A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Rational expectations theory posits that economic agents form expectations using all available information and the correct economic model, making their forecasts unbiased on average. Under rational expectations, agents do not systematically make repeated forecasting errors and expectations respond immediately to new information. This framework contrasts sharply with adaptive expectations and has profound implications for the effectiveness of monetary and fiscal policy.
Study the Lucas critique to understand how policy changes can invalidate historical relationships when expectations adjust. Compare rational expectations predictions to empirical forecasting errors to appreciate both the theory's power and its limitations.
Rational expectations does not mean perfect foresight or that agents have unlimited cognitive ability—it means efficient use of available information. It also does not imply markets are always in equilibrium or that policy is ineffective; only that expectations are unbiased.
Before rational expectations, the dominant framework for modeling expectations was adaptive expectations: agents look backward, updating their forecasts by adjusting toward recent forecast errors. If inflation was 5% last year and you predicted 3%, you revise your forecast upward a bit. This rule is simple and intuitive, but it has an embarrassing property — agents can be systematically wrong for long stretches. If inflation is persistently rising, an adaptive agent perpetually under-predicts it, period after period.
Rational expectations, developed by John Muth in 1961 and brought into macroeconomics by Robert Lucas and Thomas Sargent in the 1970s, replaces this backward-looking rule with a stronger assumption: agents use the correct model of the economy and all available information to form their expectations. Their forecasts are the best possible given what is known. Critically, this means forecast errors are random — they have no predictable pattern that a clever agent could exploit to do better. This is the key property: *unbiasedness*, not perfect foresight.
The most influential application is the Lucas critique. If a government repeatedly runs expansionary policy to exploit the short-run Phillips curve tradeoff (lower unemployment, higher inflation), agents will eventually learn the pattern and build expected inflation into their wage demands — eliminating the real effect. More generally, any historical correlation between policy instruments and outcomes was estimated under a specific policy regime. If you change the regime, rational agents will change their behavior, and the historical relationship breaks down. Only "deep parameters" — things like preferences over consumption and leisure, or production technologies — remain stable across regime changes and can safely be used for policy evaluation.
The rational expectations revolution had stark policy implications. Under strong versions of the theory, anticipated monetary policy has no real effect on output, because agents adjust prices and wages in advance, leaving only nominal variables changed. Only *surprises* — unexpected policy moves — can move output, and even those are transitory. This was intellectually bracing, but empirically contested: prices and wages do appear to adjust slowly, and expected policy changes seem to have real short-run effects.
Modern macroeconomics has settled on a middle ground. Dynamic Stochastic General Equilibrium (DSGE) models incorporate rational expectations over the model's structure while also including nominal rigidities (sticky prices and wages) that give policy room to have real effects. The rational expectations assumption is not about claiming superhuman cognitive ability — it is a modeling discipline that prevents economists from building in systematic, exploitable mistakes as a free parameter.