TED日本語 - ダニエル・ゴールドステイン: 「現在の自分」と「未来の自分」との戦い

Do you remember the story of Odysseus and the Sirens from high school or junior high school? There was this hero, Odysseus, who's heading back home after the Trojan War. And he's standing on the deck of his ship, he's talking to his first mate, and he's saying, "Tomorrow, we will sail past those rocks, and on those rocks sit some beautiful women called Sirens. And these women sing an enchanting song, a song so alluring that all sailors who hear it crash into the rocks and die." Now you would expect, given that, that they would choose an alternate route around the Sirens, but instead Odysseus says, "I want to hear that song. And so what I'm going to do is I'm going to pour wax in the ears of you and all the men -- stay with me -- so that you can't hear the song, and then I'm going to have you tie me to the mast so that I can listen and we can all sail by unaffected." So this is a captain putting the life of every single person on the ship at risk so that he can hear a song.

And I'd like to think if this was the case, they probably would have rehearsed it a few times. Odysseus would have said, "Okay, let's do a dry run. You tie me to the mast, and I'm going to beg and plead. And no matter what I say, you can not untie me from the mast. All right, so tie me to the mast." And the first mate takes a rope and ties Odysseus to the mast in a nice knot. And Odysseus does his best job playacting and says, "Untie me. Untie me. I want to hear that song. Untie me." And the first mate wisely resists and doesn't untie Odysseus. And then Odysseus says, "I see that you can get it. All right, untie me now and we'll get some dinner." And the first mate hesitates. He's like, "Is this still the rehearsal, or should I untie him?" And the first mate thinks, "Well, I guess at some point the rehearsal has to end." So he unties Odysseus, and Odysseus flips out. He's like, "You idiot. You moron. If you do that tomorrow, I'll be dead, you'll be dead, every single one of the men will be dead. Now just don't untie me no matter what." He throws the first mate to the ground. This repeats itself through the night -- rehearsal, tying to the mast, conning his way out of it, beating the poor first mate up mercilessly. Hilarity ensues.

Tying yourself to a mast is perhaps the oldest written example of what psychologists call a commitment device. A commitment device is a decision that you make with a cool head to bind yourself so that you don't do something regrettable when you have a hot head. Because there's two heads inside one person when you think about it. Scholars have long invoked this metaphor of two selves when it comes to questions of temptation. There is first, the present self. This is like Odysseus when he's hearing the song. He just wants to get to the front row. He just thinks about the here and now and the immediate gratification. But then there's this other self, the future self. This is Odysseus as an old man who wants nothing more than to retire in a sunny villa with his wife Penelope outside of Ithaca -- the other one.

So why do we need commitment devices? Well resisting temptation is hard, as the 19th century English economist Nassau William Senior said, "To abstain from the enjoyment which is in our power, or to seek distant rather than immediate results, are among the most painful exertions of the human will." If you set goals for yourself and you're like a lot of other people, you probably realize it's not that your goals are physically impossible that's keeping you from achieving them, it's that you lack the self-discipline to stick to them. It's physically possible to lose weight. It's physically possible to exercise more. But resisting temptation is hard.

The other reason that it's difficult to resist temptation is because it's an unequal battle between the present self and the future self. I mean, let's face it, the present self is present. It's in control. It's in power right now. It has these strong, heroic arms that can lift doughnuts into your mouth. And the future self is not even around. It's off in the future. It's weak. It doesn't even have a lawyer present. There's nobody to stick up for the future self. And so the present self can trounce all over its dreams. So there's this battle between the two selves that's being fought, and we need commitment devices to level the playing field between the two.

Now I'm a big fan of commitment devices actually. Tying yourself to the mast is the oldest one, but there are other ones such as locking a credit card away with a key or not bringing junk food into the house so you won't eat it or unplugging your Internet connection so you can use your computer. I was creating commitment devices of my own long before I knew what they were. So when I was a starving post-doc at Columbia University, I was deep in a publish-or-perish phase of my career. I had to write five pages a day towards papers or I would have to give up five dollars.

And when you try to execute these commitment devices, you realize the devil is really in the details. Because it's not that easy to get rid of five dollars. I mean, you can't burn it; that's illegal. And I thought, well I could give it to a charity or give it to my wife or something like that. But then I thought, oh, I'm sending myself mixed messages. Because not writing is bad, but giving to charity is good. So then I would kind of justify not writing by giving a gift. And then I kind of flipped that around and thought, well I could give it to the neo-Nazis. But then I was like, that's more bad than writing is good, and so that wouldn't work. So ultimately, I just decided I would leave it in an envelope on the subway. Sometimes a good person would find it, sometimes a bad person would find it. On average, it was just a completely pointless exchange of money that I would regret. (Laughter) Such it is with commitment devices.

But despite my like for them, there's two nagging concerns that I've always had about commitment devices, and you might feel this if you use them yourself. So the first is, when you've got one of these devices going, such as this contract to write everyday or pay, it's just a constant reminder that you have no self-control. You're just telling yourself, "Without you, commitment device, I am nothing, I have no self-discipline." And then when you're ever in a situation where you don't have a commitment device in place -- like, "Oh my God, that person's offering me a doughnut, and I have no defense mechanism," -- you just eat it. So I don't like the way that they take the power away from you. I think self-discipline is something, it's like a muscle. The more you exercise it, the stronger it gets.

The other problem with commitment devices is that you can always weasel your way out of them. You say, "Well, of course I can't write today, because I'm giving a TEDTalk and I have five media interviews, and then I'm going to a cocktail party and then I'll be drunk after that. And so there's no way that this is going to work." So in effect, you are like Odysseus and the first mate in one person. You're putting yourself, you're binding yourself, and you're weaseling your way out of it, and then you're beating yourself up afterwards.

