Tim Harford The Undercover Economist

Articles published in August, 2010

What’s your approach to picking toilet stalls?

Dear Economist,

When I travel I am faced with a difficult choice: which of the toilet booths to use in offices or hotels. I always try to guess which one is the least used. Probability theory should tell me that all the booths should be equally used, but perhaps human psychology plays a part in this. My boyfriend believes the last one is the least used – contrarily, I believe the opposite. Perhaps this proves that, on average, all are equally used. Which one do you use?

Jan Lucan, Prague

Dear Jan,

The efficient market hypothesis says that stock and bond traders will swoop on underpriced assets, driving their prices up. If so, an ignorant investor may pick and choose at random, knowing that her selections will be priced as keenly as any others.

Let us assume that everyone, like you, wishes to use the stall less travelled. Can we apply the hypothesis to public lavatories? If so, you may pick any stall you wish.

But the efficient market hypothesis is not universally accepted even in financial markets, and in public toilets it is less convincing. After all, there are no highly paid “toilet traders” engaging in arbitrage – or if there are, the term conventionally applies to a different profession altogether. So perhaps there are systematic errors to exploit.

Internet research uncovers one poll that suggests that many people, like your boyfriend, head to the farthest cubicle. Another site claims that the nearest stall is the least used.

But all this is a little puzzling. Your stall-picking strategy reminds me of Warren Buffett’s comment that “diversification is a protection against ignorance”. Presumably you just want a clean toilet with some paper. Would it kill you to do a little research and check a couple of stalls? That’s what I do.

Also published at ft.com.

28th of August, 2010Dear EconomistComments off

You can’t afford to get signals crossed in the underworld

“A wiseguy sees things if there are wiseguy things to see,” wrote Joe Pistone, the FBI agent better known as Donnie Brasco – the name under which he managed to infiltrate the mob. But what are the wiseguy things to see? And how is a wiseguy to know he isn’t dealing with the likes of Joe Pistone?

Such questions are among those that fascinate Diego Gambetta. Professor Gambetta, an Italian sociologist based at Oxford University, has managed to wrap himself in the language of economics as capably as Pistone wrapped himself in the language of organised crime. Gambetta is an authority on the Sicilian mafia, but deploys the tools of an economist to understand them and other criminals.

A key concept in modern economics is the “signal”, an idea developed by the Nobel laureate Michael Spence. A signal is an action that distinguishes one type of person from a would-be mimic because it would be too costly for the mimic to carry out. Spence suggested that the decision to acquire a degree might be a signal. The degree may be of no practical value but employers may still value it and quite rationally pay higher salaries to graduates. Why? Because a degree will distinguish good applicants from bad – if bright, energetic candidates are willing to go to the trouble of acquiring one, while dim, lazy candidates are not. The degree serves no educational purpose but the employer uses it to separate the wheat from the chaff.

For a criminal, the stakes are higher and the dividing lines sharper. However similar the boiled-down textbook model might seem, employing a graduate who turns out to disappoint is not the same as plotting an offence with a colleague who turns out to be an undercover cop. But while it is no easy matter to study criminal signals, the danger and purity of the signalling problem that criminals face makes them a tempting group for Gambetta to study in his new book, Codes of the Underworld.

Some signals can be crude. Drug dealers in New York sometimes force customers to use heroin or cocaine at gunpoint. Paradoxically, this has resulted in the signal being weakened: even police can legally take heroin if their lives are in danger.

Other signals seem perverse. Gambetta describes a convention in Italian academia for some established professors to celebrate the poor quality of their published work. This, he says, is a credible signal that they cannot jump ship for somewhere like Harvard – so will remain in Italian academia as powerful patrons.

Joe Pistone managed to avoid such expensive signals and send many smaller and more subtle ones. He didn’t buy drugs, for example, but he did sell stolen jewellery. He turned down a plum job in criminal accountancy, partly for personal reasons, and thus acquired credibility: everyone assumed that no undercover cop would turn down such a job. Above all, he spent years just hanging around in the right company, acquiring credibility by association. It was the long accumulation of many small signals that convinced criminals that he could be trusted.

An interesting implication of Michael Spence’s model is that if degrees really were largely signals, the world would be a better place if universities were closed down. There is an interesting parallel in Gambetta’s work: he points out that prison time provides a wonderfully credible signal. Few undercover police are likely to sign up for four or five years in jail, so an extended prison sentence can be an asset to any criminal trying to establish his credentials. Prisons are sometimes called universities of crime, but surely this is a parallel nobody expected.

Also published at ft.com.

Is garlic bread really fungible?

