Tom Stafford:

A slow day at the museum, and the receptionist is sitting at their desk as a stranger approaches, perhaps a tourist.

‘Hi’, the stranger begins, ‘I found this on the street just around the corner.’ He puts a wallet on the desk and pushes it over to the receptionist. ‘Somebody must have lost it. I’m in a hurry and have to go. Can you please take care of it?’. Without waiting for a response, the stranger turns and leaves.

The wallet is a clear plastic envelope, which the receptionist picks up and turns over. Inside they can clearly see a key, a shopping list, some business cards and a couple of bank notes. 

So what happens to the envelope? That’s the question asked by Alain Cohn, a researcher at the University of Michigan, and his team — who have dropped off similar envelopes all over the world. Stafford continues: 

For me, the most interesting finding is not whether most wallets were returned (they were), or whether the wallets with more money in were more likely to be returned (also true), but the differences around the world in the rates at which the wallets were returned. These differences were huge, varying from 4 out of every 5 wallets being returned (in Scandinavia, of course, and Switzerland), to fewer than 1 in 5 being returned (in places including China, Peru and Kazakhstan).

Trying to understand this range led Cohn and his team to the World Values Survey, an international research effort that polls citizens around the globe about their beliefs and attitudes. One question, in particular, asks about faith in other people: ‘Generally speaking, would you say that most people can be trusted or that you need to be very careful in dealing with people?’. In Sweden, for example, most people agree with this statement. In Peru, fewer than 10 per cent do. Cohn’s analysis suggests that it is this generalised trust in other people that drives the cross-cultural differences seen with the lost wallets. 

Stafford, following Cohn, thinks that the key variable here is “views of human nature,” but that’s not obviously correct. What if the key variable is “views of social responsibility”? I wonder if Cohn et al. asked any questions about that.