Week 14 · Alon Cohen

Aggregate Supply & the Short-Run Tradeoff Between Inflation and Unemployment

Aggregate Supply & the Short-Run Tradeoff Between Inflation and Unemployment

Part of: Macro-Economics II | Key concepts: Sticky-Price Model, SRAS Curve, Phillips Curve, Adaptive Expectations, Rational Expectations, Sacrifice Ratio, Hysteresis, Inflation Inertia, Natural Rate of Unemployment


What This Chapter Is About

Previous chapters assumed the price level PP was completely "stuck" in the short run — giving a horizontal SRAS curve. That is a useful simplification, but the real world is more nuanced: some prices are sticky, others adjust freely. Chapter 14 builds a richer model of aggregate supply that generates an upward-sloping SRAS curve, then derives the famous Phillips curve — the empirical relationship between inflation and unemployment that sits at the heart of short-run macroeconomic policy.

The chapter answers three big questions:

  1. Why does the SRAS curve slope upward in the short run?
  2. What is the Phillips curve and where does it come from?
  3. Can policymakers exploit the inflation-unemployment tradeoff, and at what cost?

The Sticky-Price Model

Why Are Some Prices Sticky?

In the real world, firms don't instantaneously adjust prices when demand changes. Reasons include:

  • Long-term contracts between firms and customers fix prices for extended periods.
  • Menu costs — the physical and reputational cost of changing a price list.
  • Customer relations — firms fear losing loyal customers if prices fluctuate frequently.
Key assumption

Firms in this model set their own prices (as in monopolistic competition). They are price-setters, not price-takers.

A Firm's Desired Price

Each firm would like to set its price pp according to:

p=P+a(Y−Yˉ)\boxed{p = P + a(Y - \bar{Y})}

where:

  • PP = the overall price level (firms want to track the market)
  • Y−YˉY - \bar{Y} = output gap (actual minus natural/normal output)
  • a>0a > 0 = sensitivity parameter

The intuition:

  • When Y<YˉY < \bar{Y}: demand is weak → firms cut prices to attract buyers.
  • When Y>YˉY > \bar{Y}: demand is strong → firms raise prices, knowing buyers will pay.

Two Types of Firms

Now introduce heterogeneity:

Type Share of all firms Price-setting behaviour
Flexible-price firms 1−s1 - s Set p=P+a(Y−Yˉ)p = P + a(Y - \bar{Y}) after observing PP and YY
Sticky-price firms ss Must set price before PP and YY are known; use expectations

Sticky-price firms set:

p=EP+a(EY−EYˉ)=EPp = EP + a(EY - E\bar{Y}) = EP

(assuming they expect output to equal its natural rate, so EY=EYˉEY = E\bar{Y}).

Deriving the Overall Price Level

The overall price level is a weighted average of what each type of firm charges:

P=s⋅EP⏟sticky+(1−s)⋅[P+a(Y−Yˉ)]⏟flexibleP = s \cdot \underbrace{EP}_{\text{sticky}} + (1-s) \cdot \underbrace{[P + a(Y-\bar{Y})]}_{\text{flexible}}

Solving for PP:

P−(1−s)P=s⋅EP+(1−s)⋅a(Y−Yˉ)P - (1-s)P = s \cdot EP + (1-s) \cdot a(Y - \bar{Y})
sP=s⋅EP+(1−s)⋅a(Y−Yˉ)sP = s \cdot EP + (1-s) \cdot a(Y - \bar{Y})
P=EP+(1−s)s⋅a⋅(Y−Yˉ)\boxed{P = EP + \frac{(1-s)}{s} \cdot a \cdot (Y - \bar{Y})}
What this equation tells us
  • The overall price level rises above its expected level whenever output exceeds its natural rate.
  • The effect is stronger when there are fewer sticky-price firms (smaller ss) — more firms respond to output conditions.

Extreme Cases

Scenario Result
All prices sticky (s=1s = 1) P=EPP = EP → horizontal SRAS (Chapter 10 assumption)
All prices flexible (s=0s = 0) Y=YˉY = \bar{Y} always → vertical LRAS (classical model)
Mixed (typical SR) Upward-sloping SRAS

The SRAS Curve

Deriving the SRAS Equation

Rearranging the price level equation to express output as a function of the price surprise:

Y=Yˉ+α(P−EP)\boxed{Y = \bar{Y} + \alpha(P - EP)}

where α=s(1−s)a>0\alpha = \frac{s}{(1-s)a} > 0.

This is the Short-Run Aggregate Supply (SRAS) curve. It says:

  • Output equals its natural rate when the price level meets expectations (P=EPP = EP).
  • Output exceeds natural rate when prices are above expectations (P>EPP > EP).
  • Output falls short when prices are below expectations (P<EPP < EP).
Common mistake

Students sometimes think "higher prices cause higher output." That's backwards. The causation runs through unexpected price movements: when firms see prices higher than they anticipated, they infer demand is strong and expand production. It is the surprise in prices — not the level — that drives output deviations.

The SRAS Curve in the P-Y Diagram

     P │       LRAS   SRAS (given EP)
       │        │   ↗
   EP  │────────┼──●────────── (equilibrium at Y̅)
       │      ↗
       │────────────────────── Y
            Y̅
  • The SRAS intersects the LRAS at (Y,P)=(Yˉ,EP)(Y, P) = (\bar{Y}, EP).
  • When EPEP rises, SRAS shifts up-left (every output level is now associated with higher actual prices, because sticky-price firms set higher prices based on higher expectations).

Why the SRAS Curve is Bow-Shaped in Reality

The linear equation Y=Yˉ+α(P−EP)Y = \bar{Y} + \alpha(P - EP) assumes a constant slope. In reality:

  • At low output (deep recession): many idle resources available. Expanding production is cheap. Firms don't need big price increases to produce more → curve is relatively flat.
  • At high output (boom, u<unu < u_n): capital and labour near full utilisation. Further expansion is costly. Firms demand large price increases to expand more → curve is steeper.
Reagan's 1982 tax cuts

When Reagan cut taxes in the early 1980s, the economy was emerging from a severe recession — on the flat part of the SRAS. Result: output rose a lot, inflation barely moved. Contrast this with tax cuts during a boom (steep part): the same demand stimulus would be mostly inflationary.


The Phillips Curve

From SRAS to the Inflation-Unemployment Tradeoff

The SRAS curve Y=Yˉ+α(P−EP)Y = \bar{Y} + \alpha(P - EP) relates output to price surprises. We can translate this into inflation and unemployment via two identities:

  1. Okun's Law: Y−Yˉ≈−(u−un)⋅YˉY - \bar{Y} \approx -(u - u_n) \cdot \bar{Y} (output gap ↔ unemployment gap, with opposite sign)
  2. Inflation relationship: π=P−P−1P−1\pi = \frac{P - P_{-1}}{P_{-1}} and similarly for EπE\pi.

Working through the algebra (Steps 1–7 in the lecture):

π=Eπ−β(u−un)+ν\boxed{\pi = E\pi - \beta(u - u_n) + \nu}

where:

  • π\pi = actual inflation rate
  • EπE\pi = expected inflation
  • uu = actual unemployment rate
  • unu_n = natural rate of unemployment
  • β>0\beta > 0 = sensitivity of inflation to cyclical unemployment
  • ν\nu = supply shock term (Greek letter nu)

Comparing SRAS and the Phillips Curve

SRAS Phillips Curve
Output variable Output YY (or output gap Y−YˉY - \bar{Y}) Unemployment uu (or u−unu - u_n)
Price variable Unexpected price level P−EPP - EP Unexpected inflation π−Eπ\pi - E\pi
Relationship Y>YˉY > \bar{Y} when P>EPP > EP u<unu < u_n when π>Eπ\pi > E\pi
Intuition Strong demand → high prices Low unemployment → high inflation

The Phillips curve is simply the SRAS curve rewritten in terms of unemployment and inflation rather than output and the price level. They are two faces of the same coin.

Why unemployment and output move oppositely

By Okun's Law, when output is above potential, unemployment is below its natural rate. So higher-than-expected prices (SRAS) ↔ lower-than-expected unemployment (Phillips curve). The inverse relationship between π\pi and uu follows directly.

What Determines Inflation? Three Forces

The Phillips curve π=Eπ−β(u−un)+ν\pi = E\pi - \beta(u - u_n) + \nu shows that inflation depends on:

1. Expected Inflation (EπE\pi)

If workers and firms expect higher inflation, they build it into wages and prices — causing it to actually occur. This is why inflation expectations must be anchored for price stability.

2. Cyclical Unemployment (u−un)(u - u_n)

  • u<unu < u_n (boom): labour market tight, wages rise, costs rise, prices rise → demand-pull inflation ↑
  • u>unu > u_n (recession): slack demand, wages stagnate, costs fall → inflation ↓

The coefficient β\beta measures how sensitive inflation is to this gap.

