BA112: Intermediate

Macroeconomics I
Unit 3
O. Blanchard, Macroeconomics, 4th
Edition, Chapter - 8 and Chapter - 9
C.L.F.Attfield, D. Demery and N.W.
Duck (1991, 2nd edn. ), p1– p28
Steven Sheffrin, (1996, 2nd edn.),
p25– p40
 Chapter – 8
The Natural Rate of
Unemployment and the Phillips
Curve
 Summary of the Chapter
 In 1958, A.W.Phillips plotted inflation rate against
unemployment rate for the United Kingdom using the data
from 1861 to 1957.
 And he found a negative between unemployment rate and
inflation rate i.e. when unemployment rate was low, inflation
rate was high; and when unemployment rate was high,
inflation rate was low.
 Two years later in 1960, Paul Samuelson and Robert Solow
replicated the Phillips’s exercise for the United States using
the data from 1900 to 1960. They too found a negative
relation between unemployment rate and inflation rate.
 This negative relation between unemployment rate and
inflation rate was then called as Phillips Curve/Original
Phillips Curve. From this point on, Phillips curve rapidly
became central to Macroeconomic thinking and policy.
 Summary of the Chapter
 However, in 1970s, the Phillips curve broke down. During
1970s, in the United States and in many major countries,
there was both high inflation rate and high unemployment rate
- clearly contradicting the original Phillips curve.
 And, later, the relation reappeared as a negative relation
between the unemployment rate and the change in inflation
rate. This means if unemployment rate is high, there will be
decrease in inflation rate over time; and if unemployment rate
is low, there will be increase in inflation overtime.
 This reappeared relation, showing negative relation between
unemployment rate and change in inflation rate is known as
Modified Phillips curve or Expectations-augmented
Phillips curve or Inflation adjusted Phillips curve.
 In this chapter, we are going to discuss everything explained
above in detail.
 Relation Between Inflation, Expected
Inflation, and Unemployment
 Substituting wage determination Eqn into price determination
Eqn from Chapter-6 of Blanchard, we get
(-) (+)
P = Pe (1+μ) F(u, z)
 Assume a specific form of F(u, z) = 1 – αu + z. The captures
the notion that higher is unemployment rate, lower is the
wage; and the higher is z, the higher is the wage. The
parameter α captures the strength of the effect of
unemployment rate on the wage.
 Substituting F(u, z) = 1 – αu + z in the top Eqn, we get
P = Pe (1+μ) (1 – αu + z) (1)
 Finally, let π denote the actual inflation rate and πe denote the
expected inflation rate. Then Eqn (1) can be rewritten as:
 Relation Between Inflation, Expected
Inflation, and Unemployment
π = πe + (μ + z) – αu (2)
 Eqn (2) has been derived from Eqn (1) in terms of rate of
change. See appendix of Chapter – 8 for the derivation. Eqn
(2) shows the relation between inflation (π), expected
inflation (πe), and unemployment (u).
Eqn (2) has three properties:
a) An increase in expected inflation (πe) leads to an increase in
inflation/actual inflation (π). Just like an increase in expected price
level (Pe) leads to increase in actual price level (P) [as we have
seen in previous chapter while talking about AS curve], increase in
expected inflation (πe) will lead to increase in actual inflation (π).
b) Given expected inflation (πe), an increase in markup (μ) or an
increase in factors that affect wage determination (z), leads .......
 Relation Between Inflation, Expected
Inflation, and Unemployment
……… to an increase in inflation/actual inflation (π). We know
that an increase in μ (meaning increase in price over the cost) or
an increase in z (by increasing the nominal wage and hence by
increasing the cost of production of firms) leads to increase in
actual price level (P). Hence, it will lead to increase in
inflation/actual inflation (π) as well.
c) Given expected inflation (πe), an increase in the
unemployment rate (u) leads to a decrease in inflation/actual
inflation (π). We know that a higher unemployment rate (by
reducing the bargaining power of workers) leads to fall in nominal
wages paid to the workers. As a result, firms’ costs of production
fall and they reduce the price charged for their produce, causing
price level (P) in the economy to fall. This fall in price level/actual
price level (P) translates into fall in inflation/actual inflation (π).
 Relation Between Inflation, Expected
Inflation, and Unemployment
 Using time indexes (in order to relate variables like inflation or
expected inflation or unemployment, to a specific year), Eqn
(2) can be rewritten as:
πt = πte + (μ + z) – αut (3)

