Intelligence spread across many people

No single person knows everything needed to supply a city: what people want, which materials are available, what workers can do, or which alternatives might succeed. Markets let people act on their own pieces of that knowledge and respond to one another through exchange. They can function as a collective intelligence engine, coordinating decisions without a single mind directing them all. But the signals they respond to are selective. Understanding what a market can discover also means understanding what it cannot see.

A price can carry a message

Imagine that a storm interrupts a major supply of timber. As available stocks tighten, its price rises. A furniture maker looks for ways to use less wood. A builder considers another material. A supplier decides that bringing timber from farther away is now worth the cost.

These people do not need to know one another, or even know about the storm. The price carries enough information to change their decisions. Each combines that signal with knowledge of their own circumstances: what can be substituted, which orders can wait, and where more supply might be found.

This is the remarkable economy of knowledge coordinated through prices. A change in one place can prompt adjustments elsewhere without anyone having to understand or direct the entire response.

A way to discover what works

People can also try different ways of meeting a need. Someone introduces a repair service, a cheaper material, or a more convenient delivery arrangement. Customers’ choices provide feedback, while the costs of labor and materials reveal what it takes to keep the offer going. Revenue and costs help determine which experiments can continue.

The coordination can work even when participants see very little of the whole. In early experimental markets, buyers and sellers who knew their own assigned values or costs—but not everyone else’s—traded toward prices and quantities close to the competitive prediction.

Those experiments demonstrate a real capacity under particular trading rules. They do not establish that every market finds the right answer. To understand its intelligence, we also have to ask what information it receives and what its feedback rewards.

Some needs speak louder than others

A market responds to demand backed by purchasing power. Wanting something urgently is different from being able to offer enough to obtain it.

Imagine two people competing to rent a house. One needs a home near their job; the other wants somewhere to spend a few weekends each year. If the second person can pay much more, the market may allocate the house to them. The transaction reveals who can and will pay the most. It does not establish whose need is greater.

This makes the distribution of money part of the mechanism. People with more purchasing power can exert more influence over what gets produced and who receives it. A market can coordinate responses to their preferences effectively while leaving other people’s pressing needs unmet. Distribution and market performance cannot simply be treated as unrelated questions.

What never enters the price

Some people affected by a transaction never get to participate in it. A factory and its customers may agree on a price that covers production while leaving the people downstream to bear the cost of polluted water. The product looks cheap because part of its cost has been passed to someone else.

These are externalities: effects missing from the transaction’s price. They can also be benefits. Research may help many people who never pay its funders, leaving less incentive to support it than its wider usefulness would justify.

Shared benefits create a related difficulty. An effective flood-warning system can protect people whether or not they contribute to its upkeep. That makes it valuable, but can make charging for it difficult. Without another way to organize contributions, something people collectively need may remain underfunded.

The intelligence of buyers and sellers cannot fully account for consequences that their decisions give them no reason to consider. Future generations face this problem especially sharply: they cannot enter today’s transactions to defend their interests.

The signal can mislead

Even participants may lack information that matters. A used-car seller knows about a recurring fault; a buyer cannot distinguish that car from a reliable one. Buyers who anticipate this risk offer less. Owners of good cars may then stop selling, making the remaining selection worse. Hidden quality can undermine the market itself.

Power also affects the signal. A dominant supplier may restrict output or keep competitors out. A high price can therefore reflect control over access as well as genuine difficulty in producing more. Counting participants is not enough to establish that competition is effective.

In markets for assets, people may buy because they expect to resell at a higher price. Rising prices can attract further buying, feeding the very expectation that drove the increase. Experimental asset markets have produced bubbles and crashes. Many people participating does not guarantee that their errors cancel out; they can reinforce one another.

We can change what the market responds to

A market depends on other social technologies: ownership rules, contracts, standards, ways to settle disputes, and rules about who may trade what. Change these arrangements and you change the information and incentives the market works with.

Inspections and warranties can help buyers assess quality. Pollution charges or enforceable limits can make harmful effects harder to ignore. Public funding can support shared benefits that are difficult to sell. Rules that protect competition can make it easier for alternatives to emerge. Each response has its own costs and practical difficulties; improving the rules requires examining the particular problem.

The useful question is what we want this market to discover and coordinate—and what it needs help to see. Its ability to draw on dispersed knowledge is powerful. Choosing whose interests count, which harms are acceptable, and what everyone should have access to still requires collective judgment.

Real cases