In 2006, Melissa Bateson and Daniel Nettle—two researchers at Newcastle—published, along with one other colleague, the results of an ingenious experiment involving their coworkers and the honesty box. Next to the instructions for payment they placed a banner that alternated each week between an image of a pair of eyes and an image of flowers. The eyes varied in sex and in head orientation, but all were chosen such that they were pointing directly at the person making the drink. The tea and coffee drinkers who served as guinea pigs in this experiment had no idea why the banner was up or why the images were switched every week; they probably thought it was some kind of game made up by colleagues with too much time on their hands. Bateson and her colleagues made sure that supplies of tea, coffee, and milk were maintained to keep up with demand, and every week they measured the volume of milk consumed as a proxy for total beverage consumption. Then they calculated the ratio of money collected in the box to the volume of milk consumed in that week to take into account the fact that people didn’t always drink the same amount of coffee or tea with milk every week. The researchers discovered that the amount of money people put in the box fluctuated from week to week, as the images alternated. People paid nearly three times as much for their drinks when eyes were displayed on the banner—the eyes on the wall seemed to exert an automatic and unconscious effect on the drinkers’ perception that they were being watched while they paid for a drink, which made them more generous.
Figure 5.1. Cues of Being Watched Enhance Cooperation in a Real-World Setting. Photo by Dr. Melissa Bateson. Modified after Figure 1 in Bateson, Nettle, and Roberts (2006). Reprinted with permission of the Royal Society of London.
The honesty box experiment at Newcastle University was actually inspired by a study conducted a few years earlier by two researchers at the University of California–Los Angeles named Kevin Haley and Daniel Fessler.2 Haley and Fessler asked over two hundred undergraduate students to come to their research laboratory and play a computerized economics game known as the Dictator Game. In this game, player 1 receives a particular amount of money and is told that he can give player 2 any fraction of this amount. Player 1 has full control over the decision to share the money (hence the name), and player 2 has no active influence on the decision. It’s not a real game, like the Prisoner’s Dilemma, but simply a decision-making test that taps into people’s generosity.
You might think that selfish people would keep all the money and give nothing to the other player, but in reality most people playing this game choose to share some of their money. Haley and Fessler paired each student with another student and randomly designated them as either player 1 or player 2. Player 1 received $10 and had twenty seconds to give player 2 part of it—any amount between $0 and $10—while keeping the remainder. The players were told that no one would know the decisions they made and that the game would be played only once. Neither student could see or communicate with the other, so the game was played under conditions of strict anonymity. Unbeknownst to the students, the desktop background of the computers used during the experiment was being manipulated by the researchers: half the time the players used a computer whose desktop background showed two stylized eyelike shapes (eyespots), and half the time they used a computer in which the word CASSEL was displayed across the same portion of the screen, using the same color scheme as the eyespots.
It turned out that the students who served as player 1 donated 55 percent more money to the other player when the desktop background had the eyespots (on average, they donated $3.79 of their $10) than in the other condition (in which they donated, on average, $2.45). As with the experiment at Newcastle University, the authors of this study suggested that the unconscious perception of being watched made people more generous with their money. In the last few years, many more studies have been conducted to examine the effects of “eyespots” on honesty, generosity, and cooperation. With a few exceptions, the studies have confirmed the results obtained by Haley and Fessler and by Bateson and her collaborators.
Figure 5.2. Nobody’s Watching? Subtle Cues Affect Generosity in an Anonymous Economic Game. Photo by Drs. Kevin Haley and Daniel Fessler. Modified after Figure 1 in Haley and Fessler (2005). Reprinted with permission of Elsevier.
So why is it that photos of eyes—or even stylized images of eyes—result in the perception of being watched? Can’t people tell that they’re just pictures and that no one is really watching? Of course they can—if they think about it consciously. But if they barely notice the eyes, the images automatically activate brain responses that unconsciously influence their decisions. Many animals, including humans, are instinctively attuned to eyes and eye gaze direction.3 For some animals, detecting the eyes of a predator means the difference between life and death. Fish, for example, are more likely to flee from objects that resemble the eyes of a predator than from similar objects that don’t. Some butterflies have large spots on their wings so that if they are about to be eaten by a bird, they can spread their wings and startle the bird with what looks like the eyes of a large animal. For highly social animals, such as domestic dogs, monkeys, and apes, detecting the eyes and assessing the eye gaze direction of a dominant individual can play an important role in avoiding aggression. Dogs avoid stealing forbidden food if they see that a person is watching them, but do it promptly if the person has her eyes closed, has her back turned, or is otherwise not looking at the dog. When rhesus macaques detect the gaze of a higher-ranking group member or a human on them, they bare their teeth and rapidly smack their lips to signal submission. Low-ranking chimpanzees avoid taking food they want if they can see another individual’s face and eyes.
