Go to the Prior Tip The Coming Generational Storm by Kotlikoff and Burns
Go to the Next Tip Antifragileby Nassim Taleb
Return to MaxValue Home Page
2011, Farrar, Straus and Giroux, New York, 499 p.
(paperback version available from Amazon.com for under US$11.)
This remarkable book summarizes Daniel Kahneman's learning and contributions across long and distinguished career. He shared in the 2002 Nobel Prize in economics. http://www.nobelprize.org/nobel_prizes/economics/laureates/2002/press.html
The press release says, in part: “for having established laboratory experiments as a tool in empirical economic analysis, especially in the study of alternative market mechanisms”. We know him best for study of judgments and biases.
Kahneman's best-known collaborator, Amos Tversky, almost surely would have shared in the Nobel prize except for his earlier death. Somewhat more technical is their well-known book (as editors and contributors, with Paul Slovic) Judgment Under Uncertainty: Heuristics and Biases, 1982, Cambridge University Press. Tversky and Kahneman's 1974 article by the same name in Science is still one of the most-frequently cited articles in the social sciences. It is reproduced as an appendix in Thinking.
First, a note of difference. Kahneman's research is about the ways in which people make decisions. This is descriptive decision analysis and is useful in predicting how people will choose. He offers prescriptive decision analysis, tools for helping people make better decisions. In my consulting and teaching, I try to guide clients toward normative decision analysis, the way a rational person should make decisions.
Here are just a few of the ideas I found highly interesting.
Most people have two modes of thinking:
System 1. Automatic, fast, little effort. Examples: comparing size or distance, detecting distress in a voice, driving a car, understanding simple sentences.
System 2. Requires effort and time, often involving computations. Examples: Searching for something or someone, telling someone your phone number, filling out a tax form, determining the validity of a logical argument. If a task deserves System 2's attention, then rest, nutrition, and focus (no multi-tasking) are needed.
Most of us are over-confident about our judgments, even when the problem is in a subject where we have many years of experience and training. As a young (21) officer in the Israeli military with a bachelor's degree in psychology, Kahneman was assigned to redesign the army's process for testing new draftees. A system already in place involved psychometric tests and interviews. The aim was to determine personality and assign people to various branches: infantry, artillery, armor, etc. Follow-up evaluations demonstrated that the process had almost no predictive validity. Kahneman had read a recent book (Meehl, c. 1954) that argued that simple, statistical rules are often superior to intuitive "clinical" judgments. He redesigned the process so that the interviewers were primarily to focus on eliciting relevant facts about the soldier's past and score each personality dimension. A scoring model based on just six personality traits vastly outperformed the earlier, all-intuitive approach. When Kahneman visited Israel 45 years later, we was shown that the same model and process were still in use.
When can we rely upon expert judgments. The most-useful situations are when the judgment is recognition. There are two basic conditions for acquiring a skill (p. 240): a) situations that are sufficiently regular to be predictable; and b) time to learn these regularities through long practice and feedback. Otherwise, intuition is unreliable.
Many of us are involved in enterprise planning or project planning. Most plans:
are close to best-case scenarios (most plans underperform)
could be improved by reference to similar cases. (Ask, for example, how many business start-ups like this succeed? How long have similar projects taken?)
Decision makers hate budget overruns, so they will often insist that a reserve be built-into the estimate. Danish planning expert Bent Flyvbjerg is quoted as saying (p. 252): "A budget reserve is to contractors as red meat is to lions, and they will devour it." Even for tasks performed internally, common behavior is to waste reserve, safety, and slack in project management. Project management professionals may wish to read Tip026.
Behavioral economist Richard Thaler (the leading researcher in "behavioral finance) coined these terms to describe two species:
Econs: rational, selfish, and consistent tastes (preferences). In the terminology of decision analysis, these are expected utility maximizers.
Humans: not fully rational, not completely selfish, and their tastes are unstable. Tversky and Kahneman are perhaps best known for their "prospect theory." This describes human behavior as it has been observed to differ from utility theory. Humans vs. Econs.
Several elements of prospect theory:
In situations where both gain and loss are possible, most people exhibit extreme risk-aversion (see the figure). Typically, for a 50:50 gamble to be attractive, the success outcome must be at least twice the amount of the failure outcome.
Most people do not weight probabilities in proportion, especially near the ends of the range 0-1. Probabilities the extremes are over-weighted. For example the weighting reduction from 1.00 to .95 probability is much more than .05.
People are more sensitive to changes of wealth than to states of wealth.
Where all outcomes are bad, people comparing a sure-loss to a possible larger-loss exhibit risk-seeking behavior.
The value of an item is different, whether buying or selling (endowment effect).
A person's risk-taking behavior is affected by the source of the money.
Figure from Wikipedia, Wikimedia commons
I read parts of the book with some frustration. Kahneman is describing behavior. The rational ("Econ") person easily avoids these issues with expected value calculations and the exponential utility function. We are wired to be "Humans." I advise people to work toward being "Econs," though clearly this takes additional System 2 effort.
This fascinating book offers numerous human behaviors that you will recognize in yourself. Many of the human decision-making shortcomings can be alleviated by basic methods in decision analysis.
Thinking represents outstanding book design. There are 38 chapters, each about 10 pages long. For me, that was a nice-size amount of reading for a sitting. The chapters comfortably continue to the next topic.
The reader is often invited to participate in an experiment (not unlike some of the clinical experiments). For example, the reader is often asked to choose between alternatives. The involvement both illuminates the topic as well as engages the reader.
A smaller, simpler and also-entertaining book about human shortcomings is: Inevitable Illusions: How Mistakes of Reason Rule our Minds by Massimo Piattelli-Palmarini, 1994, John Wiley & Sons. This is translated from Italian. There is a short chapter on Bayes' rule, where he says, In the context of 'subverting our intuitions' so as to avoid cognitive illusions, “Bayes’ law is truly one of the most important discoveries of the human mind.”
Some of the behaviors observed by Kahneman are explained by human brain physiology by Dean Buonomano in Brain Bugs (Tip 130).
—John Schuyler, Mar 2013. Revised Feb 2014.
Copyright \A9 2013-2014 by John R. Schuyler. All rights reserved. Permission to copy with reproduction of this notice.