Out of Character Part 3

How many black bears live in the forest?

Steve and Phil strolled into the lab. Steve was an undergraduate partic.i.p.ating in the experiment for course credit, and as far as he knew, so was Phil. But-surprise-Phil was actually a research a.s.sistant playing the role of a partic.i.p.ant. The experimenter entered the room and told Steve and Phil that the first thing they would have to do was fill out a questionnaire that would categorize them as one of two personality types: someone who habitually overestimates or someone who habitually underestimates things. But in fact the questions we asked them had little to do with their personality, and we weren't interested in their answers to the questions at all. Who would be? We purposely asked them to estimate trivial things like the length in miles of the Ma.s.sachusetts Turnpike, the number of black bears that lived in Ma.s.sachusetts forests, and the height of the John Hanc.o.c.k building in downtown Boston. The goal was simply to create a sense of similarity or difference-"us" and "them"-based on a completely new and meaningless criterion. After answering the questions, Steve and Phil waited as their respective computers "calculated" the results (in reality they were random). Half the unwitting partic.i.p.ants were told that Phil was an over- or underestimator just like them, while the other half were told that Phil was the opposite type of person. Nothing more. No "You two have the same taste in music" or "You both have a pa.s.sion for American history." In fact, no personal information whatsoever was given. This was on purpose. We wanted to ensure that our manipulation of similarity was as trivial as possible.

The idea, of course, was to see if perceptions of similarity about something that didn't have any preexisting significance attached to it would have an effect on how much compa.s.sion partic.i.p.ants would feel for a victim and how much they would be willing to help that victim. More specifically, how would Steve react if he witnessed a transgression against Phil? Would his actions vary depending on this arbitrary measure of similarity?

To find out, we brought back our bad guy, Alex (he was just so good at it), and had our real partic.i.p.ant, Steve, play the role of the "secret watcher" described in Chapter 2. In other words, Steve secretly observed Alex as he a.s.signed poor Phil to a long and difficult task while keeping the easy and fun task for himself. As in the earlier experiment, Alex had the option to flip a virtual coin to decide fairly who got what, but he didn't even try. Now came the important part. How much compa.s.sion would Steve feel for Phil, and how willing would he be to help Phil with his distasteful (and unfairly a.s.signed) task? While Phil began working, Steve responded to a series of questions tapping his sympathies for Phil's predicament. Then, just as Steve was about to leave the experiment, the following message popped up on his computer screen: You have now completed the experiment. Please go to the experimenter to receive your credit. As you know, one of the other partic.i.p.ants in the experiment has a long and difficult task to complete. It's not important to the experimenters who complete this task, it is just a quant.i.ty of material that needs to get done. So, if you'd like to help out in any way, indicate as much to the experimenter on your way out.

At this point, Steve could either hightail it out of the lab and go about his day or find the experimenter and offer to help. There was no social pressure since no one asked Steve to help-not Phil and not us. If Steve wanted to help, he had to take the initiative; otherwise, he could be on his merry way.

Now, if any of our partic.i.p.ants did seek out the experimenter and tell him they'd like to take some of Phil's burden, they would be escorted down a long hallway and around a corner and placed at a desk, where they would be presented with a stack of about thirty math GRE problems. The experimenter would then tell them to do just as many as they wanted-whatever they didn't finish, Phil would complete later. Furthermore, once they were done, they were just to leave everything on the desk and take off. In other words, it was made clear that they would never again interact with the experimenter or with the person they were helping. So, in the case of people like Steve, any help that they did lend would have to be motivated by a legitimate desire to relieve Phil's suffering, as opposed to an attempt to gain any social rewards from the experimenter or Phil for the actions. At this point, the experimenter left them to their good deed, but secretly timed how long they spent working on this difficult task.

What did we find? Were our partic.i.p.ants more likely to help Phil with the problems when they believed Phil shared this meaningless label as a fellow over- or underestimator? Not only was our suspicion correct, but correct to a larger degree than we ever imagined. A mere 16 percent chose to come to Phil's aid when he was dissimilar to them, but 58 percent-more than three times as many-chose to help him in the exact same situation when he was perceived as more similar. What's more, not only did many more people choose to help when Phil's "estimating type" matched theirs, they also spent significantly longer periods of time doing so than did the 16 percent who believed Phil was different from them but agreed to help anyway.10 It may seem a bit disconcerting that we were able to manufacture compa.s.sion and altruism in the lab with no more than a silly little tale about whether people tend to make similar types of guessing errors. But if this is all it takes to tip the scale of character one way or the other, it certainly goes a long way toward explaining how acts of beneficence or cruel indifference can fluctuate, even in the same person, in the blink of an eye. Still, one might argue that this experiment wasn't a perfect replica of a real-life situation. After all, the manipulation we used, though subtle, was explicitly defined. We told people who was in their group and who wasn't. Given how many social norms there are surrounding group membership (such as "take care of your own"), people might have just been responding in the way they thought they were supposed to act ("This guy is like me, so I'm supposed to feel bad for him and help him more"). We thought this unlikely, but just to be sure, we decided to rule out this possibility in our next experiment. To do this, however, we needed a way to make people perceive similarity on their own, without us applying artificial labels, and then show that this perception affected how much compa.s.sion they would feel and how much altruism they would display.

Moving as one.

In describing his experience in basic training after being drafted into the army in 1941, historian William McNeill writes of the long, grueling hours he and his fellow soldiers spent marching single file about the dusty plains of Texas.11 Such drills might seem idiotic to the onlooker. After all, if a group of soldiers were to march in unison during battle, a machine gun would quite easily mow them down. So why were drill sergeants so keen on having their soldiers practice day after day, until they were moving perfectly in sync with one another? It wasn't until later in life that McNeill began to see a possible reason for such an exercise. In hindsight, despite the heat and the fatigue, his recollection of the marching drills was one of pure enjoyment and camaraderie. The act of moving in time with others, he recalled, led to a "strange sense of personal enlargement"-a feeling of connectedness with those around him. Lest you think perhaps McNeill had been out in the desert sun too long, research has begun to suggest that moving in synchrony can actually make people feel closer together. It acts as a kind of social glue, binding individuals into a larger whole.12 This was just the feeling we wanted to create in our experiment. We thought the reason being physically in sync with another person forges a bond is because it makes people feel more similar to each other. If this were the case, then we should be able to make people actually feel more similar to another person (rather than us just telling them they were similar) simply by having them mirror the person's movements. No questions about black bears and no explicit labels would be needed. The mind's attention to synchrony would be enough.

