One party proposes how to split a fixed pie. The other accepts or rejects. If rejected, both get nothing. Rational self-interest says the responder should accept any positive share — something beats nothing. In practice, people reject low offers. They pay to punish unfairness. The ultimatum game exposes the gap between economic rationality and human behaviour: fairness norms override pure payoff maximisation.
The proposer holds structural power — they set the split — but the responder holds veto power. A 90/10 offer might maximise the proposer's expected payoff if responders were rational. They aren't. Cross-cultural experiments show rejection rates spike when the responder's share falls below roughly 20–30%. The proposer who ignores fairness risks getting zero. The game forces both sides to account for the other's willingness to burn the deal.
Werner Güth, Rolf Schmittberger, and Bernd Schwarze ran the first ultimatum experiments in 1982. The result contradicted game-theoretic prediction. Later work showed that rejection is partly about signalling — refusing bad deals maintains reputation and deters future lowball offers. In one-shot anonymous play, rejection rates stay high. That suggests intrinsic aversion to unfairness, not just strategic reputation-building. The lesson for negotiators: lowball offers can be rationally rejected. The "rational" move is often to offer enough that the other side feels the deal is fair.
In M&A, the acquirer who lowballs may trigger target management to reject out of principle or to signal they won't be pushed around. In salary negotiations, an offer that feels insulting gets rejected even when it's above the candidate's BATNA. In partnerships, the side that insists on lopsided terms may get a no and both sides walk away from value. The ultimatum structure appears wherever one party makes a take-it-or-leave-it offer and the other can kill the whole deal.
Repeated play and reputation change the calculus. In one-shot anonymous games, rejection is costly for the responder — they give up a positive payoff to punish. In repeated or networked settings, rejecting bad offers can improve future offers by signalling that you won't accept exploitation. The proposer who has a reputation for fair splits may need to offer less to get acceptance because the responder expects fairness. Culture also shifts the fairness threshold: some societies reject at 30%; others at 20% or 40%. The strategic takeaway is invariant — calibrate to the human and the context, not to the rational-agent model.
Section 2
How to See It
The ultimatum game shows up when a single party names the split and the counterparty can accept or reject with no counteroffer. Look for take-it-or-leave-it offers, final proposals, and contexts where rejection destroys value for both.
Business
You're seeing Ultimatum Game when a vendor sends a final price with "this is our best and final offer." If the buyer rejects, the deal dies. The vendor has proposer power; the buyer has veto power. A price that feels unfair — even if above the buyer's walk-away — may be rejected. The vendor who leaves too little on the table risks a rejection that hurts both sides.
Technology
You're seeing Ultimatum Game when a platform sets rev-share terms for developers (e.g. 70/30) on a take-it-or-leave-it basis. Developers can reject by not publishing. Apple's App Store terms function as an ultimatum. Developers who find the split unfair may still accept because the alternative is no distribution — but the ones with options (e.g. web, Android) can reject, and the platform's reputation for fairness affects long-run participation.
Investing
You're seeing Ultimatum Game when a lead investor presents a term sheet as non-negotiable. The founder can accept or walk. A term sheet that feels extractive may be rejected even when the company needs the capital. The investor is the proposer; the founder holds the veto. Overly aggressive terms trigger rejections that leave both without a deal.
Markets
You're seeing Ultimatum Game when a government or regulator sets a one-time settlement (e.g. antitrust, tax) on a take-it-or-leave-it basis. The target can accept or fight. Offers perceived as punitive get rejected, leading to long litigation. The proposer's offer must clear not just legal rationality but the target's fairness threshold.
Section 3
How to Use It
Decision filter
"Before making a take-it-or-leave-it offer, ask: is my proposed split above the other side's fairness threshold? If I'm the responder, am I rejecting for strategic reasons or burning value to punish unfairness? Calibrate to the human in the chair."
As a founder
When you're the proposer — pricing, partnership splits, acquisition offers — leave enough on the table that the other side doesn't feel exploited. Lowball offers get rejected in practice. When you're the responder, know your BATNA but also know that accepting a deeply unfair deal can poison the relationship. The ultimatum game says: offer or accept splits that clear the fairness bar, or expect the deal to die.
As an investor
Term sheets are ultimatums. Aggressive terms may be rejected by founders who have options or who prefer to signal they won't accept bad deals. The "rational" founder might accept; the human founder often rejects. Structure offers so the founder's share feels fair. When on the receiving end of an exit offer, distinguish between rejecting for leverage and rejecting to punish — the latter can leave money on the table for everyone.
As a decision-maker
Any time you present a final offer, you're in the proposer role. Size the offer so the counterparty can say yes without feeling cheated. When you receive an ultimatum, separate the economic value of saying yes from the emotional cost of accepting an unfair split. Sometimes rejecting is the right move for reputation or principle; sometimes it's a costly signal. Choose deliberately.
Common misapplication: Assuming the other side will accept any positive payoff. They often won't. Fairness and reputation drive rejections. Proposers who optimise only for their own share trigger rejections that destroy value for both.
Second misapplication: Overcorrecting and leaving too much on the table. The goal is to offer just enough to clear the fairness threshold, not to give away the surplus. Calibrate to the counterparty and the relationship.
