It is a truth grudgingly acknowledged that mixed ethnic communities are not as mixed as they appear. In the school playground I find myself talking to the other white middle-class mums and dads, in spite of the fact that in a Hackney school there are plenty of parents who are neither. We know the white couple at number four but have had little contact with the African family at number two. It’s not something I am proud of, but there it is.
Many people fear that ethnic diversity can bring out the worst in us. Most famous is the American political scientist Robert Putnam, author of Bowling Alone, who in 2006 published a study of ethnic diversity and community trust in the US. He told the Financial Times that “the effect of diversity is worse than had been imagined. And it’s not just that we don’t trust people who are not like us. In diverse communities, we don’t trust people who do look like us.”
But while many people will agree with Putnam, few of us ask precisely why ethnicity and co-operation might be connected. Macartan Humphreys, an Irish political scientist at Columbia University, offered me a list of possibilities as to why we seem to get things done more easily in homogenous communities.
First, there are explanations based on preferences. Perhaps “coethnics” want to help each other; or simply like the same things; or perhaps, even if they don’t like the same things, they like working together for its own sake. A second group of explanations suggests that “coethnics” work together more effectively because they understand each other better or can ostracise shirkers.
So far, so imponderable. But Humphreys and three co-authors of a new book, Coethnicity, have been conducting an ambitious series of game theory experiments, not in the psychology labs of the Ivy League but in a neighbourhood of Uganda’s capital, Kampala, which is home to members of many different tribes. The games were designed to distinguish between these explanations. Some are staples of the literature in behavioural economics, others are new. (My favourite is “Lockbox”. First, one subject was trained to open a combination lock, which is an unfamiliar technology in Uganda. The game paired a trained subject with an untrained one and gave the pair 10 minutes to get the lock open without the trained player touching it.)
The results are fascinating. There was good evidence that people from the same ethnic group worked together more effectively (although not at Lockbox, alas) but no evidence that their preferences differed from other ethnic groups or that they harboured any distaste for working with them.
The most striking result came from “dictator” games, in which a player decides whether to assist others who have no way of responding with either reward or revenge. And here’s the remarkable thing: in an anonymous dictator game, people seem to treat those of their own tribe in exactly the same way that they treat people of other tribes. It was only when the researchers lifted the veil of anonymity that people’s behaviour changed and they treated coethnics more generously.
Coethnics, it seems, co-operate in part because that is what they have come to expect from each other – just as the British expect each other to drive on the left – and they are aware of the unpleasant consequences if they do otherwise.
What is true in Kampala may not be true in Hackney – or Jerusalem. But the research is encouraging. It suggests that if we struggle to do business with people who look different, that may not be because we dislike them, but because we simply don’t know quite how to begin.
Tim Harford’s latest book is ‘Dear Undercover Economist’ (Little, Brown)