The Undoing Project by Michael Lewis *** (of 4)


Amos Tversky and Danny Kahnemann, were a pair of exceptionally brilliant Israeli thinkers, who forged a friendship and academic partnership that led to the invention of behavioral economics and a Nobel Prize. Before the pair started questioning, and then testing, centuries-old dogma about economic behavior, the fundamental assumption by economists was that all humans acted out of self interest. People maximized gain, prioritized happiness, avoided pain, weighed benefits against costs, and made decisions with the best odds for the most advantage.

Only, Tversky and Kahneman when they gave people a variety of choices whose outcomes were mathematically proven to have a better and worse outcome, discovered that a surprising number of people made the wrong decision. Not just ordinary people made mistakes; even statisticians with PhDs who should have known better could be induced to select a path that was demonstrably worse than a clearly correct answer.

For example, a terminally ill patient who knows he is going to die soon is given the option to plan the remainder of his life as if he has only a week to live. Or he can plan as if he has a year to live. Which should he choose? 

The correct response is to plan as if he is going to live for a year, because that includes the option of living for a week, whereas the plan to live for a week forecloses any planning beyond seven days.

The book is filled with so many examples that demonstrate humans do not as a rule act rationally nor even perceive the world around them accurately, it made me call into question the validity of everything from job interviews to elections and spy reports in recognizing that human nature has an overwhelming tendency to fit observations into preconceived frameworks.
The genius of Tversky and Kahneman was to interpret data as it really exists and not as they wished it to be. The quality of The Undoing Project is its attention to the personalities of two such different and fundamentally interesting individuals.


Comments