Rational Expectations vs Adaptive Expectations
Rational Expectations and Adaptive Expectations are two Economic Systems & Schools of Thought concepts in AP Economics that students often mix up. Rational expectations is the assumption that people form forecasts using all available information, so their errors are random rather than systematic. Adaptive expectations is the assumption that people predict future inflation from recent past inflation, adjusting only after they are proved wrong. Here is how they compare side by side.
Rational expectations treats people as forecasters who use everything they know, including how the central bank and Congress usually respond to conditions. They can still be wrong, because news arrives that nobody had, but they are not wrong in the same direction over and over. The exam consequence is the policy ineffectiveness result: if a monetary expansion is announced and believed, workers and firms raise wage and price demands right away, so the price level rises and real output stays at potential. Only surprises move output. That is the sharp break from adaptive expectations, where people extrapolate from recent inflation and therefore keep underpredicting it while it rises, which is what produces a usable short-run tradeoff.
Adaptive expectations makes people backward looking: their forecast for next year's inflation is built from what inflation has been, often just last year's rate or a weighted average of the last few. The consequence is that a rising inflation rate is underpredicted year after year, which is a systematic error, not a random one. That lag is what gives demand-side policy real bite in the short run: prices rise faster than the wages set under old expectations, real wages fall, and firms hire more. Once expectations catch up, wages reset, short-run aggregate supply shifts left, and output returns to potential at a higher price level. The contrast to draw on an exam is with rational expectations, which builds the forecast from today's information rather than from yesterday's outcomes.
Rational vs Adaptive Expectations: How Each Forecast Gets Formed
| Rational Expectations | Adaptive Expectations | |
|---|---|---|
| Information the forecast uses | Everything available, including announced policy and how the economy works | Past values of the variable itself, weighted toward the most recent |
| Pattern in the forecast errors | Random, with no systematic bias in either direction | Systematically too low while the variable is rising, too high while it is falling |
| Reaction to a believable policy announcement | Immediate, before the policy has done anything | None, until realized inflation actually moves |
| Cost of bringing inflation down | Small if the public believes the plan | Large, because expectations trail the outcome |
| Effect of fully anticipated policy on output | None, since wages and contracts already reflect it | Real, because contracts were signed on stale forecasts |
| Where the short-run Phillips curve moves | On the day the news arrives | Only after the inflation data have shifted |
| The standard objection to it | People are assumed to know more than they plausibly can | People are assumed to ignore information sitting in front of them |
The two rules produce different forecasts from exactly the same history
Use illustrative numbers. Inflation ran 3 percent two years ago and 6 percent last year, and the central bank has just announced a plan to bring it to 2 percent. An adaptive forecaster who puts weight 0.7 on last year and 0.3 on the year before predicts 0.7 times 6 plus 0.3 times 3, which is 4.2 plus 0.9, or 5.1 percent. A rational forecaster who finds the plan believable predicts 2 percent, because the announcement is information and throwing it away would produce an error that was avoidable. Now suppose the bank delivers and inflation arrives at 2 percent. Write the short-run Phillips curve so that actual inflation equals expected inflation minus the gap between unemployment and its natural rate, and set the natural rate at 5 percent. Wages under the adaptive forecast were built around 5.1 percent, so unemployment has to climb to 5 plus 3.1, that is 8.1 percent, before actual inflation can sit that far below what people expected. Under the rational forecast, expectation and outcome match at 2 percent and unemployment stays at 5. Same bank, same announcement, and the entire recession is manufactured by the forecasting rule. The curve doing the work here is set out at /macro/unemployment-inflation.
Adaptive expectations are wrong in a way anyone could predict, and that is the real objection
Rational expectations does not say people forecast correctly. It says their mistakes have no pattern, so they are neither too high nor too low on average. Adaptive expectations fails that test visibly. If inflation climbs for four years running, someone who always predicts last year's number is too low four years running, and a child could improve the forecast by adding a little each time. Assuming nobody notices is the part that is hard to defend. The stakes show up whenever a central bank tries to disinflate. On the adaptive rule the bank has to create slack first, wait for realized inflation to fall, and let expectations follow it down, which is slow and expensive in lost output. On the rational rule, a plan the public believes moves expectations the moment it is announced, so the cost can be small and credibility becomes the central asset. Neither extreme describes a real economy, which is why working models usually let some wages and prices be set on backward-looking rules while the rest look ahead. The same reasoning warns against reading old data too literally, since a relationship measured under one policy regime need not survive a change in the regime, an objection known as the Lucas critique. The curve these assumptions shift is defined at /glossary/phillips-curve.
Frequently asked questions
What is the difference between rational and adaptive expectations?
Rational expectations assumes people use all available information, including announced policy, so their errors are random, while adaptive expectations assumes people extrapolate from recent past values, so their errors run in one direction whenever the variable is trending. The practical consequence is that expectations jump the moment news arrives under the first assumption and only catch up with a lag under the second.
Does rational expectations mean people are always right?
No, it means people do not repeat the same mistake in the same direction, because information that would fix a predictable error eventually gets used. Any individual forecast can still miss badly when something genuinely new happens, since nobody can predict a surprise.
Why does adaptive expectations make disinflation expensive?
Because expected inflation falls only after actual inflation has already fallen, so the central bank must hold output below potential long enough to drag the realized number down first. Wages and contracts signed on the old expectation stay too high through that period, and the mismatch appears as unemployment.
Live Phillips Curve graph. Drag the curves, or open the full version.
Get AP Econ exam tips in your inbox
Occasional emails with study tips, new interactive graphs, and exam-season reminders. Free, no spam.
No spam. Unsubscribe anytime. Read our privacy policy.
Keep track of what you have studied
A free EconLearn account adds progress tracking, your quiz history, and achievements. Studying here is free either way, and there is nothing to pay for as a student.
Create a free accountAlready have one? Sign in
Last updated