backpack #1. The first chip you draw is red. That bumps up the probability that this backpack is the mostly red one. How much? The correct answer is simpler than you might think. It’s 70 percent exactly. But this wasn’t intended as a math puzzle. Most decisions are made by gut, and Edwards wanted to see how accurate these gut instincts were. He found that guesses tended to be less than the correct value. People failed to appreciate that a single red chip could be as informative as it actually is. This confirmed Edwards’s suspicion that people are not especially good at making decisions under uncertainty. But that is exactly what von Neumann and much of the economic profession were taking as a given.
In a 1954 Psychological Bulletin article, Edwards sketched the von Neumann–Morgenstern model—few of his psychologist readers would have known much about it—and posed the rhetorical question of whether it had the slightest thing to do with reality. “The method of those theorists who have been concerned with the theory of decision making is essentially an armchair method,” Edwards complained. “They make assumptions, and from these assumptions they deduce theorems which can presumably be tested, though it sometimes seems unlikely the testing will ever occur.” High among the untested assumptions was that humans behave like the fiction known as an economic man ( Homo economicus ) or rational actor or rational maximizer . This is a worker/capitalist/consumer/game-player concerned exclusively with personal gain. In Robert Heilbroner’s words, economic man was “a pale wraith of a creature who followed his adding machine brain wherever it led him.” That adding machine brain enabled economic men to calculate expected utility for decisions big and trivial. “Von Neumann and Morgenstern defended this model and, thus, made it important,” Edwards wrote, “but [by] 1954 it was already clear that it . . . does not fit the facts.” That year of 1954 was not chosen casually. It was the date of Edwards’s pivotal Psychological Bulletin paper, and it must also allude, in part, to what we now call Allais’ paradox. That deserves a chapter of its own.
Nine Lunch with Maurice
In 1952, Leonard “Jimmie” Savage had one of the most excruciating lunches of his life. Savage was a thirty-five-year-old American in Paris attending an academic conference. Seated across the table from him was a man with the expression of a startled terrier. He was Maurice Allais, a forty-one-year-old French economist. Allais had his hair trimmed up the sides of his head, leaving a bushy flat top. Between the trick haircut and the tight smile that might be a frown, Allais’ face evoked one of those odd pictures that becomes a different face when turned upside down. Allais had told Savage he had something to show him. It was a little test he wanted him to take. The important thing is that Savage failed the test. Savage was a brash statistician, then at the University of Chicago. He had gone into statistics on the advice of John von Neumann himself. Visually, the most remarkable thing about him was his eyeglasses. Their lenses packed enough diopters to reveal the space behind his head. At Chicago, Savage had acquired a second mentor, Milton Friedman—founding father of the Chicago school of economics, future Nobel laureate, and veritable saint to Reagan-era capitalists. Friedman knew quite a bit of statistics for an economist. He and Savage had begun a peripatetic collaboration. Savage was attempting to devise a theory of how people make decisions. The decisions that concerned him tended to be about money. He was interested in how people assign prices to goods and services and how they make choices between them. Savage wanted to show that decisions about money were (or could be) completely logical. Friedman desired just such a theory. It would supply a firm foundation to his utopian economics of the free market. There was one big