Given this theory, we conducted more or less the same experiment as before, but this time we simply told people that the first part of the experiment involved rhythm perception. The idea was that we would make some partic.i.p.ants tap their hands in sync with Phil, the "victim," while others would tap to a different rhythm. We suspected, of course, that the synchronous tapping would create enough of a sense of similarity to make partic.i.p.ants feel more compa.s.sion and offer more a.s.sistance to Phil.

And indeed it did. First of all, those who tapped in sync with Phil readily reported feeling significantly more similar in personality to him on a survey than did those who tapped to the different beat. For reasons that they couldn't possibly articulate, simply moving their hands in unison was making our partic.i.p.ants feel more connected. Next came the big question. Would this be enough to change the level of compa.s.sion people felt toward Phil? It was. Forty-nine percent of those who tapped in sync with Phil volunteered to come to his aid, compared to only 18 percent of the asynchronous tappers.13 Plus, the more similar they felt, the more compa.s.sion they experienced, the more willing they were to help, and the longer they spent helping him with the onerous task.

The implications couldn't have been clearer. Feeling similar to another person appears to trigger our humanity. It signals to us that these are the people who likely will be there for us in the future, tagging a person as someone we should care about-someone we need to care about in order to ensure our long-term success. It says that this is someone the ant should be particularly interested in interacting with, and for whom it should fight hard against the gra.s.shopper, which is urging us to avoid the short-term costs a.s.sociated with helping. This seems a highly adaptive system all around, right? Our future success is optimized and the other person reaps the benefit of our compa.s.sion. But it raises an interesting question. What happens when the switch gets flipped the other way?

On the outside looking in.

So far, we've looked at what happens when we perceive someone as being "like us." But what happens when we judge another person to be "unlike us"? According to our theory, dissimilarity signals that a person is someone you need not go out of your way for, at least from an evolutionary perspective. It's an ancient signal that your efforts and care might well be wasted upon this individual, who is unlikely to reciprocate your help or compa.s.sion in the future. Enter the gra.s.shopper. But in protecting us from wasting our compa.s.sion on those unlikely to return the favor (which can be a good thing), the gra.s.shopper occasionally renders us capable of committing callous or horrific acts of aggression or cruelty.

The tools the gra.s.shopper has at its disposal are powerful. It will go to great lengths to convince you that all that matters is your immediate well-being; that any cost of caring for or helping others is to be avoided. You know that voice. It's the one telling you that those few dollars you were about to put in the homeless person's cup could buy you a latte. But how does the gra.s.shopper switch off that nagging voice in our heads (i.e., our conscience) telling us that the homeless man probably needs food more than we need our morning Starbucks fix? By convincing us that the person in need of our help is not like us, or not even a human at all. Dehumanizing someone, stripping them of their ident.i.ty as being capable of thinking and feeling and reacting as we do, makes it particularly easy to ignore and transgress against them. It seems almost unconscionably cruel, but there's a growing body of research suggesting that when we perceive another person as "not like us," this is exactly what we do.

History is rife with examples of this. The writers of the U.S. Const.i.tution defined slaves as three-fifths of a person. The n.a.z.is described Jews as "vermin," and the Rwandan Hutus described the Tutsi as "c.o.c.kroaches."14 Almost every time one group has treated another horribly, they've found some way of dehumanizing their victims. And while these examples might seem so extreme to suggest that dehumanization is confined to the realm of madmen and sociopaths, that a.s.sumption would be incorrect. On smaller scales, any of us are capable of it. Consider the following story.

On a cold December night in 2005, patrons were waiting in line at the Starbucks near Nineteenth and Cambie Streets in Vancouver, Canada. As the scent of smoke began to overwhelm the powerful aroma of ground coffee beans, most people inside kept chatting, quite indifferently. Outside the coffee shop, it was much the same, with several customers calmly shivering over their lattes and chatting on their cell phones as black smoke began to billow above their heads. Just one customer, concerned about the blaze, peered around the corner and noticed an unconscious homeless man wrapped in a comforter that had somehow caught fire. As the flames crept higher and higher toward the man's face, another customer, who happened to be a nurse from nearby St. Paul's Hospital, tried to wake him, to no avail. The nurse then tried to recruit others to help her get the man to a hospital, but no one seemed to care enough to interrupt their conversation or their newspaper. One woman turned to the nurse and said with disdain, "Just leave him alone, he's a homeless person. Forget it." Another said, "Don't call the hospital. They don't want him."15 Certainly this is a particularly egregious display of cruelty, and it's tempting to a.s.sume that these callous folks are exceptions, rather than the rule. But is their cruel indifference to the well-being of the homeless really that abnormal? Consider what you feel when you pa.s.s a homeless person in the street. Do you always give cash? Or do you a.s.sume that giving won't matter, or that your money will probably go to feed a drug habit anyway, and so you walk right past? What if that homeless person was seriously hurt or looked unconscious? Do you call the police, or do you keep on walking, knowing that you'll likely be late to work if you wait for an officer to arrive? Most of us have done both-sometimes we helped, but other times we didn't. How could we be expected to never do the latter? If you walk down a city block wanting to help every needy person you see, you wouldn't make it very far and your bank account would quickly dwindle. By the same token, it simply isn't possible to contribute to every charity, join every cause, or even expend mental energy feeling compa.s.sion for everyone who needs it.

Given this fact, we need some sort of mechanism that turns these feelings of compa.s.sion off, lest they completely overcome our lives. Dehumanization seems to do this. In order for us to absolve ourselves for our callousness, our inner gra.s.shopper, in looking out for our pleasure and resources in the moment, tricks us (albeit subconsciously) into seeing dissimilar others as objects instead of human beings. The callous woman in that Starbucks most likely didn't consider the homeless person to be much different in kind from the comforter in which he was wrapped. When we see others as objects instead of fellow humans capable of feeling and experiencing the world as we do, their welfare becomes inconsequential. And if you think this phenomenon is limited to some coldhearted scrooges, we've got some bad news for you. The tendency to dehumanize seems to be a fundamental part of our psychology, and a necessary one at that.

