Bryan Caplan  

Applied Economics Assumes Selfishness, and Rightly So

In Praise of Facebook... Some Health Care Reform Discus...
Yoram Bauman has responded to my review of his Cartoon Introduction to Economics in detail.  One of my main complaints:
Klein and Bauman shouldn't have run away from self-interest in chapter 1.  Yes, I know that textbooks love to claim that economics assumes "optimizing behavior," not "self-interest."  But whenever economists do applied work, they quickly slide to self-interest.  You know why?  Because although people aren't perfectly selfish, they're shockingly close.  That's why economics tells us so much about the world.
He responds:
Caplan is right that we start with the basic assumption of "individual optimization" and not "individual self-interest". And the reason is--geez, Bryan, do we really have to rehash this tired old argument?--that (1) as a theoretical matter, "individual self-interest" is impossible to define or defend in a scientifically rigorous way, i.e., I can always say "The soldier threw herself on the hand grenade because she wanted to" or "Bill Gates is giving his money away because it makes him feel better about himself"
I agree that "individual self-interest" is not mathematically precise.  Interesting concepts seldom are.  But if you speak English, you can reliably categorize many actions as "self-interested" or "not self-interested."  Asking for a raise?  Self-interested.  Giving all your money to a stranger?  Not self-interested.  Cutting ahead in line?  Self-interested.  Donating a kidney to a stranger?  Not self-interested.  Yes, you "can always say" that soldiers jump on hand grenades out of self-interest, but only by torturing the English language.

Yoram continues:
(2) as a practical matter, you get into a terrible knot when you start thinking about spouses and children and friends &etc.
Family provides the best counter-examples to a naive self-interest theory, but there's a more general view that accounts for them:  Dawkins' "selfish gene" theory.  In most economic contexts, however, the two theories make identical predictions, so I didn't press this point in my review. 

Why do I make such a big deal about this?  Simple: Because despite Yoram's protests, applied economists habitually assume that people are selfish in the ordinary language sense of the word.  Consider Yoram's congestion example.  Every applied economist says, "Raise the price!"  But if drivers were unselfish in the right way, all of the following would be equally economically plausible solutions:

1. Ask everyone to drive less "because they're inconveniencing others."

2. Tell people they're contributing to global warming.

3. Announce that if traffic doesn't fall by 20%, we'll abolish foreign aid to Senegal.

4. Denounce materialism so people quit their jobs and stop commuting.

Now you could say that applied economists are being closed-minded.  But I think we're correct to focus on congestion charges.  Why?  Because the assumption of human selfishness is roughly true.  Almost everyone cares a lot about the price they personally have to pay to use a road, and getting their weekly paycheck.  Most people don't care very much about the effect of their driving on other drivers, global temperature, or the people of Senegal.

It's easy to multiply examples.  Rent control leads to shortages and/or declining quality - if landlords are selfish.  If they loved their tenants as themselves, it's a different story.  Printing tons of money wouldn't cause inflation if people were happy to build up unlimited cash balances for the "good of the country."  Yadda yadda yadda.

I admit that Yoram's not alone.  Even though applied economics almost always relies on self-interest, most economists prefer to tell their students about preference maximization instead.  So why am I singling Yoram out?  Because a cartoon intro to economics is a great opportunity to (a) defy silly academic conventions, (b) bluntly tell the world what economists really think, and (c) use humor to make the hard truth go down easy.  Contrary to Yoram, giving his book a B+, then finding fault, isn't "passive aggression."  It's my way of saying, "You've done a very good job, but I want you to shoot for greatness."

Comments and Sharing

COMMENTS (16 to date)
dWj writes:

An interesting post from the author of "The Myth of the Rational Voter".

bbb writes:

>> i.e., I can always say "The soldier threw herself on the hand grenade because she wanted to" or "Bill Gates is giving his money away because it makes him feel better about himself"

It is because individuals are not fully rational in the utility maximizing sense, but rather exhibit program-based behavior because of imperfect knowledge, that such seeming departures from self-interest occur.

Most economists interpret them as a deviation from self interest because they are too bound up in their maximizing-paradigm, trying to modify the assumed preferences of the individual to fit the observed facts.

