Predictably Irrational

Predictably Irrational by Dan Ariely1

Ariely's book is about how our minds really make decisions. Instead of being rational most of the time, we consistently make irrational decisions. If we understand those behaviors, we can 1) combat them to make better decisions; 2) exploit them in business to influence people's behavior; 3) design ways to constrain this irrational behavior to make everyone better off. This last point is especially interesting for economists. If everyone always makes optimal decisions, then there is no way to constrain people's decisions without making some people worse off. But if people make suboptimal decisions, there may be ways to make everyone better off. This is a fundamental issue in the ideological and political debate between "free markets" and "market intervention." Of course, in positive economics we don't have ideology. We just analyze "what if", to see what would happen if we created institutions designed to influence behavior.

Behavioral economics is based on experiments, and each chapter of Ariely's book is based on experiments testing a particular aspect of the way people make decisions. Here are bullet points summarizing the behaviors and the examples and experiments used. Each major bullet point summarizes one chapter:

  • We tend to focus our decisions on options that are easily comparable (role of context and framing in decisions, tendency toward low cognitive effort)
    • Economist pricing anecdote

For example, suppose you want to subscribe to the magazine The Economist, and there are two options: 1) an electronic-only subscription (at for $59/year; or 2) print and electronic subscriptions for $125/year. Given these two choices, 68% of people will choose the former option, while 32% will choose the latter. In standard economic terms, it means that a majority of people considers buying an electronic-only subscription to the journal’s articles for the reduced cost of $59 to be “better value” than having to pay more than twice as much for a combination of print subscription and electronic subscription. So far so good.
But then let’s say there are three options now: 1) an electronic-only subscription for $59/year; 2) a print-only subscription for $125/year; and 3) print and electronic subscriptions for $125/year. Given these three options, 16% of people will choose the first option, 0% of people will choose the second option, and 84% of people will choose the third option. That nobody chooses the second option makes perfect sense; why would you pay $125 to get a print-only subscription when you can get both print and electronic subscriptions for the same price? But here’s the puzzle: Now a vast majority of people considers print and web subscriptions for $125/year to be “better value” than an electronic-only subscription for $59/year, in clear contrast to their expressed preference when there were only two options. If people thought that A was better value than B when there were only two options, why does throwing in a third option C, which nobody chooses, entirely reverse their preferences?2

* We judge relative rather than absolute value
* That can lead us to be confused by "decoys", where we choose an option that is clearly superior to a "decoy" choice, while ignoring a third option that may be better because we have nothing to compare it to
* Manifesting in the following ways:
* Most people choose the middle option
* You can increase the price that people will pay in a restaurant by adding a higher-priced "decoy" menu item
* Bread makers did not sell until the company introduced a higher-priced version
* Go on dates with someone who is slightly less attractive than you
* People compare salaries instead of looking at the absolute value
* Many people will not question a $200 addition to a $5000 bill

  • Anchoring - we tend to value something based on a mental "anchor", even if it is arbitrary, such as a random number we hear
    • Bids influenced by random numbers (last two digits of the social security number)
    • People's spending on houses in a different city is influenced by the price they would pay in their home city, even if the market is different
    • Gosling follow the first thing they see (imprinting)
    • Context for Starbucks coffee - fancy coffee presses - influences willingness to pay
    • Demand is not completely separate from supply - prices set by suppliers can influence willingness to pay
  • The cost of zero cost - loss aversion makes us overwhelmingly prefer free options
    • "The difference between two cents and one cent is small. But the difference between one cent and zero is huge!"
    • Free shipping on Amazon induced people to spend a lot more money on a book they don't really want in order to get the "free" shipping!
  • Social norms can determine behavior too
    • Social norms vs. market norms
    • People will do things for a favor they will not do for money e.g. lawyers offering free services to retirees
    • If you mention the price of a gift people will apply market norms
    • "Just thinking about money makes us behave as most economists believe we behave" (page 75)
    • Gives famous example of the day-care center in Israel3
    • Social rewards can strongly motivate behavior (e.g. open-source software)
    • My own experience at work is that recognition and being treated with respect can matter more than money
  • The influence of arousal - surprise! We make different decisions when we are in a "cold" emotional state and when we are aroused
  • Procrastination
    • People with external deadlines tend to perform best
    • Offering tools to precommit improves performance
    • Example of periodic servicing for automobiles - avoids procrastination over scheduling services
  • Ownership (the endowment effect)
    • When people own something they tend to value it more than other people do, even if the ownership is the result of a random process
    • Tend to value something the more work you put into it
    • Bidders in auctions tend to get "invested" the longer they bid
  • Keeping options open
    • Note: can be more successful by sticking to one course of action, but people tend to want to keep options open
    • People in a computer game could not bear to see options disappear
    • Can fight it by considering the consequences of not making a decision
  • Expectations and stereotyping bias
    • People look at the same event but interpret it differently, to support their biases
    • People tend to find evidence to support their expectations
    • People are influenced by ambiance in judgment of restaurants
    • Idea is that information can actually change experience
    • Coke vs. Pepsi taste test example - giving the Coke name changed outcomes
    • Our own behavior can be influenced by stereotypes we hold of ourselves
    • Ariely says we need to acknowledge that we are all biased, and that we might be able to present facts without affiliation to get closer to the truth
  • Placebo effect and the power of price
    • Tend to believe that a more expensive product is better
    • People who stop and reflect about the relationship between price and quantity are less likely to be bamboozled.
  • Dishonesty
    • People tend to cheat "a little bit"
    • "…our internal integrity honesty monitor is active only when we contemplate big transgressions" (page 203)
    • You can invoke the idea of honesty e.g. by asking people to write down the ten commandments or with honor code statements
    • People will take Coke but not dollar bills from common-area refrigerators
  • Sequential choice
    • People's decisions are influenced by others' choices in a sequence
  • Getting "free lunches" from behavioral economics
    • e.g. Thaler's idea of getting people to commit to increasing the percentages of future pay rises that they save ("save more tomorrow")

Comments and Observations

I am dubious about "self-herding", where you follow the previous decisions you made because you made them before. Might the decision be a rational one, which it makes sense to repeat?

The issue with all these models is that they could be overcome by concentrated thought. It's not obvious that people end up with the rational choice, but they may do. I suspect that there are different "modes" of decision-making: rational analysis and irrational or instinctive decision-making.

The other problem is that these experiments produce a long list of possible behaviors, but it's difficult to predict which behavior will turn up when. However, we can anticipate some of the problems, and impose solutions such as a "cooling-off" period before committing to a big financial investment.

How to Incorporate Behavioral Economics into Introductory Microeconomics?

My thinking right now is that economics uses models of human behavior. We need to choose the models based on the circumstances, just as in physics we choose the model based on what we are trying to observe. We can use the "rational" model as a default, but behavioral models in circumstances where behavior is likely to diverge from rationality, such as when people are under strong emotion.

Notes on Writing

Stories are much more easily remembered than concepts. Vivid illustrations are important in teaching.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License