ECON5002 · Macroeconomic Theory
Inflation, Unemployment and the Phillips Curve
This chapter explains the apparent trade-off between inflation and unemployment — why the original Phillips curve looked like a stable policy menu in the 1960s, and why stagflation destroyed that reading in the 1970s. The repair is the expectations-augmented Phillips curve: there is a short-run trade-off for a given inflation expectation, but the long-run curve is vertical at the natural rate of unemployment (the NAIRU), so demand management cannot permanently lower unemployment. The exam-critical skills are diagnosing demand-pull versus cost-push inflation, drawing the short-run and long-run curves from memory, and computing the sacrifice ratio of a disinflation and the Fisher real rate.
What this chapter covers
- 011. Original Phillips curve — a stable inverse relation between inflation and unemployment, read (wrongly) as a 1960s policy menu
- 022. The 1970s breakdown — stagflation: rising inflation with rising unemployment kills the stable-trade-off story
- 033. Demand-pull inflation — excess aggregate demand (AD > Yₙ) bids prices up; responds to restrictive policy
- 044. Cost-push inflation — exogenous cost shocks (wages above productivity, oil, depreciation, GST) shift supply up; tight policy mostly raises unemployment
- 055. Expectations-augmented Phillips curve — π = πᵉ − β(u − uₙ); each short-run curve is drawn for a given expected inflation
- 066. NAIRU / natural rate — the long-run Phillips curve is vertical at uₙ; no permanent inflation-unemployment trade-off
- 077. Adaptive vs rational expectations — adaptive (πᵉ = π₋₁) makes disinflation costly; rational and credible can make it near-costless
- 088. Disinflation cost — the sacrifice ratio (excess-unemployment point-years per 1% of inflation reduction) and hysteresis
Sacrifice ratio of a disinflation
- +2Excess unemployment each year is uₜ − uₙ: Year 1 = 8.5 − 4.5 = 4.0, Year 2 = 7 − 4.5 = 2.5, Year 3 = 6 − 4.5 = 1.5, Year 4 = 4.5 − 4.5 = 0.0 point-years.
- +1Total excess unemployment-years = 4.0 + 2.5 + 1.5 + 0.0 = 8.0 point-years.
- +1Inflation fell from 12% to 6%, a fall of 6 percentage points.
- +1Sacrifice ratio = 8.0 ÷ 6 ≈ 1.33 — each one-point reduction in inflation cost about 1.33 point-years of extra unemployment.
- +1Hysteresis: a long period of high unemployment can permanently raise the natural rate uₙ (skills and labour-market attachment erode), so the true long-run cost can exceed the simple ratio.
Key terms
- Original Phillips curve
- A downward-sloping, stable inverse relation between inflation and unemployment, treated in the 1960s as an exploitable policy menu — discredited by 1970s stagflation.
- Demand-pull inflation
- Inflation caused by aggregate demand exceeding the economy's capacity (AD > Yₙ), bidding up prices and wages; it responds to restrictive monetary or fiscal policy.
- Cost-push inflation
- Inflation driven by exogenous cost increases — wages above productivity, dearer imported inputs, depreciation, or higher indirect taxes — that shift supply up, raising prices and unemployment together; tight demand policy does not cure it.
- Expectations-augmented Phillips curve (EAPC)
- Friedman's relation π = πᵉ − β(u − uₙ): each short-run curve is drawn for a given expected inflation, and the curve shifts up as expectations rise, so there is no permanent trade-off.
- Natural rate of unemployment / NAIRU
- The frictional-plus-structural unemployment rate (uₙ) at which inflation neither accelerates nor decelerates; the long-run Phillips curve is vertical here, and only supply-side reform can lower it.
- Adaptive vs rational expectations
- Adaptive expectations extrapolate the recent past (πᵉ = π₋₁), so the curve adjusts slowly and disinflation is costly; rational expectations use all information, so a credible disinflation can be near-costless.
- Sacrifice ratio
- The cumulative excess-unemployment point-years (or lost output) needed per one-percentage-point reduction in inflation; larger under adaptive expectations, near zero under credible rational expectations.
- Hysteresis
- The risk that a prolonged disinflation permanently raises the natural rate of unemployment, because long spells of unemployment erode skills and labour-market attachment — making the true cost exceed the simple sacrifice ratio.
Inflation, Unemployment and the Phillips Curve FAQ
Is the Phillips curve on the ECON5002 exam?
Yes — it is Topic 5, examined only in the 50% closed-book final (which covers Topics 4–8), not the mid-semester test. A Section-B essay typically asks for a labelled short-run and long-run Phillips diagram plus either a demand-pull/cost-push discussion or a sacrifice-ratio or Fisher computation.
What is the difference between the short-run and long-run Phillips curve?
The short-run curve slopes down: for a given expected inflation, lower unemployment comes with higher inflation. The long-run curve is vertical at the natural rate uₙ, because once expectations catch up to actual inflation there is no trade-off — any inflation rate is consistent with u = uₙ.
How do I compute a sacrifice ratio?
Sum the excess unemployment (uₜ − uₙ) over the disinflation period to get the total point-years, then divide by the percentage-point fall in inflation. If the question gives output gaps instead, convert them to unemployment using Okun's Law first.
Why did the original Phillips curve break down?
It described a relationship between the levels of inflation and unemployment and was treated as stable. When policymakers ran the economy hot to exploit it, inflation expectations rose and shifted the whole curve up, producing stagflation in the 1970s — high inflation together with high unemployment — which the level-trade-off reading could not explain.
Does restrictive policy always reduce inflation?
No. Restrictive monetary or fiscal policy works on demand-pull inflation (excess demand), but against a cost-push shock it mostly adds to unemployment without curing the underlying cost increase. Diagnosing the source of inflation comes before choosing the instrument.
Is this guide official or affiliated with the University of Sydney?
No. This is an independent AskSia study resource created to help students revise. It is not produced, endorsed by, or affiliated with the University of Sydney. Always check Canvas and the official unit outline for current dates, weights and rules.
Exam move
Treat this topic as a single argument that moves from a discredited idea to its repair: start with the original downward-sloping Phillips curve, explain why stagflation broke it, then build the expectations-augmented version with its vertical long-run curve at the NAIRU. Be able to draw the short-run-curve family plus the vertical long-run curve from memory, labelled, because the final-exam essays award marks for the correct diagram and the named assumption. Drill the two quantitative cores against the clock — the sacrifice ratio (excess-unemployment point-years divided by the inflation fall) and the Fisher real rate (r ≈ i − πᵉ) — and always state which expectations regime you are in, since adaptive expectations make disinflation costly while credible rational expectations can make it nearly free. Finally, rehearse the demand-pull versus cost-push distinction and the policy that matches each, and add a sentence on hysteresis whenever you cost a disinflation.