Humans, of course—being both highly visual and highly social creatures—are more attuned to eyes and eye gaze direction than all other animals put together. Human babies who are a few days old prefer to look at drawings of faces—even if these faces only have eyes—than at any other images. In our everyday social lives—say, when we ride in an elevator with another person—we continuously monitor the gaze direction of other people and process information about their identity, facial expression, and even emotions and motivation. It turns out that in both monkeys and people certain brain cells are automatically activated by the perception of eyes and their orientation or by objects that resemble eyes. For example, some cells show spikes of electrical activity in response to a picture of an individual with opened eyes. Some cells in the brain of a macaque respond even more strongly when the monkey is looking at a picture of another monkey gazing at the camera directly than when the monkey in the picture is looking somewhere else. Most of these cells are located in areas of the brain called the inferior fusiform gyrus, the superior temporal sulcus, and the amygdala. Brain activation in response to eyes and eye gaze is largely automatic and involuntary. This explains how even the image of eyes on a wall or on the desktop background of a computer can lead to an unconscious perception of being watched.
And why does the perception of being watched make people more generous or less likely to cheat? A quick answer to this question is that the areas of the brain that process the perception of being watched are closely interconnected with other brain areas that are involved in making decisions—like decisions about whether to be honest or to cheat, or about how much money to donate to someone else. So upon seeing eyes on the walls or on our computer desktop, the being-watched parts of our brains send electrical signals to the decision-making parts.
The long answer will take up the rest of this chapter. In the first half, I show that when people know or think that their identity is known, they tend to help others and act in ways that establish or strengthen a reputation for being generous, cooperative, or trustworthy. In the second half, I show that when people know or think that their identity is anonymous, they become more competitive than cooperative and tend to hurt rather than help others.
As in the previous cases, evolutionary biology and economics help explain why we behave the way we do. Social influences on decision-making processes,
however, are something that economists have only recently incorporated into their models. Economists used to think that people always make rational choices that maximize their gains, and that they make choices in isolation from their social context, without regard for the consequences of their behavior. Researchers studying animal and human social behavior from an evolutionary perspective, however, have discovered that maximizing one’s fitness—like maximizing one’s earnings—often depends on taking others into account. As we’ll see, the integration of economic models with evolutionary theory and the findings of animal and human behavioral research can result in more sophisticated and more predictive models of human decision-making and ultimately help bridge the gap that still separates economic and biological explanations of behavior.
The Altruist in the Spotlight
HOW TO CHOOSE TRUSTWORTHY PARTNERS FOR COOPERATION
In the last chapter, I mentioned that in both nonhuman primates and humans issues of trust are very important when choosing allies. Imagine a middle-ranking macaque male who solicits the help of the alpha male in a coalitionary attack against the beta male in the group (an example of a bridging coalition). After weeks of planning and scheming, the middle-ranking male finally launches his attack on the beta male, and the alpha male initially helps him. If, however, the alpha male changes his mind midfight and switches sides, the consequences of his defection for the middle-ranking male are disastrous. The beta male will naturally want to retaliate, and with the alpha male now on his side, he will make dog meat of the erstwhile challenger. So, whether you are a macaque or a human, choosing a trustworthy coalition partner is crucial for the success of any political enterprise. The question is how to find one.
Our family members generally make good coalition partners—think about how the Bush family members have helped one another’s political careers—because we have shared genetic interests and we can estimate, knowing how they behaved in the past, whether or not they can be trusted in the future. But how can we know whether an unrelated individual we’ve just met is trustworthy? To answer this question we need to revisit our discussion of cooperation and the Prisoner’s Dilemma in Chapter 2.
If we play a single round of the Prisoner’s Dilemma with a stranger, knowing whether the stranger has previously cooperated or defected with others can influence our move. Remember that, in the absence of any information, it’s always safer to defect, but if we have information that leads us to think that the other player will cooperate, then it’s better to cooperate because both players can then win big. So when trying to establish whether a potential cooperation partner is trustworthy, his or her reputation can make a big difference. Our own reputation, obviously, is equally important to the other player. Anyone with ambitions for a political career—which, by definition, requires cooperation with other individuals—should make every possible effort to establish a good reputation before running for office.
In the Prisoner’s Dilemma, two people accused of committing a crime can both avoid jail time if they cooperate and tell a consistent story when interrogated separately by the police. The importance of reputation for cooperation has also been demonstrated with other economic games—such as the Dictator Game used by Haley and Fessler—in which people need to make decisions about sharing money with another person.4 A player is more likely to be generous if he or she is told that the other player has a reputation for generosity. A player is also more likely to donate money to another player when by doing so he or she can establish or strengthen a reputation for generosity, which may be useful for future cooperation. In fact, studies have shown that in both the Prisoner’s Dilemma and the Dictator Game players are more likely to cooperate or be generous if they know each other’s identity than if identities are kept anonymous. As we’ll see later, having a good or bad reputation can play a major role in all human affairs that require cooperation, from joint business enterprises to politics to romantic relationships.