These seeming deviations can be much more consistently explained, without the complete loss of predicitive power that Bauman's shift towards mere "optimizing behavior" entails, within a rule-following-behavior model of human action. in this paradigm, these outcomes result from following a rule that is to the individuals own interest in other contexts (e.g. a small group of hunter gatherers, or a family) without considering wether following the rule in this particular instance is in the individuals behavior. I.e., the individual may commit a classification error in identifying the nature of the situation and the appropriate behavioral rule. However, not having to think through the ultimate consequences of one's own actions is precisely the advantage that can be gained from following rules.

The predictive power of the economic model is in this way preserved. Individuals are at their core still self interested, what rules individuals follow (and how these rules evolved) becomes an empirical question, and economic science transforms from an empty science of choice to an empirical science of human behavior.

SydB writes:

"I agree that "individual self-interest" is not mathematically precise. Interesting concepts seldom are."

Yeah, like all those boring concept from physics, e.g. the uncertainty principle. Or the schroedinger equation. Special relativity. The mandelbrot set--totally boring.

Oh wait, those are mathematically precise.

But then again, Godel's incompleteness theorem is not mathematic...oh wait, it is too.

The problem with the concept of self-interest is not that it is interesting. It's boring. Because it's useless.

Andrew T writes:

I am narrowly addressing the definitional issue brought up in this discussion.

I have seen a few times where there is a desire to define the term selfish (or self-interest) that is outside of the common understanding of the term. Merriam-Webster's defines selfish as "Concerned exclusively or excessively with oneself: seeking or concentrating on one's own advantage, pleasure, or well-being without regard for others". It should be difficult for someone to categorize jumping on a grenade as a selfish act without, as stated in the original post, "torturing the English language". The Merriam-Webster definition for self-interest is similarly "a concern for one's own advantage and well-being".

By definition self-interest is not simply pursuing what I am interested in (e.g. the man was interested in saving his friends, so jumping on the grenade was a selfish/self-interested act). Economists already have a term which is possibly a better fit ... utility. The lists the economic definition of utility as "Economist-speak for a good thing; a measure of satisfaction". This term does not carry the "without regard for others" baggage that selfish does. Optimizing utility can include both selfish and selfless acts. Again, understanding that there are measurement difficulties when comprehensively addressing utility, I am only trying to address definitional issues in this comment.

Redland Jack writes:


I would agree that strict adherence to mathematical precision is very importanat in the physical sciences. I'm less certain that it is always useful in the social sciences.

Philo writes:

"Family provides the best counter-examples to a naive self-interest theory, but there's a more general view that accounts for them: Dawkins' 'selfish gene' theory."

Dawkins' view is not "more general" than psychological egoism; it just generalizes over genes rather than people. And it is far from "accounting" for actual behavior within families, which in fact displays enormous variation. Note, by the way, that I share almost as many genes with you, a total stranger, as I do with my own children.

fundamentalist writes:

Andrew, I agree completely. It's a cheap trick to torture words until they surrender and affirm your argument. But that's the way most people argue. If we make self-interest and selfish synonyms, then we are left without any word for actions that keep us alive, such as eating. What word describes the motivation for eating, if not self-interest. You might choose survival, but is not survival selfish since you are eating food that someone else could have eaten and survived on?

Selfishness should be viewed as an extreme form of self-interest, not as identical. Otherwise, all action of any kind is selfish and we have killed communication.

Yoram Bauman writes:

I think Bryan and I mostly agree with each other: "narrow self-interest" is a reasonable characterization of the way most people act most of the time. Where we disagree is the appropriate starting point.

Bryan prefers a take-no-prisoners approach centered around a statement ("people act out of narrow self-interest") that is arguably too strong, which means he has to backtrack when confronted with people who give money to charity or let others merge in front of them in traffic.

I prefer a modest approach centered around a statement ("people are optimizing individuals") that is arguably too weak, which means I have to explain why appeals to conscience are a bad way to tackle traffic congestion or climate change. And I think the book does a good job of providing that explanation, both in the opening chapter and in the chapter on the prisoners' dilemma.

'Nuff said :)

agnostic writes:

"And it is far from "accounting" for actual behavior within families, which in fact displays enormous variation."

Read what you're critiquing first. Dawkins is popularizing Hamilton's rule in this context, which says that an individual will behave altruistically provided that:

c [less than] r * b

Where c is the cost the altruist incurs, b is the benefit they confer, and r is the probability that the recipient shares the altruist's gene. The key is that the gene in one individual is helping another copy of it -- in a different individual -- to keep going.