Dear Economist,
Having bought identical garlic breads for my wife and myself, I proposed that we should equally share one immediately and equally share the second later. She was unhappy with this, proposing that she eat half of hers and that I eat half of mine. When I suggested that half garlic breads should be fungible, she accused me of making the word up. Assuming that garlic breads could be exactly shared equally, had I been correctly using established economic terminology?
Hungry Chris, Teesside

Dear Hungry Chris,

“Fungible” is an inherently amusing word but you did not invent it and you have used it correctly, assuming – as you claim – that the tasty loaves were indeed identical. Fungibility is usually used of commodities such as Brent crude or 24 carat gold, or currencies. All these are fungible because a reasonable person wouldn’t care which particular barrel of Brent crude or which ounce of gold they received.

A new idea of fungibility has emerged in recent years, which concerns monetary transfers such as foreign aid or politicians’ expenses. If an MP claims £10,000 in expenses for something he was planning to buy anyway, it doesn’t matter whether he sends in receipts for secretarial expenses, a train season ticket or a duck island. All we know is that he is £10,000 better off when his expense claim is approved. In short, expenses are fungible.

But your story raises a deeper question than that of terminology: one of marital bliss. In the classic analysis, “Love and Spaghetti”, economist Ted Bergstrom modelled two lovers who both enjoyed Italian food but also loved to watch the joy in each other’s faces as they shared spaghetti. You and your wife, in contrast, are behaving like a pair of bond traders.

Your grasp of economic terminology is quite secure. I am not so sure about your marriage.

Also published ft.com.

21st of August, 2010Dear EconomistComments off

Illuminating advice on the dark art of ‘drip pricing’

“You inched towards the dark side,” joked one behavioural economist after he read a recent column in which I hinted that his field has some merits. It was a quip that got me thinking, because behavioural economics does indeed have a dark side. Behavioural economists study the psychology of economic decision-making, and if they are any good at their task they will discover something the unscrupulous salesman could use to his advantage.

A behavioural economist turned rogue would exploit the “endowment effect” – a tendency for people to put a higher value on something that they feel they already own. He or she would also try to create the sense that consumers would lose out if they did not buy, because people seem to hate the idea of losing £5 much more than they like the idea of gaining £5.

Third, our rogue economist would attempt to suggest an “anchor” value that was much higher than the asking price, which would make the product seem cheap. It doesn’t seem to be hard to create such anchor values: they can be produced by inviting experimental subjects to write down the last two digits of their social security number.

Fourth, he or she would make the pricing as complex as possible so that people struggled to compare one offer with a rival offer. Fifth, he or she would try to create a sense of social approval – everyone is buying this. Finally, a rogue economist would throw in something free.

Many unscrupulous salesmen have figured this advice out for themselves already. Think of infomercials. “The TimCo smokemaster doesn’t retail for £200; it doesn’t retail for £100; it doesn’t retail for £50 … ” (anchoring to a price of £200) … “if our lines are busy, please try later” (social approval) … “the smokemaster is not available in regular stores” (loss aversion) … “but wait! When you buy the TimCo smokemaster you get the TimCo soup knife absolutely free” (complex pricing and use of “free”).

The UK’s Office of Fair Trading (OFT) has been turning to behavioural economists for advice on such tactics, and has found that there is no pricing scheme more pernicious than “drip pricing”. Under the scheme, customers agree to pay a price only to discover that there is a charge for delivery; another charge for paying by credit card, and another for insurance. Drip pricing taps into the endowment effect, because customers feel that they have already made the decision to purchase; it creates loss aversion because customers commit time and effort to the search before being hit with extra charges; and it is a form of complex pricing which makes it hard to compare offers.

The OFT research, conducted by consultants and academics at University College London, was based on a laboratory experiment in which students sat at a computer and were presented with hypothetical deals from two fictional retailers. The students were beguiled with various marketing tricks and had to decide from whom to purchase, in what quantity, and after how costly a search. There was no trick quite so guaranteed to confound them as drip pricing, in which they were hit first with an extra charge for handling and then with a charge for shipping. (A two-part drip is modest: according to the OFT, one package holiday provider used four unavoidable “drips”, and two computer retailers tacked on seven optional ones.)

The OFT has been firing warning shots about drip pricing, but it will have its work cut out to regulate it – there is usually some loophole through which price drippers can slip. Buyers should remember that if they walk away when the drips start to fall, they won’t get soaked.

Also published at ft.com.

Is my dream win worth a lotto punt?

Dear Economist,

In Holland we’ve got a “postcode-lottery”: every month, people living in a certain postcode win – provided, of course, that they’ve bought a lottery ticket.

As a rule, I don’t buy lottery tickets. The chances of winning are small. In my view you only pay for the privilege to dream about what you would do with money you are never going to have in real life. And I don’t know about you, but I can dream for free.

