created videotape with demonstration by models, is cost-effective. To answer this question, Winett's research group conducted another experiment in July 1982, this time using a local-access television channel to broadcast twenty-minute videotapes (Winett et al., 1985). People in the experimental groups were telephoned and asked to watch the program, which was broadcast four times over a five-day period. Their energy reduction was around 10 percent for the rest of the summer, compared to control groups (a reduction of about 23 percent of the energy estimated to be used for cooling). In a follow-up the next summer, the experimental group was still using 5 percent less energy, compared with the controls. The researchers concluded that this method could be cost-effective on a large scale, because once the videotape had been paid for (about $40,000), the cost of the program would be about $1 per household, for the telephone contact. If one million households could be reached and each saved $14 in a summer, as these households did, a $1 million program would save $14 million in energy.
As with feedback, this program achieves reductions in energy use by behavioral change rather than by improving technology, so the results may be hard to duplicate when people have lower levels of motivation, such as when energy prices or environmental concern are low or people are affluent enough to use electricity rather than sweaters to keep warm at home.
"Framing" Messages. Another way to make information more effective involves paying close attention to how pro-environmental behaviors are described. The program developed by Richard Winett's group provides an example: It referred to energy "efficiency" instead of "conservation" because Winett and his colleagues believed that their audience would perceive energy conservation as sacrifice, but would think of efficiency as a desirable goal. Another example is the experiment Suzanne Yates (1982) conducted in Santa Cruz, California, for her Ph.D. dissertation in psychology. She provided householders with information about the benefits of insulating their water heaters. When she presented the information in terms of how much money they were wasting by not insulating them, people became much more willing to insulate them than when she presented the information in terms of how much money could be saved. Of course, the two amounts were the same. Yates's experiment was based on the principle developed by Kahneman and Tversky (1979) that people are more sensitive to the prospect of losing something than to the prospect of gaining something of equal value. We discuss these ideas further in Chapter 9.
What the methods of feedback, videotaped modeling, and framing have in common is that they present information in ways that are particularly personalized, attention-getting, or motivating for the audience. Such methods can make educational programs appreciably more powerful. But even these methods do not overcome all the internal barriers that can prevent the expression of pro-environmental attitudes. The next section describes ways to tighten the links among attitudes, information, and behavior in order to make education yet more effective.
Gardner & Stern. (1996). Environmental Problems and Human Behavior. Boston: Allyn and Bacon