🔁

Being shortchanged by a person feels different from getting the same bad deal from a broken vending machine. The money lost is identical; the urge to retaliate is not. People respond to intentions, not just outcomes.

Reciprocity and intentions

Inequity aversion judged outcomes: how unequal the final split is. But people also care why an outcome happened. A stingy offer chosen deliberately provokes far more rejection than the same offer produced by chance. Reciprocity, kindness for kindness and spite for spite, needs intentions in the model.

From earlier lessons

Two quick ones before we start.

Inequity aversion judges an outcome by how...

People reject unfair ultimatum offers because they value...

Predict, then play

The same lopsided $2-for-you,$ 8-for-them offer arrives two ways: chosen by the other player, or drawn by a random wheel. Compare how often people reject each.

Why did that happen?

Take a guess, then reveal the reason

If all that mattered were the final split, both versions would be rejected equally, since the money is the same. But a person who chose to lowball you acted unkindly, and reciprocity says unkindness deserves punishment even at a cost to yourself. A wheel has no intentions, so there is nothing to punish; you just accept the bad luck. The gap between the two rejection rates is direct evidence that people evaluate the actor’s intent, not only the outcome.

The math

Definition, reciprocity (Rabin)

Reciprocity models make utility depend on beliefs about the other player’s kindness, not just on the payoff distribution. Rabin’s fairness model is a psychological game. Let $f_B$ be the kindness B chooses to show A and $\tilde f_A$ be B’s belief about A’s kindness toward B. One version of the utility is $U_B = \pi_B + \tilde f_A(1 + f_B)$ , where $\pi_B$ is B’s material payoff. Kindness is defined from the feasible payoffs, so these beliefs and intended actions enter the equilibrium. Charness-Rabin adds evidence on reciprocity alongside distributional social preferences.

If B believes A acted unkindly, then $\tilde f_A < 0$ , so a response that is unkind to A raises B’s reciprocity term even when it costs material payoff. Rejecting can therefore beat accepting a low offer. If a wheel generated the split, there is no player intention to judge, so this reciprocal motive is absent. Identical material payoffs can then produce different choices because the beliefs about kindness differ.

Work it out

Why intent flips the response

Consider the $2-for-B,$ 8-for-A split. Under pure inequity aversion, B’s decision depends only on the numbers, so it would reject (or accept) the same way regardless of source. That predicts equal rejection rates, which is not what happens.

Add reciprocity. If A deliberately chose the split, B judges A unkind, so $\tilde f_A < 0$ . Rejecting is an unkind response, because it removes A’s $8 as well as B’s $2, and the reciprocal utility term can make that worthwhile. If a wheel produced the split, there is no player’s kindness to judge or punish, so the material $2 has more weight and B accepts. Same $x_A, x_B$ , but different beliefs about intention reverse the choice, exactly the pattern in the data.

Key idea

People reciprocate intentions, rewarding kindness and punishing unkindness, so identical outcomes get different responses depending on how they came about. Outcome-based fairness alone cannot explain this; intentions must enter the payoff.

Check yourself

Check yourselfapply

An unfair split is rejected far more when a person chose it than when a wheel produced it. What does this show?

Check yourselfanalyse

Why can outcome-based inequity aversion not explain this pattern?

Check yourselfbrings back: fairness

How does reciprocity extend the fairness models from the previous lesson?

Sources

Rabin, M. (1993). Incorporating fairness into game theory and economics. American Economic Review, 83(5), 1281-1302.
Charness, G., and Rabin, M. (2002). Understanding social preferences with simple tests. Quarterly Journal of Economics, 117(3), 817-869.
Blount, S. (1995). When social outcomes are not fair: the effect of causal attributions on preferences. Organizational Behavior and Human Decision Processes, 63(2), 131-144.