But lately I’ve started dreaming (well, it’s more like a nightmare) that my zip-code wins the lottery, so my neighbours become millionaires and I miss out on all the money.

What should I do? Join the herd and start buying tickets every month?

Rob Voorwinden, the Netherlands

Dear Rob,

Being the only pauper in a street full of millionaires does seem a miserable prospect. (A famous survey by the economists Sara Solnick and David Hemenway suggested many Harvard students would rather be the richest pauper than the poorest millionaire.) The post-code lottery pays out sums in tens of thousands, not millions. Still, that makes a difference: the economist Peter Kuhn, with three colleagues, finds that people with neighbours who win the zip-code lottery are much more likely to buy a new car in the next six months. Envy matters.

And while the zip-code format accentuates the lottery’s downside, it also seems to improve the rewards. Standard issue lottery winners tend to become isolated from their community.

I would still advise against buying a ticket, though. First, you’re not psychic: a dream is just a dream. Second, happiness research suggests that lottery winners are not made happy by their victories. But most importantly, your chance of winning remains close to zero. The zip-code format has raised the stakes, but it has not improved the odds.

Also published at ft.com.

14th of August, 2010Dear EconomistComments off

Why we have got our work cut out creating jobs that matter

My wife and I only argue about the big issues, such as whether it’s a good idea for her to leave utensils in the sink. For the record, clearly not: it means that coffee-filter cones and colanders which need nothing more than a quick rinse are infected with deposits of grease from other dishes. My wife is simply creating work.

The other day, as I was running a sink of hot, soapy water in order to clean a coffee-filter cone, I mused on an inconsistency: we celebrate creating jobs in the wider economy, but complain bitterly about creating them around the house.

We can see the obsession with creating jobs everywhere in public discourse. It seems to be easier to sell renewable energy subsidies through the idea that it will create jobs than the suggestion that it might slow climate change. The coalition’s plans to cut public spending appear to me to be more unpopular on the grounds of lost jobs than lost services. International trade – and before it, new technology, which from an economist’s viewpoint looks much the same – is also condemned because it destroys jobs.

There is much that is silly about all this, and we should pay more attention to the kitchen-sink insight that it’s not a great idea to create needless work. Even if I was inclined to hire a cleaner to wash pointlessly dirty dishes, the apparent job-creation is illusory. The money I felt forced to spend on a cleaner I might instead have spent on a night out, employing cooks and waiters. And even if I had saved it, it would have swollen the pool of savings and made it cheaper for someone to borrow money and set up a business.

Economic growth is a continual process of job destruction. Start with agriculture, which destroyed the jobs of hunter-gatherers, and keep going until you get to e-mail, mobile phones and the word processor, which have destroyed the jobs of secretaries. Historically, some of the people whose jobs have vanished find something more useful to do than the grinding task of finding enough calories: teaching, practising medicine or learning engineering.

In principle, increasing labour productivity (aka “destroying jobs”) could lead to us doing less work for the same material gains. This could be pleasant – welcome to the five-hour working week – or horrible, with an employed elite and an unemployed and marginalised majority. In fact, to the bafflement of yesteryear’s futurologists, we do not lead lives of leisure while robots handle every chore. Instead, we have chosen to enjoy the benefits of greater labour productivity as greater wealth. (We do enjoy more free time too: longer holidays, shorter hours and working lives which start later and finish earlier despite a longer overall lifespan. But we take far less leisure time than we might.)

All that said, there are circumstances in which make-work schemes might make sense. One is the situation in which we find ourselves: a weak economic climate in which public sector job cuts could depress the private sector too. The coalition has a decent argument for making cuts: tax rises would also depress the private sector, while continued borrowing is unsustainable. But the idea that the cuts themselves will help create private-sector jobs is nonsense.

And what of areas whose economies have persistently struggled to recover from the death of an industry? A simplistic economic model suggests that wages will fall, private sector companies will rush in, and growth will resume. Reality suggests a grimmer diagnosis, but not one for which either the left or the right has produced a cure. What is needed are jobs that matter. We don’t yet have a reliable recipe for creating them.

Also published at ft.com.

When should I drink my six special bottles?

Dear Economist,
I have inherited six bottles of excellent wine, which I plan to consume, over time, on special occasions. But how do I know when to open a bottle when I don’t know what occasions lie ahead? I don’t want to use up all the bottles within a few months on mediocre occasions, but neither do I still want to be hoarding them until I die.
George Stevens

Dear George,

This is known as the “spongeworthiness” problem, after an episode of Seinfeld in which the contraceptive sponge is taken off the market. The character Elaine Benes stocks up with several hundred sponges, but then faces the same problem you do: how to decide whether a man is “spongeworthy”.

