Misbehaving: a Book Review

Misbehaving: The Making of Behavioral Economics, Richard Thaler (2015). This is something of a memoir of Thaler more or less creating behavioral economics, with a lot of economics/psychology discussion thrown in. It is roughly in chronological order of what was happening to Thaler and behavioral economics in general. As an economist, he questioned many of the basic tenants of neoclassical economics, especially the underlying expectations of people ("Humans") in formal economic optimization models. These assumed that people were rational, all-knowing, unbiased, and making the right decisions. Because this is not how people normally act, he called the economists' version "Econs." It should be pointed out that neoclassical economics is useful, powerful, and explains pretty well how rational people should behave. It's just that so many shortcomings exist between rational expectations and actual behavior. This can be especially frustrating when economists are influencing (or creating) public policy [e.g., the 2008 sub-prime crisis does not happen without Alan Greenspan's idiotic interpretation of how financial markets act (and how the Federal Reserve should react)].

Thaler kept a list of examples of how people typically behave that was counter to economic theory (often based on what he called "supposedly irrelevant factors"). People might drive across town to save $10 on a $40 item, but not to save $10 on a $500 item. Then he discovered Tversky and Kahneman.

Kahneman and Tversky wrote "Prospect Theory" in 1979, a reasonable starting point for behavioral economics. It was originally called "value theory," but they decided that "prospect" was more of a nonsense word and wouldn't be tied to other perspectives (as value theory would). The first important concepts considered were risk aversion and expected utility theory (see below). KT offered an alternative theory that did not depend on rational choice. KT challenged the assumption of diminishing marginal utility of

wealth to focus on changes in wealth (and income). It becomes an s-shaped curve with gains similar to utility theory but losses roughly twice as steep, loss aversion (but diminishing sensitivity to both gains and losses). In addition, people tend to be risk averse for gains and risk seeking for losses. Experimental economics also was developing, using such techniques as markets for tokens (e.g., what happens when the market rules are changed).

These findings were challenged by "as if," that is, markets act as if people were rational. Under economists Samuelson, Hicks and Arrow economics became more mathematical and formal. Agents optimized and markets reached a stable equilibrium in the theory of the firm, including marginal analysis. Friedman focused on predictive accuracy (positive economics). Markets transform people into rational agents (e.g., market prices can still be rational).

Gambling (Chapter 10). Gambling is the main source for mathematicians to determine probabilities and statistics. Probabilities are obvious and therefore rational choices easy. KT and others found biases. Horse race tracks take 17% of wagers, meanings 17% losses on average. Losers tend to be risk seekers (to win back these losses); the odds on long shots in the last face get worse. People who get ahead early often put their original wagers away and then play on "house money." Researchers identify the 1) house money effect, 2) break even effect, and 3) risk seeking. The house money effect also is a good metaphor for why bubbles happen; people for example have complete faith in stock as the bubble peaks.

Self-control (part III). Despite Adam Smith worrying about passions (myopic choices) and Keynes' animal spirits, willpower was not an issue in traditional economics. Irving Fisher developed indifference curves to show consumption differences by time; he assumed that richer people were more patient. Samuelson focused on maximizing utility. He also developed the discounted utility model, with consumption now worth more than future consumption (present bias).

The consumption function should vary by time (today versus the future) and income. Keynes marginal propensity to consume assumed household incremental income would be consumed at a fixed percentage (that is relatively constant), but would vary by income (poor families would spend about 100%, declining as incomes rise). Friedman thought households would smooth consumption over time (permanent income hypothesis; he assumed a three-year time horizon). Modigliani created the life-cycle hypothesis, that people plan about how to consume over their lifetimes (focusing on wealth rather than income). Thaler and Shefrin proposed the behavioral life-cycle hypothesis for consumption and savings.

Chapter 12, planner vs. doer. Smith described passions vs. impartial spectator. To improve self-control: remove cues to bad behavior, develop a commitment strategy, limit choices. Mischel's marshmallow experiments, delayed gratification (predictor of later life outcomes). Also two-self system: KT thinking fast, automatic vs. thinking slow, reflexive. Thaler: forward looking planner vs. live for the present doer. Planner (as principal) can impose rules or influence through rewards and punishments (of agent)

Chapter 14, fairness. Raising prices as reasonable (Econ) vs. gouging. Framing can suggest fairness: how to sell a wildly popular car: sell above list or at "new" list price (endowment effect, here adding a surcharge vs. reducing a discount). Perceptions of fairness can explain why wages don't usually fall during recessions, although layoffs do. Note that raises less than inflation considered fair. Airlines: raising fares vs. added charges. Fairness games include ultimatum game and dictator game, splitting the pie. Cooperation games include prisoner's dilemma. Public goods game: each person can contribute say $1, total is doubled and split among the group. Econs would contribute nothing (free riding); non-economists might call them rational fools. Repeated playing deals with conditional cooperation (people continue to cooperate if enough other people do; also associated with reciprocity).

Chapter 16, mugs (endowment effect). People prefer what they already have. Experiments of lottery tickets, mugs; people prefer keeping rather than trading. Status quo bias: people stick with what they got unless their is a good reason to switch. Plus loss aversion: giving up something is a "loss." These are forces that inhibit change. Multiple experiments on boundary conditions, what are the limits for these to be observed.

