·General Thinking & Meta-Models
Section 1
The Core Idea
What people say they want and what they actually choose are often different. Revealed preference is the principle that you learn someone's true preferences from their choices, not from their statements. If a customer says they value sustainability but consistently buys the cheapest option, their revealed preference is price. If a board says it wants long-term value but fires CEOs who miss quarterly targets, its revealed preference is short-term results. Actions reveal; words can mislead. The economist Paul Samuelson formalised the idea in the 1930s: preferences are inferred from observed behaviour under constraints. You don't need to ask people what they want; you need to see what they do when they have to trade off.
The power of the model is that it cuts through stated preferences, social desirability, and self-deception. People misreport their preferences for many reasons — they want to look good, they don't know themselves, or they're answering a hypothetical. When they act, they put something at stake. The choice reveals what they're willing to sacrifice for what. In negotiation, watch what the other side does with their time, their offers, and their walk-aways. In strategy, watch how customers allocate money and attention, not what they say in surveys. In hiring, watch what candidates have done, not what they say they'll do. Revealed preference is the discipline of trusting behaviour over words.
Revealed preference has limits. Behaviour is noisy; one choice can be a mistake or a one-off. You need a pattern of choices under comparable constraints. And sometimes the constraint is hidden — someone might "choose" A over B because they never had a real chance to choose B. The model works when you observe consistent choices and you understand the constraints. Then you can infer preference, predict behaviour, and design offers that match what people will actually do.