This is an option-value problem: every sponge or bottle consumed is one that cannot be used later. It has been solved by Avinash Dixit, a renowned game theorist and former president of the American Economic Association. (For some reason, Professor Dixit waited to finish his term of office before publishing his research on spongeworthiness.) Unless you are absurdly patient, you should open them more quickly than you might think.

Assuming you are patient enough to wait for a gain of 5 per cent a year – but not more – and if you have one possibly special occasion per month, Dixit calculates that you should open a bottle if you expect an occasion in the top 21 per cent of all possible occasions. As the number of bottles remaining shrinks, raise your standards. The last bottle should be consumed on an occasion in the top 9 per cent.

Of course, there is always the chance that an occasion is so special that you are tempted to open another bottle. Dixit’s analysis allows for this. After all, Elaine once told a lover that he couldn’t have a second bout: “Sorry, I can’t afford two of them.”

Also published at ft.com.

13th of August, 2010Dear EconomistComments off

Calculators away … life’s big choices call for gut instinct

Life presents us with some very large decisions: where to live, whom to marry, whether to have children. Is there any reason to believe that we make these choices wisely?

Don’t come to an economist for the answers. When it comes to choice, classical economics leaps to the punch line and works backwards: if both Betty Sue and Sally Ann are willing to marry you and you chose Betty Sue, the economist can only conclude that you preferred Betty Sue. Whether the marriage made you happy, or you were right to choose her, is none of the economist’s business. Game theory, for instance, may tell you how to get what you choose, but not how to choose.

Economists are not surprised to hear the (perhaps apocryphal) story that circulates about the great game theorist Howard Raiffa, who is said to have been trying to choose between a large increase in salary at Columbia University or a high-status move to Harvard. Sharing the dilemma with a friend, Raiffa was told to deploy the traditional tools of economics. He is said to have retorted: “You don’t understand. This is a serious decision.”

I am not in Howard Raiffa’s league, but I feel his pain when Family Harford wrestles with the big choices in life. The true challenge is that these choices force us to imagine different possible futures – one with Betty Sue and one with Sally Ann; one living in Surrey and one in Scotland; the carefree life of the childless couple versus the nurturing joys of parenthood.

It’s quite possible that our image of these possible futures is not very good. As the psychologist Dan Gilbert points out, you might think that winning the Lottery would make you happier than being permanently paralysed from the waist down, but the empirical evidence suggests that this is just a failure of imagination: paraplegics are not, in fact, less happy than people who have won the Lottery.

To me, that fact was at first surprising, then not very surprising at all. And then, after thinking still further, I realised it kicks away the foundation of almost everything we implicitly believe about the world.

Gilbert’s own experimental work suggests that we are extraordinarily good at “synthesising happiness” – in short, convincing ourselves that we like what we chose and dislike what we rejected, no matter how agonising the choice might have been at the time. This is true even of people with severe anterograde amnesia – an inability to form any new memories. They form much stronger likes for objects which they have previously chosen, despite having no conscious recollection that they had ever chosen them.

So how should we make choices? I recently met Sheena Iyengar, another psychologist and author of a new book, The Art of Choosing – aptly titled because despite a range of recent insights from psychologists and behavioural economists, choice is not a science.

Although she did not phrase it in this way, Iyengar’s method for making big choices boils down to improving the quality of our simulation of the future. First, she says, make a note of your gut reaction. (It may change.) Then list all the pros and cons, as an economist might do. Then talk to other people who have made decisions like the one you are considering – a move to the suburbs, or having a baby. Don’t ask them whether they think they made the right choice: as Dan Gilbert will tell you, of course they do. But ask them about how they live, what they do each day, and what are the advantages and the downsides of their choice.

And then? “Then you go back to your gut feeling,” says Iyengar.

Of course you do. This is a serious decision.

Also published at ft.com.

Crisis confessions of the Undercover Economist

Ingram Pinn illustration

Illustration by Ingram Pinn

Anniversaries are a time for reflection and, as the third anniversary of the credit crunch approaches, I have been doing some reflecting on where I went wrong as an economist. My own errors, of course, are of particular interest only to me, but I fear that they are fairly representative of the economics profession. Learn More

5th of August, 2010HighlightsOther WritingComments off


  • 1 Twitter
  • 3 RSS
  • 5 Podcasts
  • 6 Facebook


  • Fifty Inventions That Shaped the Modern Economy
  • Messy
  • The Undercover Economist Strikes Back
  • Adapt
  • Dear Undercover Economist
  • The Logic of Life
  • The Undercover Economist

Free Email Updates

Enter your email address to receive notifications of new articles by email (you can unsubscribe at any time).

Join 188,947 other subscribers.