Thomas wrote: I have my reservations about the experiment you quoted below, and I have even more reservations about the inferences drawn from the evidence. I'm still trying to figure out just what irks me about it.
Wow. Almost half a year has passed since I wrote this. Since then, I read up on the issue (there was an especially good article by Robert Frank on the subject
here). I did some thinking, and I believe I now
have figured out what irks me. In fact, I'm so full of things that irk me that this post will probably become way too long for the comfort of its readers, for which case I apologize in advance.
1. I still don't fully trust the data
In the narrowest sense, the experiment described by Sozobe irked me because it didn't control for the other direction. It only measured how many people agreed to dance in public a few weeks later and then chickened out when it was their turn. It didn't control for the other directions -- the people who declined at first, then said: "Heck, I'm
doing it!" By itself this would be a relatively petty point, but it seems to be fairly typical for the field. Lots of surveys, lots of experiments that don't quite look controlled, lots of fuzziness. But Frank's article tells me that the surveys
have been calibrated against objective measures of (un)happiness like the incidence of suicide, and that they fit quite well. I have to admit that this part of my defense now looks rather weak. Still, I don't really trust the happiness researcher's data.
2. Can happiness researchers predict anything they don't know already?
When happiness researchers say that conventional economics have to be rewritten because of their findings, they are repeating an argument that's pretty old: 'Real humans aren't rational, standard economics assumes rationality, therefore standard economics must be wrong.' This argument sounds persuasive on the face of it. But in economics, like in any science, the relevant test of a theory isn't whether its premises are correct -- it is the theory's ability to predict the outcome of experiments.
As the Evil Genius will confirm if Sozobe asks him, the theory of quantum mechanics defies common sense, and its implications are peculiar to the point of insanity. So why do otherwise nice physics professors teach this insanity to unsuspecting undergraduate students? There is no reason to do this at all -- except that the insane theory happens to predict why electron microscopes magnify as they do, why atomic bombs explode as they do, and why transistors transist as they do. Throughout the early 20th century, many able physicists have tried to develop straightforward, sane theories of (sub)atomic physics, but none of them can predict the outcome of unknown experiments as well as quantum mechanics do.
A similar point applies to Richard Dawkins' 'Selfish Gene' model of biological evolution. Taken literally, it's nonsense -- genes aren't
really selfish, and they don't
reallyuse rationality to copy as many instances of themselves into the next generation. Nevertheless, the model has proven successful at predicting outcomes it wasn't designed to explain. Consequently, it is now a standard tool in the biologists' tool box and it has become common for them to use 'selfish gene' jargon in talking about evolution.
Likewise for the economist's assumptions of rationality and happiness maximization. Psychologists have long argued that this assumption doesn't hold. Milton Friedman mentioned their argument long ago, in his 1953 essay "The methodology of positive economics". And his counter-argument is the same as my argument about quantum mechanics and the 'selfish gene': theories are as good as the predictions they make, and by this test, standard, neo-classical economists beats all the elaborate psychological models.
They still do, judging from what I read in popular science articles on economics. For example, Daniel Kahnemann won half of the
2002 Nobel prize in economics for exploring the irrationalities of happiness seekers. But the other half of that price was won by Vernon Smith, who explored how these irrationalities changed the optimal design of institutions. It turned out that the changes were negligible in most cases. For another example, Eugene Fama from the University of Chicago has invented the field of "behavioral finance" -- the science of how investors make decisions in stock markets. He, too, made a lot of interesting psychological observations, but in the end, he recommends the same investment strategy as everybody else: don't buy single stocks, buy a stock index and hold it. The market isn't rational, but neither are you, so you can't out-strategize it.
