Fuller Wallets Come Back More Often

17,303 wallets in 40 countries. Empty ones came back 40% of the time, the ones holding $94 came back 72%. What that shows, and what it does not.

Ask anyone what happens to a lost wallet with a hundred dollars in it, and the answer arrives fast. Less money, better odds. It is the kind of thing that feels too obvious to test.

A team of economists tested it anyway and got the opposite.

What they actually did

Alain Cohn, Michel André Maréchal, David Tannenbaum and Christian Lukas Zünd put 17,303 wallets into circulation in 355 cities across 40 countries. Not dropped in the street, exactly — handed over. A research assistant walked into a bank, a theatre, a museum, a post office, a hotel, a police station, a court or a public office, set the wallet on the counter, said they had found it outside and were in a hurry, and left.

The wallets were transparent on one side, so whoever picked one up could see the contents without opening anything: a grocery list, a key, and three business cards carrying a name, a job title and an email address. The name belonged to nobody. That email was the entire measurement. Either a message arrived, or it didn’t.

Three versions circulated. Empty. About $13.45. And, in three of the forty countries, about $94.15.

The numbers

Empty wallets were reported 40% of the time. The ones with the small amount, 51%. The ones with ninety-four dollars in them: 72%.

The pattern held in 38 of the 40 countries. In none of them did more money produce significantly fewer reports.

Before running the experiment, the researchers asked people what they expected. Around 300 US residents predicted the ordinary thing — more money, less honesty. Then they put the same question to professional economists, who predicted it too. Both groups had it backwards.

Why it runs the other way

The authors’ explanation is not that people are secretly generous. It is that people do not want to be a thief, and the more money sits in the wallet, the harder it becomes to describe keeping it as anything else. An empty wallet you can leave on a shelf and forget. Ninety-four dollars you have to make a decision about.

That is a quieter claim than “people are good”, and it fits the numbers better. Altruism plus an aversion to seeing yourself a certain way. The second half does most of the work.

What the study does not show

Here is where the headline and the experiment part company.

Nobody in this study found a wallet on a pavement. Every one of them was handed to a person standing behind a counter, at work, on the clock, usually with colleagues within earshot and often a camera above the door — someone with an employer and something to lose. The paper says so itself: it cannot tell you what happens to a wallet lying in the street.

The big-money condition ran in three countries out of forty, so the 72% rests on a much narrower base than the headline figure.

And “reported” means an email arrived. It does not mean the wallet came back with everything still in it.

The country rankings drew the most coverage and deserve the least confidence. Roughly 400 observations per country, spread across institutions that differ enormously from one place to the next, and out of that a national number gets assembled.

One detail travelled further than it deserved. The wallets left at the Vatican and at two anti-corruption offices were never reported. It is a good line. It is also a handful of wallets and means nothing statistically. I am repeating it because it is funny, not because it is evidence.

What is left

Take away the league tables and the anecdote and something solid remains, repeated across forty countries: as the cost of keeping something goes up, so does the rate at which people hand it back.

And our guesses about strangers are wrong in a consistent direction. The same direction turns up elsewhere — people who ask strangers for help get it about twice as often as they expect, and bystanders step in far more often than the textbook says.

That is not proof of anything grand. It is a repeated finding that the pessimistic guess is the inaccurate one.

Source: A. Cohn, M. A. Maréchal, D. Tannenbaum, C. L. Zünd, “Civic honesty around the globe”, Science 365, 70–73 (2019).