ECONOMIA E POLITICA MONETARIA (Prof. Elisabetta De Antoni) Aa
2011-2012
INTRODUCTORY
LECTUREPREMISE....................................................................................................................................................................................1
I) ENGLISH
VERSION........................................................................................................
3 B. SNOWDON, H. VANE, P.
WYNARCZYK:.....................................................................3
A MODERN GUIDE TO
MACROECONOMICS:.................................................................3
AN INTRODUCTION TO COMPETING SCHOOLS OF
THOUGHT...................................3 EDWARD ELGAR,
1994....................................................................................................3
II) ITALIAN
VERSION........................................................................................................3
ETAS, 1998
.......................................................................................................................
31. THE OLD' CLASSICAL (PRE-KEYNESIAN)
MODEL...............................................................................................4
The Existence of Full Employment
Equilibrium.................................................................................................................35
2. The Keynesian
revolution...................................................................................................................................................42
3. The demand for
money.......................................................................................................................................................70
Krugmans description of the
crisis.......................................................................................................................................88
Neoclassical Synthesis
(1937-1965)........................................................................................................................................97
Monetarism.............................................................................................................................................................................138
the balance of
payments........................................................................................................................................................161
new classical (macro)
economics..........................................................................................................................................181
PREMISEIN THE LAST THREE DECADES: the financial system has grown
faster than the real economy (transactions, incomes distributed.)
The leader industrialized country (US) has accumulated
unprecedented levels of indebtedness (households, government, whole
country).
Thanks to liberalisation, financial techniques and innovations
have become more and more unscrupulous and opaque. THE RESULT was
an overall increase in financial fragility: the symptoms of a
"disaster foretold" were before everybody's eyes. This instability
clashed with the dogmas of the mainstream: The free market is
stable. Financial markets are efficient. There was NO REASON TO
WORRY!!!!!!! We had even entered a new era - the Great Moderationin
which the trade cycle and inflation had been definitely bridled.
THE CRISIS The world economy has been recently swept by one of the
biggest financial crises within the memory of man with devastating
effects on the real economy. This obliges us to rethink on the
evolution of macroeconomic theory its future perspectives. THE
COURSE The course will analyze the role of money and finance in the
economy according to the main macroeconomic schools of thought.
MAIN QUESTIONS: Why money (a simple piece of paper) is that
important? Why finance (a simple way of transferring funds from
savers to investors) is that important? Are money and finance a
source of stability/instability? The recent financial turmoil
confirms the importance of these questions. From the concrete point
of view, the course will focus on both the sides of the money
market, analyzing the theories of the demand for money and more
generally of portfolio allocation the theories of the supply of
money (the role of the central bank, of the financial system, of
monetary policy). REQUIREMENTS English courses are usually choosen
by foreign students.
In the past, many Erasmus students coming from other Faculties
(Law, Political Sciences, Engineering.). To be accessible to
everybody, the course requires: basic macroeconomics basic
mathematics. TEXTBOOK (CHAPTERS 1-7) The textbook is available
(also in the University Library) both in English and in Italian. I)
ENGLISH VERSION B. SNOWDON, H. VANE, P. WYNARCZYK: A Modern Guide
to Macroeconomics: An Introduction to Competing Schools of Thought
Edward Elgar, 1994. II) ITALIAN VERSION B. SNOWDON, H. VANE, P.
WYNARCZYK: Guida alla macroeconomia. Scuole di pensiero a confronto
Etas, 1998 FURTHER MATERIAL Lectures notes/further references will
be made available/specified on Comunit on line at the end of each
week. EXAMS The course implies written exams. Questions will be in
English. Answers can be either in English or in Italian. The
student will be able to choose between two kinds of examination.
THE SIMPLIFIED EXAMINATION will be mainly based on the textbook
will carry a mark of up to a maximum of 25/30. THE EXTENDED
EXAMINATION will be based on the textbook plus lecture notes will
carry a mark of up to a maximum of 30/30 cum laude. NB Through
ESSE3 (if not possible, by e-mail or during the lectures) students
have to inform the teacher at least a week in advance if they
intend to sit for the exam and for which kind of exam : the number
of questionnaires that have to be prepared for the exam depends on
the number of students involved. QUESTIONNAIRES will contain about
ten questions concerning the different parts of the course/chapters
of the book. questions will be of the following kind: What happens
to the variable X if money supply increases according to this
author/school of thought? Why? Explain and comment. .(5/10 rows for
the answer)
What are the effects of a given fiscal/monetary measure
according to this and that author/school of thought? Compare the
two cases and comment. .(5/10 rows for the answer)LECTURES
Monday 5-7 pm/Room 1A; Wednesday 3-4 pm/Room 2B
1. THE OLD' CLASSICAL (PRE-KEYNESIAN) MODELEnglish not revised
Keynes regarded as classical the economists who preceded him, i.e.:
-the Classics (Smith, Ricardo)-the Neoclassics (Marshall, Pigou).
From a macroeconomic point of view, these pre-Keynesian economists
generally placed complete faith in the market mechanism.
As we shall see, this faith has always dominated the field, up
to nowadays. The recent financial turmoil, however, casts doubts on
the effectiveness of market mechanisms. (Hopefully, textbook will
be rewritten!!!) Main Presuppositions of the OC Model 1. Agents are
rational; they: -have perfect knowledge; -are able to find their
possibility frontier (the set of available options); -choose the
optimal option, the one which maximizes their target 2. Markets are
perfectly competitive; -there are many small agents with no market
power -i. e. unable to affect market prices and quantities; -for
competitive agents individually considered, prices are
given.3.Prices are perfectly flexible; they instantaneously clear
all the markets.
Main Markets 1.Labor market 2.Goods market 3.Money market Every
market implies 1. A demand curve D 2. A supply curve S 3. An
adjustment mechanism towards equilibrium E
P
D E
S
Q
OCs LABOUR MARKET We shall start with the OC demand for labour.
THE OCS DEMAND CURVE FOR LABOUR This curve derives from the
production function. The production function gives the maximum
amount of output (Y) that can be produced out of any given amount
of factor inputs (K,L). Y=A Y(K,L) where: Y is the output level. A
is the total factor productivity: --it mirrors technology and input
organization --by assumption, in the short run is given and equal
to 1. K is the capital stock. --by assumption, in the short run is
given. L is the employment level. As we have jus seen, in the
short-run A e K are given (underlined) by assumption. The short-run
production function thus becomes a relationship between output Y
and employment L: if firms want to increase Y, they have to
increase L. Thus, the short-run production function can then be
represented as follows:
Y
Y=Y(K,L)
Y1 Y0
L0
L1
L
The first derivative of the production function tells us by how
much output Y rises if there is a unitary increase in L.
Y/LIt thus measures the productivity of the additional/marginal
worker. It is consequently named marginal product of labour.
