Notesale: Turn your study into money

Document Preview

Extracts from the notes are below, to see the PDF you'll receive please use the links above

---

EC201 Intermediate Macroeconomics
EC201 Intermediate Macroeconomics

Lecture 10-12: Aggregate Demand and Aggregate
Supply
Lecture Outline:
- how to derive the AD curve;
- How to derive the AS curve;
- AD-AS model;
Essential reading:
Mankiw: Ch
...
3 and Ch
...
2, 12
...
14
...
14
Introduction
The IS-LM model is a model used to describe the equilibrium in an economic system
when prices are fixed and where aggregate demand components are the main
determinants of equilibrium output
...

That identity implies that what determines Y is the right hand side, something we
called aggregate planned expenditure
...
In practice, we have implicitly assumed
that is the demand side of the economy (demand for consumption, investments and
public expenditure) that determines final output
...
To
overcome this issue, another macro model, called the AD-AS model, has been
developed, where the equilibrium output is now determined by the intersection of
aggregate demand (AD) and aggregate supply (AS)
...
The AD-AS model overcomes also this problem
...
In particular, according to

Keynes the Say’s Law1 was not true, especially in the short-run
...
According
to this law, prosperity can be increased by stimulating production and not by
stimulating the demand side of the economy as Keynes suggested
...
In general, we may assume that the Say’s law may hold in
the long run, when prices are flexible and can adjust to ensure that aggregate demand
is always equal to aggregate supply
...

The model of aggregate demand and supply
The AD-AS model is a paradigm most mainstream economists and policymakers use
to think about economic fluctuations and policies to stabilize the economy
...

Aggregate Demand
First we need to define the concept of aggregate demand
...

The aggregate demand curve shows the relationship between the price level and the
quantity of output demanded
...

We have seen at least two theories that can be useful for this purpose
...

Derivation of the Aggregate Demand from the Quantity Theory of Money
The quantity equation is given by:

MV = PY

1)

MV
Y

2)

Or, written differently:
P=

For given value of M and V, equation 2) gives us a negative relationship between P
and Y (in particular equation 2) tells us that between P and Y there is a relationship

1

Jean Baptiste Say (1767-1832), a French economist
...
Equation 2) can be interpreted as an Aggregate
Demand function
...
Therefore real balances, defined by M/P must decrease
...
In order to buy
the same amount of stuff, velocity would have to increase
...
With lower real money balances (or,
equivalently, the same nominal balances but higher goods prices), people demand a
smaller quantity of goods and services
...

If we plot equation 2), the Aggregate Demand function will look like the following:

From equation 2) we can see how a change in the money supply affects the aggregate
demand
...

This is shown in the following figure:

P

AD2
AD1

Y
For a given value of P, an increase in M implies higher real money balances
...

Derivation of the Aggregate Demand from the IS-LM model
The aggregate demand function derived from the quantity theory is not very appealing
because it considers very few determinants of the aggregate demand (money supply
and velocity of money)
...
In particular fiscal policy is an important
determinant of aggregate demand and so are consumption and investment
expenditure
...
We need to find a relationship (negative) between aggregate prices
and real income from the IS-LM curves
...
In particular prices enter in
the LM curve through the real money balances
...

Note: if prices can change in the IS-LM model, the equality between real interest rate
and nominal interest rate should not hold in principle
...
However, for the moment we will do the analysis under the assumption
that inflation expectations are constant, meaning that nominal and real interest rate
move in the same direction (if r increases, i increases as well and so on)
...

To derive the aggregate demand: if we change the price level what happens to real
income according to the IS-LM model? Prices enter only the LM curve
...

Consider the linear LM curve derived in the previous lecture notes: r = k Y − 1 M
h

h P

Suppose an increase in P:
P ↑⇒

M
↓⇒ LM shifts to the left
P

⇒ r ↑⇒ I ↓⇒ Y ↓

LM(P2)

r

LM(P1)

r2
r1

IS

P

Y2

Y1

Y

P2
P1
AD
Y2

Y1

Y

In the top graph we have an LM curve for the initial price level P1
...
On the bottom figure we
draw the relationship between price level and equilibrium output implied by the
movement in the LM curve
...
In a microeconomic context, an increase in the price of a
good reduces the demand (if the good is not a Giffen good) since there is a
substitution effect and the purchasing power of the agent is reduced (income effect)
...
This effect of prices on the interest rate
and therefore on real income is called the “Interest Rate Effect” or “Keynesian

effect”
...
Note: the “Keynes effect”
does not work if we are in the Liquidity Trap that is the closest case to Keynes’s ideas
about a macroeconomic system
...

