Homo economicus in his place

Emmy at Gendergeek doesn’t like Freakonomics:

The underlying premise of the freakonomics method is that human beings respond only to incentives and that their behaviour is consistent and can be modelled. This is the Robinson Crusoe conception of human behaviour that underpins all modern economic theory and applications. The prototypical economic agent is detached, rational, calculating, infinitely predictable and lives in a utopia of free choice and luxuriant autonomy. While there's difficulty in applying this abstraction to anyone in the real world, it's obvious this simpleton is 'male'. The adjectives associated with homo economicus are very firmly on the macho side of cultural coding.

I can see how she got this impression. Levitt and Dubner do seem to make an exaggerated claim for economics:

The typical economist believes the world has not yet invented a problem that he cannot fix [emphasis added] if given a free hand to design the proper incentive scheme (p20).

However, they go on to define incentives so widely as to threaten to make this claim almost tautological:

When it comes to crime, people also respond to moral incentives (they don’t want to do something they consider wrong.)

I find this unsatisfactory. People’s “moral incentives” cannot be designed. If we confine our understanding of incentives to conventional economic ones, that word “fix” pitches things too strong, doesn’t it? I’d prefer to say “ameliorate” or “reduce.”

Not do I think sensible economists make the grand claims Emmy attributes to us about human nature. Economists might be stupid, but we’re not so stupid to believe all people are pure rational maximizers. Loads of motives other than economic ones affect our behaviour: habit, love, social pressures, religious beliefs, sheer stupidity, and so on.

Surely, economists’ claims are more modest. We say:

1. At the margin, some people respond to economic incentives. If you raise prices, some people buy less. Not all, just some. And not necessarily much less. And incentives might take years to work. There’s an element of homo economicus in people at the margin. That’s all.

2. Although people aren’t rational, they often behave as if they were. This was the point made by Milton Friedman in his Methodology of positive economics. Sure, he said, it’s unrealistic to assume people are rational. But this assumption yields predictions that do (roughly) accord with reality.

A good example comes from financial economics. The efficient market theory says that share prices embody all available information and so you cannot earn higher than average risk-adjusted returns by studying companies.

Now, this seems to rest on absurd assumptions – people aren’t so rational or knowledgeable as to process all information quickly.

But the prediction of this theory is borne out by reality. Most professional fund managers, over the long-run, do under-perform the market.

Perhaps if economists were more modest about what their discipline could do, and didn't make unnecessarily grandiose claims about human nature, we would get more respect from non-economists.

Comments

[Emmy]

Are you assuming I'm a non-economist Chris?

[rjw]

Agreed. There's nothing wrong with smart economists looking at non-conventional topics through an economists lens and trying to see what they can make of it. Economists are pretty good "system thinkers" usually, and can often come up with interesting non-conventional explanations for things. And incentives do matter. No harm in repeating that.

But there seems to be a sort of disdain that some economists seem to feel for disciplines such as sociology that don't share the same pseudo scientific approach to modelling behaviour.

This disdain can even extend to other schools of thought within economics. I still remember when new classical economists were arguing that Keynes's general theory was inadequate as he did not have proper microfoundations. Proving not only that they had not read Keynes, but that they had no sense of irony. As if representative agents counted as adequate microfoundations.

Economists generally show all the signs of having shown insufficient interest in methodological issues or the philosophy of science, in that they are extraordinarily dogmatic at times about the what are "proper" methods. As a result they tend not to endear themselves with other social scientists. (I say this as an economist.....)

Friedman didn't help much with that very cunning but ultimately very damaging article, which was instrumental in allowing generations of economists to get away with thinking that having assumptions that were flat wrong didn't matter too much as long as the numbers came out right and the maths worked. Never mind that the causal mechanisms wasn't the right one.

[rjw]

Actually, this general sloppiness in thinking over methodological issues links to your points about tautologies. Only an economist would classify behaviour that is based on moral precepts as being "incentive driven". As if everything people do can be explained by the rational maximiser model if only you can define the right set of well ordered preferences. Philosophers would call all that making category mistakes. And they'd be right.