Y/L=MPL Analogous considerations obviously hold for the marginal
product of capital. Y/K=MPK Let us now analyze the properties of
MPL and MPK. The basic assumptions underlying the production
function are the following: -its first derivatives are positive
Y/L=MPL>0 Y/K=MPK>0 -its second derivatives are negative
MPL/L0 -on the right of Ld0 are not profitable: MPLMPK and thus 0
I=I(r) with I/rG-T+I This excess of savings implies an excess
demand for bonds EDB. S-[G-T+I]=EDB The bond price pb will increase
and consequently the interest rate r will fall.pb r By aligning the
demand and the supply of financial flows, the interest rate clears:
-the goods market, -financial markets The aforementioned
coincidence between goods and financial markets equilibrium focuses
exclusively on financial flows. The underlying assumption is that
financial stocks are in equilibrium: people are perfectly happy
about the stock of money and bonds inherited from the past. Stocks
equilibrium does not interfere with flow equilibrium. This
assumption, as we shall see, is all but granted. A DIGRESSION ABOUT
FISCAL POLICY
What happens if the government increases government expenditure
G (and government deficit G-T)?The increase in G-T implies higher
issues of government bonds.
The rise in the supply of new bonds G-T + I(r) causes fall in
the bond price and an increase in the rate of interest. From
equilibrium point 0 we move to equilibrium point 1 below.
G-T + I(r)=Bs
S(r)= Bd
r1
1 0
r0
S0=G0-T0+I0
S1=G1-T0+I1
The following figure shows the fiscal expansion (the same story)
in terms of Ys=Yd. The increase in G raises aggregate demand,
moving the Yd curve to the right. The issue of new government bonds
raises the interest rate. This reduces private demand (C&I)
along the new Yd curve.In the end Yd goes back to its initial full
employment level Ys=Yfe,
the only one compatible with the aggregate supply Yfe and thus
with goods market equilibrium.Ys r Yd1=C(r)+I(r)+G1 1 Y
0=C(r)+I(r)+G0 0d
Yd= C(r) + I(r) G=Yd
Yfe
Y
Let us compare the old (0) and the new (1) equilibrium.
Equilibrium Ys comes from the labour market: it remains at its full
employment level Yfe. Equilibrium Yd=C+I+G has thus to remain too
at its full employment level Yfe. The new government expenditure G
has to crowd out an equal amount of (C+I).
It does this trough the rise in the interest rate provoked by
the issue of new government bonds.
Yfe = C(r)+I(r)+GIf output is at its full employment (its
maximum) level government expenditure inevitably crowds out an
equivalent amount of private expenditure. Crowding out is:
-real (the constraint is full employment output) -total or
complete (each euro of public expenditure crowds out a euro of
private expenditure)MONEY MARKET
The fulcrum of the OCs money market is the famous Quantity
Theory of Money. There are two versions of it. i) Fisher's version
F ISHER ' SIDENTITY OF EXCHANGES
In the real world, goods are exchanged against money.
---Money is a medium of exchange. ---Money is an input in
transactions. However, we can say more on this issue! Individual
goods are exchanged against an equivalent amount of money. The
value of individual goods is equal to the amount of money paid for
them.
cigarettes
1
P=4
Let us write this equivalence at the aggregate level. The value
of the goods exchanged in the economic system is approximately
equal to nominal income PY. The value of the money exchanged does
NOT coincide with the existing quantity of money M. In any given
period, every coin/banknote flows from hand to hand. In the example
below, -I buy 4 euros of cigarettes -the tobacconist buys 4 euros
of bread.
4
4
I
Tobacconist
Baker
If money does 2 rounds (finances 2 transactions) in a period, 4
euros of money finance 8 euros of transactions. M times 2 = value
of transactions PY Generalizing, let us assume that money does V
(rather then two) roundsM times V=value of transactions PY We thus
come to the Fisher's identity of exchange: MV=PY
V is the of circulation of money, equal to the value of
transactions per unit of money (in our previous example=2) V=PY/M
THE QUANTITY THEORY OF MONEY
If we introduce the following assumptions: i) M is exogenously
determined by the central Bank ii) V is a given institutional
constant iii) Y is given at its full employment levelwe get the
quantity theory of money When V and Y are given, an exogenous
increase in M turns into an increase in P. MV=PY Let us divide both
the terms by P, transforming the previous equation in real terms: M
= (1) Y
P
V
The right hand side Y/V is given by the real sector. In
equilibrium, also the left hand side has thus to be equal to Y/V.
If this is true, increases in M cause equi-proportional increases
in P, leaving the real quantity of money M/P unchanged at Y/V. If M
doubles, the price level P too doubles. More generally: %M=%P
Remember that labour market equilibrium gives W/Pfe. If money
supply and the price level P doubles, the money wage W also
doubles, leaving the real wage at its equilibrium level W/Pfe.
%P=%W THE NEUTRALITY OF MONEY The important implications of the
quantity theory of money are the following: i) any given percentage
variation in money supply M implies an equi-proportional variation
in nominal variables (P and W). %M=%P=%W ii) the real variables
(W/P, Lfe, Yfe, C, I, r, M/P..) are all invariant to the quantity
of money M. Money supply M does not affect them. iii) This means
that money is -neutral, it only affects the nominal scale of the
economy leaving real variables unchanged -a veil, that covers the
real variables without affecting them. iv) Inflation is a monetary
phenomenon. If prices rise too much, this means that money supply
rises too much. The responsibility for inflation pertains to the
central bank! v) The central bank controls M but not M/P, which in
equilibrium is equal to Y/V!
ii) The Cambridge version (Marshall and Pigou)Money is not an
input in transactions Money is a particular good necessary to
finance the purchases of all the other goods. As every other good,
money will then have its own market, with
-a demand curve -a supply curve -an equilibrium condition
THE DEMAND CURVE FOR MONEY Let us start from Fisher's equation
of exchange: MV=PYBy isolating M we get M=1 PY V If we put l/V=k
the result is M=k PY This equation can be interpreted as a demand
for money function:
Md=k PY Money is necessary to finance transactions. The
transactions demand for money Md is then a multiple k of the value
of transactions which take place in the economic system PY.THE
SUPPLY OF MONEY The assumption is that the supply of money is
exogenously given Ms=M
MONEY MARKET EQUILIBRIUMThe equilibrium condition of the money
market is: M=k PY
If we adopt the following usual assumptions: i) M is exogenously
determined by the central Banck ii) K=1/V is a given institutional
constant iii) Y is given at its full employment levelonly P can
clear the money market. MV=PY This leads to the Quantity Theory of
Money. An increase in money supply M turns into an
equi-proportional increase in nominal expenditure PY. Since real
output Y is given, the result is an equi-proportional increase in
P. M =k PY The same story can be told below in real terms The real
demand for money (k Y) is given by the real sector. Equilibrium
real supply M/P has also to be equal to k Y. Variations in the
nominal quantity of money M only determine equi-proportional
variations in the price level P. M =k Y P
Money is again -neutral, it only affects the nominal scale of
the economy leaving real variables unchanged -a veil, that covers
the real variables without affecting them. The Fisher's and the
Cambridge versions of the quantity theory are perfectly equivalent.
The difference is that money -in the first case is only an input in
transactions; -in the second case is a good, with its own
market.
Graphical representation of the money marketThe equilibrium
condition of the money market in real terms is: M/P=k Y where:
--the left hand side is the supply of real money balances M/P.
--the right hand side is the demand for real money balances kY
which grows with real transactions and thus with real income Y. The
money market equilibrium condition is shown by a straight line in
the Y,M/P space.M/P M/P=kY
Y
As real income Y grows, the transactions real demand for money
M/Pd=kY also grows. In equilibrium the real money supply M/Ps=M/P
has to grow by the same amount. Y M/Pd=kY M/P As known, P is the
variable that clears the money market. The increase in M/P is thus
due to the fall in the price level P. If we want to buy more goods
Y with a given money supply M, the price level P has to fall. Y
M/Pd=kY M/PP Let us now connect the money market with the rest of
the economic system.
