The Upside of Irrationality


Another?pleasurable summer read for me was Dan Ariely‘s?The Upside of Irrationality.? Put simply, the book is an impressive achievement.? It interweaves Ariely’s compelling personal narrative with what seems like dozens of his own super-interesting academic experiments.? Ariely explains how his own struggle with being severely burned as a youth put him on the path to being one of the world’s premier behavioral economists.

His previous book,?Predictably Irrational, relied a bit on his burn story to motivate some of his academic studies.? But Upside is more revealing.? It lets you see what makes Ariely tick, and how he comes up with testable hypotheses.? The fact that he could write this book without the help of a gifted writer like Dubner makes the accomplishment all the more stunning.? It’s one thing to have somebody else profile how you think.? But it’s a higher degree of difficulty to write about yourself – especially if you’re an economist.? Gearheads normally can’t write interesting analytic autobiographies.? But Ariely has.

I’m not, however, a fan of the title.? Ariely says that the book is going to show how irrationality is sometimes a good thing.? He argues, for example, that we irrationally mispredict the true trauma of divorce:? “[A] divorce is often less devastating to a married couple than either member might anticipate.” (Kindle 2410)? If we were more rational, more people would split up.? The upside of irrationality to society, he suggests, is that it keeps us together more than hyper-rationality would.

The book isn’t a sustained or thorough exploration of circumstances where irrationality leads us to make better decisions.? There are still plenty (I’d even wager a majority) of stories where irrationality leads us astray. The best reason to read this book is to experience how an interesting mind works and read about experiments that might change the way you organize your life.? If the hidden objective of?SuperFreakonomics is to?save lives, the hidden objective of Upside is make us happier.

For example, Ariely draws practical conclusions from fascinating experiments that were conducted by?Leif Nelson and Tom Meyvis.? Since humans’ happiness quickly adapts to new situations (like divorce), these scholars asked: how does interrupting unpleasant or pleasant tasks disrupt the adaptation process?

They ran a lab experiment in which one set of subjects had to listen to a noisy vacuum cleaner for five seconds.? A second set of subjects had to listen to the vacuum for 40 seconds, and a final set of subjects had to listen for the same 40 seconds “followed by a few seconds of silence and then an additional burst of five seconds of the annoying sound.” (Kindle 2468).

The researchers then asked each group to assess how annoyed they were by the last five seconds of the noise.? Somewhat surprisingly, they found that the group that had to listen to just five seconds was a lot more annoyed (by the last five seconds) than the group that was able to adapt for 40 seconds.? But the kicker was that the annoyance rushed back for those in the third group who took a break and had to listen again.

Here’s an example where humans are again predictably irrational.? “You may think that taking a break during an irritating and boring experience will be good for you, but a break actually decreases your ability to adapt, making the experience seem worse when you return to it.? When cleaning your house or doing your taxes, the trick is to stick with it until you’re done. ” (Kindle 2494).

The flip side of the interruption principle is that we should look for opportunities to interrupt pleasurable experiences.? Ariely describes a parallel experiment (of Nelson and Meyvis) in which one group of subjects received a continuous three-minute massage, while a second group was massaged for 80 seconds followed by a 20 second break, and then was massaged for another 80 seconds.? The second group was massaged 20 seconds less than the first group but said they enjoyed the total experience more:? “As it turns out, those who underwent the shorter massages with the break not only enjoyed their experiences more but they also said they would pay twice as much for the same interrupted massages in the future.” (Kindle 2495)

To be sure, Ariely?is not the only one to suggest that you might be happier spending money on things where adaptation is less likely (because it is more interrupted).? That’s why some people argue you’d be happier spending your money on a few vacations instead of a new car or a couch.? The idea is that a nice couch is too easy to get used to.? Because it provides uninterrupted pleasure, it drifts into our cognitive background. In contrast, a short vacation is almost by definition an interruption of the routine. This simple idea that you should interrupt pleasure and avoid interrupting the necessary annoyances in life is news that, on the margin, I have been trying to put into practice.

Beyond adaptation, Ariely has important things to say about the effectiveness of incentives (a topic of a future post), and why we value things we make more than things we buy, and the role of looks in the dating market.? In each and every case, his observations are driven by controlled experiments that serve to simultaneously motivate his theories and ground them in real-world observations.

I’m also attracted to Ariely’s innovation of adding biographies of his coauthors as a special appendix.? With social science studies, which often have multiple coauthors, it’s a bit unwieldy to repeatedly mention “X and Y showed this” or “X, Y and Z then calculated that.”? But Ariely pulls off such references with a grace at which I can only marvel.? As a sometimes writer in the general genre, I could do worse than going to school on this fine book.