In one particularly compelling study, Lasana Harris and Susan Fiske at Princeton University used brain scanning technology to investigate how people would respond to different kinds of social groups.16 Interestingly, there are distinct areas of the brain that are activated when evaluating humans, and others that are activated when evaluating objects. While in the fMRI machine, partic.i.p.ants were shown pictures of people from a variety of different groups: the elderly, the disabled, Olympic athletes, the homeless, drug addicts, the rich, the middle cla.s.s, and so on. The researchers were interested not only in how people described their emotional responses to these groups but also in whether the subjects' brains would process images of certain individuals differently.

What Harris and Fiske found was somewhat shocking. When people saw images of those who belonged to what sociologists consider extreme out-groups (such as drug addicts and homeless people-those who we think are most unlike us), the social categorization areas of their brains (the ones that are involved in making judgments about humans) were quiet, while the areas involved in processing objects lit up like fireflies. Their minds, in essence, responded to these people not as if they were people but rather as if they were things. Even more surprising, this wasn't just an intuitive response. People actually reported strong feelings of disgust upon viewing the images of these outgroups, and, when asked to pick objects that best represented how they felt about the drug addicts and the homeless, they chose images like vomit and overflowing toilets. Moreover, the same areas of the brain that responded to the pictures of vomit and overflowing toilets responded to the pictures of the homeless and drug addicts. Given this, it's no wonder that many people, even those who at times seem the most caring, don't always feel the pain of and help those who may need it the most. When we perceive others to be so dissimilar from us, the parts of our brain that are responsible for treating others with humanity can turn off, allowing us, for better or worse, to numb ourselves to their plight.

Again, the evolutionary calculus behind this is simple. The less similar another person is to you, the less likely he or she is to care about your well-being and thus the less likely to reciprocate your kindness. The less likely the person is to reciprocate, the more appealing the urgings of the gra.s.shopper become.

Consider the following vignette:.

Two days ago I broke up with my (girlfriend) boyfriend. We've been going together since our junior year in high school and have been really close, and it's been great being at FSU together. I thought (s)he felt the same, but things have changed. Now, (s)he wants to date other people. (S)he says (s)he still cares a lot about me, but (s)he doesn't want to be tied down to just one person. I've been real down. It's all I think about. My friends all tell me that I'll meet other (girls) guys and they say that all I need is for something good to happen to cheer me up. I guess they're right, but so far that hasn't happened.

When people read this kind of story they tend to express some feelings of compa.s.sion or sympathy for the person. But in one interesting study, a group of researchers led by Roy Baumeister at Florida State University found that if you made people feel socially isolated before exposing them to the story, it would decrease their sensitivity to the plight of those around them. To demonstrate this, they created a clever (though somewhat harsh) experiment. They had partic.i.p.ants complete a bogus personality questionnaire and then told some of them that, based on the results, they were the type of person who most likely would not be able to develop any meaningful relationships later in life and thus would end up alone. Ouch.

Turned out that the people led to believe that they would become socially isolated did indeed care less about the plight of the girl in the story. Not only that, it also made them less likely to engage in any prosocial behavior in general, and even made them less sensitive to emotional and physical pain.17 In short, it numbed them. It seems that when the possibility of developing beneficial long-term relationships is removed, either because the person in need doesn't appear to be the type of person who is worth your efforts (i.e., is dissimilar to you) or because you have reason to believe that you are unlovable and so your efforts would be fruitless, the scales tip toward the gra.s.shopper and your impulse to care about the suffering of others switches off. If you can't count on anyone besides yourself, you might as well live only for yourself, right?

Again, this may seem as if the ant governs what's "good" (compa.s.sion) and the gra.s.shopper governs what's "bad" (indifference), but remember, nothing is that cut-and-dried, and both cla.s.ses of responses are absolutely essential to a successful social life. There are instances when seeing others as different may be necessary, such as when it protects us from being taken advantage of by others who would bleed us dry, from feeling others' pain so intensely that we are sad all the time, or from being unable to aggress in an armed conflict when much is at stake. By the same token, unfettered compa.s.sion can lead us astray; indiscriminately feeling for and helping all those around us is a one-way ticket to being the biggest sucker on the block and is simply not a tenable strategy over time. What's important to remember, though, is that your conscious mind isn't always the one that is making the decision of who is worthy of your compa.s.sion or of your disdain. The ability to turn our n.o.ble feelings on and off is a fundamental property of the human mind. All it takes is a lightning-fast a.s.sessment of another person to determine whether we'll care deeply about him or her or be callously indifferent to the person's misfortune-an a.s.sessment that can be guided by all sorts of seemingly meaningless contextual variables, with our character hanging in the balance.

6 / FAIRNESS AND TRUST.

The surprising elasticity of the Golden Rule.

Mohammad Sohail was closing up his convenience store in Shirley, New York, much as he does every night. But the night of May 21, 2009, is one he will never forget. As Sohail was tidying up behind the counter, a hooded man menacingly brandishing a baseball bat burst into his store, demanding cash. But as Sohail dropped his hands below the counter, he wasn't reaching for the cash box. To the thief's surprise, Sohail pulled out a 9 mm rifle and pointed it directly at the a.s.sailant's head. The man, apparently fearing for his life, collapsed to his knees, sobbing. He pleaded, "I'm sorry. I have no food. I have no money. My whole family is hungry." In that moment Sohail suddenly saw the would-be robber as someone who needed help-help he could provide. Sohail asked the man to promise that he would never rob again, and when the man did so, he gave him $40 from the cash register and a loaf of bread. Sohail then went to the back of the store to get the man some milk, but on his return the man had disappeared. Feeling good about having done a n.o.ble deed, Sohail a.s.sumed that would be the end of it. He was wrong.1 In early December of that same year, Mohammad Sohail was opening his mail when he came across a letter without any return address. As his eyes skimmed the page, he quickly realized the sender could only be one person. The letter read: First of all I would like to say I am sorry at the time I had [no] money, no food on the table, no job, and nothing for my family. [It] was wrong but I had [no] choice. I needed to feed my family. When you had that gun to my head I was 100% that I was going to die.... Now I have a new child and good job, make good money staying out of trouble and taking care of my family. You gave me forty dollars. Thank you for sparing my life. Because of that you change my life.