Aggregate Demand and Wealth Effect
However, there is another possible explanation for the negative relationship between
Income and P implied by the AD curve
...

Pigou effect: if money is part of households’ wealth, then a decrease in P makes the
households wealthier in real terms, this will increase consumption and therefore
aggregate income:
P ↓⇒

M
↑⇒ C ↑⇒ IS shifts to the right
P

Y ↑

The Keynesian and the Pigou effects (and in general the fact that the AD is negatively
sloped) imply that a deflation (a decrease in the general level of prices) can have
stabilising effects on output
...
Given a low demand, the aggregate price level should fall and this
will increase output, helping the economy to recover from the recession without any
external intervention
...
The idea that deflation can be good in terms of stabilising the
equilibrium income level has been challenged by some economists
...
In
particular, we assume that the demand for money depends on the nominal interest rate
while the investment depends on the real interest rate
...

1) The destabilizing effects of expected deflation:
The idea is: suppose you start with no inflation expectations, meaning π e = 0 , and so
i = r
...
3

2

There is a third explanation for the negative relationship between Y and P implied by the AD
...

3
Suppose you start with i = r = 5%
...
Now the real interest rate becomes: r = i − π e = 5% − (−2%) = 7%
...

2) The destabilizing effects of unexpected deflation: debt-deflation theory
Suppose a decrease in P that is unexpected (this means that your inflation
expectations can be wrong)
...
Suppose borrowers lose (part of their
real income is transferred to lenders because of unexpected inflation) compared to
lenders
...

Fiscal Policy and Aggregate Demand:
An increase in G or a decrease in T will shift the AD curve to the right: this implies
that at each value of P, Y must be larger
...
Therefore Y increases
...

Obviously a decrease in G or an increase in T will shift the aggregate demand to the
left
...
Using the IS-LM
model: an increase in M will shift LM to the right
...
In terms of
the aggregate demand, given the price level, this increase in M will shift the AD curve
to the right
...


Classical and Keynesian AD-AS Models
We start by describing the AD-AS model in two particular (extreme) cases
...
Later we
discuss a more general AD-AS model which is a mix of those two cases and allow for
prices to be sticky (so not completely fixed) in the short-run
...


The Classical Case
Similarly to the concept of aggregate demand, an aggregate supply function is a
relation between aggregate prices and aggregate output produced
...

While classical economists argue that this true in the short and in the long run,
nowadays economists think that this idea that output is only determined by the
technology and the factors of production is true ONLY in the long run
...
We
normally denote the natural level of output (or the full-employment level of output) as

Y


...
Later in the module we will discuss
economic growth theory that tries to explain how, over a long period time, the full
employment level of output does increase (by a long period of time we mean, 20, 50
or even 100 years and more)
...
Independently of the prices, the economy uses
all its resources (inputs of production) efficiently all the time and so it always
produces at the natural level
...
If money supply increases, AD shifts up (it

increases), output does not change while prices increase
...
As I said now economists believe that the classical case is not the right one
to describe the short-run but it is a good representation of the economy in the long
run
...

3) Assumes complete price flexibility
...


The Keynesian Case
In the short run prices are completely fixed
...
However as
we reach the full employment, according to Keynes, any further increase in demand
will be reflected in an increase in the prices only
...
However
prices cannot decrease below P*
...
Notice also that in the

short-run we may be well below the full employment as Keynes argued in the General
Theory
...

For a Keynesian economists we have:
1) Prices are fixed in the short-run
...


The General AD-AS model

Price Stickiness in the short run
The previous two examples of the AD-AS model were quite extreme in the
assumptions behind them
...

However, the assumption that prices are completely fixed in the short-run may not be
very realistic
...
Inflation from
one month to another may change and that means that prices do change
...
If there is a change in the economy, say a
decrease in aggregate demand (so aggregate demand shifts to the left), the prices of
some goods may decrease quite soon, but the prices of some other goods don’t
...

To see this: suppose there are only two goods in the economy, A and B
...
The aggregate price level is P =

1
(PA + PB ) = 15 that is
2

the average of the two prices (this is a simple example and we can assume that each
good has the same weight in total output)
...
If prices
are fully flexible (so we are like in the long run) both prices will decrease by 50%
(think about the quantity theory of money) so we should have Pnew = 7
...