It turns out that reputation matters in animal affairs as well. Take the example of cooperation between the large coral reef “client” fish and the tiny “cleaner” fish that swim inside their mouths to clean them. To evolutionary biologists who study animal behavior, this is a textbook example of mutualism—an interaction that benefits both parties. The large fish get their teeth and gums brushed without paying exorbitant dentist’s fees, and the little cleaner fish get free food without being eaten by their clients. As is always the case when there is cooperation between two parties, however, there is also the potential for conflict of interest, which may lead to cheating. For example, the cleaner fish could cheat by biting pieces of healthy tissue off the walls of the client’s mouth instead of just eating the dead tissue and parasites, and the clients could cheat by closing their mouths and swallowing the cleaners as soon as they are done with their oral hygiene session. According to University of Cambridge biologist Redouan Bshary—now at the University of Neuchatel in Switzerland—cleaners do sometimes cheat, and this becomes clear when the client makes short jolts in response to being bitten. It’s sort of like saying, “Ouch!” when your dental hygienist pokes your gums—unintentionally, I hope—with one of her sharp instruments.
Bshary showed that clients choose cooperative rather than cheating cleaners, not on the basis of personal experience, but on reputation.5 They observe cleaners at work on other clients and make a mental note of whether the clients jolt from a bite. When the time comes, they then approach cleaner fish that behave themselves and avoid cleaners that they have observed cheating. Cleaners that cooperate or cheat, therefore, acquire a good or a bad reputation, respectively, among their clients. According to Bshary, however, the cheating cleaners find ways to cheat about their reputation and therefore eventually beat the system. They behave themselves in the mouths of small client fish that don’t have fleshy mouth tissue, only to cheat the larger clients after being chosen by them as oral hygienists. In the jargon of evolutionary biologists, cheating cleaners cooperate in low-payoff interactions to build a reputation that allows them to exploit other individuals in high-payoff interactions. I discuss the cleaner fish and their clients, and more generally the issue of finding good partners for cooperative enterprises, in more detail in Chapter 8.
WHY IT’S SO DIFFICULT TO PROTECT THE ENVIRONMENT
As nicely illustrated by the honesty box experiment conducted at Newcastle University, in everyday social situations people’s tendency to contribute to a common enterprise, such as the shared cost of purchasing coffee for the espresso machine, may be influenced by events or factors that they perceive as improving or damaging their reputation—such as being watched while being generous or cheating. Although the issues involved in deciding to cooperate or defect in the Prisoner’s Dilemma and to make or not make an honest contribution to the Espresso Club are similar, there is an important difference between these two scenarios. In the former, the dilemma lies in the potential conflict between the interests of the two prisoners. As we saw earlier, this model applies to all situations in which cooperation is required between two individuals, such as in the formation of an alliance against a third party or the mutualistic interaction between a client fish and a cleaner fish. The dilemma involved in deciding whether or not to contribute money to the Espresso Club, however, lies in the conflict between the interests of an individual and those of the group. Obviously, the group does best when all players cooperate and everybody pays for their coffee. However, economists and evolutionary biologists tell us that, unless the individuals who don’t cooperate are identified and punished, nobody should ever contribute anything to a public enterprise. To prove that their game theory models are correct and their cynicism is justified, they point to the results of experiments conducted with so-called public goods games.
The classic public goods game consists of four or five players who are given the options of investing a token in either a private good—for example, keeping the token for themselves and saving its entire value—or a public good such as contributin
g it to a public pool. After contributions are made, the content of the pool is doubled—this provides an incentive for cooperation—and is then divided by the number of players and evenly paid out, irrespective of each player’s contribution. If all of the players contribute their tokens to the public pool, they all win and make the maximum possible return from their investment. However, if one or more players behave selfishly and keep the token for themselves (in the language of game theory these individuals are referred to as free riders), then contributing to the pool is no longer recommended for anyone. In fact, mathematician John Forbes Nash—the man with a Beautiful Mind, played by actor Russell Crowe in the film—developed mathematical models showing that each token paid into the pool yields a return of only a half-unit to the contributor and therefore nobody should ever contribute anything to the pool.
Consistent with the predictions of the models, experiments have shown that when multiple rounds of the public goods game are played, players initially cooperate but then gradually become more selfish as they are increasingly tempted to cheat and take advantage of the system. People’s failure to contribute to public goods explains why managing a limited public resource, such as pollution-free air or water, is so difficult. An American ecologist named Garrett Hardin aptly termed this phenomenon “the tragedy of the commons” in an influential article published in the journal Science in 1968.6 The tragedy is that, when it comes to protecting the environment and other similar matters, the individual’s interests prevail over the public ones and in the end everyone loses.
Games Primates Play: An Undercover Investigation of the Evolution and Economics of Human Relationships Page 14