But the letter wasn't the only thing in the envelope. The reformed thief had also enclosed $50-the $40 Sohail had given him plus an extra bonus. Now, no one had been holding a gun to the man's head forcing him to repay Sohail's "loan." In fact, not even Sohail himself expected the money to be repaid, and certainly not with interest. Sohail's kind act-his taking a chance to help without any certainty that it would be valued or repaid-had not only brought about a major change in the recipient but also resulted in a profit. "When you do good things for somebody, it comes back to you," Sohail proclaimed.2 It seemed the robber had decided to follow the Golden Rule.

What makes people go from stealing to repaying (with interest), or from acting selfishly to treating others as they would hope to be treated in return? Where does the Golden Rule come from, and what makes us decide whether or not to follow it? The answer strikes at the heart of what it means to be human. No matter where we live or what culture we belong to, human beings depend on the exchange of time, resources, and social support with each other. Whether it's asking a friend to help us move our furniture, a loan officer to give us a chance to fund a new business venture, or a neighbor to watch our kids for a minute so we can run down the street to the pharmacy, we rely on the a.s.sistance of others far more often than we think. Yet, even though we don't often consciously consider it, each of these transactions is characterized by risk-the risk that a favor or resources we provide won't be repaid, or that an act of generosity won't be reciprocated. Sure, sometimes the risks are greater than others-it's more costly when someone fails to repay a $1,000 loan than the ride we gave them to work-but there are always risks nonetheless. In the short term, it logically makes the most sense to take the money and run, so to speak. If you don't pay back the $40 someone loaned you or give back your time, you're ahead. As the saying goes, the one who dies with the most toys wins! But in the long term, this strategy is not a good one. If you act this way too often or to too many people, you'll get a reputation as a cheat or thief and be ostracized, and then no one will give you any "toys"-a big problem in terms of survival in the long run.

It's the cla.s.sic battle between the gra.s.shopper and the ant, with short-term gains on one side vs. long-term stability on the other. So what tips the scale toward the ant and pushes us to be fair even in the face of the temptation to do otherwise? One answer to this question is simple and straightforward: whether we're consciously aware of it or not, acting fairly is a strategic decision. We know intuitively that even though it would be great to keep the money, gifts, or other favors people give us without repaying in kind, if we don't repay, those favors will very quickly come to an end and we'll be doomed to die a cold and lonely death. But if this were purely a rational decision, certain behaviors just don't make sense. As our colleague the economist Robert Frank frequently points out, people are often fair and reciprocate the generosity of others even when there are no clear long-term benefits. For instance, he cites statistics that most people tip waitstaff the same amount when traveling as they do at their regular neighborhood haunts.3 From a purely self-interested perspective, why would you do this? You'll likely never see the nonlocal waiter or waitress again, so there's nothing really to lose by stiffing them, whereas the server at your local joint might respond by spitting in your food, or worse, the next time you come around. Yet not to tip, a.s.suming the service was fine, just feels wrong.

By the same logic, why would our would-be robber ultimately repay Sohail? He'd already made away with the $40, and even if Sohail had been expecting to be repaid, there's no way he could have tracked the anonymous thief down. What's more, the robber clearly wasn't a wealthy man. He surely could have used the $50 he sent back to his benefactor. So what rational reason is there for repaying a gift from someone whom you will never see again? Well, there really is no rational reason, but we never said the mechanisms that shape our character were always rational.

It is true that at a reasoned level, we can strategically gauge which actions will maximize our self-interest by weighing their trade-offs in the long and short runs. The problem with such a.n.a.lyses, however, is that we generally have incomplete information. For example, you might think you won't be in this restaurant again, so maybe you don't need to tip-but then again, your boss did say there was a small chance she'd send you back to this town for one more client meeting, so you can't be entirely sure. Or you're pretty sure you can ask your neighbor Jack for help moving your couch, as you'll be away on vacation when he is moving out of his apartment and might ask you to return the favor-but what if he decides to cash in by asking for your help with something else? The point is that these calculations, to the extent that we consciously make them, hinge on a.s.sumptions that may or may not be correct. Sure, these what-if scenarios are unlikely, but remember that while the conscious mind is quite adept at imagining many different scenarios, it often does so in a way that can suit our interests in the current moment (as we saw in the discussion of moral hypocrisy). The intuitive mind, at least when it comes to issues of fairness and reciprocity, usually decides more quickly and bluntly. It operates on simpler rules that don't involve weighing the likelihood of what-ifs. It doesn't take the time to create a story or rationalization. Which is why we suspect that when it comes to fairness and reciprocity, the systems of the ant usually have a leg up.

Why would the ant tend to be the winner at the intuitive level? We think it's because the risks of being selfish are simply too high here, even when a rational a.n.a.lysis would tell us we can get away with it. After all, being a cheat or a "taker" is one of the worst reputations anyone can have. Nothing leads to social isolation more quickly. And social isolation not only will make you miserable but also, evolutionarily speaking, will tend to drastically reduce your chances of survival. Which is why, on the intuitive level, playing fair and following the Golden Rule always feels safest in the long run, even if you can't quite see why at the time.

Paying back and paying forward: Grat.i.tude and the Golden Rule.

"d.a.m.n it! It's all going to h.e.l.l!" This was a sentiment commonly uttered by Monica Bartlett when she entered Dave's office. Monica was a member of our team during her graduate school days and was the person who headed up some of the lab's most creative work on grat.i.tude. Monica believed that grat.i.tude was the psychological mechanism central to our seemingly universal desire for fairness. Grat.i.tude, she'd argue, was what made us do the right thing and reciprocate kindness even when it rationally seemed there would be no benefits for doing so (or consequences for failing to do so). In this view, she had good company. The sociologist Georg Simmel referred to grat.i.tude as the "moral memory" of humankind. The economist Adam Smith believed it to be a G.o.d-given moral sentiment that was necessary to ensure human cooperation. And even the biologist Robert Trivers, who, as we noted earlier, developed one of the most famous biological models of altruism, theorized that grat.i.tude underlies many instances of cooperation. Now, these theories make sense-we've all been in situations where we couldn't shake that feeling of grat.i.tude no matter what-but there is a big problem: no one had ever been able to prove them. Why? Because studying grat.i.tude in an experimental setting is really hard.

