Economics•Chapter 9•5 min read•Updated September 24, 2026

Macroeconomics — Expectations: From Adaptive to Rational

O
OiyoContributor
9/23

Macroeconomics — When Expectations Change, the Same Unemployment Rate Brings Different Inflation

In 1958 Phillips found in British data that wages rose faster when unemployment was low. In the 1960s this relationship was read as a policy menu: accept a little more inflation and unemployment could be lowered permanently. In the 1970s, when high inflation and high unemployment appeared together, the menu collapsed. The missing variable was expectations.

1. The expectations-augmented Phillips curve

Expectations-augmented Phillips curve
π=πe−β (u−u∗)+ε\pi = \pi^e - \beta\,(u - u^*) + \varepsilon
π^e is expected inflation, u* the natural rate of unemployment and ε a supply shock. When unemployment equals the natural rate, inflation equals expected inflation. This is the short-run aggregate supply curve of Chapter 8 rewritten in terms of unemployment and inflation.

Let u∗=4%u^*=4\% and β=0.5\beta=0.5. With expected inflation of 2%, inflation is 3% at 2% unemployment and 1% at 6% unemployment. If expected inflation rises to 5%, the whole short-run Phillips curve shifts up by 3 points, so inflation is 6% at 2% unemployment.

The short-run and long-run Phillips curves
u (%) π (%) 25 SRPC (πᵉ=2%) SRPC (πᵉ=5%) LRPC A B C

Lowering unemployment to 3% from A moves the economy to B in the short run, but as expectations catch up with higher inflation the curve shifts up. In the long run only inflation is higher, at the natural rate of 4%.

In the long run, when expectations equal actual inflation (π=πe\pi=\pi^e), unemployment returns to u∗u^*. The long-run Phillips curve is vertical at the natural rate. Friedman and Phelps reached this conclusion in the late 1960s, before stagflation arrived.

2. Adaptive expectations and accelerating inflation

Suppose people take last year’s inflation as this year’s expectation (πte=πt−1\pi^e_t=\pi_{t-1}). The Phillips curve becomes πt−πt−1=−β(ut−u∗)\pi_t-\pi_{t-1}=-\beta(u_t-u^*).

If the government tries to hold unemployment at 3%, inflation rises by 0.5 points every year: 2%, 2.5%, 3%, and so on. To keep unemployment below the natural rate, inflation must keep accelerating. That is why the natural rate is also called the NAIRU (non-accelerating inflation rate of unemployment).

Conversely, bringing inflation down from 5% to 2% (disinflation) requires holding unemployment above the natural rate. To get Δπ=−3\Delta\pi=-3 we need ∑(u−u∗)=3/0.5=6\sum(u-u^*)=3/0.5=6: 6 point-years of excess unemployment. If by Okun’s law 1 point of unemployment equals 2% of output, the cumulative output loss is 12% of GDP.

Sacrifice ratio
Sacrifice ratio=Cumulative output loss (% of GDP)Fall in inflation (points)=123=4\text{Sacrifice ratio} = \frac{\text{Cumulative output loss (\% of GDP)}}{\text{Fall in inflation (points)}} = \frac{12}{3} = 4
Lowering inflation by 1 point costs 4% of GDP. Actual estimates vary greatly across countries, periods and the credibility of policy.

3. Rational expectations and credibility

Rational expectations is the hypothesis that people form expectations using all available information and their understanding of how the economy works, so they do not repeat systematic errors. They look not only at past inflation but also at what the central bank will do.

A strong conclusion follows. If the policy rule is known and prices are flexible, anticipated monetary policy is immediately reflected in expectations and changes only prices, not output (the policy ineffectiveness proposition). Only unanticipated policy affects output in the short run.

There is also an implication for disinflation. If people believe a central bank’s commitment to tighten, expected inflation falls at once and the Phillips curve shifts down, so inflation can be lowered without large unemployment. The sacrifice ratio depends on credibility. Sargent interpreted in this light the rapid ends of several 1920s European hyperinflations through abrupt changes in fiscal and monetary regimes.

How expectations are formed and what it implies for policy
ExpectationsFormationPolicy implicationWeakness
Staticπ^e fixedThe short-run trade-off looks permanentPeople learn
AdaptiveFollow past inflationAccelerating inflation, slow disinflationCannot anticipate policy changes
RationalUse all information and the modelAnticipated policy is neutral; credibility lowers the sacrifice ratioInformation costs, price stickiness

4. The Lucas critique

Lucas pointed out that macroeconomic relationships estimated from past data (such as the slope of the Phillips curve) change when policy changes, because people’s expectations and behaviour respond to the policy regime. The stable Phillips curve of the 1960s breaking down in the 1970s is an example. His argument that policy should be evaluated with models built on preferences and technology that do not change with policy transformed macroeconomic methodology.

Check your understanding

In an economy with u∗=5%u^*=5\%, β=0.4\beta=0.4 and adaptive expectations, how much cumulative excess unemployment is needed to bring inflation from 6% to 2%? Δπ=−4\Delta\pi=-4, so ∑(u−u∗)=4/0.4=10\sum(u-u^*)=4/0.4=10 point-years. Doing it over two years would require unemployment of 10% each year. If credibility were high and expectations fell to 2% at once, the cost would be much smaller.

References

  • Milton Friedman, “The Role of Monetary Policy,” American Economic Review (1968)
  • Robert Lucas, “Econometric Policy Evaluation: A Critique,” Carnegie-Rochester Conference Series (1976)
  • Thomas Sargent, “The Ends of Four Big Inflations” (1982)
  • N. Gregory Mankiw, Macroeconomics, ch. 14
O

Oiyo

Editorial Desk

The OIYO editorial desk researches money, law, lifestyle, and self-understanding topics against primary sources and public statistics. Every piece carries source notes and is reviewed on a regular cycle for accuracy and usefulness.