The mere suggestion that others are watching can put people on their best behavior, and a new study finds that concern for reputation is more powerful than cash payments in getting neighbors to do the right thing.
Researchers collaborated with the Pacific Gas and Electric Co., a California natural-gas-and-electricity provider, to understand how best to increase participation in a program to prevent blackouts. The plan involved installing a device to cut individual power use at peak periods. Enrollment could cause minor inconveniences by lowering the use of appliances like air conditioners during heat waves, but it would also prevent the much larger hassle of a system-wide power failure. Widespread participation could potentially reduce the need for new generators by 38% over 20 years and save $129 billion during that time, according to a report by the Edison Foundation.
Three times as many people participated in the program if they thought their neighbors would know who signed up; knowing that others might learn about their participation or avoidance was more than four times as effective as a $25 reward in increasing enrollment.
“Observability really matters,” says study co-author Moshe Hoffman, a postdoctoral researcher at the University of California, San Diego. “It has a much bigger impact than offering financial incentives in terms of promoting good behavior.”
The experiment presented its 2,413 participants with a classic dilemma in cooperation: everyone benefits if some accept small costs to help society, but those who don’t sacrifice at all can still benefit if others do. Throughout evolutionary history, social species that cooperate have faced such problems, and it makes the concept of cooperation appear to be a highly unlikely development. In theory, the cooperators would lose out to those who were profiting without participating.
“Evolutionary theory points out that there is real puzzle: How is it possible that people and other animals evolved to behave cooperatively?” says Hoffman.
One part of the answer is reputation. If individuals can be identified and punished for cheating and rewarded for cooperating, most will do their fare share — or at least try to appear to have done so. If humans evolved cooperative behavior in this way, people should have unconscious impulses to demonstrate good behavior when they know others are watching.
(MORE: Should We Give to Get Ahead?)
To test these ideas, the researchers compared the number of people whose enrollment involved writing an anonymous number on the sign-up sheet — so neighbors couldn’t tell who was participating — to the number who signed up when they had to use their real names. The use of a real name tripled participation.
The authors also compared participation among people who lived in apartment buildings, where the sign-ups sheets were in common areas like mailrooms, with that among people who lived in row houses, where the names of participants would only be seen by a minority of other residents. “To extent that reputations are driving results, you would expect the biggest effect to occur when people would be observed by more of their neighbors,” Hoffman says.
Owners — who are likely to have stronger ties to a community — were also compared with more-transient renters. The results showed those with deeper roots to the area also had a greater concern about reputation.
And, when fliers inviting participation were sent to an additional 1,005 Californians, the study found that being observed didn’t matter when the participation was not presented as an altruistic act. In other words, if participating didn’t involve personal sacrifice to benefit society, people didn’t care about what others thought of their choice.
“What ties it all together is [the question of] when should observability matter,” says study co-author David Rand, an assistant professor of psychology at Yale. “[It’s] when people you care about can see whether you are doing [it] or not and everyone thinks it’s something you should do.” Facebook badges saying “I voted” or pins touting a recent blood donation also take advantage of people’s pursuit of a good reputation to increase participation.
However, an earlier study found that another power-company program — one that let people see how much they consumed compared with their neighbors — lowered consumption among liberals, but may have raised it in conservatives. This suggests that reputational concerns vary depending on values. The conservative customers may have wanted to show that they are big consumers, while liberals focused on preventing global climate change, for example.
It’s not clear from the study whether people participated because they were afraid of shame and embarrassment if they didn’t, or if they sought to bolster their reputations as good neighbors. But the authors suspect shame isn’t a factor: since the sign-up sheets were removed every few days, no one could be sure who didn’t participate — they could only know definitely that someone did.
The research also offers another factor to consider in the debate over Internet privacy. “There’s been a lot of discussion about the trade-offs between privacy and security,” says Rand. “What our study shows is that there is another, similar trade-off in the dimension of prosocial behavior and cooperation, where a cost of privacy is increased selfishness.”
As in the debate over security, the answers probably don’t lie at either extreme. “I don’t want to live in a society where everything is totally observable, but I also don’t want to have no idea what anyone else is doing,” Rand says.
The human drive to cooperate is real — but it may sometimes hinge on concerns about how we are seen.