where πt denote inflation/actual inflation rate in period t; π te

denote expected inflation rate in period t; u t denote unemployment
rate in period t.
Original Phillips Curve - Incarnation
 Before 1970s, inflation was positive in some years; negative
in some years, and hence on average (over the years)
inflation was zero.
 With average inflation equal to zero in the past, wage setters
e
 Original Phillips Curve - Incarnation
 Substituting πte = 0 in Eqn (3), we get the relationship
between unemployment rate and inflation rate as:
πt = (μ + z) – αut (4)
 Eqn (4) is the Equation of Original Phillips curve showing
negative relationship between unemployment rate (ut) and
inflation rate (πt).
 This is precisely the negative relation between unemployment
rate and inflation rate that Phillips found for the UK, and
Solow and Samuelson found for the US.
 The logic for negative relation: lower unemployment rate
leads to a higher nominal wage. The higher nominal wage (by
increasing the cost of production of firms) leads to a higher
price level this year relative to last year’s price level i.e. to
higher inflation. This mechanism has sometimes been
 Mutations in Original Phillips Curve
 In 1970s, the original Phillips curve which says there is
negative relation between unemployment rate (ut) and inflation
rate (πt), broke down.
 Now the question is why did the original Phillips curve
broke down in 1970s?
Answer:
This happened because there was change in the behaviour of
inflation in 1970s. To be specific, in 1970s, inflation became
consistently positive and more persistent: High inflation in
one year became more likely to be followed by high inflation next
year.
So, wage setters changed the way they formed their
expectations. This change in expectation formation changed
the nature of the relation between unemployment rate …………
 Mutations in Original Phillips Curve
……….and inflation rate and thereby resulted in the break down
of the original Phillips curve.
 Since a high inflation in one year became more likely to be
followed by high inflation next year, the change in expectation
formation can be summarized by the following equation:
πet = θ πt-1 (5)
where θ captures the effect of last year’s inflation (πt-1) on this
year’s expected inflation (πet). Higher θ means inflation expected
for this year (or this year’s expected inflation) is highly
responsive to last year’s inflation.
[Note: As long as inflation was low and not very persistent, wages
setters ignored the past year inflation (as happened before 1970s)
while determining current year’s expected inflation and θ was
close to zero. But as inflation became consistently positive……….
 Mutations in Original Phillips Curve
……….and more persistent, wage setters changed the way they
formed their expectations (as happened during 1970s and
thereafter). They started assuming that if inflation had been high
last year, inflation was likely to be high this year as well. Hence, the
parameter θ which captures the effect of last year inflation on this
year’s expected inflation, increased and became positive ].
 Substituting Eqn (5) into Eqn (3), we get
πt = θ πt-1 + (μ + z) – αut
 In many countries including US, θ was found to be 1. So,
substituting θ = 1 in above Eqn, we get
πt - πt-1 = (μ + z) – αut (6)
 Eqn (6) is the modified Phillips curve/expectations-
augmented Phillips curve/inflation adjusted Phillips
curve, showing negative relationship between ………….
 Mutations in Original Phillips Curve
………. the unemployment rate (ut) and the change in inflation
rate (πt - πt-1).
Introducing natural rate of
unemployment (un) in the modified
Phillips Curve
 Milton Friedman and Edmund Phelps redefined the modified
Phillips curve/expectation augmented Phillips curve (6) by
introducing the role of natural rate of unemployment (un).
 We know as per modified Phillips curve (6), there is a trade off
between unemployment rate and change in inflation rate i.e.
government can sustain lower unemployment by accepting
higher inflation.
 Friedman and Phelps argued that if government attempts…….
 Introducing natural rate of
unemployment (un) in the modified
Phillips Curve
………. to sustain lower unemployment by accepting higher
inflation, the trade-off will eventually disappear because
unemployment rate cannot be sustained below a certain level - a
level known as natural rate of unemployment (un).
 Hence, there is a need to introduce the role of natural rate of
unemployment in the modified Phillips curve relation (6).
 For that, first we solve for the value of natural rate of
unemployment (un) in this case.
 We know natural rate of unemployment is a unemployment
that exists in the long run and hence at the natural rate of
unemployment, actual price level (P) will be equal to expected
price level (Pe). Equivalently, actual inflation rate (π) ………..
 Introducing natural rate of
unemployment (un) in the modified
Phillips Curve
 
