|February 17, 2005/New York TIMES
The Theory That Self-Interest Is the Sole Motivator Is Self-Fulfilling
By ROBERT H. FRANK
A NEW YORKER cartoon depicts a well-heeled, elderly gentleman taking his
grandson for a walk in the woods. "It's good to know about trees," he
tells the boy, before adding, "Just remember, nobody ever made big money
knowing about trees."
If the man's advice was not inspired directly by the economist's
rational-actor model, it could have been. This model assumes that people
are selfish in the narrow sense. It may be nice to know about trees, it
acknowledges, but it goes on to caution that the world out there is
bitterly competitive, and that those who do not pursue their own
interests ruthlessly are likely to be swept aside by others who do.
To be sure, self-interest is an important human motive, and the
self-interest model has well-established explanatory power. When energy
prices rise, for example, people are more likely to buy hybrid vehicles
and add extra insulation in their attics.
But some economists go so far as to say that self-interest explains
virtually all behavior. As Gordon Tullock of the University of Arizona
has written, for example, "the average human being is about 95 percent
selfish in the narrow sense of the term." Is he right? Or do we often
heed social and cultural norms that urge us to set aside self-interest
in the name of some greater good?
If the search is for examples that contradict the predictions of
standard economic models, a good rule of thumb is to start in France.
During my recent sabbatical in Paris, I encountered many such examples,
but one in particular stands out. One mid-November afternoon, I asked my
neighborhood wine merchant if he could recommend a good Champagne. It
was the week before Thanksgiving, and my wife and I had invited a few
American friends to our apartment for a turkey dinner.
He just happened to have an excellent one on sale for only 18 euros.
(Normal price, 24 euros.) Fine, I said, and then asked if he could also
recommend a bottle of cassis, since I knew some of our guests would want
a kir royale - a cocktail of cassis and Champagne. In that case, he
said, I would have no need for the high-quality Champagne, because no
one would be able to tell the difference once it was mixed with cassis.
Well, then, what should I buy? He brought back a bottle that he said
would be just right for the purpose.
That particular Champagne, however, was not on sale. When he told me it
was 20 euros per bottle - 2 euros more than the better one - an awkward
pause ensued. Though I thought I knew the answer, I felt I had to ask
whether a kir royale would taste worse if made with the better
Champagne. He assured me it would not. And because I knew that some of
us would be drinking our Champagne straight, I bought several bottles of
the better one. He did not protest, but I could feel him reclassify me
as yet another American barbarian.
For many French, the logic of the self-interest model is trumped by an
aesthetic principle about what Champagnes are right for specific
applications. This particular principle leads to a better outcome over
all, because it directs the best Champagne to the uses in which quality
matters most. So even though I personally was better off for having
ignored the merchant's advice (because I got to drink a better Champagne
and spent less), at least some of the better Champagne I bought was
France is, of course, not the only place in which the self-interest
model's predictions fall short.
Most Americans, for example, leave tips even after dining in restaurants
they will never visit again. We take the trouble to vote in presidential
elections, even though no single individual's vote has ever changed the
outcome in any state. We make anonymous donations to charity. From
society's perspective, our willingness to forgo self-interest in such
instances leads to better outcomes than when we all act in a purely
Does what we believe about human motivation matter? In an experimental
study of private contributions to a common project, two sociologists
from the University of Wisconsin, Gerald Marwell and Ruth Ames, found
that first-year graduate students in economics contributed an average of
less than half the amount contributed by students from other
Other studies have found that repeated exposure to the self-interest
model makes selfish behavior more likely. In one experiment, for
example, the cooperation rates of economics majors fell short of those
of nonmajors, and the difference grew the longer the students had been
in their respective majors.
My point is not that my fellow economists are wrong to stress the
importance of self-interest. But those who insist that it is the only
important human motive are missing something important. Even more
troubling, the narrow self-interest model, which encourages us to expect
the worst in others, often brings out the worst in us as well.
Perhaps the theories of human behavior embraced by other disciplines
influence their practitioners in similar ways. A core principle of
behavioral biology, for example, is that males are far more likely than
females to engage in "extra-pair copulations." Does teaching this model
year after year make male biologists more likely to stray?
Several years ago, I attended a dinner with a group of biologists that
included a married couple. After describing the research about how
economics training appears to inhibit cooperation, I asked whether
anyone had ever done a study of whether males in biology were more
likely than scholars from other disciplines to be unfaithful to their
partners. The uncomfortable silence that greeted my question made me
wonder whether I had stumbled onto a data point relevant for such a
But if biologists are like economists in being influenced by their own
theories, they are different from us in another respect: Their most
cherished hypothesis is much less likely than ours to be contradicted by
Robert H. Frank is an economist at the Johnson School of Management at
Cornell University and the author, most recently, of "What Price the
Moral High Ground?"