Peter Zhang



Richard Thaler’s legendary Misbehaving is personal biography, academic journey, and field survey wrapped into one. He recounts his struggles as a fledging researcher, his partnerships with other great minds, and their projects throughout the years. It’s an insider’s view of world of behavioral economics that you can’t find anywhere else.

Here’s the paperback and audiobook.

Utility: ★★★★★ (5/5)

Writing: ★★★★★ (5/5)

Thaler has a innate playfulness and sense of humor that animates the book. Every chapter (of the book and his career) is driven by a genuine curiosity and tendency to misbehave. I love his writing and his content. Compared to Kahneman’s Thinking Fast and Slow, Thaler’s book reveals much more about the actual making of economics research and tends to use specific experiments and examples as focal points for ideas.

If you’re at all interested in psychology or economics, I think you’ll love the book. It’s a must read!


This time, I’m writing these notes after reading the book. My theory is that having the entire book as context will let me focus on the important parts, and hence decrease the amount that I need to write. I’ll section them off by topic and omit the first-person narration.

The Endowment Effect

Humans value their own stuff more highly—it’s called the endowment effect. Someone willing to drink an old bottle of wine may not be willing to purchase the wine at market price. Similarly, customers would much rather be told that cash customers get a discount than that credit card users pay a fee. Paying “out of pocket” imposes an additional disutility that isn’t rational.

Accounting Shenanigans

Humans experience transaction utility. Marked up prices make us feel robbed, while sales make us delighted. It’s why retailers are constantly on sale or offering us to save a penny ($9.99 instead of $10!).

We also feel sunk costs. If I already paid for a year-long gym membership, I’ll keep using the gym, even when I’m sore or have better things to do with my time. It’s also why, when I’ve started writing a post, I refuse to delete it, keeping it instead in my “drafts” folder.

In addition, humans bucket their cash. Objectively, money is worth the same whether in the form of cash, assets, or savings accounts; but, we’re much more reluctant to touch our savings than we are to spend our cash. An example is the house money effect, where people are much more willing to make risky gambles with money they’ve already won, since they don’t perceive it as “their” money.

Intertemporal Utility

People experience discounting, so that each yeah utility is delayed, it shrinks by a constant factor. But they also place an additional premium on present utility, resulting in time inconsistent preferences. These ideas—including the beta-delta model—are expressed mathematically in my Econ 101A notes.


Unlike Econs, Humans get pissed off when firms take advantage of emergency-induced demand (like hiking the price of shovels during a blizzard, or jacking up the price of COVID-18 vaccines). They’re also susceptible to framing: taking away a discount hurts a lot less than raising the price. It’s why wages are sticky—people really dislike lower wages, so employers fire workers instead. Unsurprisingly, they also aren’t great at accounting for inflation.

In games with real money, research subjects demand equal splits. They’ll even spend their own money to rectify inequality. In variations of the prisoner’s dilemma, subjects are surprisingly willing to cooperate. They practice conditional cooperation, starting out nice but becoming mean when others fail to reciprocate.

Inefficient Markets

People don’t like risk and volatility, and it can’t be explained by loss aversion. Practically, it means people underinvest in high-risk, high-return assets like stocks, favoring safer bonds instead.

The efficient market hypothesis is compromised of 1) the “price is right” and will reflect intrinsic value, and 2) the “no free lunch principle”—there’s no way to beat the market. The following posed challenges:

  • Stocks with high price/earning ratios were overvalued, such that “Losers”—stocks with poor P/E ratios—outperformed “Winners” in the long term.
  • Prices are too volatile to actually reflect changes in intrinsic value, especially during events like Black Monday.
  • Closed-end funds sell equities are different functional prices than the market, challenging the “one-price” assumption. Owners of these funds offered a greater discount when small company stocks differed in returns from large company stocks.
  • Splits between existing companies and company-held shares provide examples of companies with “negative valuations.”


Some real world examples of the above phenomena include:

  1. Selling services in bigger bundles at steeper discounts attracts consumers, even when they are unlikely to use the entire bundle.

  2. Car buyers will take a loan over a rebate even when it has lower utility because it seems to immediately subsidize a greater proportion of the car’s price.
  3. When a bank tried to encourage ATM use by charging a fee to use a teller, they faced immense public backlash and retracted the program.
  4. Uber’s pricing algorithm hiked their prices when a blizzard struck New York. The state eventually reached an agreement with Uber, capping the prices.
  5. The famous mugs experiment—once people actually have a mug, they value it disproportionately high.
  6. In law, the Coase theorem—which states that parties will reach an optimum settlement through trades—is used to claim that the same state will result regardless of who is forced to pay. The mug example disproves it.
  7. When competing to claim offices, faculty will give too much weight to the info they have—like square footage—while ignoring less obvious factors—like exposure.
  8. In football, earlier draft picks are way overpriced, since individual coaches prefer to have the best players now.
  9. Game shows provide evidence for the “big peanuts” phenomenon. If contestants were really eyeing a $1M prize pool, then they’ll care a lot less about a decision with $5,000 stake.
  10. “Save More Tomorrow” boosted retirement savings must more than any change to the “after-tax financial return on savings.” Instead of directly hiking the savings rate, the plan stashed away pay raises into the new savings rate, decreasing psychological barriers to saving.
  11. The Behavioral Insights Team or “nudge unit” have leveraged behavioral science to increase tax filings and improve insulation.

Built with Jekyll on the Swiss theme.