MONEY, MACROECONOMICS AND KEYNES phần 8 doc

Second, the aggregate demand function is assumed to be less sensitive to the

interest rate than is aggregate supply, and hence the AD schedule cuts the AS from

above. The relative insensitivity of aggregate demand to interest rate changes was,

of course, a staple of the ‘Old Keynesian’ literature (and was implicit in the

Keynesian Cross). However, in that context it was not combined with a palpable

degree of interest sensitivity of supply, as is done here. For simplicity, the AD

schedule in Fig. 15.2 is shown as completely interest insensitive. However, clearly

nothing would be changed by allowing some interest elasticity, as long as this is

less than on the supply side.

The third important point is that in Fig. 15.2 the (real) rate of interest is taken

to be a financial variable determined essentially by the monetary policy of the

central bank. It is determined outside the aggregate demand and supply nexus

itself, and in the diagram shows up as a horizontal line across the page, at a pre￾determined level, . The underlying monetary theory is therefore that of the Post

Keynesian ‘horizontalist’ school, in which the interest rate (including the real

rate) is effectively a policy instrument, and the money supply is endogenous. This

is contrasted with Barro’s version in Fig. 15.1, in which there in no theory of

money and the interest rate is taken literally to be a real (non-monetary) variable.

Victoria Chick (e.g. 1984, 1986, 1991, 1995) has written extensively on Post

Keynesian monetary theory, endogenous money, the theory of banking and

alternative views on interest rate determination. She has indeed described hori￾zontalism (e.g. that of Moore 1979)4 as an ‘extreme’ position (Chick 1986: 116),

while nonetheless making it clear that as a first approximation this is still a far

more reasonable assumption than the alternative (neoclassical) extreme. The main

objection to treating the interest rate as a purely policy-determined variable would

be the extent to which this neglects Keynes’s insights about liquidity preference

and the role of speculation in financial markets (Chick 1995: 31). The practical

implication would be that there can be occasions in which the monetary authori￾ties may not get their way in setting the interest rate.5 Keynes himself had argued

this way in the General Theory (Keynes 1936: 202–4), although elsewhere (even

as late as 1945) he had stated that ‘The monetary authorities can have any inter￾est rate they like … Historically … (they) … have always determined the rate at

their own sweet will …’ (as quoted by Moore 1988b: 128).

For our present purposes, however, the debate about the precise degree of con￾trol of interest rates by central bankers may perhaps be set on one side. There

would clearly be general agreement that the stance of monetary policy is at least

a major influence on the real interest rate. Moreover, from the perspective of the

principle of effective demand, the main point at issue is not exactly how the rate

is set, but rather that it is not taken to be determined by demand and supply in

barter capital markets, as in the neoclassical model, and is exogenous to the ‘real

economy’ in that sense.6

Note, however, that if we do proceed to take the interest rate as either an exoge￾nous or directly policy-determined variable, the issue immediately arises as to

how demand and supply could ever come into equilibrium. In neoclassical or new

r

J. SMITHIN

154

classical theory, interest rate adjustment itself is supposed to be the equilibrating

mechanism, but that is ruled out in any horizontalist approach. However, it can be

suggested here that for the SOE an obvious equilibrating mechanism does exist,

namely changes in the real exchange rate. Or, it would be more accurate to say, a

combination of real exchange rate changes and output adjustment. This issue is

taken up below.

4. A simple aggregate demand and supply model for

the small open economy

Consider the following simple aggregate demand and supply ‘curves’ (they are

actually linear) for the SOE:

, (1)

. (2)

Equation (1) represents the aggregate demand schedule. The demand for output

depends positively on autonomous spending, A(t), as in traditional Keynesian

models, and positively on Q(t), where Q(t) is the real exchange rate. The nominal

rate is defined as the domestic currency price of one unit of foreign exchange, so

an increase in Q(t) represents a real depreciation. The argument is therefore that

a real depreciation increases the demand for net exports and hence total aggregate

demand. As discussed, for the sake of argument there is assumed to be no inter￾est rate term in eqn (1), which is an extreme instance of the view that the demand

schedule is insensitive to interest rate changes.

Equation (2) is the aggregate supply schedule. This is assumed to be negatively

sloped, not positively sloped, for the reasons discussed above. Also, a real depre￾ciation is taken to have a negative impact on supply. This arises as the result of

real wage resistance on the part of the labour force, and/or because of an increase

in the real costs of imported raw materials.

We can rearrange eqn (1) to yield

. (3)

Then use (3) in (2) and set aggregate demand equal to aggregate supply:

(4)

Now solve for Y(t):

{(1) (2)/[(2) (2)]}A(t) {(2) (1)/[(2) (2)]}r(t) (5)

Y(t)
[(0) (2) (2) (0)]/[(2) (2)]

{[(1) (2)]/(2)}A(t).

Y(t)
(0) (1)r(t) [ (2)/(2)]Y(t) [(0) (2)]/(2)

Q(t)
Yd

(t)/(2) (0)/(2) [(1)/(2)]A(t)

Ys

(t)
(0) (1)r(t) (2)Q(t)

Yd

(t)
(0) (1)A(t) (2)Q(t)

AGGREGATE DEMAND

155