Do people only calculate for themselves? What experiments show about self-interest and fairness

10.10.2026

Autor: Tobias O. R. Alke · mit Claude (Anthropic)

A standpoint, not an encyclopedia entry. This article is an expressly subjective statement, dated 2026. The neutral background knowledge is in the encyclopedia entry Egoism.

No – in experiments, people cooperate more often than their material self-interest requires, and compassion changes their decisions even where self-interest would be clear. To a large extent it was economics itself that showed this: it broadened the model of pure self-interest without giving up reason. Whether fairness or self-interest tips the balance depends less on the individual than on the order that frames the interaction.

What does the model of pure self-interest mean – and what does it not mean?

Two assumptions are often confused: that people act rationally, and that they think only of themselves. Research has called the second into question, not the first. Axel Ockenfels and Werner Raub give up pure self-interest and hold on to rationality (2010: 124; see Fairness in economic decisions (Ockenfels)).

Ernst Fehr and Klaus Schmidt showed in 1999 that preferences for fairness explain behaviour in competition as well as in cooperation. Whether in the end the fair-minded or the self-interested determine the outcome depends on the economic environment. Marcel Mauss had already noted that the purely calculating human being is not an original form: “L’homo œconomicus n’est pas derrière nous, il est devant nous” – “homo economicus is not behind us, he lies ahead of us” (our translation) (1925: 176).

What do experiments show about cooperation and compassion?

In a prisoner’s dilemma, the participants learned that their counterpart had not cooperated. Without compassion, 5 per cent cooperated afterwards; with compassion aroused beforehand, 45 per cent (Batson & Ahmad 2001). This is a laboratory finding; how far it carries over into everyday life is an open question. The direction is clear: compassion changes decisions, even against one’s own advantage (see Empathy).

In the experiment by Fehr and Gächter, 84.3 per cent of the participants punished non-cooperators at their own cost at least once, 15.7 per cent never, and 9.3 per cent more than ten times (2002: 137). The punishment costs the punisher and benefits others.

Mixed motives are judged harshly: whoever does good and profits from it personally fares worse in an isolated judgement than someone who does no good at all. When both cases are compared directly, the effect disappears (Newman & Cain 2014: 648, 650).

Is punishment a principle of the market?

No. Altruistic punishment is an idea of order: coercion in the name of the community. Our view: in free exchange nobody punishes the other – one simply stops exchanging with him. A readiness to punish appears in many societies, to very different degrees (Henrich et al. 2006); it is not a feature of any profession or school of thought.

And it needs moderation: punishment that is experienced as self-serving can reduce altruistic behaviour (Fehr & Rockenbach 2003; see Altruistic punishment (Fehr/Gächter)).

Can incentives crowd out public spirit?

A well-known example: a nursery introduced a fine for collecting children late – and late pick-ups increased. The parents read the fine as a price (Gneezy & Rustichini 2000). Incentives can crowd out inner motives, but under other conditions they can also strengthen them. They crowd out when they are experienced as control (Frey & Jegen 2001), and when they signal that self-interest is expected (Bowles 2016).

Does studying economics make people selfish?

Economics students cooperate less often in games (Frank, Gilovich & Regan 1993). That their studies make them so has not been shown; it is more likely that people with this inclination tend to choose the subject (Girardi et al. 2024).

Our conclusion: whoever teaches models also teaches an image of the human being. That is why the model of pure self-interest should be accompanied by what research knows about the prosocial nature of human beings. The criticism is aimed at the image of the human being that is taught, not at those who teach or study.

What does this mean for markets and exchange?

The findings point in both directions. In a study across fifteen societies, market integration went hand in hand with greater fairness towards strangers, “although not across all measures” (Henrich et al. 2010). Other experiments show that market situations can lower moral inhibitions (Falk & Szech 2013) – and that social norms limit this erosion (Bartling, Weber & Yao 2015; Bartling & Özdemir 2023).

Our thesis: free exchange is individualistic, and it is the foundation of altruism. Our criticism is aimed at a politically imposed order that passes itself off as a market, not at exchange itself. More on this in the standpoint Who really exploits – state or market?.

More background: Egoism · The Possibility of Altruism (Nagel) · What sociology can learn from altruism