Arnold Kling  

Just Resting

International Health Care Comp... Don't Believe Behaviorism's De...

Economic behaviorism isn't dead. It's just resting. Or pining for the fjords.

Seriously, I think that behaviorism is a reaction against two propositions. One is that we need to examine whether people really optimize. The other is that we need to observe people's utility functions.

The doctrine of revealed preference, which I associate with Samuelson, says that we do not need to observe utility functions. Instead, we can infer relative marginal utility indirectly by observing choices.

A short way of putting this form of behaviorism is "watch what people do, not what they say." For example, so-called "happiness research" will come up with results that say that people get little or no incremental happiness beyond $20,000 of income. Yet their behavior contradicts that, since many people could easily cut back on work and still have $20,000 a year. I believe the behavior, not the surveys.

My co-blogger wants to bury behavioralism. I am not ready to see it interred.

Comments and Sharing

CATEGORIES: Economic Philosophy

COMMENTS (10 to date)
Bruce Cleaver writes:

I think a rapproachment between Messrs Caplan and Kling is possible:

After $20,000 the money is less important, but the social benefits of *work itself* are important. Work confers status, provides structure, and enables comparisons between individuals. It isn't always the money.

I must admit though, there is still happiness utility for each additional dollar after the first 20K.

ThaddeusMcMonster writes:

So you're saying that people would still work the same if salaries were maximized at $20,000 per year?

Lancelot Finn writes:

Someday I'd like to write a book entitled Prisoners of the Food Metaphor: Why Economists Misunderstand the World.

After laying the groundwork by reviewing the miraculous power of division of labor, specialization and trade, and economies of scale to better the human condition (an argument mimicking, and updating, that of Adam Smith); after arguing that economies of scale are more important than science in explaining improved human welfare; I then take my experience in the remote Russian republic of Tuva as a starting place for questioning whether we know that humans are better off at all.

Tuva is a highland country neighboring Mongolia, and for hundreds or perhaps thousands of years leading up to the early years of the 20th century, they were nomads, who traveled across the land herding cattle and living in yurts. From proud and free nomads, the Tuvans today have been reduced to Siberian slum-dwellers. My intuition told me that Tuvans were worse off, and the local Russian community made it amply clear that that's how the Tuvans feel. (They think the Russians "robbed them of everything.") And yet the steppes are still there, vast, empty, available. If they look back so fondly to the old nomadic life, why don't they go back (and thus "reveal their preferences?")

I got an insight from a picture in the Tuvan Museum entitled "Glory to the Hero." A very large figure (three to five times as large as the other figures in the picture) was seated, and around him were old men bringing tribute, damsels singing praises, his whole people giving him glory, all against the background of a stylized and idyllic Tuvan landscape. There was something about that picture that seemed admirable, satisfying, worth longing for.

What would this "hero" have been like in real life? I thought. Probably just a cow-herder, rather strong, brave and wise, and valued by his people because under his leadership they were able to provide for and defend themselves. A Tuvan today could go out to the steppes and live a life superficially resembling the "hero," but he would no longer seem like a hero today, and no one would give them glory. And that is the secret of how the Russians "robbed" the Tuvans: they robbed them, not in terms of the material or monetary economy, but in terms of an entirely different economy: status.

The economy of status has different dynamics than the material economy. It is likely to be more zero-sum in nature, and the good concerned is psychological rather than physical. Merely by existing and being known, the Russians robbed the Tuvans. Why? Because once the Tuvans knew that the strength, wisdom, and power to provide that their "hero" had possessed were vastly inferior to those of the Russians, they could no longer look on him as a hero, even if they wanted to. As the hero lost his heroic image, his people lost the satisfaction of (sincerely) giving him glory.

But why stop there? If there is an economy of status, there is an economy of love and sexuality; an economy of morality; an economy of belief and truth; a conversation may be considered an economy (with exchanges, gains and losses, competition); there are economies of art and beauty; and for the religious, economies of salvation. Each economy starts with a good people desire and pursue, and describes the dynamic interactions that arise from their efforts to get it.

What distinguishes these other economies from (what we will continue for the moment to call) the material economy, is that in each of these economies enjoyment is inevitably social. What I mean is a bit stronger than that enjoyment is inevitably social in practice; I mean to argue, rather, that purely individual enjoyment is actually inconceivable in these other economies. In the economy of truth and belief, our ideas are connected with words, and since a private language is impossible (Wittgenstein), we cannot form beliefs or have knowledge (in anything akin to our usual sense of the word) without language and therefore community. The economy of morality relates mostly and perhaps entirely to our dealings with other people. The economy of status consists precisely of what people think of one another. Art is written, in language, by an author to an audience; and even the rare cases where an author writes a poem he intends no one to read are exceptions that prove the rule, first because he is nonetheless using a language and a genre that he derived from others, second because the emotional response to such a lonely poet is typically one of sadness, and we're inclined to suspect that, if he found the right person to read his poem and properly appreciate it, he would be delighted.

The one economy in which purely individual enjoyment is comprehensible is food. (Perhaps also warmth, or comfortable clothing, but the point stands.) We consume food; and although we don't eat cars or concerts or plane tickets, economists describe as "consumption" what might better be generalized as the "enjoyment" of all these goods. Economists tend to think this somewhat counter-intuitive generalization of the term "consume" is innocuous. It is not, because they are generalizing from one economy, that of food, in which enjoyment is intelligible at a purely individual level, to other economies in which enjoyment is intelligible only as a social phenomenon. This is why economists are "prisoners of the food metaphor."

