- •Contact us
- •About us
- •Advertise with the FT
- •Terms & conditions
© The Financial Times Ltd 2013 FT and 'Financial Times' are trademarks of The Financial Times Ltd.
A few minutes ago, I made up my mind to toss a coin to decide whether or not to leave my wife. It was Steve Levitt’s idea.
I should explain that Levitt, an economist most famous for co-writing Freakonomics, would regard that coin toss as noise in his data. In collaboration with John List, a fellow professor at the University of Chicago, Levitt is offering the blessed release of the coin to people everywhere who cannot decide whether to quit their jobs, leave their partners, have children, move cities, quit drinking or even get tattoos. It’s all in the name of social science.
Here’s how the research project works. If you’re having trouble with one of life’s big choices, you sign up at FreakonomicsExperiments.com, choose your dilemma, fill in a short survey, promise to abide by the coin’s decision, and the website will flip the coin for you. Later, the research team will email a survey to ask whether you followed the coin’s advice and how things are working out.
It’s tempting to treat the whole thing as a joke or a publicity stunt (the website is, after all, lavishly branded). But Levitt maintains that the research intent is serious, and I am inclined to agree.
Here’s the problem for social scientists everywhere: we just don’t know much about how people make big decisions. We can analyse everyday behaviour, but that means simply observing correlations with little idea of what might be causing what. Or we can bring people into the psychology lab, but laboratories are artificial environments. Laboratory experiments tend to involve small decisions for small stakes. Sometimes they involve hypothetical decisions for no stakes at all. Levitt’s colleague, John List, has been a pioneer in developing field experiments with more realistic contexts, but even then it is hard to study the really big choices in life.
Hence, the Freakonomics coin-toss website. Levitt hopes to find people who are genuinely undecided – the “marginal” decision makers – and, as he asks, not unreasonably, “who could be more marginal than the kind of person who comes to a website to flip a coin to try to decide whether to leave his wife or not?” If Levitt and List do attract the genuinely undecided, they will be able to observe genuine causation in action: was asking for a raise the right decision or not?
You might reasonably wonder what could be learnt, even if Levitt and List do get enough people to toss the coin and follow through. But there are natural hypotheses worth testing. Do we believe the grass is always greener? Or do we prefer the devil we know?
Levitt is bullish about the project. After I admitted I couldn’t make up my mind whether to take it seriously or not, he reckoned that if the website attracted a decent number of serious users it would be “some of the best research I have ever done”. It would certainly be an original angle on an important set of problems. I asked Levitt if he’d ever read The Dice Man, but he hadn’t heard of it. Luke Rhinehart’s shocking novel is about a man who frees himself from social convention by submitting to the will of the die. (Should he rape his next-door neighbour? The die says yes.)
Feeling unnervingly like The Dice Man myself, then, I logged on to FreakonomicsExperiments.com to find out whether to leave my wife. I am, admittedly, exactly the wrong experimental subject because I have quite firm opinions on the pros and cons of the decision, but I wanted to see how the process worked. I filled in a few quick questions about our ages, races, marital history, children, stepchildren, household income; then I had to tick some boxes and write a couple of sentences reflecting on why I was considering leaving my wife and how the prospect made me feel.
And then I tossed the coin.
Tim Harford is the presenter of Radio 4’s ‘More or Less’
Copyright The Financial Times Limited 2013. You may share using our article tools.
Please don't cut articles from FT.com and redistribute by email or post to the web.