………. will be equal to expected inflation rate (πe).
 Substituting πt = πte in Eqn (3) and solving for u, we get
natural rate of unemployment (un) equal to:
un = (7)
 Rewriting the Eqn (3) as
πt - πte = – α (ut - )
 Using Eqn (7) in above equation, we get
πt - πte = – α (ut - un) (8)
 With un, Eqn (8) is the final/general Eqn for modified
Phillips curve/expectation augmented Phillips curve.
 Introducing natural rate of
unemployment (un) in the modified
Phillips Curve
 Eqn (8) shows the negative relation between unexpected
inflation (πt - πte) and cyclical unemployment/excess
unemployment (ut - un). And it is the most relevant Phillips
curve for the real world today.
 Remember πet = θ πt-1 and θ was found to be 1 in many
countries including US; so we put πet = πt-1 in Eqn (8) and we
get final modified Phillips curve as:
πt - πt-1 = – α (ut - un) (9)
 Eqn (9) gives us another way of thinking about the Phillips
curve as the negative relation between change in inflation
rate, and the excess of unemployment rate over the
 Introducing natural rate of
unemployment (un) in the modified
Phillips Curve
 Eqn (9) implies that
(a) If ut > un πt - πt-1 < 0 inflation will fall overtime.
(b) If ut < un πt - πt-1 > 0 inflation will rise overtime.
(c) If ut = un πt = πt-1 no change in inflation overtime.
 We note that when actual unemployment rate (u) is equal to
natural rate of unemployment (un), there is no change in
inflation overtime i.e. inflation rate remains constant. That is
why natural rate of unemployment (un) is also called
nonaccelerating inflation rate of unemployment (NAIRU).
 Points to Note
 The final modified Phillips curve relation Eqn (8) or Eqn (9)
works quite well in the real world under the normal situation.
 However, under extreme situations such as (a) in the periods of
very high inflation and (b) in the periods of very low inflation (or
deflation); the new Phillips curve relation as implied by Eqn (8)
or Eqn (9) breaks down/disappear. (Note: deflation means
negative inflation i.e. decrease in price level overtime)
 We know, generally, nominal wages are determined on the
basis of expected inflation (equivalently, on the basis of
expected price level). But if there is a wage indexation, then
nominal wages will be indexed to actual inflation i.e. nominal
wages will be determined on the basis of actual inflation.
 So, if ‘λ’ proportion of total wage contracts in the economy are
indexed and ‘(1-λ)’ proportion of total wage contracts are not
indexed, then in the final modified Phillips curve Eqn (8), we
will have ‘λπt + (1-λ)πet’ in place of ‘πet’.
 Examples
Example1: Chap-8 (Q.3):
Suppose that the Phillips curve is given by:
πt = πet + 0.1 – 2ut
(a) What is the natural rate of unemployment?
Now assume: πet = θ πt-1
and assume that θ is initially equal to zero. Suppose that the rate
of unemployment is initially equal to the natural rate. In year ‘t’, the
authorities decide to bring the unemployment rate down to 3% and
hold it there forever.
(b) Determine the rate of inflation in years t, t+1, t+2 and t+4 ?
(c) Do you believe the answer given in part (b)? Why or why not?
(Hint: think about how people are likely to form expectation of
inflation).
Now suppose in year t+5, θ increases from 0 to 1. Suppose that
the government is still determined to keep ‘u’ at 3% forever.
 Examples
Example1: Chap-8 (Q.3):
(d) Why might θ increase in this way?
(e) What will the inflation rate be in years t+5, t+6 and t+7?
Answer:
(a) Phillips curve has been given as: πt = πet + 0.1 – 2ut
Rewriting it as: πt – πet = – 2 (ut – 0.05) (1a)
Compare with general Phillips curve Eqn (8): π t - πte = – α (ut - un)
We get natural rate of unemployment (un) = 0.05 or 5%.
(b) Inflation in year t i.e. π t ?
With θ = 0, we get πet = 0. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t = 3% or 0.03
Substituting these values in Eqn (1a), we get inflation in year t i.e.
 Examples
Inflation in year t+1 i.e. πt+1 ?
Rewriting Eqn (1a) in terms of year t+1, we get
πt+1 – πet+1 = – 2 (ut+1 – 0.05)
Again with θ = 0, we get πet+1 = 0. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t+1 = 3% or 0.03
Substituting these values in above Eqn, we get inflation in year t+1
i.e. πt+1 = 0.04 or 4%.
Inflation in year t+2 i.e. πt+2 ?
Rewriting Eqn (1a) in terms of year t+2, we get
πt+2 – πet+2 = – 2 (ut+2 – 0.05)
Again with θ = 0, we get πet+2 = 0. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t+2 = 3% or 0.03
Substituting these values in above Eqn, we get inflation in year t+2
 Examples
Inflation in year t+4 i.e. πt+4 ?
Rewriting Eqn (1a) in terms of year t+4, we get
πt+4 – πet+4 = – 2 (ut+4 – 0.05)
Again with θ = 0, we get πet+4 = 0. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t+4 = 3% or 0.03
Substituting these values in above Eqn, we get inflation in year t+4
i.e. πt+4 = 0.04 or 4%.
This is true for every year starting from year t i.e. inflation will
be equal to 4% forever.
(c) With πe = 0 and π = 4% forever in part (b), inflation
expectations will be forever wrong. This is unlikely in real world.
(d) In year t+5, θ has increased to 1. This implies π et+5 = πt+4
θ might increase because people’s expectations will adapt to
 Examples
..…..inflation has been 4% per annum. So, expected inflation for
year t+5 i.e. πet+5 will also be equal to last year’s inflation (π t+4) ,
which means θ = 1.
(e) Inflation in year t+5 i.e. πt+5 ?
Rewriting Eqn (1a) in terms of year t+5, we get
πt+5 – πet+5 = – 2 (ut+5 – 0.05)
With θ = 1, we get πet+5 = πt+4. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t+5 = 3% or 0.03
Substituting these values in above Eqn , we get
πt+5 – 0.04 = – 2 (0.03 – 0.05)
Solving it, we get inflation in year t+5 i.e. πt+5 = 0.08 or 8%.
Inflation in year t+6 i.e. πt+6 ?
Rewriting Eqn (1a) in terms of year t+6, we get
 Examples
With θ = 1, we get πet+6 = πt+5. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t+6 = 3% or 0.03
Substituting these values in above Eqn, we get
πt+6 – 0.08 = – 2 (0.03 – 0.05)
Solving it, we get inflation in year t+6 i.e. πt+6 = 0.012 or 12%.
Inflation in year t+7 i.e. πt+7 ?
Rewriting Eqn (1a) in terms of year t+7, we get
πt+7 – πet+7 = – 2 (ut+7 – 0.05)
With θ = 1, we get πet+7 = πt+6. And government wants to keep
unemployment rate (u) equal to 3% forever i.e. u t+7 = 3% or 0.03
Substituting these values in above Eqn, we get
πt+7 – 0.012 = – 2 (0.03 – 0.05)
 Examples
Similarly, you can solve the value of inflation for all the subsequent
years t+8, t+9, t+10 etc.

Solve all the back questions of Chapter-

8 of Blanchard (Homework)
 Chapter – 9
Inflation, Activity and Nominal
Money Growth
 Nominal Money Growth, Output growth,
Unemployment and Inflation
In  this chapter, we will focus on the effects of nominal
money growth on unemployment rate, output growth rate
and inflation rate. For this, we will make use of three
macroeconomic relations:
1) Okun’s law: This law is named after an economist called
‘Arthur Okun’ who studied the relationship between output
growth rate and unemployment rate in1960s.
Okun’s law explains the relationship between output growth rate
and the change in unemployment rate. In other words, it relates
change in unemployment rate to the deviation of output growth
rate from its normal rate. It is represented by following equation:
ut – ut-1 = - Ω (gyt - y) (10)
where gyt denotes actual output growth rate of the economy……
Nominal Money Growth, Output growth,
Unemployment and Inflation
 
….....and y denotes the normal output growth rate of the

economy, which is equal to the sum of labor productivity growth

rate and labor force growth rate. The coefficient Ω measures the
effect of output growth above normal on the change in the
unemployment rate.
 Eqn (10) implies that output growth above normal (gyt > y) will
lead to decrease in unemployment rate over time (ut < ut-1)
and output growth below normal (gyt < y) will lead to increase
in unemployment rate overtime (ut > ut-1).
2) The Phillips Curve: It relates the change in inflation rate to
the deviation of unemployment rate from its natural rate as
represented by Eqn (9) in the previous chapter:
πt - πt-1 = – α (ut - un) (9)
 Nominal Money Growth, Output growth,
Unemployment and Inflation
 As
  shown in previous slide, Phillips curve relation implies that
unemployment rate above the natural rate (ut > un) leads to
decrease inflation rate overtime (πt < πt-1) and unemployment
rate below the natural rate (ut < un) leads to increase in
inflation rate overtime (πt > πt-1).
3) The Aggregate Demand (AD) Relation: it relates the
output growth to the difference between nominal money growth
and inflation.
From Chapter-7 of Blanchard, we know that the aggregate
demand (AD) relation is given by: Y = F(, ,, T)
To focus on the relationship between output, nominal money and
price level, we ignore the changes in the factors other than
nominal money () and price level (P) here, and write …………
Nominal Money Growth, Output growth,
Unemployment and Inflation
 
…….the aggregate demand (AD) relation simply as a positive
function of real money supply (/P) as :
Y=Φ , Φ>0 and is a constant

