Alfred Marshall's Principles of Economics (1890) contains the diagram that has been the first thing economics teaches: a downward-sloping demand curve, an upward-sloping supply curve, an equilibrium price where they cross. Marshall called the two curves the blades of a pair of scissors — neither alone determines the price; they cut together. The diagram was not new (Cournot drew something similar in 1838; the verbal intuition runs back to Adam Smith), but Marshall's version made it teachable and operational, and within a generation it had become the signature mental object of the discipline. It is, honestly stated, a teaching device — not a description of any actual market, where prices are set by posted prices, negotiation, auction, administered fiat, or sticky-price oligopoly far more often than by curves crossing.
The intuition the diagram captures is robust even when the diagram itself isn't literal. (1) Prices respond to scarcity. When something becomes scarcer, its price tends to rise; when more is supplied or demand falls, the price tends to fall. (2) Scarcity responds to prices. Higher prices reduce the quantity demanded and increase the quantity supplied; lower prices do the reverse. The two-way feedback is the signal that decentralized markets convey — and it works without anyone having full information about who wants what. (3) The price aggregates dispersed information. Hayek's The Use of Knowledge in Society (1945) made the point more powerfully than any diagram: when copper becomes scarce, every user of copper adjusts without needing to know why it became scarce. This is, in Hayek's phrase, one of the great intellectual triumphs of the human mind — the deepest robust contribution of economics. Where the diagram misleads: real markets are not in instantaneous equilibrium; quantities adjust before prices in many markets; demand curves are not always downward-sloping (Veblen goods); supply curves can be backward-bending (labour supply at high wages); the atomistic price-taking firm is a textbook fiction that real industrial-organization economics replaces with models of imperfect competition. The diagram is the first model a student learns; almost every subsequent course modifies or replaces it. Where it is most reliable: many buyers, many sellers, low entry barriers, information symmetry, and time enough to adjust — agricultural commodities, foreign exchange, financial markets in the long run. Where it is least reliable: monopoly markets, fundamental information asymmetry, strong externalities, or public-goods properties.
Almost every contemporary economic-policy debate involves an implicit invocation of supply-and-demand reasoning, often correctly and often badly. Rent control reduces the quantity of housing supplied at the controlled price — the textbook prediction confirmed empirically (Diamond, McQuade, Qian 2019 on San Francisco). Minimum wage effects depend on whether the relevant labour market is competitive (textbook prediction: employment falls) or monopsonistic (recent literature: employment can rise). Carbon pricing invokes the diagram as the correct framework: the market price of fossil fuels does not reflect the social cost of carbon. Surge pricing (Uber, Lyft, peak-load utility pricing) is the diagram in real time. The diagram is a teaching tool — but the intuition behind it is one of the most useful pieces of mental furniture economics has produced.