Posted by: holdenlee | February 22, 2014

Predictably Irrational

Predictably Irrational by Dan Ariely is an intriguing book that shows us how we can make better decisions by understanding the patterns in our own irrational behavior. Here’s a summary of the main points in the book (by chapter).

[Remarks in brackets are my own.]

  1. We find it difficult to assess absolute worth, so we tend to prefer objects for which there is an easy comparison.
    1. Businesses take advantage of this to lead us towards choosing a specific option. For example, the Economist had three offers: $59 online subscription, $125 print subscription, or $125 print and online subscription. In the presence of the middle option, most people chose the latter.
    2. This applies to other comparisons as well: for example, someone seems more handsome in the presence of someone who looks very similar but less handsome.
    1. Most hauntingly, it applies to people’s goals and what they base their happiness on. People want to make more money in comparison to others. People may abandon their original dreams in favor of obtaining other things that are more easily comparable, like making more money in comparison to others.
    2. What do we do? Choose the circles around us [it’s important to have a good Dunbar tribe around us], and think more broadly.
  1. The theory of “supply and demand” theory assumes that there is an intrinsic level of “demand.”
    1. But demand can be created by marketing: Assael got rich selling Tahitian black pearls that no one initially wanted by putting them in the window of a world-famous gem store with a high price tag, thus “anchoring” them to the finest gems in the world.
    1. We imprint on the first price we consciously consider for a product, and then our later valuations cohere with that first price (arbitrary coherence). For example, when students asked whether they would pay the amount in dollars of the last 2 digits of their SSN for a product, there was a correlation between the number and the amount they bid in an auction. Moreover, that first price is most influential; it can’t be easily changed.
    1. Example. People who move neighborhoods spend about the same amount, even when the cost of living changes. An ambiguous event can be turned into fun or chore depending on imprinting (like Tom Sawyer getting people to pay him to have a go at painting a fence).
    2. Thus we must pay attention to the first decision in a string of decisions.
  1. We have an urge to jump for FREE things even when we don’t really want them, because they comes without the fear of loss.
    1. Ariely sold Lindts and Hersheys at 15c, 1c, and then at 14c and FREE, and limited people to one chocolate. What people bought:
      Lindt  Hershey
      15c/1c           73%     27%
      14c/FREE   31%     69%
    2. Similarly, in trick-or-treat, most kids would exchange 2 Kisses for a large Snickers rather than 1 Kiss for a small Snickers, but would rather get a small snickers rather than exchange 1 kiss for a large Snickers.
    3. When Amazon started free shipping for orders over $25, sales increased. The only country that didn’t have an increase in sales was France because they had a reduced price for shipping rather than free shipping.
    4. We can use FREE to drive social policy: to make something (like electric cars) attractive, make registration costs free. Make medical checks free.
  1. If money is mentioned, then market norms are in effect–e.g. people will work hard only if paid enough. But otherwise social norms are in effect–ex. willingness to do someone a favor.
    1. People may work more for a cause than for cash. Experiment: Participants were given $5, 50c, or nothing to drag circles into squares on a computer for 5 minutes. (The group who got nothing were asked to do it as a favor.)
      1. The group who got nothing worked the hardest (168 circles vs. 159, 101).
      2. But when offered gifts (chocolate) worth $5 or 50c, all groups worked equally hard!
      3. When the prices were mentioned, they no longer worked hard.
    1. When AARP asked lawyers to offer less expensive services to retirees, they said no, when AARP asked if they could do it for free, they said yes. People are willing to work for free or for a good wage, but not for a small payment.
    2. In a sentence-scrambling exercise, participants who unscrambled neutral sentences asked for help sooner than those who unscrambled sentences about money. The latter group felt they had to be self-reliant, and were less likely to help others.
    3. Introducing market norms into social exchanges or vice versa hurts relationships. Don’t mention the price of the selections on a date! A school imposed a fine on parents who were late to pick up kids–turning a social contract into a market contract–and had the opposite-of-intended effect. They could decide for themselves whether they could afford to be late. Once the fine was removed they kept being late!
    1. Treating customers socially is good for business; it encourages loyalty; people are more willing to ignore minor infractions. But businesses don’t understand the nature of social relationships: one violation of the social exchange (ex. a fee) returns to the market exchange. You can’t have it both ways!
    1. By creating a social exchange, companies make their workers more passionate and hardworking. (Case in point: Google.)
      1. They blur the boundary between work and leisure (so people are willing to work when not in the office).
      2. Companies must commit to their employees in return, for example, medical benefits. By demanding high deductibles, companies undermine the social contract. Although people say they prefer cash, people respond better to an equally-priced gift, because it’s social not market.
    1. A salary alone does not motivate people to take on tough jobs like police officers, soldiers, and teachers. We must instill a sense of purpose, mission, and pride in education.
    2. Burning man is an annual weeklong, highly-attended event that rejects market norms; it works as a gift exchange economy. It convinced Ariely that “life with fewer market norms and social norms would be more satisfying, creative, fulfilling, and fun.” This may be extreme, but we need to remember the place of social norms in society.