According to the Old Classics, real income Y --comes from the
labour market --is given at its full employment level Yfe The
demand for real money balances is then kYfe. Given M, the price
level will have to be such that (M/P)e = kYfe
M/P
M/P=kY
M/Pe
Yfe
Y
MONETARY POLICY What happens if the central bank increases the
nominal supply of money M? Given the full employment level of
income Yfe, from initial equilibrium point 0 we move to
disequilibrium point 1. At the full employment level of income,
money supply now exceeds the demand. M > kYfe P
M/P M M/Pe
1 M/P=kY
0
Yfe
Y
The excess supply of money ESM=M/P-kYfe will be used to purchase
goods. Since the supply of goods Yfe is given at its full
employment level, the higher demand for goods will only raise
prices. The increase in P will decrease M/P, reabsorbing the excess
supply of money. From point 1 we shall go back to point 0.
M/P M
1 P M/P=kY
M/Pe
0
Yfe
Y
The equilibrium value of M/P -is equal to kYfe -is given by the
real sector. Variations in the money supply M only determine
equi-proportional variations in the nominal variables (P and W,
given W/Pfe) %M=%P=%W According to the quantity theory of money:
-money is neutral, -money is a veil. Inflation is a monetary
phenomenon. The increase in the price level P reflects the increase
in money supply M. The responsible for inflation is the central
bank. The Central Bank controls M but not M/P It is unable to
affect the real sector. It can only affect the price level.OLD
CLASSICAL MODEL: AN OVERALL SUMMARY
The following figure starts from the bottom and summarizes what
we have said up to now. Make sure that you know what variable
clears each market and why. Notice that the figure below only
concerns real variables; by assumption they are independent of
nominal variables M,P,W(this independence is defined OC dichotomy).
The nominal scale of the economy (M,P,W) only depends on the level
of money supply M. Fiscal and monetary policies are ineffective:
remember why. M/P=kYM/Pfe
Yd(r) rfe
Ysfe
Labour market Ld(W/P) =Ls(W/P) gives W/Pfe & Lfe Clearing
variable W. Production function Ys=Ys(K,L) gives Yfe Clearing
variable Ys. Goods market Ysfe=Yd(r) gives r Clearing variable r.
Money market M/P=KY gives P&W Clearing variable P.
Yfe
Yfe
Money mkt YYfe
Goods mkt Yfe
45
45
Yfe
Lfe
Yfe
product. fct Ld W/Pfe Ls
Lfe
labour Mkt
OLD CLASSICS AND UNEMPLOYMENT
According to historical evidence, labour market equilibrium is
an exception. The real world is generally characterized by
the excess supply of labour (involuntary unemployment). Some
people would like to work at the existing real wage. They, however,
do not succeed in finding a job. In order to explain the existence
of involuntary unemployment, Old Classics had to introduce
imperfections. Specifically, they introduced money wage rigidities.
Let us then assume that the nominal wage W -rather than being
endogenously determined by labour demand and supply -is exogenously
given (at a higher than equilibrium level) by the government/trade
unions.Ld curve =MPL curve Ls curve
The real wage rises from W/Pfe to W/P0. Only workers with
MPL> W/P0.are profitable for firms. Firms demand for labour
falls to L0.Employment falls to L0.
W/P0 W/Pfe
0E
The system moves from point E to point 0.Ld0=L0Lfe Ls0L*
Involuntarily unemployment (the excess supply of labour) now is
Ls0-L0>0
Firms have no convenience to hire those who are involuntary
unemployed. The real wage is too high: in their case W/P0>MPL
THE OLD CLASSICAL MODEL WITH GIVEN WAGES If employment L falls in
the labour market below, according to the production function
aggregate supply Ys too falls. This leads to an increase in the
interest rate in the goods market (Yd has to fall) and to the
increase in the price level in the money market (kY and M/P
fall).Yd(r) M/P=kYM/Pfe=kYfe
Ys00
Ysfe
r0 0 rfe
Labour market W W/P Ld L Production function L Ys
M/P0=kY0
Y0
Yfe
Y0
Yfe
Money market
Goods market
Goods market Ys r Yd Money market Y kY =M/P P
Yfe Y045
Yfe Y0
Yfe Y045
Y0
Yfe
L0
Lfe
production function Ld WP0 W/Pfe
Y0
Yfe
0
Ls
To sum up, we have stagflation, i.e. stagnation (L, Y, C I) plus
inflation (P) involuntary unemployment L0Lfe
labour market L 0=L0 Lfed
REMEDIES AGAINST INVOLUNTARY UNEMPLOYMENT
FISCAL POLICY Let us assume a fiscal expansion (00): G=G-T=Bs In
financial markets: pb & r In the goods markets, aggregate
demand falls: r C(r) &I(r) Yd(r) Equilibrium income Y returns
to its initial level Y0: output levels beyond Y0 are not
profitable. At the end, government expenditure only crowds out
private expenditure. Y0=C(r)+I(r)+G There are no benefits in terms
of employment and income! Activity levels beyond L0&Y0 are not
profitable according to the labour market.
Yd1(r)r1 1 Yd0(r)
Ys C&I 0 0 G
r0
Ys=Y0
MONETARY POLICY Let us assume a monetary expansion (00): M &
M/P Disequilibrium point 0 implies: ESM=EDG Output Y is given from
the labour market: the EDG turns into a higher price level: P The
real money supply falls to the initial level M/P Money is neutral.
The central bank can affect M but not M/P. Again, no benefits in
terms of employment and output! Activity levels beyond L0&Y0
are not profitable according to the labour market.
M/P1
0
M/P= k Y
M/P0=kY0
0
Y0
Ld
W/P0 W/Pfe
0
WAGE FLEXIBILITY Involuntary unemployment Ls0-L0 implies L ESL
The money wage W becomes flexible. The real wage falls from W/P0 to
W/Pfe, W/P Firms demand for labour increases. Employment and output
rise. Ld L&Y We reach the full employment equilibrium.s
Ld0=L0
Lfe
Ls0
L*
Involuntary unemployment is due to an excessive cost of labour.
Firms do not hire (and do not produce) more workers since labour is
too expensive. Wage flexibility is the only weapon against
involuntary unemployment!
THE PRE-KEYNESIAN CONCEPTION OF THE ECONOMY
The pillars of the pre-Keynesian approach are the following:
Perfect rationality: Perfect competition: Perfect price
flexibility: Markets are simultaneously in equilibrium. Everybody
buys (sells) the optimal quantity. There is no individual incentive
to change. In this idealistic Olympus the real sector spontaneously
goes to its general (full employment) equilibrium. there is no need
for economic policy.
economic policies may even be dangerous, better not to intervene
government expenditure only crowds out private expenditure money
supply only affects the price level Rigidities/imperfections The
real world is not that perfect. If the system does not reach its
general (fe) equilibrium, it is precisely because of these
imperfections (wage rigidity). The main task of economic policy
authorities is to remove them.Theoretical developments Modern
versions of general equilibrium theory are much more complex:
inter-temporal, dynamic, stochastic The underlying vision, however,
remains the one presented above. Concrete developments This is
precisely the vision that inspired the European Monetary Union: the
European Central Bank is responsible for inflation (link
money/prices) National Government Budgets have to be balanced (no
crowding out) The labour market flexibility and the increase in the
skill/productivity of labour are the only remedies against
unemployment.