[emmy]

Sometimes there is harm in turning the economics lens on topics not usually considered by the discipline. Becker got the Nobel for doing just that and, in my opinion, his strictly economic analyses of the family and household economics have undermined feminist objectives in those areas. He perpetuates the traditional notion of the male-breadwinner, female-caregiver model by a. suggesting that women are innately more productive at domestic labour and hence expend less effort on market work and b. assigning a joint utility function to the male head of household ignoring assymmetrical power relations between men and women. Bollocks to that! You can't analyse the family as if patriarchy didn't exist!

I agree that there are instances where the economics lens is the right one and, even if it's not the full story, can deliver an interesting analysis. Freakonomics was the in this category for me.

But there is a bigger picture of the use of mainstream economics in looking at social problems. The economics lens can't see everything that needs to be seen. Inevitably, what it can see, and more importantly, what it can measure, becomes prioritised over everything else. That which is de-prioritised by mainstream economics quite often pertains to women and their 'invisible' labour.

[Chris Bertram]

People’s “moral incentives” cannot be designed.

Well I suppose a bit of unpacking of this claim might clarify it, but on the face of it it is false. Firms, armies, public schools, football teams, families ... the list of institutions that design "moral incentives" with the aim of getting their members to behave this way or that seems rather a long one to me.

[Paddy Carter]

Chris B - I hope you don't mind me taking you as representative of the left wing, at least for the sake of this argument. But it has always struck me that one of the ways in which left wingers tend to differ from right wingers is the extent to which they believe people can be moulded by such things as firms and public schools. Especially advertisers. Right wingers tend to prefer to think of people as more impervious and of their own mind (which can be a mistake) and left wingers tend to worry about people being shaped by the forces around them (which can also be a mistake). There is, as right wingers are wont to point out, something perverse about this, because left wingers who are 'for' the people often don't seem to have much faith in the people. I suppose this is where the protectionist instinct comes in - the world is full of malicious forces and our job is to try and establish countervailing benign forces. Libertarian types though can turn this “have faith in the people” idea somewhat to religiously, and end ignoring how liaise-faire can leave people getting shat on. Oops I’m rambling. My tuppencewoth is a third way! (I’m such a Blairite). I reckon that people moral incentives are in part shaped by institutions and so forth, as Chris B says, but in ways that cannot be designed, as Chris D says.

[dsquared]

Of course, Levitt and Dubner only believe those things because they want to cosy up to the pop science/business books crowd and make a load of money.

(I always find it hilarious that in the fields of economics and evolutionary psychology, pretty much alone among the sciences, you can gravely insult a man by saying that you agree with his theory of human motivation)

[david]

dsquared has just supplanted John Dupre as the person best suited to point out the parallel bullshit of economics and evolutionary psychology.

Becker is a perfect example of the danger of imperlistic economics. Becker's claims about the world are obviously absurd, yet he gets a Nobel prize and is used to influence policy debates because of his science. (Dupre is pretty good on this, btw).

Also, it's never been clear to me why Friedman was taken seriously on the point mentioned here. Not a lot of evidence that economists can predict the future.

[dearieme]

The general points perhaps matter less if Levitt cocks up the arithmetic. According to Steve Sailer of that blog, Levitt's argument about Abortion and Crime is crap.

[Stacy Rosenbaum]

"Loads of motives other than economic ones affect our behaviour: habit, love, social pressures, religious beliefs, sheer stupidity, and so on."

The author of Freakonomics would say that these all have economic value and the brain weighs all this stuff when making decisions. In an irrational manner of course.

The point is, that people are motivated to do stuff because they will stuff that they want. Love, money, conforming to the social norms, doing things according to the religion, this is all stuff that people want.

[dsquared]

Becker is actually very interesting and a lot more subtle than you would think by reading summaries of his work written by his wannabes. He's rather like Karl Marx in that regard.

[Angry Economist]

I think the value of Freakonomics is popularising the notion that its good to question and analyse the evidence around popular conception and misconception. I have been involved in that as a professional economist myself. Working in public policy myself there is an awful lot of misconception, myth and baseless policy positions in government. They need challenged intellegently and articulately. Freakonomics has probably done us data geeks a favour.

[david]

Which parts of Becker are the subtle parts? Having read him, of course, and not just the summaries of his wannabes. I think Becker is excruciatingly obvious, or entirely opaque (most often when scientistic), with not much room for subtlety in between. Totally unlike Marx in that regard.