  1. People fail to predict how their attitudes will change when in an impassioned state.
    1. Students significantly underestimated how attractive certain activities would be, and how likely they were to carry out immoral behaviors when they were nonaroused vs. aroused.
    2. “Just say no” is not good advice because it assumes we can turn off passion at will. Either we have to teach teens to say no before any temptation takes hold, or be prepared for the consequences of saying yes (carrying a condom even if they think they’ll be controlled).
    1. Teens cause car accidents a lot. We could build cars that would automatically call parents, or change the music, if it exceeds a certain speed.
    1. Before a decision, simulate the effects of the decision–for example, dunking one’s hand in cold water to see if one would forgo an epidural during birth.
  1. Why can’t we save even though we think we should? Without pre-commitments, we fail in self-control.
    1. Ariely gave 1 group of students the ability to set their own (hard) deadlines for turning in 3 papers, 1 group the deadline of the last day for all papers, and 1 group spaced-out deadlines. The class with firm deadlines did best, and the one with only the final deadline did worst.
    1. Precommitment and scheduling can solve problems such as
      1. Health care: doctors could charge a deposit for regular checks, and refund it only if you show up promptly.
      2. Car maintenance: Honda lumped car maintenance into regular “engineering intervals,” and had more success in getting people to take their cars in on time than Ford.
      3. Savings: We should make a self-control credit card that lets people restrict their spending behavior, or automatically trigger consequences (such as a notification to your wife) if you overspend. Unfortunately, credit card companies are not enthusiastic about this idea.
  1. We overvalue what we have (because we fall in love with what we already have), we focus on what we may lose rather than gain in transactions, and we wrongly assume others see the same perspective that we do. Overvaluation is due in part to emotional attachment that the seller has but not the buyer.
    1. Duke students who had basketball tickets demanded about $2400 for them, while students who did not have tickets were willing to pay about $170 for one.
    2. Virtual ownership: we can get attached to something before we even own it, for instance, in bidding for something or testing a free trial. Pride of ownership also correlates with energy put into it (for example, assembly).
    3. [This applies to values as well: people prize the values that they themselves have and downplay the importance of other values, and this prevents self-improvement.]
    4. View all transactions from a nonowner’s perspective.
  1. We try to keep options open for too long, when it would be better to just choose one.
    1. Commander Xiang Yu burned his own army’s ships so that they had no other choice but to fight to victory or perish.
    1. Parents sign children up for too many activities, etc. People take too many classes/majors to keep career choices open. People double-date to keep options open.
    1. In a experiment, Ariely gave students a computer game where there are 3 virtual doors leading into 3 rooms. When you are in a room, you can click to get money from that room, or take a turn to go to another room. Each room has a different distribution of payoffs. When the rules were such that a door would disappear if you didn’t go there in 8 turns, people kept switching doors, and got less money than they would if they had if they had just picked a random room and stuck with it. This is stressful and uneconomical!
    2. We fail to realize that some things are really disappearing doors and need our attention; these tend to be things that recede slowly (ex. childhood, friendships).
    1. Choosing between 2 equally attractive options could be hard, and spend too much time deciding. We fail to take into account the consequences of not deciding.
  1. Our expectations affect our experience.
    1. For example, supporters of two different teams will support their respective teams when something comes up for review.
    2. Ariely offered free samples of beer: one normal, and one with two drops of balsamic vinegar. When participants weren’t told beforehand, they preferred the vinegar beer; when they were told, they preferred the original. (Revealing it later did not change opinions.)
    1. In a coffee shop, the presence of odd condiments in fancy containers made people enjoy the coffee more (even though nobody used the condiments).
    2. This is why fancy descriptions on menus and food presentation are important.
    3. Pepsi and Coke both claim that studies show more people prefer their brand. Pepsi did blind taste tests, while Coke did not. Brain activation as measured by fMRI was different depending on whether the brand was revealed. When they knew they would get Coke, the DLPFC (higher brain center) activated.
    1. Stereotyped people react differently when they are aware of the label they are forced to wear: African-Americans did worse on exams when they had to fill in their race beforehand.
    2. People who completed a scrambled-sentence test with words like “aggressive” were more likely to interrupt afterwards than people who did the test with words like “polite.” People who did a test with words like “Florida” walked out of the experiment more slowly.
  1. If we expect something to work, it will more likely work.
    1. The placebo effect: Several medical procedures that had long been in use were found to be no better than placebos, for example, internal mammary artery ligation, arthroscopic surgery, drugs for irregular heartbeat, antidepressants, and brain surgery for Parkinson’s. No one had thought to do the experiments before!
    2. Two mechanisms shape the expectations that make placebos work: belief and conditioning. Expectation can release endorphins and opiates that make you high.