MACROECONOMIC EQUILIBRIUM: ITS EXISTENCE, UNIQUENESS,
STABILITYITS
Let us reconsider the under-employment equilibrium situation
represented by points 0 in the goods market (right upper panel) and
in the labour market (lowest panel). The goods market is in
equilibrium (Ys0=Yd), the labour market is not. The excess supply
of labour can be defined involuntary unemployment.
M/P=kY
Yd(r) r0
Ys00
M/P0=kY0
0
Y0
Yfe
Y0 Goods market
At the ongoing W/P0, some workers would like to work, but cannot
find a firm available to hire them.
Money market
Y0
Y0
Y045
Y0
L0 production function
Y0
WP0
Ld 0Efe
Ls
According to the OCs, this situation presupposes constraints on
market mechanisms: minimum money wages W imposed by trade unions or
governments.
labour market Ld0=L0
Let us then remove the obstacles! Let us liberalise the labour
market!M/P=kY M/Pfe=kYfe M/P0=kY0Efe
Yd(r) r0 rfe
Ys00
Ysfe
0
Efe
Y0
Yfe
Y0
Yfe
Money market
Goods market
Yfe Y045
Yfe Y0
Yfe Y045
If money wages W become flexible: -the excess supply of labour
will push W downwards -the system will reach its f.e.
equilibrium.
Y0
Yfe
L0
Lfe
production function
Y0
Yfe
WP0 W/Pfe
Ld 0Efe
Ls
labour market L =L0d 0
Lfe
Old Classics (the standard) macroeconomic theory assumes that
full employment equilibrium: exists is unique is stable. As we
shall see, these three assumptions are not necessarily true.
The Existence of Full Employment EquilibriumAccording to the
OCs, goods market equilibrium can be specified in the following
alternative ways. First formulation: Y=C+I+GAggregate supply Ysfe
is equal to aggregate demand Yd.
Second formulation: S=I+G-TThe deficiencies of expenditure (S)
are equal to the excesses of expenditure I+G-T.
r
Yd
Ysferfe
G-T + I
S (Yfe)
rfe
Yfe
Y
Sfe=G-T+I
Both the formulations assume that full employment equilibrium
exists. Paul Krugman, 2008 Nobel Prize, questions this assumption.2
In his view, we can frequently face the following situation. First
formulation: Y=C+I+Gr
Second formulation: S=I+G-Tr
Yd
Ysfe
I+G-T
S(Yfe)
Yfe
Y
S, I+G-T
For positive interest rates r>0, there is no interception
between the relevant curves. This means that full employment
equilibrium does not exist. Specifically, according to Krugman, at
any interest rate r0: the corresponding level of aggregate demand
Yd0 will determine output Y0S (i.e when Yd=C+I>Ys=C+S) Y falls
when S>I (i.e. when Yd=C+I0 Money yields 0, bonds yield e>0.
Wealth is entirely held in bonds: Md=0, Bd=W ii) r=rc e=0 Money
yields 0, bonds yield e=0. M and B are totally indifferent. iii)
ru*) L=L*; u=u* E L =L ; g =0 W/P* with falling money wages (gwL ;
g >0 There is excess demand for labour EDL (Ld>Ls). La Lb L*
Money wages W consequently rise. The Ph.C associates the low
unempl. rate (ua0 with rising money wages (gw>0). E gw=0 Point E
ua ub u* There is labour market equilibrium (Ld=Ls). PhC gwtY0. Let
us assume that the financing of the initial government deficit DF
0=G0-tY0=W* considered as a whole has expansionary wealth effects.
The consequent increase in income Y and in tax revenues tY will end
with reabsorbing the initial government deficit. The system will
thus reach the stable medium-run equilibrium point 1, with no
government deficit, with no creation of wealth and wealth effects
and with a stable equilibrium income level Y1.
G,T G0
T=tY 1 G0
T0=
Y0
Y1
As we have just seen, the stability of the system requires a
balanced government budget: DF=G-tY=W*=0 In the medium-run, income
Y has to generate a fiscal revenue tY equal to the given government
expenditure G. The expression for medium-run equilibrium income
thus becomes: Y=G/t The medium-run effect of government expenditure
on equilibrium income consequently is Y/G = 1/t>0 The conclusion
is that: government expenditure G has an expansionary effect on
equilibrium income; this expansionary effect depends on the tax
rate t: with a higher t, the income level needed in order to
reabsorb the government deficit is lower; this expansionary effect
does not depend on the way (M or B) in which it is financed: the
effectiveness of fiscal policy does not require the support of
monetary authorities.14 Orthodox Keynesians went even further. In
the expression for the government deficit, let us introduce
interest payments (iB), where i is the nominal interest rate and B
is the stock of pre-existing government bonds. DF=G + iB - tY=W*
G-tY is the primary deficit iB are interest payments on previously
issued government bonds. The medium run equilibrium condition
becomes: DF=G + iB - tY=W*=0 The expression for medium-run
equilibrium income consequently becomes: Y=(G + iB)/t The
medium-run effect on income of government expenditure thus becomes:
Y/G= (1+ i B/G)/t>0 The conclusion is the following.14
In order to focus on wealth effects, the exposition deliberately
ignores the supply side of the money market. Let us now extend the
analysis. Financial wealth generally stimulates the demand both for
money and for bonds. In the case of monetary financing, the
increase in money supply thus implies an equal increase in
financial wealth that only partially turns into an increase in the
demand for money. This means that the increase in money supply
prevails over the increase in the demand considered in the text.
The overall result is a downward shift of the LM curve that
strengthens the expansionary effects of government expenditure.
Medium run equilibrium income remains the one considered in the
text. The novelty is that, in case of monetary financing, stability
is granted and does not need ad hoc assumptions.
Government expenditure has again an expansionary effect on
equilibrium income. This expansionary effect keeps depending on the
tax rate t: in the presence of a higher t, the income level needed
to reabsorb the government deficit is lower. This time, however,
the way in which the deficit is financed is no more irrelevant.
Bond financing (B/G>0) now becomes more expansionary than money
financing (B/G=0). In the case of bond financing, the rise in
equilibrium income and in tax revenues has to offset not only the
initial increase in G but also the increase in interest payments
iB. Comments The debate on the wealth effects of government deficit
seems to represent an example of bounded rationality in economic
theory: a myopic over-evaluation of the present. The conclusions on
wealth effects derive from the assumption that the economic system
is stable. In the 1950s and 1960s, this assumption was relatively
plausible. Industrialized economies performed well; government
deficits and debts were contained. According to nowadays
experience, however, stability is not that granted. Industrialized
countries are currently experiencing the worst crisis after the
Great Depression. At the beginning, instability was endogenously
generated by the financial sphere of the private sector.