So I've been working for about a decade now on finding other ways to change people's relationship to the future self without using commitment devices. In particular, I'm interested in the relationship to the future financial self. And this is a timely issue. I'm talking about the topic of saving. Now saving is a classic two selves problem. The present self does not want to save at all. It wants to consume. Whereas the future self wants the present self to save. So this is a timely problem. We look at the savings rate and it has been declining since the 1950s. At the same time, the Retirement Risk Index, the chance of not being able to meet your needs in retirement, has been increasing. And we're at a situation now where for every three baby boomers, the McKinsey Global Institute predicts that two will not be able to meet their pre-retirement needs while they're in retirement.

So what can we do about this? There's a philosopher, Derek Parfit, who said some words that were inspiring to my coauthors and I. He said that, "We might neglect our future selves because of some failure of belief or imagination." That is to say, we somehow might not believe that we're going to get old, or we might not be able to imagine that we're going to get old some day. On the one hand, it sounds ridiculous. Of course, we know that we're going to get old. But aren't there things that we believe and don't believe at the same time?

So my coauthors and I have used computers, the greatest tool of our time, to assist people's imagination and help them imagine what it might be like to go into the future. And I'll show you some of these tools right here. The first is called the distribution builder. It shows people what the future might be like by showing them a hundred equally probable outcomes that might be obtained in the future. Each outcome is shown by one of these markers, and each sits on a row that represents a level of wealth and retirement. Being up at the top means that you're enjoying a high income in retirement. Being down at the bottom means that you're struggling to make ends meet. When you make an investment, what you're really saying is, "I accept that any one of these 100 things could happen to me and determine my wealth."

Now you can try to move your outcomes around. You can try to manipulate your fate, like this person is doing, but it costs you something to do it. It means that you have to save more today. Once you find an investment that you're happy with, what people do is they click "done" and the markers begin to disappear, slowly,one by one. It simulates what it is like to invest in something and to watch that investment pan out. At the end, there will only be one marker left standing and it will determine our wealth in retirement.

Yes, this person retired at 150 percent of their working income in retirement. They're making more money while retired than they were making while they were working. If you're like most people, just seeing that gave you a small sense of elation and joy -- just to think about making 50 percent more money in retirement than before. However, had you ended up on the very bottom, it might have given you a slight sense of dread and/or nausea thinking about struggling to get by in retirement. By using this tool over and over and simulating outcome after outcome, people can understand that the investments and savings that they undertake today determine their well-being in the future.

Now people are motivated through emotions, but different people find different things motivating. This is a simulation that uses graphics, but other people find motivating what money can buy, not just numbers. So here I made a distribution builder where instead of showing numerical outcomes, I show people what those outcomes will get you, in particular apartments that you can afford if you're retiring on 3,000,2,500,2,000 dollars per month and so on. As you move down the ladder of apartments, you see that they get worse and worse. Some of them look like places I lived in as a graduate student. And as you get to the very bottom, you're faced with the unfortunate reality that if you don't save anything for retirement, you won't be able to afford any housing at all. Those are actual pictures of actual apartments renting for that amount as advertised on the Internet.

The last thing I'll show you, the last behavioral time machine, is something that I created with Hal Hershfield, who was introduced to me by my coauthor on a previous project, Bill Sharpe. And what it is is an exploration into virtual reality. So what we do is we take pictures of people -- in this case, college-age people -- and we use software to age them and show these people what they'll look like when they're 60,70,80 years old. And we try to test whether actually assisting your imagination by looking at the face of your future self can change you investment behavior.

So this is one of our experiments. Here we see the face of the young subject on the left. He's given a control that allows him to adjust his savings rate. As he moves his savings rate down, it means that he's saving zero when it's all the way here at the left. You can see his current annual income -- this is the percentage of his paycheck that he can take home today -- is quite high,91 percent, but his retirement income is quite low. He's going to retire on 44 percent of what he earned while he was working. If he saves the maximum legal amount, his retirement income goes up, but he's unhappy because now he has less money on the left-hand side to spend today. Other conditions show people the future self. And from the future self's point of view, everything is in reverse. If you save very little, the future self is unhappy living on 44 percent of the income. Whereas if the present self saves a lot, the future self is delighted, where the income is close up near 100 percent.

To bring this to a wider audience, I've been working with Hal and Allianz to create something we call the behavioral time machine, in which you not only get to see yourself in the future, but you get to see anticipated emotional reactions to different levels of retirement wealth. So for instance, here is somebody using the tool. And just watch the facial expressions as they move the slider. The younger face gets happier and happier, saving nothing. The older face is miserable. And slowly, slowly we're bringing it up to a moderate savings rate. And then it's a high savings rate. The younger face is getting unhappy. The older face is quite pleased with the decision. We're going to see if this has an effect on what people do. And what's nice about it is it's not something that biasing people actually, because as one face smiles, the other face frowns. It's not telling you which way to put the slider, it's just reminding you that you are connected to and legally tied to this future self.

Your decisions today are going to determine its well-being. And that's something that's easy to forget. This use of virtual reality is not just good for making people look older. There are programs you can get to see how people might look if they smoke, if they get too much exposure to the sun, if they gain weight and so on. And what's good is, unlike in the experiments that Hal and myself ran with Russ Smith, you don't have to program these by yourself in order to see the virtual reality. There are applications you can get on smartphones for just a few dollars that do the same thing. This is actually a picture of Hal, my coauthor. You might recognize him from the previous demos. And just for kicks we ran his picture through the balding, aging and weight gain software to see how he would look. Hal is here, so I think we owe it to him as well as yourself to disabuse you of that last image. And I'll close it there.

On behalf of Hal and myself, I wish all the best to your present and future selves. Thank you.

(Applause)