 In terms of growth rates, the above AD demand relation can

be rewritten as:
gyt = gmt – πt (11)
where gyt denotes actual growth rate of the economy; gmt
denotes nominal money growth/growth rate of nominal money
supply; πt denotes the inflation rate/rate of growth in price level.
gmt – πt denotes real money growth/growth rate in real money
supply.
Nominal Money Growth, Output growth,
Unemployment and Inflation
……. inflation (gmt > πt), real money growth is positive (gmt – πt >
0), and so is output growth (gyt > 0). And if nominal money
supply is less than inflation (gmt < πt), real money growth is
negative (gmt – πt < 0), and so is output growth (gyt < 0).
 The three relations: Okun’s law Eqn (10), Phillips Curve
relation Eqn (9) and AD relation Eqn (11) are shown in the
figure at the next slide.
 It can be noted from the figure that: Through aggregate
demand relation, nominal money growth and inflation
determine output growth.
 Through Okun’s law, output growth determines the change in
unemployment rate. And, through the Phillips curve relation,
unemployment rate determines the change in inflation.
Nominal Money Growth, Output growth,
Unemployment and Inflation
Figure showing the interlinkage between three relations :

Nominal Money Output Growth

Growth
Aggregate Okun’s
Demand Law

Inflation Unemployment
Phillips
Curve
 Effects of Nominal Money Growth
 Using
  the three relations Eqn (9), Eqn (10) and Eqn (11), we
can determine the effect of change in nominal money growth
(gm) on unemployment rate, output growth rate and inflation
rate in short run as well as in the medium/long run.
Three facts regarding medium/long run :
1) In the medium/long run, unemployment rate must be
constant; the unemployment rate cannot be increasing or
decreasing forever. Putting ut = ut-1 in Okun’s law [Eqn (10)]
implies gyt = y. This means, in the medium/long run, output must
grow at its normal rate of growth, y.
2) With nominal money growth equal to gm and output growth
equal to y in the medium/long run, using AD relation [Eqn (11)]
inflation rate in the medium/long run will be given by
 Effects of Nominal Money Growth
  In the medium/long run, we do not find expectational error i.e.
3)
actual inflation will be equal to expected inflation. Putting πt = πte
i.e. πt = πt-1 (as we know with θ=1, πte = πt-1) in the Phillips curve
relation [Eqn (9)], we get ut = un. This means in the medium/long
run, unemployment rate (u) must be equal to natural rate of
unemployment (un).
Question: Suppose that economy is initially working at
medium/long run equilibrium i.e. gyt = y , π = gm - y and ut = un.
Now suppose there is decrease in nominal money growth (gm).
How does it affect unemployment rate, inflation rate and output
growth rate in the short run and in the medium/long run?
Short run effect:
Through AD relation [Eqn (11)], the decrease in gm leads to
decrease in output growth rate (gyt) at the initial rate of inflation.
 Effects of Nominal Money Growth
 
…………[Eqn (10)] leads to increase in unemployment rate. This
increase in unemployment rate (u) though Phillips curve relation
[Eqn (9)] leads to decrease in inflation rate (π).
Medium/long run effect:
In the medium/long run, output growth rate (gyt) will return to its
normal rate (y). This means the decrease in nominal money
growth will have no impact on output growth rate in the
medium/long run.
In the medium/long run, unemployment rate (ut) will return to its
natural rate (un). This means the decrease in nominal money
growth will have no impact on unemployment rate in the
medium/long run.
However, as we know, inflation in the medium/long run is given
by π = gm - y. Since y is constant, that means there will be one
 Effects of Nominal Money Growth
………..money growth (gm). This means the decrease in nominal
money growth (gm) will lead to equal decrease in inflation rate
(π) in the medium/long run.
Summary: In the short run, monetary tightening (i.e. decrease in
nominal money growth) leads to a slowdown in growth and a
temporary increase in unemployment.
But in the medium/long run, output growth returns to normal,
unemployment rate returns to the natural rate. Nominal money
growth and inflation are permanently lower at this point.
Putting differently, the temporary increase in unemployment
(through decrease in nominal money growth) buys a permanent
decrease in inflation.
 Disinflation and Sacrifice Ratio
 Disinflation means decrease in inflation rate. From the
Phillips curve [Eqn (9)], we know that the disinflation i.e.
decrease in inflation requires increase in unemployment rate
above the natural rate of unemployment. In other words, we
can say that disinflation can only be achieved when there
is excess unemployment in the economy.
 However, the point to note here is that the total amount of
excess unemployment required for a given decrease in
inflation does not depend on the speed at which the decrease
in inflation is achieved.
 In other words, disinflation can be achieved quickly at the
cost of high unemployment for a few years Or it can achieved
more slowly with a smaller increase in unemployment spread
over many years. In both the cases, the total amount of
excess unemployment, summed over the years, will be same.
 Disinflation and Sacrifice Ratio
 The
  discussion at the previous slide can be summarized by a
term called ‘Sacrifice Ratio’. Sacrifice ratio is defined as the
total amount of excess unemployment required to achieve 1%
decrease in inflation. In other words, it tells you by how much
unemployment rate will increase when inflation rate falls by
1%. It is calculated as:
Sacrifice ratio =

