Thursday, March 13, 2008

Economist's Notebook: Irrationality

Elizabeth Kolbert of the New Yorker recently wrote a review of a book about behavioral economics which she uses as an opportunity to criticize economics - but I believe she has totally missed the point. Her errors are all too common among critics of my profession, I fear, which leads to very self-satisfied attacks on a straw man whom I don't recognize as being much related to what I do. She begins by describing a supposedly irrational decision she made: buying a Tintin book for her daughters to get free shipping on an order that originally had only one book she was buying for work. But why does she think this is irrational? By qualifying for free shipping, the marginal price of the new book is now lower and it is quite reasonable to suggest that this book might be desirable at $3, say, when it was not at $7. In fact, why does she think Amazon makes these offers? If everyone decided irrationally, there would be no reason to make such an offer. In reality it is the very fact that people act predictably in the face of such an offer that they exist at all. So how can you call that irrationality?

She says that the shipping cost would have been borne by the New Yorker anyway - but it is not irrational to care about saving the place you work money. I do it constantly - I am about to fly to Nashville to do research with a colleague at Vanderbilt. The state approved travel agent found a flight for $800. I could have just said yes and been done with it. Instead I did my own search, 'wasted' and hour, and found a flight for $300. I am happier now because I did that, so it was worth it to me (of course my flight is on Southwest and they just grounded almost 40 planes, so maybe I messed up).

This idea that economists believe that individuals are only motivated by money is common. This is nicely evidenced by this letter to the New Yorker about Kolbert's article. In it, the writer exposes the stereotype: "People’s behavior is governed by feelings. No one would even bother to buy a house or a candy bar unless he wanted to. And one important measure of a successful purchase is how you feel about it, damn the numbers." But this describes how economists view behavior perfectly. Economic theory only demands that people's decisions are internally consistent. Lets run with this example. You buy a house because you feel good about it for, say, $300,000. Now, do you feel good enough about it to spend $350,000? $400,000? $500,000? At some point the good feeling you get from owning the house is eclipsed by either the cost outweighing the good feeling or the price exceeding your budget. Aha! So, I can predict that when the house price rises, there will be fewer people who want or are able to buy it.

And sure enough, when economists have looked into all kinds of areas of human behavior you might not think as being governed by the laws of economics, we find behavior that matches exactly our predictions: marriage, divorce, sex, drugs, crime, you name it. My favorite example, written by a friend of mine, is a study of births right around the new year. You might think that there is no way economic incentives could influence that most sacred of human activities: bringing a new life into the world. You would be wrong. You see, if you have a baby born in a calendar year, regardless of when in that year it is born, you get the federal income tax credit for that child. That means that if you are in the ninth month of a pregnancy and due right around new year's day, there is a strong incentive to induce or have a c-section before the clock ticks midnight on December 31st. Sure enough, in the US there is a spike of births in the last week of December and a trough in the first week in January.

As for the irrational behavior uncovered in the book, we are learning all the time that we often make decisions that are influenced by things we might not have expected - for example framing choices (as Kolbert mentions). The strength of neo-classical economics is how flexible the theory is - these are things we can incorporate into the theory, not things that destroy the very foundations. In addition, we are studying these things within the discipline, not ignoring them. We are constantly testing our assumptions as any good academic discipline should do, so this speaks to the health of the discipline, not the sickness. Lastly, much of what we find in laboratory experiments does not coincide that well with observed behavior in real life, suggesting that it is important not to make too much of the magnitude of laboratory evidence - but we should pay close attention to the qualitative evidence. So framing might not be that severe in real life decisions, but it certainly occurs.

It is amazing to me that at a time when computing power and data are such that we are now finding evidence of behavior that corresponds to economic theory in so many areas there should be such a desire to suggest that we have it all wrong, but I think it is perhaps evidence that we all like to think of ourselves as exceptional, not beholden to the same incentives as everyone else. Economics, perhaps people feel, reduces people to simple calculators, which does not coexist with the 'humanness' we all feel defines us.

So let's play what Robert Frank calls the 'economic naturalist.' I'll leave you with an irrational puzzle and you play the economist: In my local Safeway, they constantly have "two for..." sales. The big red tag will say something like "two tubs of salsa for $5" and then in smaller print, "$2.50 ea." So really the sale is just $2.50 per tub of salsa. Why on earth do they bother with the 'two-for' ruse? Wouldn't economic theory predict that it should not matter one wit? Can economic theory explain this? You tell me...


Jeff said...

"when economists have looked into all kinds of areas of human behavior you might not think as being governed by the laws of economics, we find behavior that matches exactly our predictions: marriage, divorce, sex, drugs, crime, you name it."

Yes, but...a cynic might say that when something doesn't match our predictions (i.e., the economic explanation didn't work well), it's a 'bad result' that can't get published in a major journal, and therefore doesn't enter into the economics mainstream... we don't hear about the cases in which an economic explanation didn't work.

I once had a sociologist ask me: do you think there's anything that economics cannot explain? I don't remember what I said, but at present, I would say that just because we can think up an economic explanation for many phenonmena (i.e., an explanation involving optimization and equilibrium), this does not mean that this is the only one or the right one.... testing remains imperative.

Jeff said...

"It is amazing to me that .... there should be such a desire to suggest that we have it all wrong"

Well, there will probably be no shortage of attacks on the economic way of viewing the world because:

(a) economics is colonizing areas previously thought to be the turf of sociologists, geographers, and psychologists;

(b) economics' core concepts (optimization and equilibrium) can seem sterile and simplistic as a means of viewing human behavior.

Since most critics don't really understand how flexible economic theory is, there will probably continue to be attacks and references to the "the dismal science...". In all, however, the criticisms are a sign of an ascendant science and something that I think is healthy.