Macro Stabilization Policy (C17+):

Is there an
inflation/unemployment
trade-off?
Its 1957 and this New
Zealander, Alban William
Housego Phillips, plots some
historical data on inflation and
unemployment in Britain and
thinks he sees something
huge: when inflation goes up,
unemployment goes down! If
this negative relationship holds,
we can reduce the suffering of
joblessness by enduring a little
more inflation.
1

The
pre1970s
Phillips
Curve
(U.S.A.)

Inflation rate

68

(% change in GDP
price deflator)

Phillips curve
57
55

66
56

67

65

60

2
64

54

62
59
63

58

61

Unemployment
rate (%)

Questions: What causes this relationship? Does it exist for all

countries? Is it stable? Why does the trade-off change (i.e., why is
it non-linear)? How should we use it?

Figure 17.2

How the Phillips Curve happens: Shifts in AD along SRAS

In panel (a), the economy in 2013 is at point A, with real GDP of $14.0 trillion and a price level of 100.
If there is weak growth in aggregate demand, in 2014, the economy moves to point B,
with real GDP of $14.3 trillion and a price level of 102.
The inflation rate is 2 percent and the unemployment rate is 6 percent,
which corresponds to point B on the Phillips curve in panel (b).
If there is strong growth in aggregate demand, in 2014, the economy moves to point C,
with real GDP of $14.6 trillion and a price level of 104.
Strong aggregate demand growth results in a higher inflation rate of 4 percent but a lower unemployment
rate of 5 percent. This combination of higher inflation and lower unemployment is shown as point C on the
Phillips curve in panel (b).

But is this relationship

Note that stable?
the previous graphic
suggested that AD Real
GDP unemployment in the
short run (i.e., as we move along
SRAS)
That hints that maybe this
favorable effect on
unemployment might not survive
in the LR
But wait: will we consistently see
this effect even in the SR?
Thatll depend on how we form
our expectations

Expectations and Macro policy:

Two theories of expectations formation

Adaptive Expectations*: Individuals form their

expectations about the future on the basis of
data from the recent past. E.g., if inflation was
2% last year, itll probably be 2% this year.
Rational Expectations: Assumes people use
all pertinent information, including data on the
conduct of current policy, in forming their
expectations about the future. E.g., inflation
mightve been 2% last year, but I hear the Fed is
cutting the required reserve ratio, so

*Not discussed in H&O, btw.

Adaptiv
e
expectations
Were
easy
to fool

Actual rate
of inflation (%)
12

Actual rate
of inflation

8
4

Time
period
Expected rate
of inflation (%)

Corresponding expected
rate of inflation in next period

12
8
4
1

Time
period

Under adaptive expectations, what occurs during the most recent period (or set
of periods) determines an individuals future expectations. So, the expected future
rate of inflation lags behind the actual rate by one period as expectations are
altered over time.

Adaptive vs. rational expectations

Under adaptive expectations, people are easy

to fool an unanticipated shift to a more
expansionary policy will temporarily stimulate
output and employment.
With rational expectations, decision-makers do
not make systematic, repeated errors and
therefore the impact of expansionary policies is
unpredictableand possibly neutral even in the
short run.
But both expectations theories predict that
sustained expansionary policies will lead to
inflation without permanent increase in output and
employment (i.e., return to LR equilibrium).

SR
stimulus:
Adaptive
expectations

Price
Level

LRAS
SRAS1

P2
P1

e2
E1

AD1
YF Y2

AD2

Goods & Services

(real GDP)

Under adaptive expectations, anticipation of inflation will lag behind

its actual occurrence.
Thus, expansionary policy will aggregate demand and lead to a
temporary in GDP (to Y2) and modest in prices (to P2).
8

SR
stimulus
:
Rational
expectations

Price
Level

LRAS

SRAS2
SRAS1

P2

E2

P1

E1

AD1
YF

AD2

Goods & Services

(real GDP)

Under rational expectations, decision makers expect the inflationary impact

of a demand-stimulus policy. So, though AD (to AD2)
resource prices and production costs rise just as rapidly (thereby shifting
SRAS to SRAS2).
Result: P , but Y doesnteven in SR. Policy Ineffectiveness Thm.
9

Whos right? What do most macroeconomists think?

While expectations may not be formed perfectly

rationally, policymakers ability to make unanticipated
moves has clearly diminished in recent years.
Result: much tougher to get Y > Yf these days, and SR is
getting shorter (i.e., convergence to LRAS more rapid).
Emerging consensus views:

Best monetary policy is one aimed at price stability (persistently

low rates of inflation);
High living standards are usually the result of free market
efficienciesgains from trade, specialization & division of labor,
mass productionand technological change. Low inflation and
the smooth operation of the pricing system facilitates the
realization of these gains.
In contrast, high and variable rates of inflation create uncertainty,
distort relative prices, and reduce the efficiency of markets.

Emerging consensus views (cont.)

Demand stimulus policies cannot reduce the

unemployment below the natural rate at least
not for very long.
Wide swings in both monetary and fiscal policy
should be avoided.
In countries with substantial checks and balances
built into the political process (i.e., where budget
changes take lotsa time), fiscal policy is best
reserved for large-displacement situations, not
for fine tuning. As the unstable Phillips Curve
showed over time

Adaptive Expectations and the Phillips Curve

Begin at full-employment output YF , (pt A in both frames).

