You find a lost wallet. Do you return it? Do you think other people would return it? It is a bit of a hassle. And what if there’s a little money in the wallet? A lot of money? Who couldn’t use a little extra cash? These are the questions at the heart of the most extensive experiment in civic honesty to date, published on June 20 in Science. And the results are remarkable and encouraging.

Most of us would predict that the more money is at stake, the more likely individuals are to keep the wallet. Even economists who study incentives expect people to pocket the cash. But human beings deserve more credit than that. Over three years, a team of economists left more than 17,000 wallets containing varying amounts of money at civic institutions in 40 countries, then measured how many were reported to their owner. Rates of return varied greatly, from 14 percent in China to 76 percent in Switzerland. But strikingly, in 38 out of 40 countries, the more money a wallet contained, the more likely people were to return it. “People are a bit too pessimistic in their view of human behavior,” says economist Michel Maréchal of the University of Zurich, one of the lead authors of the study.

Even the researchers believed the opposite would be true. After exploring possible explanations for their finding in follow-up studies, they concluded that people everywhere were motivated by a combination of altruism and an aversion to viewing themselves as a thief.

“This is an amazing study and an important contribution,” says economist Marie Claire Villeval of the French National Center for Scientific Research, who was not involved in the work. She found it especially striking that more money generated stronger moral dilemmas. “The psychological cost of not returning the wallet increases at the same time as the selfish temptation to keep the money for oneself,” she says. “Interestingly, this stronger moral dilemma is resolved by a more honest, or more civic, behavior.”

Wallet contents used in the study. Credit: Christian Zünd

The study is also notable for methodological reasons. “Most research studying human honesty is conducted using experimental tasks in the lab,” says behavioral scientist Shaul Shalvi of the University of Amsterdam, who was also not involved in the research and wrote an accompanying commentary on its significance in Science. “[This] study provides a measure of civic honesty with an actual behavior that you may encounter in the real world.”

The breadth and depth of the study is also impressive. “So many wallets!” Shalvi says. The fact that the result was replicated across such a large number of countries, cities and institutions is convincing. “We can try to understand why one or two countries showed a slightly different pattern than the rest,” Shalvi says, “but if you sample 40 countries, and in 38 out of 40, the same pattern emerges, that suggests a very clear, robust finding.”

The study began much less ambitiously.  In 2013 Maréchal and Alain Cohn of the University of Michigan conducted a pilot study. They asked a student in Finland to pose as a tourist and drop off wallets at civic institutions. He said he had found each wallet and asked the staff to deal with it. When more money led to higher rates of return, Maréchal and Cohn did not believe it and told the student to triple the amount in the wallets. Nothing changed. “We thought, ‘Maybe there’s something special about Finland,’” Maréchal says. So he and Cohn set out to understand “whether this result is specific to particular cultures or whether it actually represents a more global phenomenon.”

In 355 cities, they used clear plastic wallets that revealed their contents. Each contained a key, a grocery list written in the local language, three business cards (using a name common to the country) and either no money or the local equivalent of $13.45 in U.S. dollars. The wallets were always turned in by study collaborators inside similar institutions (banks, hotels, and so on) to control who participated and allow for generalized results. (The study cannot tell us what happens to wallets found by passersby on the street.) The result was nearly always the same: “On average, adding money to the wallet increased the likelihood that it would get reported from 40 percent to 51 percent,” says co-author Christian Zünd of the University of Zurich.

Because the finding was so surprising, Maréchal, Cohn, Zünd and their co-author David Tannenbaum of the University of Utah went to extraordinary lengths to verify it. They controlled for the age of the person receiving the wallet, access to computers and the presence of witnesses or security cameras. None affected the results. They substantially increased the money to a more tempting $94.14 in three countries. Reporting rates increased to 72 percent.

In surveys, the researchers described the experimental setup to American adults and asked what they thought would happen. Participants overwhelmingly believed that higher amounts of money would inspire more people to keep the cash. A survey of top economists generated more muted predictions, but they still anticipated a slight decrease in return rates as the monetary value went up.

“[People] place too much weight on self-interest and too little weight on psychological factors like self-image concerns,” Cohn says. The more money that was in a wallet, he and his colleagues found, the more said it felt like stealing not to report it.

Differences in rates of return in various countries correlated to other proxies of honesty, such as measures of public corruption and tax evasion. Yet the researchers feel strongly that the different rates of return are less interesting than the consistent finding that more money motivated more honesty. Only Mexico and Peru failed to demonstrate that pattern.

The importance of the factors at work here may vary from one context to another, Maréchal says, “but we believe that people, policy makers and politicians should be encouraged to adopt a broader view of human behavior.” The authors hope increasing awareness of the negative impact people’s behavior has on others might increase honesty—so might making it more difficult for people to persuade themselves that they are honest when they do something wrong. “Reduce the moral wiggle room,” Maréchal says.