Stumbling on Happiness

by Daniel Gilbert

  Yet if we measure the actual satisfaction of people who have children, a very different story emerges. As figure 23 shows, couples generally start out quite happy in their marriages and then become progressively less satisfied over the course of their lives together, getting close to their original levels of satisfaction only when their children leave home.17 Despite what we read in the popular press, the only known symptom of “empty nest syndrome” is increased smiling.18 Interestingly, this pattern of satisfaction over the life cycle describes women (who are usually the primary caretakers of children) better than men.19 Careful studies of how women feel as they go about their daily activities show that they are less happy when taking care of their children than when eating, exercising, shopping, napping, or watching television.20 Indeed, looking after the kids appears to be only slightly more pleasant than doing housework.

  Fig. 23. As the four separate studies in this graph show, marital satisfaction decreases dramatically after the birth of the first child and increases only when the last child leaves home.

  None of this should surprise us. Every parent knows that children are a lot of work—a lot of really hard work—and although parenting has many rewarding moments, the vast majority of its moments involve dull and selfless service to people who will take decades to become even begrudgingly grateful for what we are doing. If parenting is such difficult business, then why do we have such a rosy view of it? One reason is that we have been talking on the phone all day with society’s stockholders—our moms and uncles and personal trainers—who have been transmitting to us an idea that they believe to be true but whose accuracy is not the cause of its successful transmission. “Children bring happiness” is a super-replicator. The belief-transmission network of which we are a part cannot operate without a continuously replenished supply of people to do the transmitting, thus the belief that children are a source of happiness becomes a part of our cultural wisdom simply because the opposite belief unravels the fabric of any society that holds it. Indeed, people who believed that children bring misery and despair—and who thus stopped having them—would put their belief-transmission network out of business in around fifty years, hence terminating the belief that terminated them. The Shakers were a utopian farming community that arose in the 1800s and at one time numbered about six thousand. They approved of children, but they did not approve of the natural act that creates them. Over the years, their strict belief in the importance of celibacy caused their network to contract, and today there are just a few elderly Shakers left, transmitting their doomsday belief to no one but themselves.

  The belief-transmission game is rigged so that we must believe that children and money bring happiness, regardless of whether such beliefs are true. This doesn’t mean that we should all now quit our jobs and abandon our families. Rather, it means that while we believe we are raising children and earning paychecks to increase our share of happiness, we are actually doing these things for reasons beyond our ken. We are nodes in a social network that arises and falls by a logic of its own, which is why we continue to toil, continue to mate, and continue to be surprised when we do not experience all the joy we so gullibly anticipated.

  The Myth of Fingerprints

  My friends tell me that I have a tendency to point out problems without offering solutions, but they never tell me what I should do about it. In one chapter after another, I’ve described the ways in which imagination fails to provide us with accurate previews of our emotional futures. I’ve claimed that when we imagine our futures we tend to fill in, leave out, and take little account of how differently we will think about the future once we actually get there. I’ve claimed that neither personal experience nor cultural wisdom compensates for imagination’s shortcomings. I’ve so thoroughly marinated you in the foibles, biases, errors, and mistakes of the human mind that you may wonder how anyone ever manages to make toast without buttering their kneecaps. If so, you will be heartened to learn that there is a simple method by which anyone can make strikingly accurate predictions about how they will feel in the future. But you may be disheartened to learn that, by and large, no one wants to use it.

  Why do we rely on our imaginations in the first place? Imagination is the poor man’s wormhole. We can’t do what we’d really like to do—namely, travel through time, pay a visit to our future selves, and see how happy those selves are—and so we imagine the future instead of actually going there. But if we cannot travel in the dimension of time, we can travel in the dimensions of space, and the chances are pretty good that somewhere in those other three dimensions there is another human being who is actually experiencing the future event that we are merely thinking about. Surely we aren’t the first people ever to consider a move to Cincinnati, a career in motel management, another helping of rhubarb pie, or an extramarital affair, and for the most part, those who have already tried these things are more than willing to tell us about them. It is true that when people tell us about their past experiences (“That ice water wasn’t really so cold” or “I love taking care of my daughter”), memory’s peccadilloes may render their testimony unreliable. But it is also true that when people tell us about their current experiences (“How am I feeling right now? I feel like pulling my arm out of this freezing bucket and sticking my teenager’s head in it instead!”), they are providing us with the kind of report about their subjective state that is considered the gold standard of happiness measures. If you believe (as I do) that people can generally say how they are feeling at the moment they are asked, then one way to make predictions about our own emotional futures is to find someone who is having the experience we are contemplating and ask them how they feel. Instead of remembering our past experience in order to simulate our future experience, perhaps we should simply ask other people to introspect on their inner states. Perhaps we should give up on remembering and imagining entirely and use other people as surrogates for our future selves.