    1. An experiment showed that the same phony painkiller, if priced higher, would prevent pain more. Another experiment showed that people who bought lower-priced SoBe did worse than those who bought regular-priced SoBe, and people given a brochure with phony claims about SoBe increasing mental functioning did the best.
    1. This results in a moral dilemma: is it immoral to give someone a placebo claiming it will help them, if it could actually be helpful psychologically? Do we insist that people get the cheapest generic drugs, or do we indulge people’s irrationality?
    1. Experiments that compare treatments vs. placebos are essential.
  1. The more removed we are from the “real thing,” the more dishonest we are. The mere act of thinking about honor codes, etc. can help us be more honest.
    1. There are 2 types of dishonesty, the planned crime, and the type committed by people who consider themselves honest (taking a pen, taking extra soda, reporting business expenses…). The total cost of robberies in the US is $525 million in 2004, but theft and fraud at the workplace totaled $600 billion.
    2. Experiment: students take a test, and 10 cents per correct answer Setups for the different groups: Do the test and then
      1. Transfer to a scoring bubble sheet to turn in.
      1. Transfer to a scoring bubble sheet with the answers pre-marked.
      1. Transfer to a scoring bubble sheet with the answers pre-marked, and shred their original.
      1. Transfer to a scoring bubble sheet with the answers pre-marked, shred all sheets, and pick up the money themselves.
      2. Scores: a<b~c~d. Honest people are tempted to cheat, and participants are less influenced by the risk of being caught as one might think.
    1. If we care about honesty, why are we frequently dishonest? Our internal honesty monitor is only active when we contemplate big transgressions.
    2. Ineffectual ethics reforms in Congress: The toothpick rule says lobbyists can only serve things legislators can eat standing up.
    3. Students cheat less after being called upon to recall the Ten Commandments, but the rate of cheating is uncorrelated with how many they recalled. Students who signed an honor code statement cheated less.
    4. We could use (nonreligious) benchmarks to raise the general level of honesty–for example, honor codes for students, professional oaths (when professions were deregulated, professionalism declined, ex. “Moral Expertise: A problem in the professional ethics of professional ethicists.”)
    5. Honesty/trust is important in business.
  1. Cheating is easier when it’s a step removed from money. Hence cheating by accounting practices, backdated stock options, executive vacations… The same people who do these things would not rob a woman on the street.
    1. Ariely planted Cokes and plates of dollar bills in MIT refrigerators. Within 72 hours all the Cokes were gone, but not the money.
    2. Experiment: Students take a test, and either
      1. Turn it in, and are paid by correct answers,
      2. Tear up worksheets and self-report scores for payment, or
      3. Same as (b) but the students are given tokens later exchanged for money.
        Reported scores: a<b<c (3.5<6.2<9.4)
    3. Insurance fraud: consumers over-report claims by 10%. Wardrobing: buying clothes, wearing them, and then returning. Fraud on expense reports depends not on cost or fear of getting caught, but ability to justify the item to themselves as a legitimate use.
    4. Example: Breaking into Skype account to make calls is easier to justify than stealing money directly. (1) You’re stealing time not money, (2) you didn’t gain anything tangible, (3) it’s stealing from Skype! So Skype would probably cover the cost. (4) It’s charged automatically via PayPal.
    5. Businesses have low standards of honesty: airplane companies lie about number of frequent-flier miles required. Two-cycle billing incurs extra credit card charges.
    6. When we deal with hard cash, we are primed to think about our actions as if we had signed an honor code–a dollar bill has George Washington, In God We Trust, etc. We need to understand our tendency to cheat in an age where cash is going obsolete.
  2. “Beer and free lunches”
    1. Ariely took orders in a bar in two ways (1) take orders out loud and sequentially, and (2) take orders privately (by paper), and surveyed customer satisfaction. When people ordered out loud,
        1. There was more variety per table,
        2. But customers were less happy, except for the first person.

      One reason for more variety is people’s need for uniqueness. People with a high need for uniqueness sacrifice personal utility to gain reputational utility. In Chinese culture the need for uniqueness is not as strong: in Hong Kong participants were more likely to order the same food.

    2. Lesson: plan your order before the waiter comes, and announce it so you won’t be swayed. Maybe restaurant owners should ask customers to write orders privately!
    1. The main difference between standard and behavioral economics is the concept of free lunches.  In standard economics everyone is trying to maximum profit, so if there were free lunches, someone would already have found them and extracted their value. But people aren’t rational.
    1. Asking questions like “why are we not saving for retirement” is meaningless from the viewpoint of standard economics, but the reality (behavioral economics) is that people have a hard time understanding the costs/benefits, starting habits, etc. Behavioral economics suggests: it is psychologically easier to sacrifice consumption in the future. When employees join a company they could be asked what percentage of their future salary raises they would be willing to invest in retirement. This increased savings from 3.5% to 13.5%.


  1. […] the product, or the “dream” behind the product, ex. Apple is more creative (See also Predictably Irrational (Ch. 9-10); our expectations influence our experiences). We dismiss advertisers as manipulative, […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s


%d bloggers like this: