I first encountered Dan Ariely on the radio show Marketplace, where he offers up little nuggets of research data from the new field of behavioral economics. Because of the individual scale of the research many of Ariely's findings have some personal finance implications. Consider the pain of paying. This is the finding that when people pay with credit as opposed to cash for dinner, they are willing to spend more. Why? Because credit cards decouple the psychic "pain" of payment from the specific act. The act of deferring reduces our pain at the damage done, and allows consumption with less guilt and discomfort. The big-picture implication of this is obvious when it comes to the credit economy, but for myself I have taken to always paying for frequent small purchases with debit or cash. I do put less frequent large purchases on credit cards on occasion, for example when I buy a plane ticket for a trip months into the future. My reasoning is that I can concentrate in a reflective and rational manner on a few large purchases to a far greater extent than I can on the habitual small weekly food purchases. Monthly automatic cell phone payments for example are naturally a different beast than going to the grocery store; not only are the payments infrequent and regular, but the interval of their expense is predictable (totally predictable if you have some sort of unlimited plan).
In any case, because of the nature of the topic which Predictably Irrational covers the chapters have an a la carte feel. Ariely attempts to stitch them together into a bigger narrative about the limits of the rational actor..but this is clearly a "they had to do research on that?" insight for the lay audience (though perhaps not for economists). Interesting Ariely reports that strangers are much more interested in his findings since the financial collapse, probably explaining why he's coming out with an expanded and revised edition later this month.
To give you a flavor of Predictably Irrational, here are summaries or examples from each chapter.
1) The Truth About Relatively
If given these choices to subscriptions to The Economist
a) $59 - internet only
b) $125 - print only
c) $125 - print & internet
vs.
a) $59 - internet only
b) $125 - print & internet
Most people choose c) in the first circumstance, but a) in the second ("people" in this case means MIT Sloan School students).
2) The Fallacy of Supply and Demand
In auction for various products students were asked to write down their bids, after entering the last 2-digits of their social security number (presumably for purposes of identification, but actually to "prime" them with those numbers as "anchors").
Here are the bids along with SSN's:
| Average bids for products by last 2 digits of SNN | ||||||
| 00-19 | 20-39 | 40-59 | 60-79 | 80-89 | Correlation | |
| Cordless trackball | $8.64 | $11.82 | $13.45 | $21.18 | $26.18 | 0.42 |
| Cordless keyboard | $16.09 | $26.82 | $29.97 | $34.55 | $55.64 | 0.52 |
| Design book | $12.82 | $16.18 | $15.82 | $19.27 | $30.00 | 0.32 |
| Neuhaus chocolates | $9.55 | $10.64 | $12.45 | $13.27 | $20.64 | 0.42 |
| 1998 Cotes du Rhone | $8.64 | $14.45 | $12.55 | $15.45 | $27.91 | 0.33 |
| 1996 Hermitage | $11.73 | $22.45 | $18.09 | $24.55 | $37.55 | 0.33 |
3) The Cost of Zero Cost
Choice between Lindt truffles & Hershey's Kisses:
Lindt truffle - 15 cents (73% prefer)
Hershey's Kiss - 1 cent (27% prefer)
Lindt truffle - 14 cents (31% prefer)
Hershey's Kiss - Free! (69% prefer)
4) The Cost of Social norms
From page 71:
A few years ago, for instance, the AARP asked some lawyers if they would offer less expensive services to needy retirees, at something like $30 an hour. The lawyers said no. Then the program manager from AARP had a brilliant idea: he asked the lawyers if they would offer free services to needy retirees. Overwhelmingly, the lawyers said yes
5) The Influence of Arousal
See this post.
6) The problem of procrastination and self-control
Students at MIT are offered a choice when to hand in their 3 papers in terms of date. They could select dates at the end of the term for all 3, or space them, or select the beginning of the term. There was a penalty for handing papers in late after your selected date, but no penalty for handing papers in early. Most students chose to space the papers, so that they would pay a penalty if papers were handed in late. The students who chose not to space the papers but allowed that they could be handed in at the end of the term did worse in terms of grades than those who self-enforced deadlines with consequences.