As you might imagine, this issue-how to study grat.i.tude and its effects scientifically-was usually the source of Monica's consternation. If we were going to show that grat.i.tude triggers a gut-level, intuitive desire that can push any of us to be fair, we first had to develop a situation that would not only evoke grat.i.tude but also let us isolate its effects. Herein lay the challenge. Making people feel grateful usually requires giving them a gift or favor. How could we know, then, if a person's subsequent actions necessarily were a direct result of feeling grateful? They might just stem from good manners or from the knowledge that people are supposed to reciprocate favors or gifts. That is, how could we show that actually feeling grateful to someone who helps you (as opposed to just feeling obligated to repay the person for helping you) is the engine that makes us act fairly? Solving this problem was going to require some creativity, but first we had to figure out how to make people feel grateful within the context of the lab.

We had tried several methods of manufacturing grat.i.tude, but most ended in the same refrain of "It's all going to h.e.l.l!" First we tried giving people small gifts, but this tactic usually didn't work, since finding a relatively inexpensive gift that most college students would appreciate enough to feel grateful turned out to be an impossible task. (Giving them beer would have been unethical.) Plus, we found that most students viewed any gift we gave them with skepticism that got in the way of grat.i.tude; some saw it as compensation for partic.i.p.ating, while others a.s.sumed it was part of the experiment and tried to figure out our true motives. So we sc.r.a.pped the giving of gifts (it was getting expensive, anyway).

Next we tried a different tack: giving nonmaterial rewards (such as letting someone do something enjoyable instead of something onerous). We set up a situation where two individuals would arrive at the lab and be told that there were two different tasks that needed to be done. Much as in the moral hypocrisy studies, one of these tasks was short and fun, while the other was long and difficult. The experimenter told the partic.i.p.ants that she was going to flip a coin to see who would be a.s.signed to which task. Before she did this, however, one of the individuals (who was a confederate working for us) would stop her. "Hey, you know what?" he'd say to the true partic.i.p.ant, "I have a lot of free time this afternoon, and you look like you're busy, so I'll do the logic problems." We had hoped that this would make our partic.i.p.ants feel grateful. It didn't. It seemed the little voice in their minds didn't say, "Thanks!" It said, "Sweet! This other guy's a sucker!"

Now, you might wonder why the partic.i.p.ants didn't feel at least a little grateful. After all, the other guy in the room had just done them a favor by agreeing to do the task they were hoping to avoid. But that's the important point: they were hoping they wouldn't have to do the long, boring task. They didn't know for certain that they would have to do it; there was still a chance they could win the coin flip. The simple fact that the a.s.signment hadn't yet been given to them-that the problem wasn't theirs yet-meant grat.i.tude was in short supply. Since it wasn't clear on either a rational or intuitive level who was going to get stuck with the bad task and, consequently, whether anything was really to be gained by feeling indebted to the guy who volunteered to take it, our partic.i.p.ants didn't feel grateful; they reported feeling lucky. Many even said they viewed the other person as weak, or even slightly strange, for making the offer he did.

Now we were getting frustrated. We were beginning to wonder whether college students just might be immune to grat.i.tude, or whether grat.i.tude was simply impossible to manufacture in a lab. Finally, after more cursing and head scratching, we settled on a way to tweak the existing procedure in a way that we were certain would prevent partic.i.p.ants from attributing their good fortune to luck. We were going to make them own a problem and eliminate any hope or expectation of escape-that is, until a "benefactor" swooped in and decided to take pity on them and help. If anything would trigger their grat.i.tude, this should, because as we've seen, people intuitively know that success in the long run depends on repaying the generosity of others.

Feeling fair.

"This sucks! When is it going to be over?" Pam wondered as she sat in our lab completing the third long-and we mean long-block of grueling word problems. We'd a.s.signed Pam the rather unenviable task of looking at strings of letters flashing on the computer screen in front of her and deciding whether they const.i.tuted actual English words.

Baddax-no.

Sinan-no.

Cabinet-yes.

Fun, right? And it got better. As Pam finally got to the end of what seemed an unending task, a message appeared on her computer screen: Trials complete-calculating score.

"Yes!" she thought. "At last!"-until, a second later, the computer screen went black. As far as Pam could tell, the computer had died, and with it all remains of her work. As she sat there alternating between fury and dread at the thought of having to start all over again, the other partic.i.p.ant, who had been completing the same experiment in the seat next to her, was hurriedly getting up to leave. It shouldn't surprise you by now to know that this other person-her name was Allison in this case-was a confederate working for us.

"What happened?" Allison asked, feigning surprise as she noticed Pam's blank screen. "Mine didn't do that. I'll get the experimenter for you." As Pam sat staring glumly at her screen, Allison reentered with Monica, who matter-of-factly informed Pam that she'd call a technician to come and reset the computer, but that Pam would have to start over from the beginning. As Monica left to make the call, Allison looked at her watch and noted she was already late for her campus job. "But," she said to Pam, "this really sucks for you. Let's see if we can figure out what happened and maybe fix it." With that, Allison began rummaging around the back of the computer, pulling on cords, and hitting keys, all the while asking if anything she was doing seemed to help. Finally Allison surrept.i.tiously hit a key that we had rigged to bring the computer back from its state of apparent death. When it did, a grateful smile broke out on Pam's face.

"That's it! Thanks!" Pam shouted.

When Monica returned to the room, there on Pam's screen were her three scores. There was no need to begin the drudgery again from scratch. With that, Pam spent a few minutes more answering some final questions about her feeling state while Allison left. Then, with the experiment over-at least as far as she knew, anyway-off Pam went.

On her way out of the building, however, Pam just "happened" to run into Allison, who had a clipboard in hand. Allison explained she was working as a research a.s.sistant for a professor who was studying problem solving. She said to Pam, "Look, I really need to collect some data. Would you be willing to help? There are probably more problems in this packet than you can finish, but completing any amount would be helpful." Even though it didn't sound very interesting or enjoyable (and believe us, it wasn't), and Pam was already pretty tapped out from all the tasks we'd just subjected her to, she agreed. After all, if it hadn't been for Allison, she'd still be redoing those computer tasks. Allison took her to a desk at the end of a quiet hall and gave her the packet. She told Pam to just do as many problems as she could and leave the packet on the desk when she was done. She then headed out to recruit more people.

"Great," sighed Pam, "word problems, letter mazes, logic problems-ugh." Yet, even though she could have left immediately with no one seeming to be the wiser, Pam still spent about twenty minutes working on the problems. How do we know? We were secretly timing her through a hidden video feed.

