Skip to content

Direct vs Indirect Mechanisms

Why It Matters

Designers instinctively want people to "just tell the truth" about preferences, but direct asking often fails (strategic lying, inability to articulate preferences, privacy concerns). Indirect mechanisms can extract the same information through behavior. Recognizing when to use indirect revelation - and that any indirect mechanism corresponds to some direct revelation game - clarifies design choices.

The Key Move

When facing a coordination problem requiring private information, ask: "Can I infer what I need from choices rather than asking?" Design options that reveal information through selection. Alternatively, if you have a complex indirect mechanism, ask: "What direct revelation game is this equivalent to?" This simplifies analysis and might reveal simpler implementations.

Classic Application

Auctions as preference revelation. Instead of asking everyone's valuation (direct), run an ascending auction. Dropping out reveals your valuation indirectly. The timing of exits communicates information without asking. English auctions are strategically equivalent to second-price sealed bids (by the revelation principle), but participants find the dynamic process more intuitive.

Surprising Application

Restaurant menu pricing. Restaurants want to know price sensitivity but can't ask directly. Instead: offer multiple portion sizes, sides, premium ingredients. Customer choices reveal willingness to pay. Someone ordering the large size with extra premium toppings reveals high valuation. The menu structure is an indirect mechanism for preference revelation. This is why "price discrimination by quality" works - customers self-select into segments by their choices.

Failure Modes

Over-engineering indirect mechanisms when direct asking works fine (sometimes people just tell the truth, especially in high-trust environments). Ignoring that indirect mechanisms require more sophistication from participants (people may not understand what their actions reveal). Forgetting that the revelation principle is theoretical equivalence, not practical equivalence (direct and indirect mechanisms may have different psychological effects, fairness perceptions, or computational requirements). Creating indirect mechanisms where the link between action and revelation is opaque (participants can't figure out what strategy to use).

Go Deeper

Myerson (1979), "Incentive Compatibility and the Bargaining Problem" - proves the revelation principle. Milgrom (2004), "Putting Auction Theory to Work" - discusses practical differences between theoretically equivalent mechanisms. Roth (2002), "The Economist as Engineer: Game Theory, Experimentation, and Computation as Tools for Design Economics" - on the gap between theory and implementation.