7) The High Price of Ownership
Duke students who didn't manage to get a hold of the tickets for a national title game were asked how much they would pay to purchase a ticket from a student who managed to get one (tickets being distributed by lottery). Those who had them were asked how much they would sell them for. The average offered purchase price was $170, the average selling price was $2400. None of the students who wanted tickets were willing to pay the prices that the students with tickets were willing to sell them for.
8) Keeping Doors Open
The experiment in this chapter is hard to describe easily, and there are no charts. So all I'll say that many times people perform better when given fewer options. If the number of options increase so that the range of outcome is greater (you could do way better or way less than the bounds of events with smaller numbers of options) it seems that people do worse than you would expect, frivolously exploring options.
9) The Effect of Expectations
MIT students are given two beers to sample. One is Sam Adams, the other Sam Adams + vinegar. Students told the nature of the options for the taste test prior to tasting rate Sam Adams better. Students told after the tasting rate Sam Adams + vinegar higher, and will add vinegar to Sam Adamsi> to have some more of the brew.
10) The Power of Price
Individuals were given a Vitamin C placebo which supposedly reduced pain in ~10 minutes. They were also placed in a waiting room before administration with brochures about the drug that listed its price as $2.50 per single dose. All reported subjective feelings of pain reduction afterward. The same experiment was run with the change that the brochure listed the price as 10 cents, instead of $2.50. Only 50% reported pain reduction.
11) The Context of our Character, Part I
Students are given a test, and allowed to cheat in a variety of ways. The likelihood of being caught varied between experiments. The results show once cheating was possible, gradually reducing the likelihood of being caught did not result in more cheating. Rather, there was a threshold effect and stasis beyond the threshold.
12) The Context of our Character, Part II
6-packs of Coke and 6 dollar bills are left in dorm fridges. After 72 hours all the cokes were consumed (by individuals who did not buy the Coke 6-packs, so they were stealing). After 72 hours all the plates of dollar bills remained.
13) Beer and Free Lunches
When people are forced to order beers out loud in sequence they choose different beers; there is a uniform distribution of preference. When people order by writing down their selection privately they have strong preferences and there is modality, as some beers are preferred and others are seldom ordered.
The book has a website with blog.
- Log in to post comments
Moral Sentiments and Material Interests by Gintis, Boyd, Bowles, and Fehl is behavioral economics in an evolutionary context, and makes a case that solidarity is a more important human motivation than economists have realized.
I believe that economic rationality was never more than a formal, counterfactual working assumption, but became fossilized in economics because it made it possible to develop elaborate theories and models, and for a long time was enforced on students in the early days of training and not questioned for decades. The assumption of a certain kind of individualist rationality made it possible to treat human beings as hard little atoms whose behavior is predictable. Pretty goddamn wishful, and with all the many forms it now takes, pretty incoherent.
Amartya Sen has developed a concept of economic rationality which is not behavioral, but which does bring the economic concept of rationality closer to the commonsense concept of rationality (as an attainment or valuable trait, rather than as a universal quality), but I can't see how you could model it (not that that would be a bad thing). He points out that the ideally rational economic man is a blind obsessive and perhaps a sociopath.
Economics is confused as to whether it's a descriptive, predictive system or an ideal model of social and economic organization. It pretends to be a real descriptive science in order to gain science cred allowing it to horn in to the front of the policymaking line, but it's so riddled with ad hoc and/or counterfactual assumptions that it's pretty fake. (Sure, physicists do that too, but there's a difference between a teaspoon of salt and a 50 gallon drum).
Normatively economists, at least the pure economists, seem to claim that a.) in a pure market system, rational individuals will be rewarded, and irrational individuals will be punished, and b.) that a system of that type will be the most productive, and the reward distribution will be the most just. That's the economic utopia, and behavioral economics doesn't harm it. But behavioral economics does damage the claim that this utopian system is also is some way rationally descriptive and predictive.
Economists have always had trouble with bubbles, like the one we just experience, and this is partly because not only are people not totally rational and not only do they not have perfect knowledge, but besides that, they communicate with one another, so the irrationality is not randomly distributed so that the irrational individuals are weeded out, but can pervade a whole population.