Now, someone might argue that Pam might have been an unusually fair-minded person. Surely most people wouldn't subject themselves to that kind of mental torture just because they felt grateful, would they? Well, yes. In fact, our results showed that the other randomly selected partic.i.p.ants like Pam (those whose computers "crashed" and received help from Allison) not only reported feeling very grateful but also more frequently agreed to Allison's requests for help-and, what's more, worked 50 percent longer on the tedious problems than did partic.i.p.ants whose computers didn't crash and who had no reason to feel grateful.4 Okay, you're probably thinking: "Well, this still isn't much of a surprise; these partic.i.p.ants knew they owed Allison for helping them a few minutes ago. They're probably just helping her because they know it's the right thing to do, Golden Rule and all." Fair enough. But what if we told you we also ran conditions of this study where the person down the hall who asked partic.i.p.ants to complete the problem solving measure wasn't Allison but a complete stranger? Everything else about the experiment was the same. Allison was still the person who sat next to them and, in the cases where their computer crashed, helped them out. But this time, the person at the end of the hall with the clipboard wasn't Allison but someone they had never met. What would Pam and her compatriots do then?

In this case, there isn't really a logical reason to help the stranger. Why would you feel obligated to help someone if you don't owe them, let alone know them or expect to see them again? But some people decided to help nonetheless. As we suspected, it was precisely the people who left our lab feeling grateful to Allison who much more frequently agreed to help the stranger. In fact, the more grateful people reported feeling before they left, the more time they spent working at that lonely desk at the stranger's request.

What was going on here? On the face of it, it seems to make no rational sense. But remember, our rational brains evolved fairly recently, evolutionarily speaking, and in the old days, when our brains didn't have the capacity to work through a reasoned cost-benefit a.n.a.lysis to decide whether to help another person, decisions were made relatively simply and automatically. If we receive a favor, we feel grateful. And if we feel grateful, we pay back. But like all simple systems, this one can be tripped up. So when we quickly replaced a true benefactor (i.e., Allison) with someone else asking for help, the system hiccupped. It knows that feeling grateful means you should be fair and follow the Golden Rule; it just doesn't stop to check to see whom you're repaying. After all, most times when we're feeling grateful, it's the person in front of us we're grateful to, so why expend the extra effort to check? This system works most of the time, but on occasion it can be co-opted.

To show how easily grat.i.tude can be misdirected, we ran one final version of this experiment. This time we ran three conditions. The first two were the same as before: the computer breaks and the partic.i.p.ant receives help (i.e., grat.i.tude condition) or the computer doesn't break (i.e., neutral condition). In all cases, the person who requests help at the end is a complete stranger. The third condition was identical to the one meant to induce grat.i.tude save for a single difference. As Monica was signing partic.i.p.ants out of the session, she asked them, "So, that other person-Allison, I think was her name-helped you out by fixing that computer, right?" They invariably said yes, and then off they went, never suspecting that they were about to be met by a stranger asking for their a.s.sistance.

This simple question disarmed their intuitive systems in two important ways. First, it reminded people that their feelings of grat.i.tude were bound to Allison; they couldn't now easily misattribute this feeling to the stranger. Second, forcing them to stop and think about how and why they were feeling grateful, and to whom, brought their rational systems back into the decision-making process. And indeed, the results showed that this simple act of reminding people to whom they felt grateful strengthened the hand of the short-term systems. This time, the people who felt grateful to Allison were no more likely to help the stranger than were those who didn't. Yes, they still felt grateful at the moment they were asked for a.s.sistance, but the rational forces of the gra.s.shopper now had time and motivation to correct the intuitive hiccup of the ant to repay just anyone who crossed their path.

From this experiment, one thing is clear: whether we are fair partners and pay back our debts stems more from automatic feelings than from reason. We can always justify why we don't have to pay back just yet, but we can't help feeling grateful. More important, we are wired in such a way that our grat.i.tude can be misdirected, leading us to repay our debts to the wrong person. The danger of this, of course, is that if we're feeling grateful, we're liable to help anyone who requests it. But in some ways this isn't such a bad thing. In fact, it can be quite adaptive if it doesn't happen too often, as it encourages people to take the chance on a stranger with whom they might end up having a mutually beneficial relationship.5 In short, it's kind of like paying it forward, driven by emotion.

Still, this fact also makes us vulnerable to the ploys of others. Think about it. When is the best time to ask someone for a favor or for money? When they're feeling grateful (even if it's to someone else). Ever wonder why sometimes those charities asking for donations stick a dollar in the envelope or give you a "gift" of stamps or stickers that you never asked for? As the results of our experiment suggest, these tactics work. So the next time you're feeling grateful and you're tempted to do someone a favor, take a minute to stop and think about whether or not the person asking you for the favor is someone who really deserves it.

That said, most of the time grat.i.tude serves a bigger and more important function in life than just upholding a quid pro quo. Grat.i.tude doesn't only help us reap favors, acquire resources, or build wealth. It builds something that may be even more valuable over the long haul: loyalty and trust.

That warm feeling isn't just the Polartec: Grat.i.tude as social glue.

On the night of December 11, 1995, Aaron Feuerstein was attending a surprise seventieth-birthday party in his honor. Surrounded by friends and family, it must have been a wonderful celebration. Feuerstein was a successful businessman. He was owner and CEO of Malden Mills, one of the biggest textile producers in Ma.s.sachusetts, recognized worldwide as the company that makes Polartec fleece. But the night didn't end as well as it began. As Feuerstein sipped champagne and accepted birthday wishes, his plant thirty miles to the northwest was beginning to smolder. By eight o'clock, it had become engulfed in a six-alarm fire that burned for hours, destroying three of its most central buildings. It took more than four hundred firefighters to put out the fifty-foot walls of flame, and by the time they did, more than 600,000 square feet of manufacturing s.p.a.ce had burned to the ground.6 Everyone expected Feuerstein to take his $300 million in insurance money and either retire or rebuild the plant overseas, where costs would be cheaper, as any shrewd businessman surely would have done. So on December 14, three days after the fire, when Feuerstein stepped to the microphone to address his workers and the local media, the majority of the crowd was hardly expecting him to announce holiday bonuses. Yet the news was far better than they could have imagined. Feuerstein was going to use the insurance money not only to rebuild the plant but to rebuild it right where it was. He was also going to use some of that money (in combination with bank loans he'd take out) to pay his employees their full salaries and benefits for the next ninety days, while the rebuilding began. This was no small gesture; salary and benefits totaled roughly $1.5 million per week. The more than a thousand people in front of him broke into hugs and cheers the likes of which are rarely seen at the site of a ma.s.sive disaster.

