Sway: The Irresistible Pull of Irrational Behavior

  • Ori Brafman &
  • Rom Brafman
Currency/Doubleday: 2008. 224 pp. $21.95 9780385524384 | ISBN: 978-0-3855-2438-4

In the Biblical parable in Matthew 25:14–29, a servant who was given five talents of money invested them and returned ten talents, whereas a servant given one talent buried it in the ground without profit. The master gave the risk-averse servant's one talent to his successful rival. The effect was elevated into a principle: “For to everyone who has, more shall be given, and he will have an abundance; but from the one who does not have, even what he does have shall be taken away.”

Sometimes named the 'Matthew Effect', marketers call this response 'cumulative advantage'. I think of it as the 'bestseller effect'. Every author and publisher knows that once a book gets a head-start in sales it signals to consumers that other people want that book, causing them to desire it and purchase more, so the richest authors get even richer.

In Sway, the brother authors Ori Brafman, an entrepreneur, and Rom Brafman, a psychologist, describe the social and psychological effects that shape our beliefs and behaviours. They hope to trigger their own Matthew Effect with this highly readable book. But predicting the next bestseller is as reliable a business as astrology. That problem affects all books, including, ironically, those about marketing and behaviour: the psychological principles may explain what happened in hindsight, but cannot be used to predict the future.

Sway is a fun read, and the brothers Brafman are compelling storytellers, pulling in the reader immediately and narrating at a breezy pace. But the book is thin on science and thick on anecdotes. The authors have a propensity for 'just-so' stories, favouring this or that behavioural principle when other explanations exist.

People find evidence for what they already believe and ignore anything contrary.

The book opens, for example, with the tragic 1977 crash of KLM flight 4805 during take-off from the tiny Tenerife airport in the Canary Islands. While motoring down the runway, the Boeing 747 slammed into Pan Am flight 1736, also a 747. The crash was the worst disaster in aviation history. What was the cause? The authors argue that it was psychological. The KLM captain Jacob Veldhuyzen van Zanten was a top pilot, featured in airline advertisements, who took pride in getting his passengers to their destination on time. That day he was behind schedule, having been rerouted to Tenerife after a bomb threat at his destination airport, and delayed on the island by fog. Captain van Zanten worried about his reputation for punctuality. “An unseen psychological force was at work,” claim the authors, “steering van Zanten off the path of reason.” This force was “loss aversion”. Behavioural economists have shown that when we make a decision, potential losses hurt twice as much as potential gains feel good. “This principle is key to understanding van Zanten's actions,” the Brafmans explain. He dreaded “the cost of putting up the passengers, the chain reaction of delayed flights and the blot on his reputation for being on time”.

Baloney. Van Zanten's plane was one of several large aircraft diverted to Tenerife. They manoeuvred tightly around the runway, the taxiway that ran parallel to it and four small connector taxiways between the two. Several spilled over onto the taxiway, so some planes had to taxi up the runway, turn around, and then take off down that same runway. Van Zanten did this, but after turning around in preparation for take-off, the fog reduced visibility to 300 metres. Unknown and invisible to van Zanten, at the same time Pan Am 1736 had been instructed to taxi down the same runway and take the third exit on its left in order to avoid the KLM flight's take-off. After clarifying which exit to take — “The third one, sir; one, two, three, third, third one” the controller emphasized — the Pan Am jet counted them off against an airport diagram. The cockpit voice recorder revealed that the Pan Am crew identified the first two connecting taxiways, but missed the third; the collision happened near the fourth exit.

Meanwhile, in the KLM plane, van Zanten's co-pilot radioed the tower for clearance. The tower did not clear them for take-off immediately. At this moment, a call from the Pan Am jet to the tower caused interference on the radio. The Pan Am crew signalled that they were still on the runway, but because of the radio interference the KLM crew did not hear the message, and began their fateful take-off sequence. The airport lacked ground radar so no one could locate the planes. By the time van Zanten saw the Pan Am plane it was too late. He throttled his engines full and pulled up the nose of the plane, but his fuselage clipped the top of the Pan Am jet, ripping it to shreds. The Pan Am pilot hit his engines and turned sharply into the exit path, but it was too little too late. Total death toll: 583.

The cause of this crash, investigators concluded, was a concatenation of conditions, none of which had anything to do with the psychology of loss aversion: bad weather, crowded conditions, big planes on a small runway, and misinterpretations and false assumptions.

Even if we grant the brothers Brafman the option of looking for an 'ultimate' instead of 'proximate' cause of the crash in the form of cognitive biases and behavioural persuaders that drove van Zanten to make his fateful decision to take off, loss aversion would be low on a causal vector list. Top of my list would be the 'confirmation bias', in which people look for and find confirmatory evidence for what they already believe and ignore evidence to the contrary. Once van Zanten thought he got the “OK” for take-off, everything else made sense. Or, perhaps it was the effect of 'inattention blindness', in which people attend to one task so intently that they miss obvious things in their visual field. Or it could be the 'self-serving bias' and the 'better-than-average bias' that made van Zanten overconfident in his abilities and thus less risk-averse than he might normally be. Maybe there was a 'priming effect', such that van Zanten's brain was primed to hear “take-off” in that garbled radio message. Or how about just the power of expectation?

The real problem here is the hindsight bias. Not for van Zanten, but for observers trying to read into a past event psychological effects that have been measured in the laboratory. The research on cognitive biases and judgemental heuristics — cleverly used in the service of reconstructing past events by the authors of Sway — is well grounded in empirical data, but the Brafman brothers face the same problem as the rest of humanity in trying to make sense of seemingly chaotic human behaviour: those very same biases operate in the process of using them to explain someone else's behaviour. Call it the 'meta-heuristic' bias.