The standard definition of utility starts with the assumption of complete, transitive, non-satiable, non-alphabetic preferences over all goods. This definition does not justify our restriction of the "utility" under consideration to what are traditionally considered "economic" goods. It is just as intelligible to ask a person "Would you prefer to have seven mansions, or seven men in love with you?" as it is to ask "Would you prefer two apples or three oranges?" The definition of utility, properly understood, should compel economists to interpret utility as across all the economies I described above (and any others that may be described), not merely over the "material and monetary economy" (which is not properly an economy in the sense described above, anyway, since much or most of people's consumption is directed to acquiring truth, or status, or attracting sexual partners, or friendship, and so on).

So why did economists ever think it justifiable to narrow their focus to the realm of goods purhcasable in "the material and monetary" economy (or whatever)? Because of money.

The definition of utility envisions an infinite questionnaire. You ask a subject an infinite number of questions about his preferences. But this infinite interview is not feasible to conduct: it would take two long, questions might be ambiguous, the subject's views would change in the course of the interview, and so on. So how is this account of utility useful at all? It is, because money conducts the interview for us! People have money, they perpetually make choices, this-or-that, this-or-that, apples-or-oranges, now-or-later, and these choices allow us to infer their preferences-- but only over things money can buy, and those are an arbitrary cross-section of goods, or means to acquire goods, from many different economies.

There is a joke that is relevant here. Walking down the street at night, I encounter a man who is groping through the dirt under a streetlamp. "What are you looking for?" I ask.

"I dropped my keys," he says.

"Did you drop them here?" I ask.

"No, over there in the bushes. But the light's better here."

When economists narrow the focus of their research to physical, sellable goods, they are like the man who looks for his keys under the streetlamp. Money is the streetlamp. It conducts the infinite interview for us. But it conducts it in a partial and tendentious manner, and there is little reason to believe that it gets at a very comprehensive or accurate picture of the truth.

And if we were to understand the economy as a whole, we have reason to believe that Pareto-improvement through consensual exchange would play a less important role. Why? Because since enjoyment in these arenas is social, there are huge and ubiquitous problems of externalities. The Russians rob the Tuvans of everything merely by existing, robbing their harsh but glorious way of life of its glory. Or positive externalities: by living morally, I set an example which others admire and follow, others perhaps who I barely meet.

I would describe the relation of contemporary economics to this imaginary but potential discipline, with its broader understanding of "utility" and "economies," as comparable to the relationship between alchemy and chemistry. Alchemists performed many experiments and learned a good deal, and prepared the way for chemistry, but ultimately they were working on the false (though not implausible) belief that base metals could be mixed with fire to produce gold. Economists today are prisoners of the food metaphor, operating on the false belief that enjoyment, utility, can be understood as a merely individual phenomenon. They have achieved wonders in terms of feeding mankind and providing for them materially, but other than that, it's not clear they have contributed much to human happiness. They may, however, lay the groundwork for a new science, once their individualist assumption is abandoned, more appropriate to an age in which physical well-being is trivially easy to obtain, which serves to organize the community of man so as to...

DeadHorseBeater writes:

Lancelot --

That was the best comment I've read in any blog, ever. If you do publish a book-length treatment of that subject, I'd buy it. I'm sure many others would as well.

There are a few works on these kinds of status-equilibrium effects (one by a French fellow whose name I don't remember), but none that have been fleshed out into multi-sector theories. Most of them just have relative social status flowing from relative economic status.

Jason Ligon writes:

I think there is something to the piece of Mr. Finn's opus about looking for keys where there is light, but I don't agree with the tone in which the message is delivered. The correct response, to me, is to acknowledge that you can only look in the light. Looking in the dark IS completely pointless.

The danger here is in loading the data underlying economic theory with self serving garbage that polls reveal about each of us. I don't know that there is a methodology to only count the 'good' responses. Further, most people don't know ANYTHING about their own motivations, and we as a species are amazingly adept at retroactive justification. The essence of the revealed preference concept is that it distills a thousand variable problem into a behavior that we can talk about. We can say, "In the net of all considerations, this person clearly chose x over y." If you ask someone why they chose x over y, you get much more interesting information about their self view than you do about their actual decision. We are supposed to believe that economists, who suffer the same foibles as the rest of us, will just somehow know when to ignore what someone says? I just don't see it as a scientifically useful process.

I can't shake the notion that what we are all concerned about is what people do. The question seems to be, can we get a better understanding of what people do by studying what they do, or can we get a better understanding of what they do by listening to them talk?

Randy writes:

Interesting read Lancelot. Just one quick comment. It is one thing to say that human beings are social creatures, but another to say that we owe anything to any society other than the ones we choose. It is true that most happiness is social. It is also true that most unhappiness is social. The individual is preeminent.

Tom West writes:

I must admit I feel a bit ambivalent towards Lancelot's posting. I've been slowly working towards the same sort of ideas that Lancelot expressed and now find that he's considered it more thoroughly and expressed it far more clearly than I ever could.

I'm both impressed and jealous.

Lancelot, good work. Your blog goes on my permanent reading list.

So why did economists ever think it justifiable to narrow their focus to the realm of goods purhcasable in "the material and monetary" economy (or whatever)? Because of money.

Your only problem is that they don't. Becker, for instance, writes about love and marriage. John Lott about guns and crime. Levitt on the long run effects Roe v Wade. Sowell about perverse incentives created by Affirmative Action laws.

Lawrance George Lux writes:

The 'happiness research' gets cramped by the average Household debt load. Happiness must assume no desire for lifestyle changes. lgl

Lancelot Finn writes:
"Becker, for instance, writes about love and marriage. John Lott about guns and crime. Levitt on the long run effects Roe v Wade. Sowell about perverse incentives created by Affirmative Action laws."

Yeah, that's true. There have certainly been some steps in the right direction.

Comments for this entry have been closed
Return to top