Just three weeks later, enough of the factory had been rebuilt that 10 percent of the workforce was back on the job. By spring 80 percent were back to work. And once a year had gone by, the new replacement mill had been opened and any employee who wanted his or her old job back had returned. The new mill was bigger and better than the one that had been destroyed, but that wasn't the only difference. Productivity was way up. Before the fire, the plant had produced 130,000 yards of fabric a week. A few weeks after the fire, it hit 230,000 yards. And it wasn't just because of the new equipment and technology either. It was because the grateful employees were working longer and harder than they ever had before. "People were willing to work 25 hours a day," Feuerstein recounted. One after another, the employees described their grat.i.tude to their boss. Angel Aponte, who had worked at Malden Mills for three years, typified the response: "The way I see it, there isn't anything Mr. A [as Feuerstein was known] could ask us that we wouldn't do. I even heard one of they guys say they'd take a bullet for Mr. A."7 It wasn't just his employees who were paying him back with their time and loyalty. As the story of Feuerstein's actions spread, he suddenly seemed to have spun a reputation of gold. Feuerstein had quickly become one of the most beloved men in America. NBC's Tom Brokaw called him the "best boss in America" and "a saint for the 1990s." Upon hearing that Feuerstein ate a dozen oranges a day, a Florida orange growers a.s.sociation sent him crates of fresh fruit. The Bank of Boston, the local labor union, and the Lawrence, Ma.s.sachusetts, Chamber of Commerce sent donations. People all over the country-people he'd never even met-seemed to feel grateful to him too. He received hundreds of supportive letters from across the United States, many of which included cash ranging from a few dollars to $500. The gifts just kept pouring in.

On the face of it, this seems quite strange, as these people had nothing to gain by sending such gifts. Nonetheless, at the intuitive level, the rationale makes good sense. It seems that just as there are benefits to being fair and trustworthy, so too are there benefits to forging relationships with those we feel we can trust. It's obvious that we admire individuals, such as Aaron Feuerstein, who seem honest and who honor their responsibilities. These are the people that we want as partners and friends. When push comes to shove, we need someone to count on, someone who won't sell us down the river to turn a profit. As we've said before, social relationships are a two-way street. These potential partners also need to know the same about us. They need to know that our short-term interests won't always win, that we're in it to share both the profit and the perils. There needs to be some sort of social glue that binds people together.

We believed grat.i.tude functions as just this type of glue. When those warm feelings of grat.i.tude well up inside us, we feel so bonded to others-at least for the moment-that we become focused on our collective welfare and willing, like those Malden Mills workers, to make sacrifices for the collective good.

To test this theory, you can guess where we headed. This time, though, there wouldn't be months of "It's all going to h.e.l.l." We had already developed a method to induce grat.i.tude in real time. All we needed now was a way to capture the dynamics of trust and tease out its impact on the desire for individual vs. communal gain. Luckily, the field of behavioral economics offered a number of experimental "games" designed to examine these very issues. One of these games, known as the Give Some game, fit our needs perfectly. Here's how it works.

Imagine you're sitting at a table and the experimenter puts down four tokens in front of you. Each of these tokens is worth $1. At the end of the game you'll be able to exchange them for their actual cash value. In a different room, another person is sitting at a table with four tokens in front of him too. You each have the option to exchange any, all, or none of your tokens with the other person. Why would you want to do this? Well, because the rules of the game are such that when you do so, the value of each exchanged token doubles. That is, each of your $1 tokens is worth $2 to your partner, and each of his $1 tokens is worth $2 to you. This exchange, however, happens simultaneously. You won't know his decision before you make yours.

This set of rules nicely pits selfish against communal interests. If you were being totally rational about maximizing your own profit, you wouldn't give any tokens to your partner. That way, you're guaranteed a minimum of your original $4, and anything he gives you just adds to it up to a possible $12. What's more, you're only playing this game once, so he's not going to have a chance to punish you for being selfish later. On the other hand, the best communal outcome is for each person to give all four tokens to the other. This way, both of you end up with $8-double what you started with. Of course, choosing to give is risky, as it works out well for you only if you can trust your partner to do the same.

In several ways, this game mirrors many everyday dilemmas we commonly face. The difficult question, though, is how we decide whom to trust. What determines whether we follow the Golden Rule and go for the communal good or screw over the other person in hopes of maximizing our own profit? What would make someone loan money to a friend or help a neighbor with yard work and trust that those favors would eventually be repaid? What made the workers of Malden Mills agree to work harder to earn the exact same amount of money they would have received without putting in the extra effort? We suspected not only that feeling grateful would have something to do with it but also that, as we saw in our earlier study on grat.i.tude and the Golden Rule, the effect of this feeling would radiate outward. People not only would be more trusting of those to whom they felt grateful but also would be more trusting of anyone. The ant would be dominant, at least for a time.

To test this theory, we ran our partic.i.p.ants through the exact same procedures as before. They came to the lab, they did the onerous task, and their computer either crashed or it didn't. If it did crash, we again evoked grat.i.tude in them by having a confederate help them fix the computer, thereby getting them out of redoing all their work. But then the procedure changed. We told them they were now going to complete a different experiment for the Behavioral Economics Lab. We then took partic.i.p.ants to a separate isolated room and explained the rules of the Give Some game. We told them that the other person who was going to play the game with them was either the person they had just met in the lab or a different person who was a complete stranger. Among the people playing with the person they had just met, half of them were feeling grateful to him (as he'd helped them fix their computer) and half had no particular feelings toward him whatsoever (as their computer hadn't crashed). Same went for the people playing with the stranger: half were grateful to the person in the lab, half weren't. We next flashed some cash to the partic.i.p.ants just so they knew we were ready to make good on our promised payments, then left them alone to decide how many tokens they wanted to give.