However suppose that while the price of good A can decrease by 50% (for any
reason) the price of good B cannot (for any reason no matter what)
...
The aggregate price level is
therefore P = 12
...
So the aggregate price has indeed decreased from 15 to 12
...
5)
...

Over time also the price of good B will decrease and in the long run the aggregate
price level will be 7
...

If the aggregate price level can change in the short run (it is sticky but it does change)
then we can have a short-run aggregate supply that is upward sloping
...

In the long run, when prices are fully flexible, the aggregate supply will be vertical at
the natural level of output as in the classical case
...
There is clearly a relationship between the outcomes in the labour market
(equilibrium employment level and wage) and the amount of real output produced in
an economy
...
It is obvious that if the level of
employment increases, also the final output will increase
...
The Sticky Wage model
assumes that in the short-run NOMINAL wages are sticky
...
Even in industries not covered by

formal contracts, implicit agreements between workers and firms may limit wage
changes, and so on
...
In particular, even if the bargaining is over the
nominal wage, what really matters for the firm and the worker is the REAL wage
...

We denote with W the nominal wage
...

Therefore, firms and workers bargain the following nominal wage:

W = ω × Pe
where

P e is the expected price level and

1)

ω

is the real wage target
...
So we need to include their
expectations about the price level in the equation
...
This implies that output equals its natural rate
(aka full-employment output)
...
If the actual price (P) is greater (lower)

e

than the expected price ( P ), then the negotiated real wage is lower (greater) than
the target
...

For example, if the firms are competitive, we know that the labour demand is given
by the condition:
MPL =

W
P

where MPL is the marginal productivity of labour
...

Now we have a link between prices and final output that goes through the amount of
labour hired by firms
...

Since the target is the real wage that clears the labour market we have that:

P e = P : Unemployment and output are at their natural rates;

a)

when

b)

when P > P : negotiated real wage exceeds its target, so firms hire fewer
e

workers and output falls below its natural rate;
c)

when

Pe < P :

negotiated real wage is less than its target, so firms hire

more workers and output rises above its natural rate;
Points a), b) and c) imply that a short-run aggregate supply can be written as:

Y = Y + α (P − Pe )
where Y is the real output,

4)

Y is the natural level of output and

α is a positive

constant that measures the responsiveness of real output to a price change
...


From equation 4) we can see the two special cases we have considered in previous
lectures:
1) A vertical aggregate supply: when

P = Pe

2) A horizontal aggregate supply: when

α →∞

Since we normally write the aggregate supply in the P, Y − space, we can rewrite
equation 4) in the equivalent way given by:

P = Pe +

1

α

(Y − Y )

5)

An important implication of this model is that the real wage should be countercyclical
...


In recessions, when P typically falls, real wage should rise
...
Indeed, it seems that real wages
are pro-cyclical
...

From the sticky-wage model it is clear that the short-run AS can be upward sloping
only if the actual price level is different from the expected price level
...

Two main models to explain expectations formation:
a) Adaptive expectations: an idea developed by Milton Friedman
...
Agents use a mechanic rule to process the
information they got about past observations
...

−
−

t −1 t

Expectation at time
t-1 of what the
price level will be
at time t

Last period’s expectation
error
Previous period’s
expectation

If a = 0 agents never adjust their expectations even if the new price level is now
available
...

Using this rule agents update their expectations by a fraction a in every period
...

The possibility that agents can make systematic errors is not particularly appealing
...


Agents base their expectation of economic variables on all available information
...


Pt e = E[ Pt Φ t −1 ]
t −1
Rational
expectation

Information
set

Where E is the expectation operator
...
Expectations become an
endogenous variable and depend on all the determining (exogenous) variables (such a
government spending and monetary policy) of the model
...

Errors can be made because of imperfect information and confusion about random
shocks
...
Each firm produces a single good
...
This implies that we are assuming some sort of inefficiency in
the market since we are not in a competitive market framework
...
Reasons for price stickiness:
- long-term contracts between firms and customers;
- menu costs;
- firms not wishing to annoy customers with frequent price changes;
The desired price level that a firm would like to set depends on two main variables:
a)

The overall level of prices P
...
So the higher the overall price level and higher is the
price a firm would like to set for its product
...
If a firm charges a higher
price than its rivals it does not lose all its demand);
b)

The level of real output Y, or aggregate income
...
Higher is the demand and
higher will be the price that particular firm would like to set
...