This allows for plenty of behavioral variation within families. Full siblings are more closely related than cousins. Also there is uncertainty from the altruist's point-of-view since they never fully know what r is -- they have to use heuristics to guess how closely related the recipient is.

"Note, by the way, that I share almost as many genes with you, a total stranger, as I do with my own children."

That's a silly thing to note, though, as any economist will tell you. If you're making a choice, it's on the margin. Sure, any two randomly chosen humans share much of their genome in common, just as any two randomly chosen car models share many features. So, it will be where they differ that competition takes place, i.e. with respect to those genes that are not shared among all of humankind.

Thus, since your children are a little more closely related to you, they win. If two car models are highly similar overall, but one has power windows and locks while the other doesn't, the first one wins.

Douglass Holmes writes:

Self interest is a lot easier to understand than greed. Alan Blinder claims that greed is back and it isn't always good. Do I have to study at Princeton to understand what greed is?

Andrew T. writes:

Fundamentalist, thanks for the reply. I do believe they are synonyms. I went to my hard copy Random House dictionary (1981 ... apparently I haven't bought a hard copy dictionary for awhile) hoping it would list synonyms ... unfortunately it did not. However, it did give another definition for self-interest. The random house definition of self-interest is: 1. regard for one's own interest or advantage, esp. with disregard for others. 2. personal interest or advantage.

As a weaker test, I typed self-interest into Microsoft word and requested synonyms. Selfishness was at the top of the list.

This website, also listed selfish as a synonym to self-interest.

Even so, synonyms, don't have to have the exact same meaning ... they can have nearly the same meaning, so one being a strong form of the other is fine. Although, the Random House definition of self-interest did seem to be on the strong side as well.

Either way my concern is with selfless acts being defined as self-interested acts. An example given in the original post was jumping on a grenade. I will add the qualifier to save the lives of others.

When I have seen a detailed defense of defining such acts as self interested, the defense goes as follows: Self-interest is doing what interests the self. Since the person was interested in saving the lives of the people around him/her, jumping on the grenade was an act of self-interest. There are probably other defenses as well.

That is where I believe self-interest is being redefined away from its common (dictionary) understanding to include acts that are selfless.

Thanks again for your comments.

Loof writes:

Loof loves to be a coach (actually a looch in loofy language) and blow the whistle while whistling in the wind. After dribbling well from “optimizing behaviour” to “self-interest” placing and grounding the pivotal principle in applied economics was perfect. Sliding the pivot foot from “self-interest” into being shockingly to being “perfectly selfish” is perfectly irrational. No solid ground; slippery floor warning. Whistle blow. Slides into selfishness and back to self-interest, then out-in, out-in, out-in will slip one up every time, let alone each time. Blow. Blow. Blow. OK. Looch sees the water on the floor, throw in the towel please and let me wipe up after you.

David Landy writes:

From the original post:

'"I agree that "individual self-interest" is not mathematically precise. Interesting concepts seldom are. But if you speak English, you can reliably categorize many actions as "self-interested" or "not self-interested." Asking for a raise? Self-interested. Giving all your money to a stranger? Not self-interested. Cutting ahead in line? Self-interested. Donating a kidney to a stranger? Not self-interested. Yes, you "can always say" that soldiers jump on hand grenades out of self-interest, but only by torturing the English language."'

I don't think it's a question of whether or not "individual self-interest" is mathematically precise - since in considering it we are not dealing with a mathematical issue. I do however think, that one can be rational, objective and scientific in approaching the matter.

Each individual has free will. As a result he may make choices that result in his happiness or his despair, his life or his destruction. The means to these results are the value judgments he makes in every minute of every day during the course of his life. He must make decisions pertaining to the course he wishes to take with his life, which when done properly integrates the long term with the short term and drives his self-esteem, pride and general happiness. These value judgments form a kind of internal "hierarchy of values" that varies widely depending on the individual and his interests.