Chapter 17, debate. Economists thought fairness a silly concept. Arrow: rationality (optimization neither necessary nor sufficient for economic theory. Requires additional assumptions such as everyone has same utility function. Modigliani & Miller irrelevance theory (corporate finance): it does not matter is firm pays a dividend, buys back its own shares, or pays down debt (investor wealth should remain the same). Note that taxes make a difference; also some investors prefer dividends (income) such as retirees. Lintner model (incorporating loss aversion) claims firms raise dividends only when they are sure earnings have gone up enough that they will not have to be cut in the future.

Asian disease: save 200 people vs. 1/3 chance to save everyone vs. 2/3 chance 600 will die (importance of framing).

Chapter 18, anomalies. Confirmation bias: people have a natural tendency to search for confirming rather than disconfirming evidence. Agency theory assumes agents have stable, well-defined preferences and make rational choices. Calendar effects on stock market, stocks up on Friday, down on Monday; January effect.

Chapter 20: narrow framing. Decision-making with bold forecasts and timid choices (inside vs. outside views); base rates: average time/cost for similar projects. Timid is based on loss aversion (big gains get modest rewards, losses get you fired. Hindsight bias plus "dumb principal" problem. "Equity premium": higher return for stocks relative to bonds. Myopic loss aversion: turning down a 2 to 1 bet.

Part VI, finance. The efficient market hypothesis claims that the stock market (and other financial markets) are efficient, meaning that prices are correct at all times--backed by "solid empirical evidence." Specifically, the price is right (true intrinsic value) and no free lunch (you cannot beat the market). However there are anomalies like seasonal effects. There is high trading volume (overconfidence, forecasts on flimsy data, small firms vs. large, value investing, mispricing risk). Fama-French came up with 3-factor & 5-factor models. Shiller: present value of dividends stable from 1871, but stock prices vriable. Long-term Capital Management: limits of arbitrage.

Chapter 27: law schooling. Richard Posner introduced formal economic reasoning into legal scholarship, including cost/benefit analysis and consumer sovereignty. Coase theorem: resources will flow to highest-value use (problems of social costs, endowment effect, transaction costs. Unfair offers, consider the ultimatum game.

Chapter 29, Football. Gary Becker: labor markets, people act as Econs in key positions. Decision making biases in football (draft) include overconfidence, make extreme forecasts, winner's curse (most extreme bidder most overvalues object), false consensus (think people share your preferences, present bias (overvalue winning now)--dumb principal problem. Becker conjecture: 10% of people that can do probabilities will end up in jobs where this matters, but Peter Principle.

Nudge (mainly section viii): improving individual choices with active/passive policies. Increase savings through 401K, such as automatic enrollment, gearing increases in savings with raises. Libertarian paternalism, asymmetric paternalism, optimal paternalism, basic idea is encourage good behavior given systematic biases, without harming others. Fly in the urinal (user centered choices), organ donations as default. Framing of bonus (pay upfront vs. promise to reward at end of period).

Here are some interesting findings and observations on decision-making (judgment under uncertainty) and biases (roughly as presented in the book):

"Identified" vs. "statistical life": people respond to a story of say a sick girl as opposed to statistics, say of mortality or education results; such as the trade-off of money to death or a college education.

Framing: how something is presented can make all the difference: an inheritance vs. death tax; a discount vs. a surcharge (e.g., cash vs. credit card).

Endowment effect: people value what they already own more highly that what they could own.

Systemic bias and predictable errors: people often do things that seem irrational and do them consistently.

Hindsight bias: after the fact we "know" what the outcome would be, although it was not obvious before. [An obvious pitfall of historians.]

Heuristics: with limited tie, brain power and information, people use rules of thumb (heuristics), which can lead to predictable errors.

Availability: a common heuristic, where decisions are made based on what you know.

Bounded rationality: people have limited ability to solve complex problems, leading to biases and heuristics (Herb Simon).

Risk aversion: starts with Bernoulli's idea of risk aversion, that happiness (utility) increases with wealth but at a decreasing rate. Von Neumann and Morgenstern incented expected utility theory based on a series of rational choices such as transitivity (e.g., if A>B and B>C then A>C). [Transitivity would be challenged by preference reversals].

Profit maximization: standard economics of the firm develops formal optimization models that maximize profits (Baumol proposed that firms maximize size). The heart of microeconomics is formal optimization models.

Loss aversion: losses hurt more than equivalent gains.

Acquisition utility: satisfaction gained less opportunity cost.

Transaction utility: price paid versus expected price (reference price).

Incentive compatibility: people are willing to pay more under certain circumstances (a beer at a bar vs. a convenience store).

Reference price: expected price for something; retailers can give the illusion of a deal by claiming a sale (less than retail price). Discount retailers on the other hand like Walmart have convinced customers that everything is a bargain.

Sunk cost fallacy: The money has already been spent and the money is gone. The fallacy is that people still want to get their money's worth; consider the escalation of of the Vietnam War; or the need to use season tickets. Consider a bottle of wine costing $20 (sunk cost, also historical cost) is now worth $75 (opportunity cost and replacement cost). How would an individual value this?

Revealed preferences: choices reveal preferences (note: self-control problems do not exist in traditional economics).

Intertemporal choice: people prefer consumption now rather than later, future consumption diminishes rapidly. Consumption differs with income (consumption function)

Theory-induced blindness: Kahneman's description of economists' focus on formal models over actual behavior.

Public goods: goods everyone can consume without diminishing the consumption of others (e.g., military, fireworks display, roads, fire and police protection are close.

Cognitive dissonance (Leon Festinger)