Now suppose there are two types of firms
...
So
for those firms the price can adjust to changes in P or in Y
...
In particular we assume that firms with sticky prices expect the
aggregate income to be at the natural level
...

Denote with s the proportion of firms with sticky prices and with (1-s) the proportion
of firms with flexible prices
...
In particular, assume that:

[

P = sP e + (1 − s ) P + β (Y − Y )

]

8)

Subtract (1−s )P from both sides of 8):

[

sP = sPe + (1 − s) β (Y − Y )

]

9)

Divide both sides of 9) by s :

P = Pe +

(1 − s ) β
(Y − Y )
s

10)

Equation 10) is exactly the same as equation 5), with:
1

α

=

(1 − s ) β
s

So even with this model we obtain the same upward sloping aggregate supply
...
Firms with flexible
prices set high prices
...
The ides is that: in the short-run
firms with flexible price will have a vertical supply curve, firms with sticky prices
will have a horizontal supply curve
...
In contrast to the sticky-wage model, the sticky-price model implies
a pro-cyclical real wage
...
Then,
Firms see a fall in demand for their products
...

This reduction will shift to the left the labour demand in the economy and
this will cause the real wage to fall
...
The Ls is the labour supply and the Ld is the labour demand
...

w
Ls

∆w
Ld

∆L

L

3) The Lucas’ imperfect information model4
The basic idea of this model is that there are many firms and each firm knows its own
price (the price of the good it produces) but it does not know (or knows it imperfectly)
the overall price level in the economy
...
If you decide to study a master in economics you are going to see it quite in detail
...


You may think at each firm located in a particular island
...
The
Lucas’ imperfect information model is also known as the Lucas’ island model
...

- Each supplier produces one good and consumes many goods (here we assume that
the supplier is a producer and also a consumer
...

- Each supplier knows the nominal price of the good she produces, but does not know
the overall price level
...
The firm
does not know whether this increase in its demand reflects the fact that consumers
started liking more the product it sells or if there was a general increase in the
aggregate demand
...

In particular assume that the supply of firm z (or island z) is given by:
y ( z ) = y ( z ) + γ ( p ( z ) − P)

11)

Where y(z) means that amount produced by firm z, p(z) is the price charged by firm z
and P is the aggregate price level (the average of all prices charged by all firms in the
economy, so p(z) is part of P as well)
...
The parameter γ is just a
positive constant
...
If there is an increase in all prices, p(z) and P increase by
the same amount5 and so nothing changes and production remains unchanged
...

Therefore we can write equation 11) like:
y( z) = y ( z ) + γ ( p( z) − P e )

12)

where Pe is the expectation of the aggregate price level
...
Is that
something specific to firm z so that p(z) increases compared to P or it is due to a

5

Suppose only two prices, p1 = 10 and p2 =10
...
Now
suppose that p1 increases to 20 but p2 remains at 10
...

The aggregate price level has increased by 5 but p1 has increased by more than that (by 10)
...
The
aggregate price has increased by the same amount as p1 and p2
...

This implies that the firm tends always to overestimate or underestimate the general
price level in the short-run (in the long run by definition expectations are in average
correct)
...
Suppose we start from a situation
where P e = P = p( z ) = 10
...
The firm
can see only the increase in p(z ) from 10 to 20
...
5 the increase in due to an increase in P and
with 1-0
...
This lead to an expected
value of 0
...
5 × 0 = 5
...

If we aggregate all the firms in our economy (= we sum equation 12) for all firms) we
get an aggregate supply that is:

Y = Y + α (P − Pe )
That is exactly the same aggregate supply function we have derived previously
...
So all the models we
have described give origin to the same Lucas’ supply function
...
It is however normally a good thing to assume that the
economy starts at the full employment equilibrium
...

Suppose an increase in aggregate demand, for example an increase in government
expenditure or an increase in money supply
...
At that equilibrium all the agents
expect the right price, so P = Pe1 and the agents expect that price to persist over time
...
The higher demand puts pressure on
prices to increase and P increases above the level people had expected
...
Given
2
2
1
the short run aggregate supply (SRAS), the level of output increases beyond the
natural level, for example at Y2
...
An increase in P e will
shift the SRAS to the left until the long-run equilibrium will be reached
...
The vertical

e
intercept of that equation is P e − 1 Y
...