Subsequently, however, it has also affected public finances. In
order to offset the dramatic consequences of the crisis, many
countries have reached unprecedented levels of government deficit
and debt. In the case of the weakest countries like Greek, Spain
and Italy, the financial systems availability to buy government
bonds has decreased. The bond-financing of their government
deficits and the refinancing of their government debts are getting
more and more expensive. This is the origin of the ongoing spread
problem. An analogous example of bounded rationality in economic
theory is offered by the recent pre-crisis experience. The nineties
were goods years for the US economy. The industrial application of
the ICT fuelled a period of sustained not-inflationary growth. The
profession (the Chairman of the FED, Ben Bernanke, included)
claimed that we had inaugurated a new era - the Great Moderation in
which economic fluctuation would have been contained and under
control. In a 2000 article, Blanchard (chief economist of the IMF)
proudly wondered: What do we know about macroeconomics that Fisher
and Wicksell did not? Slogans are often dangerous. Some years
later, we face the opposite question: What did Fisher and Wicksell
know about macroeconomics that we have forgotten? Bonds are
promises to pay concerning interest payments as well as the
repayment of the principal. Bonds will be considered as assets (as
component of wealth) only insofar as these promises are credible.
If the government has to represent a reliable debtor, its deficit
and debt have not to be perceived as excessive and uncontrollable.
This leads us to the problem of the sustainability of government
debt. The sustainability of government debt As we have seen, in a
static economy the medium-run stability of the system requires: a
constant stock of bonds a zero government deficit. G + rB - tY=
B=0
Let us now move to a growing economy. If income Y grows, the
stock of bonds B too may grow. The problem is that the ratio B/Y
has not to rise. Such a rise might undermine the confidence on
government bonds of financial markets. Let us consider the
government budget in nominal terms and in discrete time. By
assumption, T=G. DFt = (G-T) + i Bt-1= i Bt-1=Bt = Bt - Bt-1 where:
G=T are given nominal variables i is the given nominal interest
rate Bt-1 is the stock of government debt at the end of the
previous period Bt is the stock of government debt at the end of
the current period According to the previous expression, government
debt self-feeds itself (si autoalimenta). The existing stock of
government bonds (Bt-1) generates interest payments (iBt-1) which
imply new bond issues, thus increasing the stock of government debt
(Bt). Every euro of todays debt implies (1+i) euros of tomorrows
debt. The growth rate (g) of government debt (B) is thus equal to
the interest rate (i). Government debt self-feeds itself, growing
at a rate equal to the interest rate. gb=i15 The growth rate of
nominal income gPY is given by the inflation rate gP plus the
growth rate of real income gY. gpy=gp +gy The government debt
sustainability condition requires that the growth rate of
government bonds is not greater than the growth rate of income.
Such a condition can be thus formulated in two alternative ways: in
nominal terms, the nominal interest rate i has to be lower than, or
equal to, the growth rate of nominal income. i < gp +gy in real
terms, the real interest rate r=i-gp has to be lower than, or equal
to, the growth rate of real income gy. r (= i - gp) < gy This
means that government debt sustainability also depends: on the
availability of financial markets to buy government bonds (via
interest rate) on monetary policy (again, via interest rate) on the
growth rate of the economy (via real income) Specifically, the
sustainability problem can be accentuated:15
If government deficit is not balanced, government debt grows
more rapidly at the beginning but at the end its growth rate tends
to be equal to the interest rate. Dividing the expression for the
deficit by Bt-1, we get gb = Bt - Bt-1= ( G-T ) + i Bt-1 Bt-1 In
the presence of a bond financed government deficit, B automatically
grows with the passing of time and consequently the ratio G-T/ Bt-1
tends to zero. The growth rate of government debt consequently
tends to g b=i; independently of the given level of the initial
deficit.
by a crisis of confidence in government bonds by a restrictive
monetary stance by a low growth rate of the economy The
afore-mentioned sustainability condition: is often mentioned in the
economic policy debate (newspapers, television and so on). is
particularly relevant for Italy, given its high government debt/GDP
ratio The condition, however, is only a rule of thumb. It is based
on crucial simplifying assumptions, for instance: that the interest
rate is given, independently of G-T that the real growth rate gy is
given, independently of G-T. that the inflation rate gp is given,
independently of G-T. To analyze the problem properly, we should
develop a model. The results, however, would then depend on the
specification of the model itself. The historical experience of the
last decades shows that government debt sustainability can
represent a problem. Public finance is not as stable as Orthodox
Keynesians thought.NEOCLASSICAL OBJECTION TO THE EFFECTIVENESS OF
FISCAL POLICY
We shall refer to Barro 1974: Are government bond net wealth?
Barros different perspective Orthodox Keynesians considered bonds
as financial assets of the private sector. However, government
bonds also represent debts of the state! In the future, these debts
will have to be repaid by new taxes. The anticipation of these
future taxes has depressive effects on consumption As a result,
government expenditure may become totally ineffective: it may end
with crowding out private consumption. Barros basic assumptions:
there is perfect coordination, prices clear all the markets; income
is always at its full employment level; there is perfect knowledge,
agents know the future. (In such a perfect world, however, any
policy seems destined to be ineffective!) The consumer plans
her/his consumption for the whole life. Life is composed by two
periods: the present (period 0) and the future (period 1). Todays
and tomorrows decisions are strictly interconnected. If the
consumer saves sacrificing todays consumption (CYd), she/he will
inevitably have to consume less tomorrow in order to repay her/his
debt. Present and future are connected by the consumer
inter-temporal budget constraint, which requires the equality
between: the present value of consumption PVC the present value of
disposable income PVYd PVC=PVYd
Specifically C0 + C1 = Y0 + Y1 - T0 - T1 (1+r) (1+r) (1+r)1
today Analogously, 1 today (1+r) This means that: 1 today (1+r) is
equivalent to (1+r) tomorrow is equivalent to 1 tomorrow is the
present (actual) value of 1 tomorrow
By isolating C1 on the right hand side, the inter-temporal
budget constraint becomes a sort of consumption frontier. Given the
capitalized (the future) value of disposable income [], a higher C0
today implies a lower C1 tomorrow. C1 = [Y0 (1+r)+Y1-T0 (1+r)-T1] -
(1+r) C0 The intertemporal budget constraint is shown by the
downward sloping line in the figure. The optimal solution is point
O, where the budget constraint is tangent to the highest
indifference curve. In the figure, the utility function is such
that the consumer prefers a stable consumption path. The solution
thus is: C0=C1=C
C1
O C1
45
C0M
*= LM0
C0
Point A below shows the given income levels in the two periods
(Y0,Y1). Without financial system, our consumer would have had to
choose point A, on a lower indifference curve (corresponding to a
lower utility level). Specifically, she/he would have been forced
to choose: a higher current consumption C0=Y0 in period 0 a lower
future consumption C1=Y1 in period 1
0
C1
O C1 S0(1+r) Y1 S0 C0 Y0 C0 A
In the presence of the financial system, by contrast, our
consumer can reach the optimal solution (point O) (corresponding to
a higher utility level) and stabilize her/his consumption (C0=C1=C)
as desired. Specifically, in period 0 she/he can consume only C0=C
and save Y0-C0=S0 in period 1 she/he can rise her/his consumption
from Y1 to C0=C=Y1+(1+r) S0. What about fiscal policy? As known,
the consumers inter-temporal budget constraint is: PVC=PVYd i.e. C0
+ C1 = Y0 + Y1 - T0 - T1 (1+r) (1+r) (1+r) Analogously, the
governments inter-temporal budget constraint requires that the
present value of government expenditure (PVG) is equal to the
present value of taxes (PVT). PVG=PVT i.e. G0 + G1 = T0 + T1 (1+r)
(1+r) Today, the government may run a bond financed deficit
G0-T0=B0. Tomorrow, however, it will have to repay its debt
(1+r)B0=(1+r) (G0-T0) by running a corresponding surplus T1-G1.