 Since total amount of excess unemployment required to

achieve a given decrease in inflation does not depend on the
speed at which the decrease in inflation is achieved, the
sacrifice ratio will be same irrespective of the speed at which
decrease in inflation is achieved.
 Disinflation and Sacrifice Ratio
 Given
  the Phillips Curve relation [Eqn (9)], the sacrifice ratio
is given by:
Sacrifice ratio =

= =
 Since is constant at a given point in time, the sacrifice ratio
(1/ ) will also be constant.
 Lucas Critique
 Two economists, Robert Lucas and Thomas Sargent, came
at a conclusion, known as Lucas Critique. Lucas critique
argues that it is wrong to predict the effects of a major policy
change solely on the basis of empirical relationship derived
using historical data.
 For example, in the Phillips curve: πt - πte = – α (ut - un), it
was assumed that wage setters would keep expecting
inflation in the future to be equal to last year inflation and that
the way wage setters formed their expectation would not
change due to change in policy. Given this, the only way to
decrease inflation would be to accept higher unemployment
for some time i.e. by accepting excess unemployment.
 Lucas said it was an unwarranted assumption. Why should
not wage setters take into account changes in policy while
forming their expectations.
 Lucas Critique
 If wage setters believed that central bank was committed to
lower inflation, they might well expect inflation to be lower in
the future than in the past. If they lowered their expectations
of inflation, then actual inflation would decline without the
need for excess unemployment or increase in unemployment
rate above natural rate.
 For example, assume inflation in the past was 14%. Now
suppose central bank commits to reduce inflation from 14%
to 4% in the future. Suppose wage setters believe in the
commitment by the central bank.
 So, wage setters will form their expectations according to the
commitment of the central bank i.e. they will expect inflation
to be 4% in the future. Hence, actual inflation will fall to 4%
without the need for excess unemployment or increase in
unemployment rate above natural rate or recession.
 Lucas Critique
 The sacrifice ratio – the amount of excess unemployment
needed to achieve a given disinflation, will be lower in this
case than suggested by the traditional approach.
 The essential ingredient of successful disinflation as argued
by Lucas and Sargent in the previous slide (or for the
adjustments explained in the previous slide to work
successfully), what is required is the credibility of central
bank – the belief by wage setters that the central bank is truly
committed to reduce the inflation.
 Only credibility would cause wage setters to change the ways
they formed their expectations. Further, credibility decreases
the unemployment cost of disinflation. So, to increase the
credibility, central bank should go for fast disinflation i.e. it
should follow disinflation policies very aggressively.
 Nominal Rigidities and Contracts
 Two economists, namely, Stanley Fischer and John Taylor,
took a stand opposite to Lucas and Sargent.
 They said there are two reasons why actual inflation will
not fall even if policy changes are credible and wage
setters have taken policy changes into account while
forming their expectations:
1) Nominal rigidity: Wages and prices are often set for some
time and are typically not readjusted when there is change in
policy.
2) Staggering of wage contract: Not all wage contracts are
signed (hence expire) at the same time. They are staggered
overtime. Therefore, nominal wages in all the wage contract
cannot be revised at the same time. This also implies rigidity
in nominal wage (and hence in price) at the economy level.
 Nominal Rigidities and Contracts
 Because
  of the above two reasons, when central bank
reduces nominal money supply growth (to decrease
inflation), it will not lead to proportional decrease in
inflation.
 As a result, real money supply () will fall, triggering
recession and increase in unemployment rate.
 Examples
Example 1: Chap-9 (Q.6):
Suppose that the Phillips curve is given by:
πt = πet – (ut – 0.05)
and expected inflation is given by: πet = πt-1
(a) What is the sacrifice ratio in this economy?
Suppose that unemployment is initially equal to the natural rate
and π = 12%. Central bank decides that 12% is too high and that,
starting in year t, it will maintain unemployment rate one
percentage point above the natural rate of unemployment until
inflation has decreased to 2%.
(b) Compute the rate of inflation in year t, t+1, t+2, t+3, t+4?
(c) For how many years must the central bank keep
unemployment rate above the natural rate of unemployment? Is
the implied sacrifice ratio consistent with your answer to part (a)?
 Examples
 