  This idea sounds all too simple, and I suspect you have an objection to it that goes something like this: Yes, other people are probably right now experiencing the very things I am merely contemplating, but I can’t use other people’s experiences as proxies for my own because those other people are not me. Every human being is as unique as his or her fingerprints, so it won’t help me much to learn about how others feel in the situations that I’m facing. Unless these other people are my clones and have had all the same experiences I’ve had, their reactions and my reactions are bound to differ. I am a walking, talking idiosyncrasy, and thus I am better off basing my predictions on my somewhat fickle imagination than on the reports of people whose preferences, tastes, and emotional proclivities are so radically different from my own. If that’s your objection, then it is a good one—so good that it will take two steps to dismantle it. First let me prove to you that the experience of a single randomly selected individual can sometimes provide a better basis for predicting your future experience than your own imagination can. And then let me show you why you—and I—find this so difficult to believe.

  Finding the Solution

  Imagination has three shortcomings, and if you didn’t know that then you may be reading this book backward. If you did know that, then you also know that imagination’s first shortcoming is its tendency to fill in and leave out without telling us (which we explored in the section on realism). No one can imagine every feature and consequence of a future event, hence we must consider some and fail to consider others. The problem is that the features and consequences we fail to consider are often quite important. You may recall the study in which college students were asked to imagine how they would feel a few days after their school’s football team played a game against its archrival.21 The results showed that students overestimated the duration of the game’s emotional impact because when they tried to imagine their future experience, they imagined their team winning (“The clock will hit zero, we’ll storm the field, everyone will cheer . . .”) but failed to imagine what they would be doing afterward (“And the
n I’ll go home and study for my final exams”). Because the students were focused on the game, they failed to imagine how events that happened after the game would influence their happiness. So what should they have done instead?

  They should have abandoned imagination altogether. Consider a study that put people in a similar predicament and then forced them to abandon their imaginations. In this study, a group of volunteers (reporters) first received a delicious prize—a gift certificate from a local ice cream parlor—and then performed a long, boring task in which they counted and recorded geometric shapes that appeared on a computer screen.22 The reporters then reported how they felt. Next, a new group of volunteers was told that they would also receive a prize and do the same boring task. Some of these new volunteers (simulators) were told what the prize was and were asked to use their imaginations to predict their future feelings. Other volunteers (surrogators) were not told what the prize was but were instead shown the report of a randomly selected reporter. Not knowing what the prize was, they couldn’t possibly use their imaginations to predict their future feelings. Instead, they had to rely on the reporter’s report. Once all the volunteers had made their predictions, they received the prize, did the long, boring task, and reported how they actually felt. As the leftmost bars in figure 24 show, simulators were not as happy as they thought they would be. Why? Because they failed to imagine how quickly the joy of receiving a gift certificate would fade when it was followed by a long, boring task. This is precisely the same mistake that the college-football fans made. But now look at the results for the surrogators. As you can see, they made extremely accurate predictions of their future happiness. These surrogators didn’t know what kind of prize they would receive, but they did know that someone who had received that prize had been less than ecstatic at the conclusion of the boring task. So they shrugged and reasoned that they too would feel less than ecstatic at the conclusion of the boring task—and they were right!

  Fig. 24. Volunteers made much more accurate predictions of their future feelings when they learned how someone else had felt in the same situation (surrogators) than when they tried to imagine how they themselves would feel (simulators).

  Imagination’s second shortcoming is its tendency to project the present onto the future (which we explored in the section on presentism). When imagination paints a picture of the future, many of the details are necessarily missing, and imagination solves this problem by filling in the gaps with details that it borrows from the present. Anyone who has ever shopped on an empty stomach, vowed to quit smoking after stubbing out a cigarette, or proposed marriage while on shore leave knows that how we feel now can erroneously influence how we think we’ll feel later. As it turns out, surrogation can remedy this shortcoming too. In one study, volunteers (reporters) ate a few potato chips and reported how much they enjoyed them.23 Next, a new group of volunteers was fed pretzels, peanut-butter cheese crackers, tortilla chips, bread sticks, and melba toast, which, as you might guess, left them thoroughly stuffed and with little desire for salty snack foods. These stuffed volunteers were then asked to predict how much they would enjoy eating a particular food the next day. Some of these stuffed volunteers (simulators) were told that the food they would eat the next day was potato chips, and they were asked to use their imaginations to predict how they would feel after eating them. Other stuffed volunteers (surrogators) were not told what the next day’s food would be but were instead shown the report of one randomly selected reporter. Because surrogators didn’t know what the next day’s food would be, they couldn’t use their imaginations to predict their future enjoyment of it and thus they had to rely on the reporter’s report. Once all the volunteers had made their predictions, they went away, returned the next day, ate some potato chips, and reported how much they enjoyed them. As the middle bars in figure 24 show, simulators enjoyed eating the potato chips more than they thought they would. Why? Because when they made their predictions they had bellies full of pretzels and crackers. But surrogators—who were equally full when they made their predictions—relied on the report of someone without a full belly and hence made much more accurate predictions. It is important to note that the surrogators accurately predicted their future enjoyment of a food despite the fact that they didn’t even know what the food was!