Section 4
The Mechanism
Section 5
Founders & Leaders in Action
Herb CohenNegotiation expert, author of You Can Negotiate Anything
Cohen emphasised that negotiators who push for the last dollar often trigger walkaways. His approach — leave the other side feeling they won — aligns with the ultimatum game: offers that feel fair get accepted; offers that feel exploitative get rejected even when they're above BATNA. "The key is to make the other side feel good about the deal." That's the proposer's fairness calibration.
Kissinger's diplomatic practice often involved take-it-or-leave-it moments — deadlines, final positions, and ultimatums. He understood that the party receiving the ultimatum could reject and blow up the deal. His approach was to size offers so the other side could accept without losing face or core interests. The ultimatum game in statecraft: the proposer must leave enough on the table that the responder can say yes. Overreach triggers rejection and conflict.
Section 6
Visual Explanation
Ultimatum Game — Proposer offers a split (e.g. 70/30). Responder accepts (both get their share) or rejects (both get zero). Rational play: propose minimum, accept any positive. Observed play: propose ~50/50, reject low offers. Fairness and veto power drive outcomes.
Section 7
Connected Models
The ultimatum game sits at the intersection of game theory, negotiation, and social preference. The models below either formalise the structure (game theory, BATNA), explain why fairness matters (loss aversion, social norms), or extend to related dynamics (signalling, win-win).
Reinforces
Game Theory
The ultimatum game is a simple two-stage game. Game theory predicts subgame-perfect equilibrium: proposer offers minimum, responder accepts. Observed behaviour deviates because utility includes fairness. The formal structure is game-theoretic; the content of payoffs is behavioural.
Reinforces
BATNA
Your BATNA is what you get if the deal dies. In the ultimatum game, the responder's BATNA is zero. Rationality says accept any positive amount. People reject anyway — fairness and signalling matter. BATNA sets the floor; the fairness threshold often sets the effective floor for acceptance.
Tension
Loss Aversion
Loss aversion says losses weigh more than gains. Rejecting a low offer is forgoing a gain to punish unfairness. The tension: loss-averse individuals might still reject because the "loss" of being treated unfairly dominates the small gain. Fairness and loss aversion can align against accepting.
Tension
Expected Utility Theory
Expected utility assumes people maximise payoff. The ultimatum game shows they often maximise something that includes fairness and reputation. Utility is multi-argument; narrow payoff maximisation fails to predict rejections.
Section 8
One Key Quote
"In the ultimatum game, the responder's rejection of positive offers cannot be explained by standard game theory. Subjects exhibit strong preferences for fairness."
— Werner Güth, Rolf Schmittberger, Bernd Schwarze — Journal of Economic Behavior & Organization (1982)
The original experimenters identified the core puzzle: people pay to punish unfairness. That observation forced economics to take fairness preferences and social utility seriously. The quote anchors the model: when you're the proposer, fairness isn't optional decoration — it's a constraint that determines whether the deal happens.
Section 9
Analyst's Take
Faster Than Normal — Editorial View
Take-it-or-leave-it offers are ultimatums. The proposer has the structural advantage; the responder has the veto. Ignore the responder's fairness threshold and you get rejections that destroy value for both. Calibrate the offer so the other side can say yes without feeling cheated.
Lowballing is often a losing strategy. In theory, the responder should accept any positive share. In practice, they reject insulting splits. M&A, hiring, partnerships — the same pattern. Leave enough on the table that the deal closes.
When you're the responder, separate economics from emotion. Rejecting can be rational (signalling, principle) or costly (burning value to punish). Know your BATNA and the long-run value of the relationship. Sometimes yes is the right answer even when the split isn't ideal.
Fairness is a variable, not a constant. Thresholds vary by culture, relationship, and frame. Test and calibrate. The proposer who finds the minimum acceptable split captures more value without triggering rejection.
Repeated interaction changes the game. In one-shot settings, rejection is purely costly for the responder. In ongoing relationships, rejection can be an investment — it signals type and improves future offers. Proposers with a reputation for fairness may get more acceptance at lower offers; proposers known for lowballing may need to offer more to overcome distrust. Build reputation deliberately.
Summary: The ultimatum game is a two-player split: one proposes, the other accepts or rejects; rejection means zero for both. Proposers must clear a fairness threshold or risk rejection. Use it when making or receiving take-it-or-leave-it offers — price the other side's fairness into the split.
Section 10
Test Yourself
Is this mental model at work here?
Scenario 1
A VC presents a term sheet and says the terms are final. The founder can sign or walk. The founder finds the liquidation preference harsh but accepts because the alternative is no deal.
Scenario 2
Two co-founders negotiate equity split. They go back and forth with counteroffers until they agree on 55/45.
Scenario 3
A company makes a final offer to acquire a startup. The startup rejects, and both sides walk away. The startup had no other bidders.
Scenario 4
A hiring manager offers a candidate a salary. The candidate asks for more. The manager raises the offer once, and the candidate accepts.
Cross-cultural ultimatum experiments; documents variation in fairness thresholds and rejection rates across societies.
Leads-to
Signalling & Countersignalling
Rejecting a bad offer can signal that you won't accept exploitation. That signal can improve future offers. The ultimatum game thus connects to signalling: the responder's rejection is a costly signal of type (tough, fair-minded).
Leads-to
[Win-Win](/mental-models/win-win)
Win-win negotiation expands the pie so both sides get more. The ultimatum game is a fixed-pie split. Moving from ultimatum to negotiable, multi-issue deals creates room for win-win. Understanding the ultimatum logic helps you know when you're in a split-the-pie vs expand-the-pie situation.
