Designing the rules of the game so self-interested players produce good outcomes.
Mechanism design is the branch of economics that works backwards: instead of analyzing behavior given fixed rules, it asks what rules would produce a desired outcome when agents are self-interested and have private information. Often called reverse game theory, it earned Leonid Hurwicz, Eric Maskin, and Roger Myerson the Nobel Prize in 2007.
Incentive compatibility: A mechanism is incentive-compatible when every participant does best by reporting their true preferences. Truth-telling becomes the dominant strategy—not because people are honest, but because the rules make honesty the smart move.
Individual rationality: Participation must be voluntary and beneficial. No agent should be forced into an outcome worse than what they could achieve by walking away.
Implementation: Given a social choice function (the desired mapping from private information to outcomes), mechanism design asks: does a game exist where rational play produces exactly that mapping?
The revelation principle: For any outcome achievable by any mechanism, there exists an equivalent direct-revelation mechanism where truth-telling is incentive-compatible. This simplifies analysis enormously.
Gibbard-Satterthwaite theorem: With three or more alternatives, no deterministic mechanism is strategy-proof and onto unless it is dictatorial. This is an impossibility result—perfect mechanisms don’t exist in general.
Vickrey-Clarke-Groves (VCG): A family of mechanisms that achieve efficient allocation by charging each agent their marginal externality on others. They are incentive-compatible but vulnerable to collusion and budget imbalance.
Myerson’s optimal auction: When selling a single item, the revenue-maximizing mechanism involves a reserve price and careful allocation rule—explaining why eBay reserves and minimum bids exist.
Spectrum auctions: Governments use combinatorial auction mechanisms designed by economists to sell radio frequencies to telecoms—raising billions while ensuring efficient allocation.
School choice: The Gale-Shapley deferred-acceptance algorithm is used in New York, Boston, and other cities to match students to schools based on preferences.
Kidney exchange: Paired donation chains use mechanism design to match incompatible donor-patient pairs, enabling swaps that save lives no bilateral match could.
Carbon markets: Cap-and-trade systems are mechanisms designed to internalize the externality of pollution by making emission rights tradeable.