Example 1: Chap-9 (Q.6):
Now suppose that people know that the central bank wants to
lower inflation to 2% but they are not sure of the central bank’s
willingness to accept an unemployment rate above the natural
rate of unemployment. So, their expectation of inflation is a
weightage average of the target of 2% and last year’s inflation i.e.
= λ 2% + (1-λ)πt-1 , 0< λ <1
where λ is the weight they put on the central bank’s target of 2%.
(d) Let λ = 0.25. How long will it take before the inflation rate is
equal to 2%. What is the sacrifice ratio? Why is it different from
the answer in part (c).
(e) Suppose that after the policy has been in effect for one year,
people believe that the central bank is indeed committed to reduce
inflation to 2%. So, they now set their expectations according to:=
2%. From what year onward can the central…..
 Examples
Example 1: Chap-9 (Q.6):
……bank let the unemployment rate return to the natural rate?
What is the sacrifice ratio now?
(f) What advice would you give to a central bank that wants to
lower the rate of inflation by increasing the rate of unemployment
as little and for as short a time period as possible?
Answer:
(a) Phillips curve πt - πet = - (ut – 0.05) [πt - πet = – α (ut - un)]
We note that α is 1. So, sacrifice ratio = 1/α = 1
(b) πt-1 = 12% and the central bank is keeping unemployment
rate equal to 6% (1% above natural rate) until inflation falls to 2%.
Inflation in πt ? πt = 12% - (6% - 5%) = 11%
Inflation in πt+1 ? πt+1 = 11% - (6% - 5%) = 10%
 Examples
 
Inflation in πt+2 ? πt+2 = 10% - (6% - 5%) = 9%
Inflation in πt+3 ? πt+3 = 9% - (6% - 5%) = 8%
Inflation in πt+4 ? πt+4 = 8% - (6% - 5%) = 7%
(c) Central bank should keep unemployment rate above the
natural rate of unemployment for 10 years for inflation to fall to 2%.
Sacrifice ratio = total amount of excess unemployment (summed
over the years) / decrease in inflation (over the years)
= 10/10 = 1
This answer here is consistent with answer in part (a).
(d) With λ = 0.25, = (0.25*2%) + (1-0.75)*πt-1
= 0.5% + 0.75 πt-1
Substituting above in Phillips curve, Phillips curve can be
 Examples
Inflation in πt ? πt = [0.5% + (0.75*12%)] – (6% – 5%) = 8.5%
Inflation in πt+1 ? πt+1 = [0.5% + (0.75*8.5%)] – (6% – 5%) =
5.875%
Inflation in πt+2 ? πt+2 = [0.5% + (0.75*5.875%)] – (6% – 5%) =
3.906%
Inflation in πt+3 ? πt+3 = [0.5% + (0.75*3.906%)] – (6% – 5%) =
2.430%
Inflation in πt+4 ? πt+4 = [0.5% + (0.75*2.430%)] – (6% – 5%) =
1.322%
We need 4 to 5 years before inflation rate is equal to 2%.
Sacrifice ratio = 5/(12-1.322) = 0.468
We can note that this value of sacrifice ratio is different from the
one found in part (c). Reason for difference: because people are
 Examples
………. inflation rate into their expectations.
(e) The central bank can let the unemployment rate return to the
natural rate beginning at time t+1. The sacrifice ratio in this
scenario = total amount of excess unemployment (summed over
the years) / decrease in inflation (over the years)
= 1/10 = 0.1
(f) Central bank should take measures to enhance its credibility.
This will ensure central bank to achieve lower rate of inflation by
increasing the rate of unemployment as little and for as short a
time period as possible. [compare sacrifice ratio in part (c), (d)
and (e)]