When their fictitious models fail, economists atart talking about psychology ('animal spirits"). But it's not individual psychology, but group psychology, which is sociological (and cultural) and not merely psychological. (At my URL: My social psychological explanation of why the bubble happened.)
"The big-picture implication of this is obvious when it comes to the credit economy"
Really? Let us slow-pokes in on these obvious implications.
Yes individuals are irrational twits that don't look anything like homo economus. Everyone, including economists, know this. The fact is, though, market outcomes look like they are a result of rational actors (see Vernon Smith's work). Economics is concerned with market outcomes and so the irrationality of individual actors plays *zero* role in economic analysis.
As I'm happy with the division between nuclear physics and chemistry, I'm happy with the division between psychology and economics.
Yes individuals are irrational twits that don't look anything like homo economus. Everyone, including economists, know this. The fact is, though, market outcomes look like they are a result of rational actors (see Vernon Smith's work). Economics is concerned with market outcomes and so the irrationality of individual actors plays *zero* role in economic analysis.
what about systematic biases in the population? in any case, yes, i want credit to contract. i'm not going to argue with you since you have vernon smith on your side ;-)
As I'm happy with the division between nuclear physics and chemistry, I'm happy with the division between psychology and economics.
i think the analogy between nuclear physics and chemistry fails. there's a reason that general chemistry has sections with cover nuclear physics (the periodic table makes more sense, etc.). nuclear physics and ecology might be better.
There's theoretical economics that deals in psychology, but for those of us doing 'real' economics psychology doesn't really matter (for doing economics at least). Maybe a better analogy is with biochemistry and physics. My fiancée does medical research and she tells me that intro chemistry was the only time she studied atomic structure.
Minor point, but I have never understood this thing about credit card payments being less painful than cash. I use my credit cards as little as possible (I like to be able to read my statements and remember every purchase on them), but when I do use them, whatever the amount, it feels exactly like paying cash.
This may be the reason I don't have any debit cards. I just don't understand their advantage over credit cards! I lose the option (although I never use it) of deferring payment, and I can't see that I've gained anything in return.
My fiancée does medical research and she tells me that intro chemistry was the only time she studied atomic structure.
well, physical chemistry. there was a lot of quantum chemistry when i took it (it is required for biochem majors usually i assume). though i think the analogy works,you don't really need quantum chemistry for biochemistry (someone who does research in some abstruse area of biophysical chemistry will now correct me!).
that deals in psychology, but for those of us doing 'real' economics psychology doesn't really matter (for doing economics at least).
that's fine. then the government should then start hiring a lot of psychologists and cognitive scientists to supplement the economic policy analysts. i don't care what you call them.
"This is the finding that when people pay with credit as opposed to cash for dinner, they are willing to spend more. "
Credit-card issuers refer to this tendency as "spend" and expend a fair amount of effort to quantify "spend" in various situations. It's at the heart of their efforts to displace cash with plastic for all transactions. Hence the hard sell to convince people that paying cash is bad. You've seen the ads. Why look at all these happy people just so easily and quickly spending, spending, spending! Only the clumsy doofus pays cash. But the reality of course, is that only the clumsy doofus actually has cash to spend. Everyone else is now in hock up to his eyeballs.
"The fact is, though, market outcomes look like they are a result of rational actors"
Really? Like the recent real estate bubble? Or the tech stock bubble of 90s (or even the one of the 60s)?
s. I just don't understand their advantage over credit cards! I
the advantage is that it comes out of your account at that time, so you feel the "pain." it might not matter for you, but subjectively i think it makes a difference for me, so i use debit. i prefer it to cash since i can keep track of my purchases online if i use debit (my bank has a nice budget application with breaks down my spending by sector).
But the reality of course, is that only the clumsy doofus actually has cash to spend.
well, if you use a debit card it comes out immediately. the only thing though is i don't know if the physical manner of paying, the resemblance of the debit card to a credit card, has an effect. that might seem like a paranoid thought, but there's plenty of weird research that shows totally irrelevant cues can change behavior.
"then the government should then start hiring a lot of psychologists and cognitive scientists to supplement the economic policy analysts."