α

making the aggregate supply to shift to the left
...
So in the long run we have that output
is back to the natural level but prices are permanently higher than in the original
equilibrium
...
However, output increases
because the increase in the AD creates a difference between the expected price level
and the actual price level
...
However,
over time agents revise their expectations and so the aggregate supply moves to the
left and at the end the economy moves back to the long run equilibrium
...


Shocks in the AD-AS model
Shocks are unexpected temporary events (temporary in the sense that their effects last
only in the short-run given that in the long run, no matter what, we will be at the full
employment since prices are flexible) that can affect the aggregate demand (we call
them demand shocks) or the aggregate supply (we call them supply shocks)
...
For
example, there is always a positive probability that an earthquake will happen in
Japan in a given period of time, but we cannot predict with certainty when it is going
to happen
...
We normally called them exogenous
shocks
...
Exogenous shocks are used to explain economic fluctuations, booms and
recessions (business cycles or short-run fluctuations)
...
These are shocks that
that affect the IS or the LM curve (since the AD is derived from the IS-LM model) as
the ones we have seen in the previous lecture note
...
A
positive demand shock has the effect of shifting AD to the right
...
An unexpected increase in consumers
confidence in the economy
...
Etc
...

e) Negative demand shock: any shock that decreases aggregate demand
...

Example: a sudden decline (unexpected) in house prices (since houses
are part of people wealth, a decline in their values result in a decline in
people wealth and this affect people consumption)
...
etc
...
There are two
types:
a) Positive supply shock: any shock that decreases the costs of production of
the firms in the short run
...
A positive supply shock has the effect of
shifting SRAS to the right
...
Etc
...


b) Negative supply shock: any shock that increases the costs of production of
the firms in the short run
...
Example: the oil
shock in the 70s
...
etc
...
In the case
of the shock we just assume that the increase in aggregate demand is unexpected, but
the reasoning and the graph of the AD-AS model look exactly the same
...
Output decrease below the natural level and the low
demand puts pressure on prices to decrease from PA to PB
...
We are in a recession
...
Since we are in a recession,
low demand puts downward pressure on the prices to decrease
...
The SRAS shifts down
...
Output is back to the natural level but prices are now
permanently lower than at the original equilibrium we start from (point A)
...
First output decreases
below the natural level so we have a recession
...


The economy over time self-corrects after a shock so that the effects of a shock are
just temporary (short run effects)
...


Consider a negative supply shock
...


Negative supply shock
LRAS

P

SRAS
PB

SRAS1

B

PA

A
AD

PC
YB

Y

AD1

Y

Short run: The economy moves from A to B
...
A situation of low output and high prices is
known as Stagflation
...

Long run: since we are in a recession, low demand puts pressure on prices to
decrease
...
The SRAS shifts back to the original level and we move
back to the full employment level of output
...

Again, this shock has created a short-run fluctuation in output (first decreases and
then it increases)
...

Does it mean that there is no need to use economic policies (fiscal and monetary) in
the economy? Since we are going to end up at the full employment equilibrium why
should we care to affect the economy using policies?

The problem is: how long it will take for the economy that is hit by a shock to move
back to the full employment level of output
...

Suppose a negative demand or negative supply shock
...
We don’t really know how long the recession is going to last
...

The same is true for a positive shock
...
Increasing prices may be something we want to avoid and so
there is scope to intervene in the economy even during a boom to reduce the duration
and the size of the economic boom
...
Suppose that after the shock has hit the economy
the Central Bank uses monetary policy to offset the negative effects on output of the
negative shock
...
Graphically:
Negative demand shock and policy
LRAS

P

SRAS

PA
PB

A
B

AD

Y

AD1

Y

In the short run the economy moves from A to B since aggregate demand shifts down
from AD to AD1
...
We are in a recession
...
The aggregate demand shifts back to AD and the economy moves back to A
...
Prices increase and move
back to PA
...


Instead of increasing money supply we could have increased government expenditure
or decreased taxes (or both) and we will have a similar effect (the AD will shift back)
...


How do central banks form inflation expectations?
We saw two main theories that explain how expectations can be formed (adaptive vs
rational)
...
The Bank of England
for example uses a survey, the so called GfK NOP Inflation Attitudes Survey, in which
they ask people (around 2000 individuals) their inflation expectations
...


Another possible way to measure inflation expectations is look at the difference
between nominal and real yields on gilts (an index-linked gilt is a bond in which the
payments are adjusted in line with the inflation rate, in UK it is used the Retail Price
Index (RPI)) or by looking at inflation swaps (a derivative that allows you to hedge
against future price rises)
...




---