3. Supply Shocks (ν\nu)

  • Cost-push inflation: An adverse supply shock (e.g. oil price spike, crop failure, new compliance costs) raises production costs directly, pushing inflation up regardless of demand conditions.
  • ν>0\nu > 0 → inflation rises without any change in unemployment (stagflation risk)
  • ν<0\nu < 0 → inflation falls (favourable supply shock, e.g. productivity boom)
Cost-push vs demand-pull
  • Cost-push examples: Oil embargo (1973), unionised wage increases, new environmental regulations requiring costly equipment.
  • Demand-pull examples: Stock market boom causing a consumption surge; currency depreciation boosting net exports.

Adaptive Expectations and Inflation Inertia

Adaptive Expectations

Adaptive expectations assume people form their inflation forecasts based on recently observed inflation:

Eπ=π−1E\pi = \pi_{-1}

(Expected inflation this period = actual inflation last period.)

Substituting into the Phillips curve:

π=π−1−β(u−un)+ν\boxed{\pi = \pi_{-1} - \beta(u - u_n) + \nu}

Inflation Inertia

This adaptive form implies inflation inertia: in the absence of supply shocks or cyclical unemployment, inflation perpetuates itself.

Why? Past inflation shapes expectations → expectations shape wages and prices → wages and prices determine current inflation. There is a self-fulfilling feedback loop:

πt−1→shapesEπt→shapesπt\pi_{t-1} \xrightarrow{\text{shapes}} E\pi_t \xrightarrow{\text{shapes}} \pi_t
The policy implication of inertia

To reduce inflation, you cannot simply announce a lower target and hope for the best. You must either (a) engineer a recession to create slack (u>unu > u_n), or (b) change expectations credibly so EπE\pi falls on its own. This is the core tension in anti-inflation policy.


The Short-Run Phillips Curve (SRPC)

Graphing the Tradeoff

In the short run (given EπE\pi and ν\nu), policymakers face a downward-sloping tradeoff between inflation and unemployment:

  π │  (high Eπ → higher SRPC)
    │  SRPC₁ (Eπ high)
    │ ↘
    │   SRPC₀ (Eπ baseline)
    │    ↘
    │─────────────────── u
         uₙ
  • Moving along the SRPC: expanding AD → u↓u \downarrow, π↑\pi \uparrow; contracting AD → u↑u \uparrow, π↓\pi \downarrow.
  • The SRPC is not fixed: it shifts when EπE\pi or unu_n changes.

What Shifts the SRPC?

Shift factor Direction Mechanism
Higher expected inflation (Eπ↑E\pi \uparrow) SRPC shifts up Workers and firms pre-emptively raise wages/prices
Higher natural unemployment (un↑u_n \uparrow) SRPC shifts right Cyclical unemployment falls for any given uu, pushing inflation up
Adverse supply shock (ν↑\nu \uparrow) SRPC shifts up Direct cost-push pressure

Short-Run vs Long-Run: Connecting AD-AS and the Phillips Curve

Short-Run AD Shock

AD-AS: An increase in AD moves the economy along SRAS → Y>YˉY > \bar{Y}, P>EPP > EP.
Phillips curve: u<unu < u_n → π>Eπ\pi > E\pi. Economy moves along the SRPC — lower unemployment, higher inflation.

Long-Run Adjustment

Over time, people observe π>Eπ\pi > E\pi and update expectations upward (Eπ↑E\pi \uparrow):

  • AD-AS: SRAS shifts up-left; output returns to Yˉ\bar{Y}.
  • Phillips curve: SRPC shifts up; inflation is now permanently higher even though u=unu = u_n.
Long-run Phillips curve is vertical

In the long run, u=unu = u_n regardless of the inflation rate. The economy is always at the natural rate in the long run — there is no permanent inflation-unemployment tradeoff. The LRPC is a vertical line at u=unu = u_n.


The Sacrifice Ratio

What Is It?

If policymakers want to reduce inflation, they must tighten policy → raise unemployment above unu_n → move up the SRPC to lower π\pi. But this costs real GDP.

The sacrifice ratio measures:

Sacrifice Ratio=% of one year’s real GDP lostPercentage point reduction in inflation\boxed{\text{Sacrifice Ratio} = \frac{\text{\% of one year's real GDP lost}}{\text{Percentage point reduction in inflation}}}

A typical empirical estimate is 5: reducing inflation by 1 pp costs 5% of a year's GDP.

Paul Volcker's Disinflation (1981–1985)

The Fed tightened aggressively to bring inflation from 9.7% (1981) to 3.0% (1985) — a 6.7 pp reduction.