3. Differential happiness, not absolute happiness, is the right basis for decision making
Perhaps this point is best illustrated by
the example Sozobe gave about her own financial life a few pages ago. Back when she was a poor student, she was at her happiness set point. Back then, if someone had beamed her into her present, middle-class, activist, conference-organizing supermom life, that would have made her intensely happy. So happy, indeed, that she would have paid X Dollars to bring it about. Today, Sozobe is at her happiness set-point too. But if someone beamed her back into her poor-student life again, that would make her intensely
unhappy -- so unhappy that she would be willing to pay Y Dollars to prevent that from happening. I wouldn't be surprised at all if X=Y, which is what neoclassical economics would predict.
Judged from an absolute happiness point of view, there is no difference between poor-student-hood and supermom-dom. And that's a good thing, because the alternative would be permanent depression as a poor student, or permanent mania in the present middle-class life, or both. Nevertheless, the activist middle class mom life is
always preferred over the poor student life, and it is always preferred by more or less the same margin. In my judgment, this margin is a much better guide to Sozobe's, or President Bush's or the state of Ohio's decision making than Sozobe's absolute happiness.
4. Even if happiness researchers are right, this plays out both ways politically
From the "futile pursuit of happiness": "Loewenstein tells me that he doesn't see how anybody could study happiness and not find himself leaning left politically; the data make it all too clear that boosting the living standards of those already comfortable, such as through lower taxes, does little to improve their levels of well-being, whereas raising the living standards of the impoverished makes an enormous difference. "
On the other hand, here is a pretty persuasive counter-argument,
courtesy of Paul Krugman:
Quote:But before you denounce (or applaud) happiness research as leftist propaganda, be aware that it also cuts the other way. For example, how do you feel about the "living wage" movement, which in effect wants a large increase in the minimum wage? That would certainly increase the incomes of the lowest-paid workers; but it would also surely have at least some adverse effect on the number of jobs available. You may think that a price worth paying -- but the equations say that the extra unemployment would be a very bad thing for those who lose their jobs, while a higher wage would make only a small difference to the happiness of those who remain employed.
Or think of it this way. If you don't want a society in which everyone is desperately trying to get ahead in a zero-sum status game, you might advocate government policies that slow down the rat race: high tax rates, generous health and unemployment benefits, long mandatory paid vacations, maybe even a limit on individual working hours. In other words, you might want to turn America into France. But France has an unemployment rate more than twice as high as America's, largely because of those same government policies. And unemployment -- even comfortable European-style unemployment -- makes people very unhappy, because it is demoralizing.
It irks me that articles on happiness research always emphasize the aspects that play out for the political left, and downplay those that play out for the political right.
5. Even if happiness researchers are right, I don't want it to justify government intervention.
Scientists have thought of ways to make people happy before. At the end of the 19th century, they thought they had solid statistical evidence that governments could make people happier by providing for a better social environment. The political movements thus encouraged were called socialism and communism. It brought us better social insurance, which is good. But socialism also brought us housing projects that looked and felt like prisons, and communism brought us some of the most murderous regimes in the world.
Other scientists, convinced by other statistical evidence, suggested that governments could make people happier by improving their nature. In the best cases, this brought us eugenics programs in otherwise nice countries like Sweden and Britain. In the worst case, it brought us the Holocaust. That's why I'm very, very, very,
very weary of government coercion grounded in the belief that scientists know better what makes people happy than the people themselves. Which brings me to my final point. (Big sigh of relief from the audience, I'm sure.)
6. Even if happiness researchers are right, and governments can make people happy by acting on it -- I prefer liberty over happiness.
Now we're out of science territory and deeply into gut feelings. But I don't think it's just me. A good illustration of my point is Aldous Huxley's novel "Brave New World". It describes a world in which the government skillfully manipulates its citizens' nurture and nature in order to make everyone perfectly happy. With the exception of one person, which was necessary for Huxley to have a novel to write, the system works, and everybody
is more or less perfectly happy. Huxley's description of this world feels plausible and consistent to me, so he may well be describing some variant of a reasonably realistic future.
I hate every aspect of it. And I would much prefer to be unhappy on my own terms than happy on the terms of the "Brave New World"'s government.
(And again, I'm sorry if the length of this post made any of you unhappy.)
)
)