Why would policy makers care about individual behavior if it doesn't impact social behavior?
will, individual behavior which is systematic across the population does impact social behavior. of course you might disagree, and in the sciences it is normal for intuition and perception to not match up with the deep patterns of reality. but most people aren't willing to cut your discipline slack based on theory when it comes to counter-intuitive statements right now since it doesn't seem as if you had a good grasp of the local empirical topography. engineering validates the bizarre assertions of science. i obviously don't know a lot about economics, and 1 year ago i would probably have retreated from my own naively held positions based on the body of knowledge in economics if you'd pointed to it, but at this point i'm not very inclined to do that. and it's not based on caprice. just show me the $$$. to the predictors go the spoils.
razib, look at Smith's work. He shows that stupid actors aggregate into smart markets. This is not theory, this is experimental evidence.
ok, i'll check out his work. is there a specific paper you recommend as representative?
I wonder whether there are any cross-cultural studies on credit card payments. I have seen some people of Indian origin in Australia actually saving money using the time gap between purchases and the deadline to pay the credit card company.
gaddeswarup, i wouldn't be surprised if there's a big difference in how credit is used in societies which didn't have good access to consumer credit until recently.
"He shows that stupid actors aggregate into smart markets. This is not theory, this is experimental evidence."
you failed to answer the quesion on the housing market. The one which just collapsed. Why was that market smart?
The original version of the Alchian theorem explains how a bunch of idiots can still create an effective market. Doesn't mean they can't also make a defective one though.
Check out Ariely vs Harford.
As I'm happy with the division between nuclear physics and chemistry, I'm happy with the division between psychology and economics.
Except that whenever economics fails, economists start talking about psychology -- "irrational exuberance" or "depression".
The problem isn't irrationality of individual actors, it's irrationality shared by large social groups sharing a culture and communicating with one another. Economics and psychology are both individualist, and economics can handle a group of isolated individuals of varying degrees of rationality, but it can't handle social groups of individuals communicating with one another.
razib, look at Smith's work. He shows that stupid actors aggregate into smart markets. This is not theory, this is experimental evidence.
Recent history shows something quite different.
Pushmedia1, you did a great job of parroting back the orthodoxy you were taught, but you don't seem to understand that the orthodoxy has collapsed. As I indicated in the last two paragraphs of my first post.
razib, I suppose Smith summarized his research on this in Rationality in Economics. I've just read the intro but that put it on my to-read pile. I think you'll like him.
eoin, Smith's work shows that dumb agents end up converging in their collective behavior to what would be predicted by economic theory. Price movements, even severe ones, aren't precluded by economic theory. That said, I don't know anything about the housing market and so I couldn't tell in particular why housing prices change. It could be that the conditions in Smith's experiments don't hold in that market, it could be that economic theory has problems, or it could be that this wasn't an irrational bubble (that there was some method to the madness). The point of my comment is that individual irrationality doesn't translate simply to collective behavior. Its easy to imagine such a connection but experiments have shown no necessary connection.
TGGP, theory, in this case, is weaker than the data. Idiots almost always converge to rational behavior.
Idiots almost always converge to rational behavior.
Almost?
It could be that the conditions in Smith's experiments don't hold in that market, it could be that economic theory has problems, or it could be that this wasn't an irrational bubble (that there was some method to the madness)
This is really what's at stake here.
You have to quit talking about data vs. theory. What you're saying is that in experiments designed in a certain way, you get a certain kind of result. That's data all right, but if the experiment is designed too narrowly, the significance of the data is slight.
John Emerson, I'm not sure what you're point is. By "orthodoxy", I assume you are referring to neoclassical theory. By "spouting orthodoxy", I assume you think I'm just reciting theoretical results. I'm not doing that.
If I can't point out that experimental results seem to confirm theory, what can I say? I agree that the median economist thinks they're studying rational individuals when, in fact, they're not. Its not my fault that data seems to make this rhetorical stance harmless.
By "recent history" what do you mean? You mean the current recession? Neoclassical theory predicts business cycles. Do you mean neoclassical theory didn't, like Nostradamus, divine exact paths for macroeconomic variables? If so, I have the perhaps dissappointing reply that business cycle theory is a stochastic theory and only has predictions about distributions of time-series, not their exact instance paths. Sorry about that.