This latter point is important in relation to the quote of Bryan's I excerpted above. Value hierarchies are complex and sometimes difficult to delineate for the individual themselves, let alone outsiders. For some individuals - such as for example soldiers - the value they place on their self-professed duty to their country and their fellow warriors is such that they are willing to give their own life in defense of those values. Similarly, for some, giving lots of their money (I have difficulty envisaging they giving *all* of their money) to a stranger may rank highly for them - perhaps they take pride in giving promising strangers a chance to develop their talent. "Cutting ahead in line?" - you brand it as self-interested - I would not. It is not courteous to do this, one let's oneself down for having poor etiquette and being uncivilized.

It is difficult to judge individual actions as self-interested or not - in many situations. But the main point I want to make is that each individual himself can be objective and rational as he makes those decisions. He is, and should be, the only person in the position to make the judgments that pertain to his life. If he is unsure about what value belongs where in his "hierarchy" he can *still* be scientific about what to do. Being unsure does not preclude one from action - but act one must.

hacs writes:

Dawkins argues that our genes do not exist to perpetuate us; instead, we are useful machines that serve to perpetuate them (that is why the gene is selfish in Dawkins point of view; that has nothing to do with human selfishness or anything similar). For example, altruistic behavior in animals does not evolve for "the good of the species" but is really selfishness in disguise. Natural selection acts on each individual, not on any kind of group, and the gene is selfish, not necessarily the machine serving it (a person). Social sciences are full of misconceptions about this subject.

Robert Bumbalough writes:

[quote]Mr Bauman: Caplan is right that we start with the basic assumption of "individual optimization" and not "individual self-interest".[/quote]

I think Mr Bauman overlooks an obvious point. Why would I or anyone else be interested in optimizing our performance unless doing so conveyed some benefit to ourselves, and wouldn't my desire to obtain some benefit via optimizing my performance be evidence of my own self-interest. In my world view, morality answers questions of what I should do while ethics addresses why I should do those things. That I have to ask such questions at all means that I hold my own interests as core values; for if I lacked a core value to my own life and interests, I would not be able to survive and reproduce. The point of living is to live and not to serve.

Loof writes:

According to hacs:
"Natural selection acts on each individual, not on any kind of group, and the gene is selfish, not necessarily the machine serving it (a person). Social sciences are full of misconceptions about this subject."

And the social science about selfishness seems to be the fullest of misconceptions. Selfishness, established in evolutionary biology, is out of date (selfishness in economics follows suit) and misconceived. The notion that human behavior is governed by selfishness as an ultimate motive - without altruism (selflessness) and group selection - is now pet theory and off base. Group selection and altruism, ego and empathy, is at the leading edge of theoretical biology, sociobiology as well as socioeconomics. Behavioral economics “bounded selfishness” also moves away from limitless selfishness. In peer reviewed work the scientific Ego'n'Empathy Hypothesis (W. Hayes & G. Lynne, Towards a Centrepiece for Ecological Economics, 2004) is fascinating and establishes a middle ground between selfishness and selflessness.

In the 1960s altruism and group selection was successfully attacked and devastated by G.C. Williams (Adaption and Natural Selection, 1966) and supplanted with selfishness as an ultimate motive. By the middle 70s Ghiselin famous quip: “Scratch an ‘altruist,’ and watch a ‘hypocrite’ bleed” ridiculed group selection; while positively Dawkin’s The Selfish Gene (1976) was framed with William’s foundational principle. Group selection became scandalous in graduate studies in theoretical biology and selfishness theory developed the full swing until the middle 90s.

Since it has been supplanted and a swing the other way is under way (towards a middle ground perhaps) and become respectable again in higher education. Darwin is back and a robust theory of group level functionalism is developing. Though Loof has been out of the loop for years E. Sober & D. Wilson Unto Others: The Evolution and Psychology of Unselfish Behavior (1998) is an exceptional reference, iLo.

It is worth quoting Darwin’s Descent of Man (1871) at length on this subject since his reasoning is so good and clear: “It must not be forgotten that although a high standard of morality gives but a slight or no advantage to each individual man and his children over the other men of the same tribe, yet that an increase in the number of well-endowed men and advancement in the standard of morality will certainly give an immense advantage to one tribe over another. There can be no doubt that a tribe including many members who, from possessing in a high degree the spirit of patriotism, fidelity, obedience, courage, and sympathy, were always ready to aid on another, and to sacrifice themselves for the common good, would be victorious over most other tribes; and this would be natural selection. (p. 166.)

Comments for this entry have been closed
Return to top