  Imagination’s third shortcoming is its failure to recognize that things will look different once they happen—in particular, that bad things will look a whole lot better (which we explored in the section on rationalization). When we imagine losing a job, for instance, we imagine the painful experience (“The boss will march into my office, shut the door behind him . . .”) without also imagining how our psychological immune systems will transform its meaning (“I’ll come to realize that this was an opportunity to quit retail sales and follow my true calling as a sculptor”). Can surrogation remedy this shortcoming? To find out, researchers arranged for some people to have an unpleasant experience. A group of volunteers (reporters) was told that the experimenter would flip a coin, and if it came up heads, the volunteer would receive a gift certificate to a local pizza parlor. The coin was flipped and—oh, so sorry—it came up tails and the reporters received nothing.24 The reporters then reported how they felt. Next, a new group of volunteers was told about the coin-flipping game and was asked to predict how they would feel if the coin came up tails and they didn’t get the pizza gift certificate. Some of these volunteers (simulators) were told the precise monetary value of the gift certificate, and others (surrogators) were instead shown the report of one randomly selected reporter. Once the volunteers had made their predictions, the coin was flipped and—oh, so sorry—came up tails. The volunteers then reported how they felt. As the rightmost bars in figure 24 show, simulators felt better than they predicted they’d feel if they lost the coin flip. Why? Because simulators did not realize how quickly and easily they would rationalize the loss (“Pizza is too fattening, and besides, I don’t like that restaurant anyway”). But surrogators—who had nothing to go on except the report of another randomly selected individual—assumed that they wouldn’t feel too bad after losing the prize, hence made more accurate predictions.

  Rejecting the Solution

  This trio of studies suggests that when people are deprived of the information that imagination requires and are thus forced to use others as surrogates, they make remarkably accurate predictions about their future feelings, which suggests that the best way to predict our feelings tomorrow is to see how others are feeling today.25 Given the impressive power of this simple technique, we should expect people to go out of their way to use it. But they don’t. When an entirely new group of volunteers was told about the three situations I just described—winning a prize, eating a mystery food, or failing to receive a gift certificate—and was then asked whether they would prefer to make predictions about their future feelings based on (a) information about the prize, the food, and the certificate; or (b) information about how a randomly selected individual felt after winning them, eating them, or losing them, virtually every volunteer chose the former. If you hadn’t seen the results of these studies, you’d probably have done the same. If I offered to pay for your dinner at a restaurant if you could accurately predict how much you were going to enjoy it, would you want to see the restaurant’s menu or some randomly selected diner’s review? If you are like most people, you would prefer to see the menu, and if you are like most people, you would end up buying your own dinner. Why?

  Because if you are like most people, then like most people, you don’t know you’re like most people. Science has given us a lot of facts about the average person, and one of the most reliable of these facts is that the average person doesn’t see herself as average. Most students see themselves as more intelligent than the average student,26 most business managers see themselves as more competent than the average business manager,27 and most football players see themselves as having better “football sense” than their teammates.28 Ninety percent of motorists consider themse
lves to be safer-than-average drivers,29 and 94 percent of college professors consider themselves to be better-than-average teachers.30 Ironically, the bias toward seeing ourselves as better than average causes us to see ourselves as less biased than average too.31 As one research team concluded, “Most of us appear to believe that we are more athletic, intelligent, organized, ethical, logical, interesting, fair-minded, and healthy—not to mention more attractive—than the average person.”32

  This tendency to think of ourselves as better than others is not necessarily a manifestation of our unfettered narcissism but may instead be an instance of a more general tendency to think of ourselves as different from others—often for better but sometimes for worse. When people are asked about generosity, they claim to perform a greater number of generous acts than others do; but when they are asked about selfishness, they claim to perform a greater number of selfish acts than others do.33 When people are asked about their ability to perform an easy task, such as driving a car or riding a bike, they rate themselves as better than others; but when they are asked about their ability to perform a difficult task, such as juggling or playing chess, they rate themselves as worse than others.34 We don’t always see ourselves as superior, but we almost always see ourselves as unique. Even when we do precisely what others do, we tend to think that we’re doing it for unique reasons. For instance, we tend to attribute other people’s choices to features of the chooser (“Phil picked this class because he’s one of those literary types”), but we tend to attribute our own choices to features of the options (“But I picked it because it was easier than economics”).35 We recognize that our decisions are influenced by social norms (“I was too embarrassed to raise my hand in class even though I was terribly confused”), but fail to recognize that others’ decisions were similarly influenced (“No one else raised a hand because no one else was as confused as I was”).36 We know that our choices sometimes reflect our aversions (“I voted for Kerry because I couldn’t stand Bush”), but we assume that other people’s choices reflect their appetites (“If Rebecca voted for Kerry, then she must have liked him”).37 The list of differences is long but the conclusion to be drawn from it is short: The self considers itself to be a very special person.38