What did they do? Well, in the condition where they were playing with the person they knew, those who weren't feeling any grat.i.tude gave about two tokens on average, whereas those who were feeling grateful gave an average of three tokens-or 50 percent more. So far so good. This fit with our prediction, but, as with the fairness experiment, it might be equally likely that they were giving more simply because they knew they owed the person who helped them, not necessarily because they trusted the other person to be generous. But now let's look at the condition where they were playing a stranger. As it turned out, people acted the exact same way. Grateful partic.i.p.ants were more trusting than ones who were not grateful-they again gave 50 percent more money. These people couldn't possibly feel they "owed" their partner, nor could they have any idea about his or her trustworthiness based on past behavior, since they'd never met. Nonetheless, trust they did. It seemed that, in this moment, the grat.i.tude we induced triggered intuitive systems that made people more trusting of everyone, and thereby more willing to take a chance for communal gain. It made them, in other words, behave like the kind of people you'd want as your friend or partner in the long run.8 This notion that grat.i.tude helps to foster successful long-term relationships and social bonds has been borne out in a number of real-world studies as well. Psychologists Sara Algoe, Jon Haidt, and Sh.e.l.ly Gable have shown that grat.i.tude among new friends is a strong predictor of whether they intend to spend additional time together.9 When these researchers followed new pledges at a sorority where "big sisters" were a.s.signed to give gifts to new "young sisters," they showed that feelings of grat.i.tude for the gifts (irrespective of the size of the gift itself) were directly a.s.sociated with how close new members felt to their sponsors. Similarly, work by Nathaniel Lambert, Margaret Clark, and their colleagues revealed that the more grat.i.tude partners in relationships expressed, the more responsibility each person felt for the other's welfare. When they instructed partic.i.p.ants to reflect on events that made them feel grateful to their friends, and then express these feelings to them over a three-week period, they found that both individuals attached more strength to the relationship.10 In other words, grat.i.tude made the bonds feel stronger.

A note of caution is important here, however. It may seem as if we're putting the systems of the ant up on a pedestal when it comes to issues of fairness and trust, and to some extent that may be true. But that's because, as we've noted, very little is as damaging to a person's social standing and success as to be branded a cheat or a welsher, and our long-term systems know it. Yet, as we've also noted, these systems, especially on the intuitive level, are not perfect. Their regular success depends on getting an accurate read of the situation and thereby making sure that when you're feeling grateful, it's to the right person and for the right reasons. Remember, we demonstrated that grat.i.tude can make you more likely to agree to help anyone-not just your benefactor. And while this can be a good thing in limited doses, it can also get you into a tight spot if someone who is seeking to take advantage of you knows just when to ask you for a favor.

Similarly, the amount of grat.i.tude you're feeling needs to be appropriate if the systems are to work well. At times some of us can experience grat.i.tude too intensely. When this happens, we may misjudge how much we need to repay and open ourselves to big losses. We may work too hard or give too much to someone who helped us only minimally in the past. The intensity of grat.i.tude we feel should be commensurate with the costs and benefits of what we received. If the equation becomes too unbalanced, we can run the risk of becoming a doormat. And, really, the same goes for all the other psychological forces that shape our character. Take guilt, for example. When experienced appropriately, guilt can keep us in line, but when it's experienced for the wrong reasons, it can also keep us in chains. For these reasons, we shouldn't dismiss the gra.s.shopper as an evil influence. These systems looking out for short-term self-interest have to act as a balancing force on our decisions. They are what can prevent our long-term systems from turning us into suckers.

Monkey see, monkey do: Monkey don't see, monkey do even more.

Honor students: the word brings to mind studious teens toiling for hours in the library. Honest kids who are putting in hard work now as an investment for a bright future, right? Well, that might describe some honor students, but unfortunately not the majority. A survey of more than three thousand top academic performers conducted by the folks at Who's Who Among American High School Students as part of their twenty-ninth annual poll revealed that 80 percent of the nation's best students have cheated on a.s.signments to get where they are. These disturbing figures closely match the results of a similar survey by the Management Education Center at Rutgers University. What's more, when the Rutgers researchers asked students why they cheated, they plainly stated they believed it was just what those who wanted to succeed did. In short, honesty wouldn't cut it.11 If everyone else was cheating, then it was a necessary evil.

This sad state of affairs demonstrates the important truth that the decision about whether or not to act honestly or fairly isn't always an internal one. External factors, such as whether or not others are acting honestly or fairly, clearly play a role. What's more, these decisions also seem to be swayed by whom the cheating is affecting: it's easier to cheat on something, be it a test or your taxes, when it seems the only one who might be harmed is some faceless ent.i.ty, like a school or "the system." Why? Because when you don't have a face to put to the victim, you intuitively know that you're less likely to confront social consequences in the future.

Take the honor students as an example. Cheating appears rampant, and from what these kids say, it seems that they cheat in school because they believe it is commonplace, practically expected. They know that cheating in general is wrong, but in the specific instance of cheating on tests or homework, they do it nonetheless. One way to explain why this happens is to look at the old proverb "Monkey see, monkey do." Broadly speaking, there are two ways that people decide what const.i.tutes correct behavior: what they learn and what they see. The problem is the two ways don't always go hand in hand. And which we choose to let sway us at any given moment often hinges on what is at stake in the long run.

Seeing is believing.

To see just how malleable honesty is and how easily it's shaped by the honesty (or dishonesty) of others, Francesca Gino from Harvard Business School and her colleagues concocted a clever experiment.12 They brought groups of eight to ten partic.i.p.ants into a lab under the pretense of studying math ability. Each partic.i.p.ant was given a brown envelope containing $10 in dollar bills and coins, along with a white envelope containing a score sheet. Each partic.i.p.ant also received a packet of worksheets that contained twenty confusing math problems. Partic.i.p.ants would have five minutes to complete the worksheet-not nearly enough time for anyone to possibly complete all the problems-and were told they would be allowed to keep 50 for every correct answer. At the end of the five minutes, partic.i.p.ants were to record their own scores on a slip of paper, place it in the white envelope, and take the correct amount of money from the brown envelope.

Now for the really interesting part. The experimenters also varied whether or not the partic.i.p.ants' answers could be checked, and whether there were cheaters in their midst. In one condition, partic.i.p.ants had to bring their solutions to the experimenter, who double-checked their scores and the amount of money they took. In a second condition, no one checked their score-it was all on the honor system (the students were even instructed to destroy their work in a shredder, thereby removing any possibility that cheating could later be uncovered). In the final two conditions, the researchers made it clear that another person in the group was cheating. After about sixty seconds of working on the problems, a confederate would stand up and loudly proclaim that he had finished everything, which was obviously an impossibility given the time allotted. Af