With adaptive expectations, a shift to an expansionary

policy will increase prices, expand output beyond fullemployment, and reduce the unemployment rate below its
natural level (move to pt B in both frames).
Price
Level

LRAS

Rate of
inflation
PC1

(stable prices
anticipated )

SRAS1
8%
P104
P100

B
A

YF Y2

4%

AD2

AD1

Goods &
Services
(real GDP)

B
A
1% 3% 5% 7%

Rate of
unemployment

Expectations and the shifting Phillips Curve

Decision makers eventually anticipate rising prices and incorporate them into their
decision making (shifting SRAS1 to SRAS2, returning Y to YF and unemployment to
the natural rate (pt C in both frames).

Question: At C, is there an inflation/unemployment trade-off because

the Phillips Curve has shifted, or will people anticipate further expansion
and inflation, so that AD & SRAS both shift upward without increases in
output & employment (leading to a vertical long run Phillips curve)?
Price
Level

LRAS

P112

SRAS3
SRAS2
SRAS1

P108

P104
P100

YF Y2

Long-run Phillips curve

(natural rate of unemployment)

PC1

(stable prices
anticipated )

8%
B

Rate of
inflation

AD3
AD2

AD1

PC2

(4% inflation
anticipated )

4%
Goods &
Services
(real GDP)

A
1% 3% 5% 7%

Rate of
unemployment

Real Inflation rate

in GDP
shifts (change
price deflator)
in the
Phillip
s
Curve

10 %

80 81

74

75

79

8%
6%
4%
2%
0%

78
77

73
69
68

71
70
89

66
00

67

76

90 72
88
95 87

91

84
85
92
93
86

65
62 94
97
96 61
99
64
63
98

82

PC3 (1974-1983 )
83

PC2 (1970-1973,
1984-1993)

PC1 (1961-1969,
1994-2000)

Unemployment
3 % 4 % 5 % 6 % 7 % 8 % 9 % 10 % rate

Source: Economic Report of the President, 2001

Monetary restraint in 1984-1993 unexpectedly decelerated inflation, raising

unemployment until people adjusted their inflationary expectations downward.
Low rates of inflation were maintained, reducing inflationary expectations, and
the 1994-2000 Phillips curve appears to be in a position similar to the 60s.

14

Figure 17.3 Conclusion: A Vertical Long-Run Aggregate Supply Curve Means

a Vertical Long-Run Phillips Curve

Milton Friedman and Edmund Phelps argued that there is no trade-off between
unemployment and inflation in the long run.
If real GDP automatically returns to its potential level in the long run,
the unemployment rate must return to the natural rate of unemployment in the long run.
In this figure, we assume that potential GDP is $14 trillion and the natural rate of
unemployment is 5 percent.

Figure 17.6
A Short-Run Phillips
Curve for Every
Expected Inflation Rate

There is a different
short-run Phillips curve
for every expected
inflation rate.
Each short-run Phillips
curve intersects the
long-run Phillips curve
at the expected
inflation rate.

Last but not least:

Real business cycle models

An argument that shocks on

the supply side such as
technological changes or
changes in availability of
productive resources can yield
changes in Real GDP even if
expectations are rational
See, e.g., mid-70s oil shock,
when OPEC more than tripled
the world price of oil

Prescott & Kydland

Figure 17.9

The 70s Oil Shock Shifts the SRAS and the Short-Run Phillips Curve

When OPEC increased the price of a barrel

of oil from less than $3 to more than $10, in
panel (a), the SRAS curve shifted to the left.
Between 1973 and 1975, real GDP declined
from $4,917 billion to $4,880 billion,
and the price level rose from 28.1 to 33.6.

Panel (b) shows that the supply shock

shifted up the Phillips curve.
In 1973, the U.S. economy had an
inflation rate of about 5.5 percent and an
unemployment rate of about 5 percent.
By 1975, the inflation rate had risen to
about 9.5 percent and the unemployment
rate to about 8.5 percent.

The Phillips Curve verdict

So the inflationary (and eventually

stagflationary) 70s were a disaster
for believers in the Phillips Curve
Phillipss discovery ultimately
proved to be an unstable, dangerous
conception: we bought temporary
(and inefficient?) reductions in the
unemployment rate with ever-higher
inflation
Howd we recover? With Paul
Volcker, tight money and a
recession

Paul Volcker and Disinflation

Figure 17.10
The Fed Tames Inflation,
19791989

The Fed, under Chairman

Paul Volcker, began
fighting inflation in 1979
by reducing the growth of
the money supply, thereby
raising interest rates.
By 1982, the unemployment
rate had risen to 10 percent,
and the inflation rate had
fallen to 6 percent.
As workers and firms lowered
their expectations of future inflation,
the short-run Phillips curve shifted down,
improving the short-run trade-off between unemployment and inflation.
This adjustment in expectations allowed the Fed to switch to an expansionary monetary policy,
which by 1987 brought the economy back to the natural rate of unemployment,
with an inflation rate of about 4 percent.
The orange line shows the actual combinations of unemployment and inflation
for each year from 1979 to 1989. Note that during these years,
the natural rate of unemployment was estimated to be about 6 percent.