I'm also sorry to have to predict for you that we'll never have a deterministic theory of business cycles. This is because if we could predict them, they wouldn't happen.
Ariely's book is interesting... although I disagree with the implications that being irrational is "bad" (in my book Neuro Web Design I discuss that this is just the way our brain works and has evolved)... and that if we know we are being "irrational" we can change it! Jonah Lehrer's book How We Decide is a good one too.
"I'm also sorry to have to predict for you that we'll never have a deterministic theory of business cycles. This is because if we could predict them, they wouldn't happen."
However that is what we want from economics. That's what would make it useful - economists who can predict a recession, or a bubble based on certain actions - warn us about them, and let the political classes fix the issue in a timely matter. We don't care that it doesn't happen, we are human agents who can change our actions.
Similarly we need the predictive powers of the CDC (Centre for Disease Control) to predict a possible swine flu pandemic which then doesn't happen precisely because they predicted it, and action was taken.
Frankly I am not convinced that Economics is that useful if it does not have predictive powers, and - and we see - is theoretically flawed anyway.
The problem is not that the economic world is stochastic but that you probably don't have the right models to begin with.
With the right models we could work out if new financial instruments - like Credit Default Swaps - would have the effect of a unsustainable bubble which had a chance of bringing the world into recession. Or we could add as variables the extension of credit to sub prime candidates, and securitization. Pump that in to a computer model, and get a result. Even if the result depended on different computer runs and models - like weather forecasting a week ahead - we would get some idea as they converge. Jesus, look! A boom followed by a recession say 80% of the models.
The reason we cant do that is we don't have the proper mathematical models of human behaviour to such stimuli. Either the economic world is too complex for humans to model, or it is too complex for economists to model - which is not quite the same thing.
@eoin and @pushmedia1: I have done quite a bit of modeling of mortgage prepayment/default, and from everything I've looked at, the housing market was definitely not efficient. I am not terribly familiar with Smith's work, so at this point I can't point to which of his specific assumptions and/or conditions of his experiments the housing market does not follow, but I do think there are some barriers to the market's efficiency purely from a capital perspective, and one of the requirements for any market to be efficient is liquidity.
Basically, it comes down to the fact that a) people need to live somewhere and b) with the kind of down payments people were putting down, they were generally only able to participate in the market effectively (i.e. without going bankrupt or walking away from their home instead of selling it for a lower price) if their home went up in value. This created an inherently very skewed distribution of potential prices, i.e. near term upward bias with potential for large downward movement of home-prices in part due to the local feedback effects (i.e. if people start defaulting on their mortgages in your neighborhood, your home becomes less desirable).
In addition, people can't stay in their homes because of rate resets and penalties, and penalties clearly severely affect liquidity. To be clear, this doesn't mean the MBS/ABS markets are necessarily inefficient, just the home price market, since MBS/ABS have the potential to be very liquid.
For what it's worth, Robert Shiller started a company to try to create an exchange traded trust to create a liquid instrument tied to the Case-Shiller index; unfortunately, it never took off. Home Price Futures are traded on the CME, but their spreads, especially for the derivatives with the longer maturities, are so high as to be practically useless (i.e. they are not very liquid either).
So, if you took the current housing market and made it more liquid, could it be efficient and behave ala the markets that Smith analyzed? I don't know, but it's definitely a bad example since even if people did behave rationally, a bubble could very easily have happened.
This is not cyclic, it's a globally-disastrous bubble collapse caused by irrational behavior not corrected or disciplined by the market until about ten years in. It was preceded by the dot-com bubble collapse less than ten years earlier.
You keep talking about the data in toy experiments, and claim that this proves that individual irrationality makes no difference, but the real-world data is showing something different. The scope of the experiments was too small.
There are specific contexts where individual irrationality does not make any difference, but the world as a whole is not one of them. You can set up an experiment to prove that individual irrationality makes no difference. You can even find places in the real world where individual irrationality makes no difference.
I recognized above that in certain contexts the market punishes most of the irrational actors and rewards most of the rational actors, and in these contexts individual irrationality is indeed insignificant.
The problem is group irrationality among intercommunicating individuals, and this is as much sociology, culture, or history as it is psychology in the limited sense.
