Picture this: you find a wallet on the street, containing £200 in cash, a debit card, and an ID and nobody saw you pick it up. Classical economic theory predicts that you keep the money, as all people are rational, self-interested agents. However, most people return the wallet. Why?
What Classical Economics Got Wrong
Standard economic models are built around a figure economists call Homo economicus, which portrays people as perfectly rational, narrowly self-interested, and endlessly driven to maximise either utility or profit. This assumption is enormously useful for modelling how markets behave at scale. Aggregate price movements, supply and demand dynamics, and firm behaviour under competition can all be modelled reasonably well without worrying too much about individual psychology.
However, when this is applied to individual human behaviour, the model breaks down. People donate to charity, leave tips in restaurants they will never visit again, vote in elections where their individual ballot is statistically meaningless, and return wallets full of cash, making none of this make sense under the strict Homo economicus framework.
The Prisoner's Dilemma
The Prisoner's Dilemma is the most famous thought experiment in game theory, and it captures a tension that appears everywhere in social and economic life.
Two suspects are arrested and held in separate cells with no means of communication. Each is offered the same deal: betray the other and go free while your partner receives a heavy sentence, or stay silent. If both stay silent, both receive a light sentence. If both betray each other, both receive a moderate sentence; worse than mutual silence, but better than being the only one who stayed silent.
The rational choice, according to classical game theory, is always to betray. Whatever your partner does, you are better off defecting: if they stay silent, betraying them means you go free; if they betray you, defecting limits your sentence. Defection is what game theorists call a dominant strategy, as it produces a better outcome for you regardless of what the other player does. The Nash equilibrium, the point at which neither player can improve their outcome by changing their strategy alone, is therefore for both players to defect.
The paradox is that if both players follow this logic, which is individually rational, they both end up worse off than if they had cooperated, meaning that individual rationality produces collective irrationality. This dynamic appears throughout real economic life: arms races between countries, price wars between competing firms, and the failure of international climate agreements all follow the same underlying structure. Every party would be better off with mutual cooperation, but the incentive to defect unilaterally is always present.
Repeated Games and the Power of Reputation
In the 1980s, political scientist Robert Axelrod ran a famous computer tournament in which strategies for repeated Prisoner's Dilemma games competed against each other. The winning strategy, submitted by the mathematician Anatol Rapoport, was called tit-for-tat: cooperate on the first move, then simply mirror whatever the other player did in the previous round. It was the simplest strategy entered and it won convincingly.
Tit-for-tat works because it makes cooperation self-reinforcing. If both players adopt it, mutual cooperation is maintained indefinitely. Defection is immediately punished in the next round, removing the long-term incentive to free-ride. The study revealed that when interactions are repeated and players care about future payoffs, cooperation becomes individually rational, not because of altruism, but because reputation has value. This explains cooperation in long-term business relationships, small communities, and international alliances. But it leaves an important question unanswered: why do people cooperate with strangers they will never encounter again?
Fairness, Altruism, and the Limits of Self-Interest
The answer requires moving beyond game theory into behavioural economics, using the idea that people have preferences over more than just their own material payoff.
The Ultimatum Game helps to picture this. One player is given £10 and must propose a split to a second player. The second player can either accept, in which case both receive the proposed amounts, or reject, in which case both receive nothing. Classical theory predicts that the second player should accept any offer above zero as something is better than nothing. In practice, offers below 30% are rejected roughly half the time, as people would rather receive nothing than accept an outcome they perceive as unjust.
This reveals something that classical economics struggle to consider: the fact that people have fairness preferences. They derive utility not just from their own payoff but from the perceived equity of the outcome. Economists Ernst Fehr and Klaus Schmidt formalised this in their 1999 model of inequity aversion, which demonstrated that a significant proportion of people are willing to sacrifice personal gain to punish unfair behaviour, even when punishment is costly to themselves and yields no material benefit. This helps explain phenomena that Homo economicus cannot: why employees work harder when they feel fairly compensated, why consumers boycott firms perceived as exploitative, and why people leave tips for waitstaff they will never see again.
What the Evidence Shows
Decades of laboratory experiments have put these ideas to the test, and the results consistently challenge the self-interest assumption.
In public goods games, each player receives an endowment and decides how much to contribute to a shared pot, which is then multiplied and divided equally among all players. The individually rational strategy is to contribute nothing and free-ride on others' contributions. Yet in practice, average contributions in initial rounds typically sit at 40-60% of the endowment, which is far above the self-interested prediction.
Contributions do tend to decline over repeated rounds as participants observe free-riding and adjust their behaviour accordingly. But through introducing a punishment mechanism, players are allowed to pay a small cost to penalise free-riders, sustaining cooperation at high levels even over many rounds. Crucially, players use punishment even when it offers no personal material benefit, a finding that points toward an intrinsic preference for fairness enforcement rather than purely strategic behaviour.
Perhaps the most striking evidence for genuine altruism comes from the dictator game, a variant of the Ultimatum Game in which the second player has no power to reject the offer. Since rejection is impossible, any offer above zero must reflect the first player's actual preferences rather than strategic calculation. Classical theory predicts the first player keeps everything. In practice, most players voluntarily give away 20-30% of their endowment, suggesting that at least some degree of concern for others' outcomes is real, not performative.
Cross-cultural research adds further nuance. Joseph Henrich's studies across 15 small-scale societies found that fairness norms vary significantly across cultures, from highly egalitarian (the belief that all people are equally important and deserve equal respect) societies where offers below 50% are routinely rejected, to others where generous offers are themselves viewed with suspicion, suggesting that cooperation is not a fixed biological constant, but is shaped more by social institutions, economic structures, and cultural norms.
So Why Do People Actually Cooperate?
Reciprocity, the expectation of future interaction, sustains cooperation where reputation matters. Fairness preferences lead people to cooperate and punish defection even in anonymous, one-time settings. Genuine altruism, however modest, means some people simply care about outcomes for others. And social norms, the shared expectations of behaviour within a community, also coordinate cooperation in ways that individual incentives alone cannot.
No single theory fully explains human cooperation. What is clear is that Homo economicus, the coldly rational self-maximiser, is a useful abstraction for some purposes but a deeply misleading one for others.
Linking back to the found wallet. The reason most people return it is not because they have miscalculated their self-interest, it is because self-interest, defined narrowly as immediate material gain, was never the whole story as people care about fairness, reputation, and inner beliefs.
Emily Jong