Ultimatum & Dictator Game

In the Ultimatum Game a proposer proposes how to split a fixed sum; the responder can accept (the split stands as proposed) or reject (both get zero). In the Dictator variant the responder cannot reject — the proposer decides alone. The game is the classic testbed where rational-choice theory and empirical fairness behaviour diverge.

The rational-choice prediction is clear: any offer above zero leaves the responder better off than rejecting, so they should accept any positive offer — and the proposer should therefore offer the minimum. Empirically the opposite holds: responders routinely reject “unfair” offers (typically below ~20% of the sum), absorbing their own loss to punish a norm violation. Proposers anticipate this and usually offer much more (often near 40–50%). The Dictator Game isolates pure giving: without the ability to punish, generosity falls but does not vanish — evidence of intrinsic fairness preferences rather than mere strategic caution.

In behavioral economics and anywhere fairness norms shape the outcome: wage-setting and salary negotiations, collective-bargaining disputes, splitting synergies in M&A, bonus allocation, pricing perceived as “fair” or “gouging”. Whenever an objectively advantageous offer can still be rejected on principle.

Two levers. First, manage information asymmetry: what the responder knows about the total shifts the fairness benchmark — if they don’t know the full sum, lower offers look acceptable. Second, repetition and reputation: in repeated play responders reject low offers early to build a reputation that unfairness doesn’t pay, forcing the proposer into higher offers in later rounds.

An employee who turns down a salary offer they find disrespectful even though it beats their alternative. Customers who boycott a product whose price they perceive as outrageous. Negotiators who blow up a deal that would profit them because the other side claims what they see as an unfair share.