The big problem I see with economics is that it doesn't seem to know or care whether it's trying to describe actual behavior, or to design (and propagandize for) an ideal economy. In the latter case you don't have to care about individual irrationality as long as it doesn't infect the whole and as long as only the irrational are harmed. But it seems that if you have enough individual irrationality in a group of intercommunicating individuals, it becomes group irrationality, and then BOOM!
A whole different question is whether economic rationality is a ood standard. That's what came up in the books by Sen and by Gintis et al, with regard to group loyalty and "fairness" vs. individual utility-maximizatio.
The book "Kluge" discussed some of these questions, but didn't seem to distinguish between economic and other forms of rationality.
"This is not cyclic, it's a globally-disastrous bubble collapse caused by irrational behavior not corrected or disciplined by the market until about ten years in."
John, are we to just trust your instinct that this is the case? Or do you have a model that connects systematic irrationality to macro-economic variables? Or is it enough to talk in slogans (e.g. "crisis", "global bubble", etc)?
I don't know if "economics" cares about explaining aggregate behavior or not. I know I do and most practicing economists I know, do. Also, its funny to go after experiemental evidence in the context of this post lauding it.
eoin, you're experiencing availability bias. The current recession was unpredicted. We don't experience the predicted ones. Heading off recessions is the job of the Fed and its been pretty good at it until recently. This is why this is the first severe recession in over 20 years. I'm pretty sure the Fed is learning from their mistakes even as we type.
Pushmedia1, this isn't between me and you. I've been following the discussions on the econoblogs, and your opinion is not universally held. Talking about the bubble burst as cyclic was definitely a big mistake. Most economists seem very uncertain about what's happening, and talk about animal spirits and irrational exuberance and the like.
You're an economist and I'm not, and probably nobody else here is, but you're pulling rank and bluffing. You're far more confident of your conclusions than the rest of the profession is. There's no one here to prove you wrong, but no one here should take what you've said so many times at face value.
until i read the smith paper i'm not going to say much, but i'm a little surprised at will's 'tude after greg clark's confessions. it's obvious economists and the variegated finance related professions (which are not economics, but lean upon economic theories, thank you dr. scholes and black) know a lot more than the rest of us, but after the past year one wonders what the fuck exactly they do know of any relevance to the rest of us :-) social science is really hard, and it is obviously not going to be easy to have both tractable and non-trivial models and inferences, and i am still really interested in economics. but if bridges started collapsing all across the world i would give a civil engineer who signed off on those bridges a good kick to the face if he tried to pull rank.
John, I can only express my opinions not the profession's.
Razib, the problem with your analysis is that bridges aren't falling all over the world. Some people, and many prominent economists, are using bridge falling narratives, but what we really have currently is a deep recession not very much unlike most post-WWII and pre-1985 recessions. Of course, we're in the middle of this one and it could get worse, but it could also get better. If, in the end, all the bridges come tumbling down, I'll be the first to throw all of economics in the trash heap. It hasn't happened (yet (thank god)).
Clark... well he likes to throw stones and we love him for it.
will, my analogy used an exaggeration. that being said, i used to read a lot about 'the great moderation'....
Pushmedia1: you expressed your opinion with undue confidence, and you also tried to speak for the profession: "for those of us doing 'real' economics psychology doesn't really matter (for doing economics at least)".
People really should know that there are a lot more big open questions in economics, these days than there were a couple of years ago, including questions about what irrationality and about the the real-world effects of individual and group "irrationality" on economic phenomena.
There's a lot of soul-searching going on in econ these days. People know that something has gone wrong, and they're looking a lot deeper than they used to. To my mind this is long overdue. (I'm a long-time econ skeptic, but recent events startled even me; the problems were in areas where I had thought that econ was pretty much on top of things.) Even the blackballed schools (Marxists and Austrians) might be able to get into the action.
The links below let you eavesdrop on the internal argument. It's a pretty mainstream list, with the partial exception of the last two. Someone else's list would be very different.
http://delong.typepad.com/
http://economistsview.typepad.com/
http://rodrik.typepad.com/
http://www.ft.com/comment/columnists/martinwolf
http://www.nakedcapitalism.com/
http://econospeak.blogspot.com/