Notes on The Upside of Irrationality by Dan Ariely

I recently enjoyed listening to the audio version of Dan Ariely's The Upside of Irrationality. He's also the author of Predictably Irrational, which I hope to read soon.

The author, a behavioral economist and psychologist, begins by introducing an irrational behavior that affects all of us: procrastination. He says that procrastination is universal. When he sought a special treatment for a rare condition he had, he needed to inject himself with very uncomfortable medications on a weekly basis. In order to get himself to not procrastinate, he motivated himself with watching movies on his injection days, thereby associating good with bad.


As humans, we prefer short-term, not long-term goals. We routinely fail to take short-term pains for long-term goals. He explains that there is a lot of human-incompatible technology out there that doesn't take into account our fallibility (like the stock market). While his first book (Predictably Irrational) is about the downsides of irrationality, this book is about its upsides.

The first part of the book is about defying logic at work. Throughout the text, the author explained in detail the scientific studies used to come up with the conclusions presented. It was neat to understand how the experiments were designed.

• Paying more for less (bonuses): The author ran experiments on rats about incentives to understand how animals respond to different strengths of incentives. The rat experiments showed that increasing the magnitude of an incentive does not markedly change performance (above a certain medium level of incentive). Therefore, paying larger CEO bonuses doesn't help either.
  
  Medium motivation turned out to be better than small or large motivation. The author also found that social pressure (performing in front of others as they watch) worsens performance. High payments turned out to be only useful for simple, mechanical tasks.
  
  Therefore, the author advocates paying smaller, more frequent bonuses. He says the emphasis should be on straight salary, and you could even try paying a bonus for performance averaged over the past 5 years to smooth it.
  
  
• Animals prefer to work for food: Work without meaning or which doesn't find effect on the world is annoying even if it's well compensated. From the author's experiments, he found that even if you receive lots of money, if your work is immediately destroyed or unused, you are less satisfied than getting less money and having your work have effect on the world.
  
  In addition, the act of seeing meaning in one's work produces more productivity. The simple acknowledgement of work by others makes a big difference to your productivity and satisfaction. This means that division of labor, when it separates people from seeing results, can be counterproductive.
  
  
• Ikea effect: We overvalue what we make. Cake mixes that are fully automatic sell worse than cake mixes where you have to add eggs and water. People feel more attached to those things into which they make an investment of money or labor.
  
  This investment of labor increases attachment and causes us to overvalue the items compared to outsiders. From experiment results, it turns out to be important that one completes a project in order to overvalue it (and not simply start it). In fact, if we have to overcome some obstacle, we will love the results even more.
  
  
• Not invented here bias: "My ideas are better than others'." Organizations and individuals routinely dismiss others' ideas and overvalue the worth of their own.
  
  
• Consumers engage in vengeful behavior: We don't care whom we punish, agent or principal, as long as we punish someone if we feel like we've been harmed. Apologizing does lessen this revenge instinct.
  
  
• We adapt more quickly than we imagine: We are horrible at predicting how badly we'll feel if something unwanted will happen; we'll overestimate the effect on us and underestimate how quickly we'll recover. Humans can even get used to pain, and the adapted response lasts for a long time (based on many experiment subjects he studied that had traumatic experiences many years ago and yet their pain adaptation remains). He found that the severely injured actually associate pain with hope for recovery rather than suffering.
  
  We can even get used to joy. He calls the emotional leveling out of joy "hedonic adaptation." It turns out that job satisfaction is connected to raises, not pay level. Happiness is temporary, and one's emotional state reverts to the mean after good or bad situations. However, we are bad at predicting hedonic adaptation.
  
  Small interruptions delay hedonic adaptation; you can use this to extend enjoyment. However, don't break up annoying experiences (try to get those done all at once); do break up pleasurable ones to extend the enjoyment. You'll be happier if you do intermittent purchases, not one large shopping spree. When cutting back expenses, though, do so on many types all at once so you can quickly get used to the lower quality of life rather than cutting back one benefit at a time.
  
  Inject serendipity and uncertainty into your life to enjoy it more.
  
  
• Hot or not: The author spent a lot of time studying online and offline dating behavior. He found that like attracts like, which he calls "assortative mating." If you're somehow handicapped, it turns out there is no adaptation of what you like based on your own limitations; people just change the priority of other attributes they care about.
  
  
• When a market fails -- online dating: Without an efficient market maker in dating, like a yente, the market is really inefficient. The problem with online dating systems is that people can't be described like products with simple parameters. People are experience goods; you can only learn about others by seeing how they are in the real world.
  
  The author experimented with creating virtual dating systems, where users could have online date simulation experiences, like looking at art together or playing a game. The presence of any external object helps stimulate discussion and allow people to learn more about each other.
  
  
• Why people will help one person but not many: People feel agony over one child who fell in a hole (a famous news story from a few years ago), but not for genocides killing hundreds of thousands or millions. There are several reasons for this: identifiable victim effect, closeness (physical or emotional or similar in some way), vividness, and specificity. Another related issue is the drop in the bucket effect, where people will donate more for one named person in Africa than to help many people in Africa; they think they can make a difference for one person and are just a drop in the bucket when contributing to many (even though this is irrational). The best way to counter this is to agree to certain rules and stick to them.
  
  
• The long term effects of short term emotions -- why not to act on negative feelings: We forget our emotional states but remember our decisions in order to be self-consistent later on. Don't make quick decisions when emotional as this can drive future behavior even when the initial emotional state is no longer relevant. Slowly count to 10 when you're under duress.

The book concludes with two general conclusions. First, we need to know our irrational limitations but also use our irrational strengths to be effective. Second, we need to experiment to learn how to make better decisions.

The book was informative and entertaining from the experiment-design standpoint, and as someone who likes learning about psychology, I enjoyed the book very much.

---