Thursday, July 16, 2009

Psychologists to the Rescue!


With Danny Kahneman's Nobel prize in economics every other week has a big-name author or journalist writing a breathless article about how the new 'behavioral economics' will fix sterile financial theory. Alas, way back in 1951 this was a tired theme. George Stigler noted "each decade, for the past nine or ten decades [ie, back to 1851!], economists have read widely in the then-current psychological literature. These explorers have published their findings, and others in the field have found them wanting—wanting in useful hypotheses about economic behavior."

Take, for example, the anchoring bias, where people do not adjust their prior beliefs sufficiently when presented with new data. On the other hand, there's the case where people do not sufficiently account for base rate information, as for example when they are told a woman is quiet and assume she's a librarian, not saleswoman, even though there are more saleswomen than librarians. Thus, over, or underreaction to information is a common 'bias', and so Kahneman's classic work Judgment Under Uncertainty: Heuristics and Biases surveys papers from the 1970s! After a generation, the low hanging fruit has been picked, and where are we? Momentum and mean-reversion are part of the 'new' finance, and George Soros proudly notes that his theory of markets has markets biased—though it could be in either direction. Yet, these are not really discovered by behavioral economics, but explained by it on a case by case basis. Prospect theory teaches us that people overweight, or underweight, extreme observations in various contexts. These insights are no more useful than saying the effect of X on Y 'could go either way, depending on a bunch of other information we probably won't notice until after the fact'.

A new book about psychiatry, Doctoring the Mind by Richard Bentall argues that the science of the mind is hardly a successful science. He argues that mental illness is on the increase and sufferers in the developed world with access to psychiatric care actually fare worse than patients in poorer countries. It wasn't too long ago, many psychologists thought a lot of anxiety came from wanting to kill one's father in order to have sex with one's mother, an ambition so absurd only a scientist could think it made sense. Then there were lobotomies, homosexuality as a mental illness, criminals having insufficient self esteem (we now know they have too much), autism caused by bad mothering, and on and on. Lastly, have you ever thought that psychologists are happier, more well adjusted people than average? I don't think so, yet that's their specialty. So these guys are now going to help economics? I doubt it.

This is not to say I think economics is especially fruitful. My book addresses a rather central issue that I think is flawed in economics. I'm just saying that 'cognitive biases' are an embarrassment of riches that lead everywhere and nowhere, which Bentall's books suggests is hardly surprising.

6 comments:

Anonymous said...

Who cares about the pop sci books?

Mainly, people who do not want to, or can't, read the studies and analyse the experiments critically.
With practice I believe anyone can do it. Head over to Google Scholar, or ISI Web of Science if you've got an account, and give it a go.

I've noticed certain people are annoyed by these books:

- other authors who are jealous

- other potential authors who are jealous

- people who love science and don't like to see it watered down

Do you even know the difference between pscyhology and psychiatry? If you don't, then you have some more reading to do before you can start intelligently comment.

There is a hierarchy of science, outlined by Nobel physicist in the 70's:

Physics
Chemistry
Biology
Social sciences

( a very general synopsis)

Each one inherits from the other. That is the beauty of his organisation.

Once one gets to the level of social sciences things start to get murky.

People do not normally challenge the finding of experimental biology, chemistry or physics however, UNLESS they know something about the subject.

If one knows noting about experimetal chemistry and biology, one is not in a position to evaluate the safety and efficacy of psychiatric drugs. do you want an economist doing this work at the FDA? there is already enough economics involved before it even gets to the FDA.

and with respect to experimental psychology, one is not in a position to critique experiments if one has never themselves designed one. you can't just look at the conclusion and disagree. the conclusion is the leats important aspect of the paper. you have to support your argument by looking at the methods and the data.

it seems to me that economists may think that by just saying something negative about various things, they will apppear intelligent. or at least, not stupid.
or maybe they are truly just very pessimistic about all things.

but is this is any better than the person with the proverbial rose colored glasses? the economist's glasses are charcoal black. everything looks unsuccessful, even those things the economist cannot hope to understand, because he keeps away from experiments., and retreats to pure theory.

and economics is billed as a "way to get the most out of life"?

how? by looking at each and every action and criticising it? even those outside one's field of expertise?

the way to get the most out of life is to take risks. i'm sure any economist would agree.
when you do experiments you risk being wrong, unlike with pure theory.
moreover, you have to attempt to figure out what results mean, even if they were not as expected. this is no easy task.

you think designing experiments is easy?

i applaud kahnemann for taking risks. he has put his theories to test and whatever credit he gets he's earned it by taking those risks, by DOING THE WORK.

there is much more to his work than just prospect theory. that was just the beginning.

ecomomists can only thank him for doing the work they don't have the motivation or skills to do themselves.

this post is all about jealousy.

who cares about pop sci books?

Anonymous said...

we should put an economist in one of kahnemann's experiments and watch him behave just like every other human.

oh wait, human behaviour has totally changed in 30 years, so all those studies are no longer valid.

head over to Edge.org if you want to see some more intelligent people try to disagree with Kahnemann's work.
you really need to go back and read jon bargh's work to begin to understand anchoring.

it's uncomfortable to learn some things about ourselves perhaps, but that doesn't make these things invalid.

as for utility, were they intended to be used for anything, except to better understand our own behaviour?
i don't know.

but guess who wants to use them? economists!
grasping for straws.

if you don't like the results these guys got, or you disagree with their conclusions, then you have to do the experiments yourself and show that you did not get the same results. otherwise you are just wasting your breath, or keystrokes.

do the work.

Anonymous said...

Anonymous 11:33 said "if you don't like the results these guys got, or you disagree with their conclusions"

Anonymous 11:05 said "one is not in a position to critique experiments if one has never themselves designed one. you can't just look at the conclusion and disagree"

True, but you can have an opinion about whether the results could potentially be useful. I think that's the point here.

By the way, good blog Mr.Falkenstein, thanks!

ilene said...

Hi, do you have an email address or any way to get in contact with you? My email's ilene@philstockworld.com - if you could send me an email, I'd have it. Thanks! Ilene

Anonymous said...

Anonymous 12:37

re: question of utility

If that is indeed the point of this comment on Kahnemann, psychiatry and psychology in general, then it raises the questions:

What are you aiming to do?

How intelligent are you?

Once we know what you're after, and your level of ability to use information, then we can assess and opine how useful something may be to you.

I think the fact that someone reads a study that clearly focuses on individual human behaviour and then attempts to extrapolate the findings to something non-human, an abstraction such as "a market", says a great deal about their ability to interpret and use information.
Not to mention paying so much attention to what is written in the popular media, as if it will bring "useful" information.

I've never tried to apply economics to explain or predict human behaviour, and I never will.

I simply do not understand the pessimism of economics. Of course I could say economics is continually "unsuccessful". All I have to do is be upset about the way things are, no matter what they are, and say, "See, it failed. Again."

Brilliant.

Maybe the real reason economists keep looking to psychology is the hope it will cure their pessimistic views. In the long run, they just can't get happy.



Again, answer the simple question "What are you aiming to do?" and then we'll talk about usefulness.

Meanwhile, take a tip from psychology, get happy and stop complaining.

Robin Hanson said...

Yes, behavioral econ can look like "it all depends" from the outside. I do think they know some things about which biases happen when, but mostly about the lab; it is much harder to map that into which biases go which way in the field.