T1-G1=(1+r)B0=(1+r) (G0-T0) By substituting the governments into
the consumers budget constraint, we get C0 + C1 = Y0 + Y1 - G0 - G1
(1+r) (1+r) (1+r) First implication; the tax and bond financing of
G are perfectly equivalent
-a tax financed government expenditure G0 implies T0=G0 taxes
today. -a bond financed government expenditure G0 implies
T1=G0(1+r) taxes tomorrow, -the present value of taxation is the
same in both the cases (PVT=G0). This leads us to the famous
Ricardian equivalence theorem: bonds are not wealth, they are
future taxes. Tax financing and bond financing are perfectly
equivalent. Second implication: government expenditure crowds out
private consumption We have seen that a new government expenditure
G0, however financed, raises the present value of taxation by G0.
From the consumers point of view, it thus implies an equivalent
fall in the present value of her/his disposable income an
equivalent fall in the present value of her/his consumption.
Government expenditure crowds out an equal amount of private
consumption. Third implication: a bond financed government
expenditure stimulates current saving In the case of a new
government expenditure G0 financed by bonds, the new taxes
T1=G0(1+r) will be collected tomorrow, when government bonds will
have to be repaid. The consumer has to save more in order to be
able to pay these new future taxes. Comments: Barros results
obviously depend on his initial assumptions; first of all, the
assumption of a full-employment income Y which by itself implies
the ineffectiveness of government expenditure. 3rd extension: the
long-run IS-LM model According to the OKS, in the long-run money
wages and prices become perfectly flexible. The labour market
reaches its equilibrium; outputs Y goes to its full employment
level Yfe. The long-run aggregate supply curve LRAS is a vertical
line at Yfe. Let us analyze the long-run working of the IS-LM
model.LRAS IS E
LMe(M/Pfe)
rfe, rfe
Yfe
The long-run equilibrium of the real sector is given by the
intersection IS/LRAS Yfe, rfe The LM curve is passive: its position
(i.e. M/P) has to adapt itself to the intersection IS/LRAS.
M/Pfe=L(Yfe, rfe) This adaptation is based on the price level P, at
the given level of M M Pfe (& Wfe)
There is dichotomy. Real equilibrium does not depend on the
nominal scale of the economy
rIS
LRAS
The labour mkt gives the LRAS curve and output Yfe.LMe(M/Pe)
The goods mkt (IS curve) gives rfe, the interest rate at which
Yd(r)=Ysfe. (a real variable). The money mkt gives M/Pfe. Given
Yfe&rfe, M/Pfed and consequently M/Pfes are given. Point E
represents a real equilibrium (Yfe, rfe, M/Pfe) independent of the
nominal scale of the economy (M, P, W).
E
Yfe
Monetary policy is ineffective: money is neutral. Money supply M
does not affect M/P; it only determines the price level P (and
money wages W). The figure below shows the case of a monetary
expansion. LRAS From 0 to 0: monetary expansion r LM0 IS M M/P r
(LM) I&Yd (on the IS curve) Point 0: excess demand for goods
Yd0>Yfe 0 Adjustment mechanism: P M/P LM1 Point 0: we go back to
the initial position.0 -The LM is passive: its position has to
adapt itself to the ISASLR intersection. -Equilibrium real money
balances M/Pfe (the LM position) is consequenly given by the real
sector. -An increase in M creates an equi-proportional increase in
P.
Yfe
Yd0
Fiscal policy too is ineffective Let us assume a bond financed
increase in government expenditure.r IS0 IS1 LRAS 1 LM(M0/P1)
LM(M0/P0) 0 Point 0: initial equilibrium. Point 0: G Yd (IS):
excess demand for goods Yd0>Yfe Adjustment mechanism: P M/P r
(LM) Point 1; new equilibrium at Yfe again. Equilibrium income does
not change The crowding out of I is -total -real (the constraint is
Y=Yfe)
0
Yfe
INFLATIONARY EXPECTATIONS Before we have considered a non
inflationary regime, with once and for all changes in M and P. Let
us now move to an inflationary regime with: rising money and prices
gM=gP>0 an expected inflation rate equal to the effective one
gPe=gP >0 Expected inflation introduces the distinction between
the nominal (i) and the real (r) interest rate: i=r+gPe According
to the IS-LM model, the wealth holder has the choice between money
and bonds. Inflation decreases the real value of both of them
(M/P&B/P).
The cost of money holdings (i=r+gPe) is given by the real
interest rate on bonds (r) plus the expected loss in purchasing
power due to inflation. The traditional LM curve will thus be
defined in terms of the nominal interest rate i. The new LM* curve
in the figure (in r= i- gPe) gives the real interest rate r. The
vertical distance between the two curves is obviously equal to
expected inflation gPe.i,r LM (in i) LM* (in r=i-gPe) gPe
Y
For investors, what matters is the real interest rate r.
Inflation reduces the real value of loans. It thus represents a
gain for firms who borrow in order to invest. The real cost of
borrowing is thus r= i- gPe For the IS curve, what matters is the
LM* curve (in r). The equilibrium real interest rate (rfe) is given
by the interception LRAS/IS. The relevant LM curve (the LM*) has to
intersect that point (via fluctuations in P). Point R thus
represents the equilibrium of the real sector (Yfe, rfe). Point N
represents the equilibrium of the monetary sector
(ife=rfe+gPe).LRAS ife rfe
LM gPe
(in i= r+ gPe) LM* (in r)
N
R
IS Yfe
The figure below analyzes the full employment effects of
inflation. In the initial situation of monetary stability where
gP=gPe=0 the nominal interest rate coincides with the real one
(i0=r0) the LM0 (in i) curve coincides with the LM0* (in r) the
real and the nominal equilibrium of the system coincide with point
R0 Let us move to an inflationary regime where gM=gP=gPe>0 real
equilibrium remains at R0 the real interest rate remains at r0 the
LM0* curve (in r) coincides with the initial one. the LM curve (in
i) moves upwards to LM1 the nominal equilibrium of the system moves
to N1 the nominal interest rate i raises by an amount equal to
expected inflation (i1=r0+ gPe)
LRAS N1
LM1 gPe
(in i)
i1 i0=r0 =
R0=N0
IS0 Yfe
Adjustment mechanism Inflationary expectations initially imply a
corresponding fall in the real interest rate (r) at any given
nominal interest rate (i). The LM* curve (in r) falls below the
unchanged LM curve (in i). This has expansionary effects on
investment and aggregate demand (on the unchanged IS curve). Since
output is already at full-employment, however, the higher demand
will only increase the price level P. The consequent fall in M/P
will raise the nominal and the real interest rates. The LM* curve
(in r) and the LM curve (in i) will both shift upwards, At the end,
the real interest rate (r) will go back to its initial level. The
only effect of inflation will be an equivalent increase in the
nominal interest rate (i) in the figure.16.LRAS N1 R0=N0 0 IS0 Yfe
LM1 LM0= LM*0=LM0=LM*1 LM*0
i1 i0=r0 =
Who cares about the increase in the nominal interest rate? To
answer, let un focus on the money market. Inflation decreases the
purchasing power of money. Expected inflation gPe consequently
raises the cost of holding money (i). The demand for money L thus
falls in favour of the demand for goods. The excess demand for
goods raises the price level P. The real supply of money M/P
consequently falls, aligning itself to the lower demand. Money
market equilibrium moves from point 0 to point 1. M = L(i, Yfe)
P
16
Looking at the LM1, however, this reflects a rise in the price
level P and a fall in M/P. We shall come back to this below.
