The term comes from game theory. In a zero-sum game, the payoffs sum to zero: what one player wins another loses. Poker is zero-sum before the house takes its cut, and negative-sum after.

Most real situations are not zero-sum, and the assumption that they are has a name — the fixed-pie fallacy — because it is so common.

Why trade is not zero-sum

Money moving one way makes trade look like a transfer, and the error is in ignoring what moved the other way.

A buyer pays $5,000 for work they value at $12,000. A seller does work costing them $3,000 in time. The buyer gained $7,000 of value, the seller gained $2,000, and no one lost. The gain came from the difference in what each party could do — the same mechanism as comparative advantage.

This is why voluntary exchange is the default positive-sum situation, and why an argument that treats a supplier's margin as money taken from the buyer has dropped one side of the ledger.

Where each framing belongs

Genuinely zero-sum. Market share within a fixed market in a given quarter. A single contract with one winner. Positional goods, where the value is in being ahead of others — which is what status games are, and why they are exhausting.

Positive-sum. Voluntary trade. Specialization. Most negotiations with more than one term. Long-term relationships of almost every kind. Knowledge, which the giver does not lose.

Negative-sum. Litigation. Price wars. Most retaliation. Situations where the contest itself consumes more than is at stake — which is the real reason to avoid a fight rather than any principle about conflict.

The negotiation case

Almost every negotiation that feels zero-sum contains positive-sum structure, because the parties value the terms differently.

Price feels like a fixed pie. But timeline, scope, payment terms, exclusivity, references and risk allocation are all on the table, and the two sides rarely weight them identically. A client who needs speed and a supplier who needs predictable cash can both improve by trading one for the other, and neither has conceded anything.

The practical move is to put more terms on the table rather than fewer. A single-issue negotiation is zero-sum by construction, and most negotiations are made single-issue by the participants rather than by the situation.

The reason it matters

The framing determines behavior, and it is self-confirming. Treat a supplier relationship as zero-sum and you will negotiate hard, they will protect themselves, and you will get a transactional relationship that really is zero-sum. Treat it as positive-sum and the terms that make both sides better off become findable.

Which also means the framing is worth checking explicitly, because it is usually inherited rather than chosen.