Imagine you’re handed ten dollars and sat across from a complete stranger.
The rules are simple. You can keep all of it, give all of it, or split it however you like. Whatever you give will be tripled on its way across the table. The stranger then decides, alone, how much to send back to you. Maybe nothing. Maybe all of it.
How much do you send?
Economic theory — at least the textbook version that shaped a century of thinking — has a clean answer. You should send nothing. A rational stranger, receiving any amount, will keep it all. And since you can predict that, you’d be a fool to hand any of your money across. The textbook calls this equilibrium. It’s mathematically tidy. It’s also, it turns out, almost nobody’s actual behaviour.
What happens when you actually run the experiment
In 1995, three economists named Joyce Berg, John Dickhaut and Kevin McCabe ran exactly this scenario at the University of Minnesota. They called it the Trust Game, and the result was so consistent that it has since been replicated thousands of times across dozens of countries.
The typical pattern: senders give about half. Receivers send back, on average, slightly more than the sender gave. Nobody is behaving rationally in the textbook sense — and yet almost everyone is ending up slightly better off than if they’d stuck to the maths.
A related experiment, called the Dictator Game, strips the scenario down even further. Here, the receiver has no power to reciprocate. You’re just handed the money and told you can give any amount — or none — to a stranger you’ll never meet. Rational self-interest predicts you keep everything. In reality, about two-thirds of people give something, and the average gift is around 20 to 30 per cent.
These aren’t wealthy participants feeling flush. The experiment has been run with people in developing economies where the stakes equal weeks of income. The generosity doesn’t disappear. If anything, in some settings it grows.
Something is going on that the textbook missed.
The Ultimatum Game — where people pay to punish
If the Trust Game shows that people give more than theory predicts, the Ultimatum Game shows something stranger. Here, one person proposes how to split a sum of money. The other person can either accept the split — in which case both parties get what was proposed — or reject it, in which case nobody gets anything.
Rational theory predicts something counterintuitive: the proposer should offer the smallest possible amount, and the responder should accept it. After all, even one dollar is better than zero.
What actually happens: offers below about 30 per cent are routinely rejected. People walk away from real money to punish what they perceive as unfairness. They pay, out of their own pocket, to deny someone else a greater payoff. This finding is so robust that researchers have used it as one of the clearest pieces of evidence that humans carry something theory has struggled to name — a deep, active, sometimes costly preference for fair treatment.
Across cultures, across different species of fairness
One of the most illuminating chapters of this research came from an anthropologist named Joseph Henrich, who did something most economists hadn’t bothered to do: he ran these games in societies very different from the American universities where they’d been designed.
What he found was striking. The Machiguenga people of the Peruvian Amazon, who live in relatively self-sufficient family units with limited market exchange, made very low offers in the Ultimatum Game — and responders almost never rejected them. They played close to the textbook version. The Lamalera of Indonesia, who hunt whales cooperatively and must share large catches across the village, made offers that were often above 50 per cent. Some offered 60 or 70 per cent, presumably because in their daily lives, generous offers build the kind of cooperation that feeds a village through lean years.
Henrich’s work suggests something important: fairness isn’t one universal human instinct, applied the same way everywhere. It’s a set of social technologies, each calibrated to the economic life of the community that built it. The rules we follow about money are often the rules our society has needed money to do.
What this means for money
Put these experiments beside each other and a picture emerges that the textbook never drew.
Money is not simply a resource. It’s also a signal — about trust, about fairness, about who you are in relation to the person across the table. When you hand money to someone, you’re not only transferring value; you’re making a statement about what you think they’ll do, about what you believe is reasonable, about how you want to be seen. Receiving money does the same thing in reverse. The person who hands you cash at a dinner table has said something very different from the one who quietly picks up the tab, even if the dollar amount is identical.
This matters in ordinary life far more than most financial advice acknowledges. Splitting rent with a flatmate, paying a tradesperson, tipping, lending money to a friend, paying for a date, negotiating a salary, asking for a refund, giving a wedding gift — every one of these transactions carries information beyond the number. People who are fluent in reading that information generally do better socially and financially. People who aren’t often feel, accurately, that they’ve miscommunicated something important and can’t quite name what.
The Trust Game and its cousins are not really about ten-dollar stakes. They’re about what we are doing whenever money moves between us.
The counter-view worth hearing
Not all economists have been persuaded that these findings overturn rational self-interest. The most careful critique comes from John List, an economist at the University of Chicago, who has argued that the behaviours seen in laboratory experiments shrink substantially in real-world markets. Once people are experienced, once the stakes are meaningful, once they’re operating under social scrutiny they can’t control, the generous behaviours dim. List’s field experiments have shown that the “generous” responses in lab games can depend heavily on people’s sense that they’re being watched, that the experimenter expects them to give, that there are no real long-term consequences.
This is a genuine caveat. It doesn’t entirely undo the findings — even List’s work still shows humans deviating from the pure self-interest model — but it suggests the picture is more complex than a simple “people are more generous than economics assumes.” We are generous in particular conditions. We are less generous in others. Reading which conditions we’re currently in is part of the skill.
The takeaway
The textbook economic model of a selfish rational actor turns out to describe almost nobody. Real humans hand money across tables. They pay real costs to punish unfairness. They adjust what fairness means depending on what their community has needed. They use money to say things that aren’t captured in the number — and they expect others to read what’s being said.
This doesn’t mean you should throw out self-interest as a framework. People do respond to incentives. Markets do work, most of the time. But if you want to understand what’s actually happening when money changes hands between humans, the tidy model is not enough on its own. You need a model that includes what economists now sometimes call social preferences — the preferences people have not just for their own outcomes, but for the outcomes of the people around them and for what their transactions signal about them.
The next time you find yourself in a financial interaction that feels awkward — the split-the-bill moment, the unexpectedly generous gift, the friend who insists on paying back too much — notice that what you’re feeling is not confusion about the numbers. It’s information about a deeper game being played underneath. The question worth carrying is this one:
When money moved between you and someone else today, what did it say, and did you mean for it to say that?
Key research referenced: Berg, Dickhaut and McCabe (1995) on the Trust Game; Joseph Henrich’s cross-cultural work on the Ultimatum Game (Henrich et al., 2001); the foundational Ultimatum Game research by Güth, Schmittberger and Schwarze (1982); John List’s field experiments on behavioural economics.