M/Ps1 i1 r1=r0=i0 1 gPe 0
M/Ps0
Inflation tax
M/Pd
M/P1
M/P0
The conclusion is that inflation implies an inflation tax, with
tax rate gPe and tax base M/P1. As a consequence of this tax, real
money balances M/P fall from M/P0 to M/P1. Since money is wealth,
this is the welfare loss due to inflation. Conclusion In a regime
of flexible money wages and prices: Money is neutral Changes in the
level of money supply M imply equi-proportional changes in the
level of P, leaving M/P and all the other real variables unchanged.
The level of M does not really matter. Money is not super-neutral
Changes in the growth rate of money supply M imply an inflation tax
on money holdings that in turn decreases M/P. Real money balances
M/P are real wealth, this is thus a real welfare cost. The growth
rate of M really matters. LEIJONHUFVUDS HETERODOX VIEWS ABOUT
INFLATION The empirical value of the inflation tax is negligible.
Thus, according to neoclassical theory, anticipated inflation has
no relevant real costs. Nominal variables grow at the same rate as
M, real variables remain (essentially) the same. The real world
however, is radically different. In a 1995 book, Leijonhufvud and
Heymann analyze the experience of high inflation in Argentina.
Inflation implied a great contraction in economic activity (not
only a negligible inflation tax). How to explain this??
Leijonhufvuds and Heymanns answer is the following. Under an
inflationary regime, uncertainty increases. The result is that:
agents and economic policy authorities loose the ability to
forecast the future. They have no idea about future costs, future
price levels, future profits, future interest rates, and so on.
Between the phone call and her/his arrival, the taxi driver has
already increased the price. When uncertainty is so high, economic
decisions become more difficult to take. medium and long-term
contracts disappear. -in financial markets, the segments beyond
twelve months disappear; only very short-term assets are exchanged.
-in the real sector, investments drastically fall. All of this
entails heavy depressive repercussions on the real economy.
Monetarism(English to be revised) The Monetarist School
flourished from mid-1950s to mid-1960s. This school represented the
neoclassical component of the Synthesis. Its aim was to restate the
pre-Keynesian Quantity Theory of Money. Milton Friedman was the
leader of the Monetarist School; he got the 1976 Nobel Prize. The
following presentation of his contribution will focus on the two
issues raised by the Synthesis: the nature of aggregate demand the
role of aggregate demand.THE NATURE OF AGGREGATE DEMAND
Basic Questions Is aggregate demand mainly a real or a monetary
phenomenon? Do shocks mainly come from the real or from the
monetary sector? To be more effective is fiscal or monetary policy?
Orthodox Keynesians answered that: aggregate demand is a real
phenomenon; the basic equation for Y is the expression below of the
Keynesian multiplier; income fluctuations are mainly due to real
disturbances: expectations, and thus the investment component of
autonomous expenditure A, are highly volatile. fiscal policy is the
most efficient weapon for the control of A and consequently of
economic activity. Y = [A - d i] Monetarists objected that
aggregate demand is a monetary phenomenon; the basic equation for Y
is the one at the basis of the Quantity Theory; monetary shocks are
the main source of fluctuations in nominal income; monetary policy
is the most efficient weapon for the control of economic activity.
(M) V = (PY) The crucial role assigned to money explains the label
Monetarism In order to present the monetarist view about the nature
of aggregate demand, we shall start with: Friedman (1956), The
Quantity Theory of Money: a Restatement. This article was published
in the same year as the 1st edition of Patinkins book, which
completed the Synthesis. As the title claims, its aim was to
restate the Quantity Theory of Money. The IS-LM model adopted by
Orthodox Keynesians focused on the allocation of financial wealth
between money and bonds. Friedman extends the concept of wealth; in
his view, wealth W* includes all what yields
future income streams either in monetary or in non monetary
terms. The main components of wealth - and the corresponding income
streams - are shown in the following table. AssetsMoney M Bonds B
Equities E Durable goods G Human capital H
Pecuniary or not pecuniary income streamsTransactions,
precautionary and speculative non-pecuniary services, whose amount
depends on the purchasing power of the stock of money (ceteris
paribus, on P) Nominal interest payments, in percentage equal to
the nominal interest rate on bonds rb Dividends d (a share of firms
real profits which in percentage ensure a real return re ) plus
expected inflation gep (an expected capital gain on real assets)
Non pecuniary services (car, washing machine, flat). Their value is
expected to rise with expected inflation gep Labor incomes,
corresponding to a percentage return h on human capital H.
ReturnsP rb re+gep gep h
On this basis, Friedman deduces that the nominal demand for
money Md is: a demand for real money balances M/P, i.e. a positive
function of P; a negative function of the rates of return on
non-monetary assets; a positive function of nominal wealth W*.
