KEYNES AND THE CLASSICS: THE ISLM, ASAD BACKGROUND

1
KEYNES AND THE CLASSICS THE ISLM, AS/AD
BACKGROUND
1. Classical model closed economy version For
the moment, assume full employment output,
although there will be a story to justify this.
So
Goods market equilibrium
r is the real interest rate, and its equilibrium
value follows from the above equation, given full
employment output. C is consumption, T is taxes,
I is investment and G is government expenditure.
Signs above the equation are signs of partial
derivatives.
2
Loanable funds interpretation Y C S
T subtraction from the preceding equation and
rearrangement gives us
Private savings finance investment plus
government borrowing r balances the supply and
demand for LOANABLE FUNDS.
IS
r
r0
Y
3
Money and the Classical Dichotomy
How is money to be defined? The underlying idea
is of money as a transaction medium, but in
advanced economies there is a smooth spectrum of
liquidity, from cash through checkable deposits
and so on. Ignore this problem for now. Think in
terms of Money cash Bonds (infinite
maturity) perpetuities that pay a fixed sum per
year to the holder. By definition
where i is the interest rate, x the coupon
payment, and Q the price and the ? refers to the
infinite maturity. A key feature of perpetuities
is capital risk, since i and Q are inversely
proportional (Burda/Wyplosz 4th ed, chapter 5,
box 5.2)
4
The Classical model assumes
Money supply equals money demand equals price of
output multiplied by real output divided by the
income velocity of money. Therefore with output
at its full employment level, changes in money
supply cause equiproportionate changes in P (and,
in the background, money wages, W) The nominal
rate of interest, i, in the Classical model In
steady state equilibrium, this is determined
simply as the real rate of interest plus the
inflation rate. The inflation rate depends on the
growth rate of M, since from the above equation
and assuming v constant, we have
With real output growth exogenous, this equation
determines the inflation rate
5
ISLM representation
i
LM
i0
IS
Y
How do we know that IS will conveniently
intersect LM above the x-axis? We dont! The
fundamental idea of the Classical model is that r
regulates aggregate demand, so that - subject to
the above caveat - there cannot be such a thing
as deficient aggregate demand.
6
2. Keyness critique
Essential is Keyness critique of the Classical
Theory of Interest. In Keyness model, money
demand is assumed to depend on i, so
Why? One effect is via the transactions demand
if the opportunity cost of holding money is
lower, then agents will hold more of it,
other things equal. A second effect is via what
Keynes calls the speculative demand. Recall that
Keynes is simplifying by supposing that
individuals are allocating wealth between just
two assets money and bonds, which are
capital-risky perpetuities (Keynes is, in fact,
also thinking of equities as lumped in with
bonds). Then agents may hold idle speculative
cash balances because (a) they are averse to
risk, and/or (b) they believe that Q is going to
fall, i.e. that i is going to rise
7
For (b) Keynes needed a theory of bond price
expectations. He assumed that expectations were
regressive. What this means is that there is
a normal value of Q, say Q0 to which it is
assumed that Q will tend to revert. So that when
Q is above Q0, Q is expected to fall and when Q
is below Q0, Q is expected to rise. Translated
into interest rate terms, abnormally low
interest rates go with expectations of future
interest rate rises and therefore falls in Q.
This will raise money demand as agents get rid of
assets (bonds) on which they expect to make
capital losses. The end result of this reasoning
is a non-vertical LM curve. If the speculative
demand is very important, we may end up with a
floor on the rate of interest, as illustrated in
the figure.
LM
i
Y
8
As soon as we have a non-vertical LM curve,
either because of interest dependence of the
transactions demand for money or because of the
speculative demand for money, the Classical view
that aggregate demand does not matter ceases to
hold.
IS0
LM
i
IS1
A
B
C
Y
Shown in the figure is the effect of a fall in,
say, investment spending. IS shifts left and the
equilibrium moves from A to B. In the Classical
analysis, the LM curve would have been vertical
through point A and so the demand fall would have
taken the economy to C. In that case the interest
rate is performing its Classical role of
regulating aggregate demand.
9
Interim assessment Keyness basic insight, lost
in most standard textbooks, is that in asset
markets where stocks are large in proportion to
flows, the price expectations of holders of the
stocks are going to matter. He is surely right,
and this means, as his model illustrates, that we
cannot rely on the real rate of interest to act
as an automatic regulator of aggregate demand.
Whether Keyness particular model of price
expectations is always or ever correct is quite
another matter! Still, if we view the economy as
having a long run steady state (questionable)
with a natural long run equilibrium real rate
of interest, then it would not seem completely
irrational for agents to expect the rate of
interest to move towards that natural value. So
far we have said almost nothing about wage and
price adjustments. That is the next task.
10
3. Wages and prices the Classical model, the
Neoclassical Synthesis and some deflation
dynamics.
The Classical model We can amend the classical
model so as to make full employment a result
rather than an assumption of the model. If we
assume that P (and W) fall when we are below full
employment output and rise when we are above, and
if we consider only effects of this on the LM
curve, then it is clear that the LM curve will
come to rest sitting above full employment
output, as in the figure below. This mechanism is
known as the Keynes effect.
LM0
LM1
LM2
i
IS
11
The Neoclassical Synthesis What happens if we
take Keyness model and apply the reasoning from
the preceding slide? We end up with the figures
below. The key idea is that what is known as
balanced deflation, i.e. equiproportional falls
in W and P, will restore full employment after an
adverse demand shift. We get exactly the same
interest rate fall as in the Classical model, but
by an indirect route.
Construction of AD closed economy, Keynes
effect only
Neoclassical synthesis adverse IS shift (it
shifts AD left)
LM0
IS0
IS1
LM0
LM1
i
i
a
LM1
IS
b
Y
AS
P
a
AD0
b
AD
AD1
Y
12
4. Critique of the neoclassical synthesis Two
things to consider (a) whether the model has
special cases that prevent full employment, (b)
whether the model adequately analyses the effects
of balanced deflation. (a) IS curve intersects
the x-axis to the left of full employment
IS
The unlabelled lines are LM curves as P falls
the LM curve is shifting rightwards. Once
deflation has reduced i to zero, no further
increases in output are possible. This is a
modern liquidity trap, exemplified by the case
of Japan. The AD curve is kinked, as shown.
i
P
AD
13
In a liquidity trap, monetary policy becomes
useless (given the assumptions we are currently
working with). Keynes originated this idea,
although he envisaged an interest rate floor
above zero, arising from the speculative demand
for money. If agents are sufficiently convinced
that bond prices can only fall (the rate of
interest rate can only rise), then the LM curve
will be flat and monetary expansion (or price
deflation) will shift it rightwards not
downwards.
LM2
LM0
LM1
Open market purchases of bonds by the Central
Bank will simply result in agents willingly
holding more money and fewer bonds, with no
upwards pressure on bond prices.
i
IS
Y
14
b. Effects of balanced deflation There are three
other effects to consider, apart from the Keynes
effect. The discussion here is very brief, but we
will look at some details in class. Pigou
effect When P falls M/P rises and this raises
wealth and may therefore raise consumption
expenditure. Fisher redistribution effect When
P falls, real wealth is redistributed from
debtors to creditors. It cannot be assumed that
the effects on the spending of those two groups
will cancel out. Fisher real interest rate
effect Expectations of deflation, which may be
caused by deflation itself, raise the real
interest rate for any given nominal interest
rate. There is therefore an effect on the IS
curve.