Solve all the back questions of Chapter-

9 of Blanchard (Homework)
C.L.F.Attfield, D. Demery and
N.W. Duck (1991, 2nd edn. ),
p1– p28:
Expectations in
Macroeconomics
 Expectations in Macroeconomics
 Expectations
  are very important in Macroeconomics because
in Macroeconomics, many decisions such as consumption
decisions, investment decisions, wage-setting decisions etc.
are taken in advance wherein the expectations play a very
important roles.
 And the way people form their expectations affect the
outcome in the economy. For example, in the general Eqn
of final modified Phillips curve Eqn (8), if πte = πt , we get ut =
un whereas if πte πt , we get ut un.
There are two theories/types of expectations
formation:
1) Theory of adaptive expectations: Expectation of a variable
is adaptive if it is determined by past values of that variable.
More specifically, under adaptive expectations, a person will
 Expectations in Macroeconomics
……….of the difference between variable’s actual value last year
and its expected value last year.
 For example, if we are forming expectation about the inflation
for year ‘t’, then as per adaptive expectation theory, we must
have
πet - πet-1 = ρ (πt-1 - πet-1) (12)
where 0<ρ<1
 Eqn (12) implies that under adaptive expectations, each year
expectations are revised only by a fraction (ρ) of the
difference between actual value last year and expected value
last year, implying expectations will catch up with the actual
value in the long run.
 Major drawback of adaptive expectations theory: under
adaptive expectations theory, agents ignore current/new
information and only use past information to form their………
 Expectations in Macroeconomics
……….their expectations about a variable and therefore they are
likely to commit systematic error. [Note: error means there is a
difference between actual value and expected value]
 To be specific, under adaptive expectations theory, agents
follow the rule like Eqn (12) to form their expectations while
ignoring any new/current information which would help them
form better expectations. In this way, they commit
systematic error. (Think about the locus critique).
2) Theory of rational expectations: Expectations will be
rational if agents use all the available information (past and
current) relating to the process determining a variable while
forming their expectation of that variable.
 Since under rational expectations, agents use all the available
information to form their expectations, they do not commit
any systematic error.
 Expectations in Macroeconomics
 Error under rational expectation theory may arise only when
some information was unexpected/not accessible while
agents were forming their expectations. This type of error is
called random error.
 Problem with rational expectation theory: First problem is -
rational expectation requires use of all the available
information while forming expectations about a variable.
However, in reality, not all agents (because of different mental
ability) can be sensible enough to use all the available
information while forming their expectations.
 Second problem is – all the relevant information might not be
available and if available, it might not be accessible.
(Refer the relevant reading/book for details)
Steven Sheffrin, (1996, 2nd
edn.), p25– p40
Policy Ineffectiveness
Proposition (PIP)
Policy Ineffectiveness Proposition (PIP)
 Policy ineffectiveness proposition (PIP) is a proposition
associated with rational expectation theory and has been
given by Robert Lucas and Thomas Sargent.
 Policy ineffectiveness proposition states that, in the short run,
any predictable/anticipated change in monetary policy will
have no effect on output/employment/any other real variables
in the economy. Only the unpredictable/unanticipated
monetary policy change can affect output/employment/any
other real variables in the economy.
 Reason: As people are rational, so any predictable/anticipated
change in monetary policy will induce them to revise their
expectations immediately and hence there will be no impact
on output/employment/real variables. This implies only the
unpredictable/unanticipated monetary policy change can
affect the output/employment/real variables in the economy.
Policy Ineffectiveness Proposition (PIP)
 Suppose central bank follows expansionary monetary policy
i.e. increases the money supply. There can be two cases:
Case I: Policy change is In case of anticipated policy
change, wage setters know that
predictable/anticipated.
expansionary monetary policy (by
AS2 shifting the AD curve right) will
P
lead to Y>Yn, hence P > Pe.
AS1
Hence, they will immediately
P2 e3
revise their expectation of price
e2 level upward, causing AS curve to
shift upward immediately. As a
P1 e1 result, economy will immediately
move from e1 to e3 and hence
there will be no change in output.
AD2
Thus, predictable/anticipated
AD1
policy change has no effect on
Yn Y output/employment/any other real
variable in the economy in the
short run.
Policy Ineffectiveness Proposition (PIP)
 Suppose central bank follows expansionary monetary policy
i.e. increases the money supply. There can be two cases:
Case II: Policy change is
First, expansionary monetary
unpredictable/unanticipated
policy will lead to rightward shift
in AD curve. However, AS curve
P
will not shift up immediately
AS1 because here policy change is
unanticipated. As a result,
P2 e2 economy will move from e1 to
e2 and hence there will be
P1 e1 increase in economy’s output.
This shows that only the
unpredictable/unanticipated
AD2 policy change will have effect on
AD1 output/employment/any other
Yn Y2 Y real variable in the economy in
the short run.
 Rough 1
  Y = Φ
 logY = log Φ + log – log P
 dlogY/dt = (dlog Φ/dt) + (dlog/dt) - (dlogP/dt)
 Growth rate of Y = growth rate of Φ + growth rate of
- growth rate of P
 gyt = 0 + gmt – πt
 gyt = gmt – πt
Rough 2
 Real
  money supply = /P
 Log (real money supply) = log – log P
 Differentiate with respect to time ‘t’
 dlog(real money supply)/dt = (dlog/dt) - (dlogP/dt)
 Real money growth rate = growth rate of – growth rate of P
= gmt – πt