Year uu unu_n Cyclical u=u−unu = u - u_n
1982 9.5% 6.0% 3.5%
1983 9.5% 6.0% 3.5%
1984 7.4% 6.0% 1.4%
1985 7.1% 6.0% 1.1%
Total 9.5%

By Okun's Law (1% unemployment = 2% GDP):

  • Lost GDP = 9.5% × 2 = 19% of a year's real GDP
  • Sacrifice ratio = 19 / 6.7 = 2.8

The realised ratio (2.8) was far below the historical average (5). Why? Possibly because Volcker's credibility caused EπE\pi to fall faster than under adaptive expectations alone — rational expectations at work.


Rational Expectations

The Concept

Rational expectations assume people use all available information — including knowledge of current and future policy — when forming inflation expectations. This contrasts with adaptive expectations, which only use past data.

Adaptive Expectations Rational Expectations
Information used Past inflation only All available information + policy announcements
Implied inertia High Low
Cost of disinflation Potentially large Potentially small

Painless Disinflation?

If the central bank credibly announces a disinflation plan, and if people have rational expectations:

  1. EπE\pi drops immediately — without waiting for a recession.
  2. Phillips curve: π=Eπ−β(u−un)+ν\pi = E\pi - \beta(u - u_n) + \nu → if Eπ↓E\pi \downarrow by the same amount as the inflation target, uu need not rise above unu_n.
  3. Inflation falls with no sacrifice in GDP or employment.
Why this may not work in practice

Painless disinflation requires perfect credibility. If the public doubts whether the central bank will follow through, adaptive expectations dominate, and the sacrifice ratio remains large. The Volcker episode suggests partial credibility: the sacrifice ratio was lower than the historical average but not zero. True painless disinflation has never been fully observed.


Hysteresis

The Natural-Rate Hypothesis

The natural-rate hypothesis says: changes in aggregate demand affect output and employment only in the short run. In the long run, the economy always returns to its natural rates (Yˉ\bar{Y}, unu_n). This is the assumption underlying all preceding analysis.

Hysteresis: A Challenge to the Natural-Rate Hypothesis

Hysteresis refers to the possibility that negative shocks can permanently raise the natural rate of unemployment. The economy may not fully recover.

Mechanisms:

  1. Skill atrophy: Workers unemployed during a recession may lose skills or fall out of professional networks. When recovery comes, they may be structurally unemployable — raising unu_n permanently.
  2. Insider-outsider dynamics: Employed "insiders" bargain for higher wages for themselves. Long-term unemployed "outsiders" lose bargaining power and become structurally unemployed, even after the recession ends.
Why hysteresis matters for policy

If hysteresis is real, the costs of a deep recession are not temporary — they permanently worsen the economy's supply side. This strengthens the case for aggressive counter-cyclical stabilisation policy (preventing the recession in the first place) rather than accepting unemployment as a short-run cost of disinflation.


Summary

Concept Key equation/formula Key insight
Sticky-price model p=P+a(Y−Yˉ)p = P + a(Y - \bar{Y}) Firms want prices to reflect demand; sticky-price firms use expectations
SRAS curve Y=Yˉ+α(P−EP)Y = \bar{Y} + \alpha(P - EP) Output rises above Yˉ\bar{Y} only when prices exceed expectations
Phillips curve π=Eπ−β(u−un)+ν\pi = E\pi - \beta(u - u_n) + \nu Inflation-unemployment SR tradeoff; derived from SRAS
Inflation inertia π=π−1−β(u−un)+ν\pi = \pi_{-1} - \beta(u - u_n) + \nu Past inflation perpetuates itself via adaptive expectations
Sacrifice ratio ≈5\approx 5 (typical) % GDP lost per 1 pp reduction in inflation
Rational expectations EπE\pi uses all info May allow painless disinflation if credibility is high
Hysteresis Recession → un↑u_n \uparrow Deep recessions can permanently raise the natural unemployment rate

Chapter Takeaways

The SRAS curve slopes upward because some prices are sticky and only flexible-price firms can respond immediately to output conditions. Rewriting this in terms of unemployment and inflation yields the Phillips curve, which shows policymakers a short-run menu: lower unemployment at the cost of higher inflation (or vice versa). But this menu is not fixed — it shifts with inflation expectations and supply shocks. Reducing inflation is costly in terms of lost output (the sacrifice ratio), unless expectations adjust rapidly under rational expectations and credible policy. Finally, hysteresis warns that tolerating unemployment may have permanent consequences, not just transitory ones.