Creative Commons License

Why People Make Bad Decisions

Wednesday, June 24, 2009 at 09:08 AM EDT

A while ago, when I wrote a review of Freakonomics, I got a comment from a reader who made it quite clear that they didn’t think I knew the hell what I was talking about: “As for rationality. People don’t make decisions rationally, or for that matter irrationally. Just not how we’re wired.”

freakonomicsI think that Cian is sort-of right, and I don’t think that Cian being sort-of right means that I’m totally wrong. It’s not a zero-sum game. Most people make most of their decisions in the hope of achieving the best outcome – but because the amount of social, economic and cultural data to be parsed is too huge for most people to survey, many decisions come out pretty eccentric. “Our brains work like big coincidence detectors and use improbable coincidences to make decisions about what is real”, as psychologist Tom Stafford explained. If a coincidence is neither big enough nor improbable enough, it may not register at all.

MMR is a good example. From a medical outcome point of view, the rational decision would be to take the jab and have done. But what if you’ve been made aware of the vaccine damage hypothesis, but have no experience of the diseases the vaccine is protecting against? Then you might be more fearful of the vaccine than the disease, and the rational response to that (however irrational the data) would be to avoid vaccination.

MMR is safe (image from

And what if non-vaccination had some other benefits that don’t show up in the epidemiology? Like a powerful feeling of community with other anti-vaxxers and a belief that you are part of an important narrative of repressed knowledge? Those social factors could be very powerful incentives, and might push the decision-maker firmly away from vaccines.

Why am I dwelling on a months-old blog comment today? (Hint: not just because I am slow-brained and resentful of correction.) Because I came across two excellent blog posts lately which used analogies with irrational economics to rationalise apparently illogical behaviours.

Tom Ewing parses some data on boom-and-bust in baby names:

the rapid spikes of popularity and unpopularity in some baby names look very much like the inflation and bursting of market bubbles. And the driver of unpopularity is the sudden increase in perceived risk (social risk, in this case). Would it be true to say that the more people’s ‘network perception’ plays a role in decision making, the more likely rapid popularity spikes are?

Blackbeard Blog, “Baby bubbles”

Aaronovitch Watch compares David Aaronovitch’s ongoing apologia for the Iraq war to the behaviour of a scam victim:

What clearly went on in 2002 was either that there was intentional deception, or that the government believed that Saddam had WMDs, and therefore because it believed this, thought it was a gamble worth taking to portray the evidence as much more conclusive than it was. That’s the sort of thing that people go to jail for if they do it in a set of accounts; if this isn’t “lying”, then there were no liars in the executive suite at Enron.

Aaro himself, notoriously, was persuaded by the government case to make a massive investment of credibility points into a decidedly subprime vehicle (the parallels between the September dossier, in which poor quality underlying material was layered, structured and given the imprimateur of a supposedly neutral agency to create the illusion of AAA status, and the CDO market, are perhaps fertile ground for someone more desperate for a column than myself). Unlike the investors in Bernard Madoff’s funds, however, he seems determined to defend the very people who swindled him. Nice one.

Aaronovitch Watch, “What is this, the missing chapter of Voodoo Histories?”

To which the only thing I feel like adding is that people quite often continue to poor funds into a duff investment they’ve been persuaded to make, because the social cost of admitting the mistake seems greater than the financial cost of continuing it. Similarly, the social cost of changing your child’s name is usually even greater than the cost of sticking with something that’s turned out to be a bit common. Even irrationality can have a rational outcome.