Ian Kemmish

Time to play Devil's Advocate: it seems that you can label pretty much whatever you want as "irrational" these days.

Take the divorce question: if you anticipate beforehand that your divorce will be worse than the average, then you're just planning for the worst. You're taking into account the "long tail" risk that your divorce will be really bad. You can call that irrational if you want, and you may or may not be right.

Then again, bankers and regulators are currently being lambasted for failing to plan for the worst during the noughties, for failing to take the "long tail" risk into account. You can call that irrational if you like, as many fashionable commentators and authors have done, and you may or may not be right.

What you can't do, however, is call both these polar opposite behaviours irrational. Either considering the risks is irrational or ignoring them is, but not both!


Barrett Jacobsen


I believe the going definition of an irrational act is any act where I'd say I'd have done things differently (regardless of if I would have actually done things differently, and regardless of any asymmetry in knowledge).


I don't see the problem. Social science doesn't use the term rationality in the same sense as a more philosophical approach does. They use the term rationality in the sense of the modeled rational man. Real man diverges from the assumptions of rationality as modeled.

In some cases, this means man is irrational in the philosophical sense. In other cases, this means man is rational in the philosophical meaning of the term but not in the more heavily constrained and precisely defined term used in modeling behavior. In this sense, man can be rationally irrational. This form of irrationality could theoretically be modeled using a different set of assumptions about what defines rationality. It would have a different subset of rational behaviors, which may or may not be as analytically useful as current rational assumptions.

That the current definition of rationality has been used so successfully for so long hints to me that the current definition is still the best one to use as a default. Which isn't to say that in particular contexts a different model will be more helpful and we're increasingly learning enough to tell when to use each model.



I thnk I will take the first opportunity to use this principle in the greatest and most pervasive of all irrational acts: gambling. Casinos want you to enter their hyperstimulant, windowless realm to lose for long periods of time. They want you to leave with nothing, and many do. At base, gambling is pleasurable; a series of games in which a player has the opportunity to win in any given iteration. For most, after a couple of hours it produces abject regret, completely overshadowing the small victories that are essentially the reason for participating. Next time I gamble, I'll stop to have a bite to eat, talk up the old ladies at the penny slots, treating run of games as a good experience that I want to extend, renew, and repeat. Too often I take the approach of constant betting, only exhausting more of my savings and expediting my (inevitable) demise.

Fritz Mills

Many years ago, I won a sales contest that involved a trip to Miami Beach in February (from Chicago). As a life-long Chicagoan, I had never taken a trip south in the winter, and Chicago's winter weather had never really bothered me. But after that long warm sunny weekend, the return to wintery weather was pure hell. Since then, I have never taken another winter trip to a warm climate before mid-April. I have had to make mid-winter sales calls in southern California, but my grizzled expectations (and perhaps the fact I was focused on business rather than pleasure) kept the post-trip annoyance factor down.


I just want to say that I also am enjoying this book and I really like the use of the Kindle reference points.

Joel Upchurch

This raises an interesting notion, that you might enjoy more taking ten three day weekends, rather than taking a two week vacation. From my personal experience, a couple of months of four day work weeks is a lot relaxing than a formal vacation.

John Vidale

I agree entirely with Ian's article, but the vacuum cleaner experiment seems ill-constructed, so far as I can tell from a brief reading of the article.

Participants were only "told that they would be listening to a brief sound clip of a vacuum cleaner". I imagine the irritation with restarting the noise is more uncertainty when it will next end, given that the first time it lasted 40s, than the claimed accumulation of listening time.

I thought it was well documented that irritations over which one has control, such as doing chores, is less annoying than irritation over which one has no control, such as when the noise starts and stops in this experiment.

I'd hesitate to extrapolate such logic to doing my chores in one go rather than breaking them up into whatever chunks seem less tedious to my un-social-science educated intuition.

Sebastian Villalobos Moltedo

I've read both of Ariely's books. I must say they 're astonishing. Nevertheless, Predictably Irrational is a much better piece of "gorgeousness".
I've even passed the book to some colleagues, so that we build somewhat like The Power of Free and the Decoy club.
Thumbs up!
What I'd really like is to have Dan's suggestion for further reading.


opportunities to interrupt pleasurable experiences - UH, I BELIEVE THIS IS CALLED WORKING.