Md=f(P, rb, re+gep, gep, h, W*) As we have seen, wealth has the
property of yielding future incomes streams (pecuniary or not) If
we multiply nominal wealth W* by the average interest rate r, we
get the average future nominal income stream. Friedman defines it
as permanent nominal income Ynp. Ynp=r W* Without explanations,
Friedman then assumes that current income is equal to permanent
income. The superscript p thus disappears. Yn=r W* This means that
W*= Yn /r The equation for the nominal demand for money then
becomes: Md=f(P, rb, re+gep, gep, h, Yn/r) This function can be
further simplified: the twofold negative effect of gep can be
unified. the average interest rate in Yn/r can be removed; its role
is captured by individual rates of return. Md=f(P, rb, re, gep, h,
Yn) Finally, we have to consider that what is relevant is the
amount of money in real terms M/P. If nominal variables (P and PY)
double, the nominal demand for money Md too doubles. 2Md=f(2P, rb,
re, gep, h, 2Yn) More generally, if P and PY grow by , Md too has
to grow by . Md=f(P, rb, re, gep, h, Yn) Given nominal income Yn
=PY, by setting
=1/Yn =1/PY we get (1/PY) Md= f(1/Y, rb, re, gep, h, 1) This
means that Md= f(1/Y, rb, re, gep, h, 1) PY In equilibrium, Md is
equal to the exogenously given money supply M. The money market
equilibrium condition consequently becomes: M= f(1/Y, rb, re, gep,
h, 1) PY Friedman highlights that everybody (even Keynes himself)
would agree on this result. What is then the distinguishing feature
of Monetarism? According to Friedman, Monetarism is characterized
by the following assumptions: the supply of money M is exogenous
the f() function behaves like a constant. This can be due to many
reasons: -the variables inside f() come from the real sector and
are consequently given for the monetary sector; -the variables on
which f(..) depends do not vary with time by a significant amount;
-the variables on which f(..) depends vary with time, but their
effects on f(...) offset each other. -f(..) has a very low
sensitivity to its determinants: more generally, f(..) is a stable
and predictable function of a limited number of variables. Under
these assumptions: nominal income PY becomes a stable and
predictable function of the given money supply M; put otherwise,
nominal income is a monetary phenomenon; economic fluctuations
reflect nominal shocks; money supply is the best control weapon. PY
= f(1/Y, rb, re, gep, h, 1)-1 M When money supply M increases, the
excess supply of money will turn into a demand for alternative
assets. Partly through the fall in interest rates and partly (as we
shall see, above all) through wealth effects, this will stimulate
the value of expenditure on goods and services. M PY To sum up,
Monetarism (the modern Quantity Theory proposed by Friedman) is
essentially a monetary theory of nominal income. It is the equation
above (not the expression for the Keynesian multiplier) that we
have to estimate in order to explain and to forecast the time
behaviour of nominal income. With regard to the debate animating
the Synthesis, Friedman (1956) proudly adds that Monetarism
represents THE general theory: Old Classics took real income Y as
given at its full employment level;. Keynes (according to the
mainstream interpretation) took money wages W and thus P as given;
Monetarism focuses on nominal income PY, without introducing ad hoc
assumptions in order to distinguish between real income Y and
prices P. In his (1958) article, The supply of money and changes in
prices and output, Friedman claims
that the relationships between M and PY is strongly confirmed by
econometric evidence. Money and nominal income have a very similar
cyclical behavior. The peaks (troughs) in the money cycle, however,
anticipate the peaks (troughs) in the nominal income cycle. This
confirms that causality runs from money M to nominal income PY:
nominal income is a monetary phenomenon. M PYM, PY
M PY
peaks troughs
time
The debate between Friedman and Tobin In a 1970 article entitled
Money and income: post hoc ergo propter hoc, James Tobin - leader
of the Orthodox Keynesian School - objected that: post hoc does not
mean propter hoc put otherwise, afterwards does not mean as a
consequence of. Firms have generally to get bank credit before
investing, and bank credit in turn stimulates money supply. The
expansion in bank credit and in money supply thus anticipates the
expansion of expenditure and income. Nevertheless, it is a
consequence of the higher propensity to spend. More generally -
according to Tobin - money supply M is endogenous.17 The existing
amount of money depends on the behavior of the economy. As a
consequence, it is outside the control of the central bank. Let us
see why. As mediums of payment, we use: currency CU (coins and
banknotes in our pockets) bank deposits D. Money supply can be
consequently defined as: M = CU + D According to the budget
constraint of the banking system, bank deposits D are equal to bank
reserves RE plus bank credit CR. D = RE + CR By substituting above,
we get that: M=CU+RE+CR By definition, the coins and banknotes
issued by the central bank17
According to Tobins (1963) article entitled Commercial banks as
creators of money, however, the distinction between banks and the
other financial intermediaries in not sharp. For a recent debate on
this issue, see the web site uneasy money.
represent the monetary base or the high powered money (H), which
in its turn is held partly by the non bank public (as currency CU)
and partly by banks (as bank reserves RE). H=CU+RE The result is
that M=H+CR Money supply has thus two components: the high powered
money - or monetary base H, which is issued by the central bank
bank credit CR, created by the banking system by buying bonds
(promises of payment) or by granting loans. To conclude, the
banking system too creates money. When you get a credit from a
bank, you exchange a promise of payment (a non monetary asset)
against banknotes or deposits (a monetary asset). Thanks to bank
credit, the quantity of money available to the non-bank public
rises. In Friedmans view, the most important component of money
supply is the high powered money or monetary base H issued by the
Central Bank. Bank credit (and thus the whole money supply) can be
considered as a multiple of H. 18 The control of H by the central
bank consequently ensures the control of the whole M. The supply of
money is exogenously determined by monetary authorities. In the
relationship between money and nominal income, causality thus runs
from the former to the latter. M PY According to Tobin, by
contrast, the main component of money supply is represented by bank
credit. This variable is determined by banks and by their
customers, not by the monetary authorities.19 This means that bank
credit and money supply are endogenously determined by the economic
system. In the relationship between money and nominal income,
causality runs from the latter to the former. PY CRd CRs M=H+CR To
conclude, Tobins objection to Friedman is that: the quantity theory
equation explains money supply M, not income Y. income is
determined by the Keynesian multiplier (1/1-c) and by autonomous
expenditure A.18
Let us assume that the desired amount of CU and RE is a given
fraction of bank deposits D, i.e. that CU=cu D, RE=re D, CR=(1-re)
D. From H=CU+RE we have that: D=[1/(cu+re)] H Since M=CU+D=(1+cu)
D, we can conclude that: M=[(1+cu)/(cu+re)] H The Keynesian
multiplier [1/(1-c)] focuses on the interdependence between
expenditure and income, coming to the conclusion that in the goods
market income Y is a multiple of autonomous expenditure A. The
money multiplier [(1+cu)/(cu+re)] focuses on the interdependence
between bank deposits and bank credit according to which the
monetary based deposited into the banking system comes back to the
non-bank public through bank credit and is deposited again. The
result is that the stock of money M is a multiple [(1+cu)/(cu+re)]
of the monetary base H created by the central bank. In Friedmans
view, the parameters cu and re (and consequently the whole money
multiplier) can be considered as given. The control of the monetary
base H by the central bank consequently ensures the control of the
whole supply of money M. According to Tobin, by contrast, cu and re
are not given. They reflect the choices of banks and of their
customers and consequently depend on the behavior of the economy.
19 The endogenous nature of money is one of the tenets of the Post
Keynesian School. This issue is crucial. As we have seen, if money
is endogenous, the fall in money wages and prices is unable to lead
the system to its full employment equilibrium.
Extending Tobins approach, we may notice that money supply can
be endogenous also in the absence (and thus independently) of the
banking system, i.e. when M=H. This can happen as a consequence of
the strategy adopted by the central bank. Given the demand for
money function (L), the monetary authority has two options: it can
control the amount of money (M); in this case, money supply is
exogenous whilst the rate of interest (i) is endogenously
determined by the demand for money. it can control the interest
rate (i); in this case, the latter becomes endogenous whilst money
supply is endogenously determined by the demand for money.i
Md=L(i)PY Ms
Ms
M
The choice between the two intermediate targets (the quantity of
money M and the interest rate i) will depend: on their
controllability by the central bank; on their ability to affect the
final targets (output, prices..) of the central bank.Strategy of
the Central Bank Instruments(open market operations)
Intermediate Targets(money supply, interest rates)
Final Targets(expenditure, output, prices.)
Until the 1990s, the strategy of monetary authorities was
generally based on the control of money supply. In line with the
Quantitative Theory, this was meant to ensure the control of
expenditure and prices. With time, however, this strategy proved
increasingly impracticable and ineffective. The stock of money
proved increasingly difficult to control. The link between money
and expenditure (the velocity of circulation) became increasingly
unstable and unpredictable. Given the impracticability and the
ineffectiveness of the control of money supply, monetary
authorities moved to the control of the interest rate. Currently,
central banks set the interest rate and offer all the amount of
money that the system requires. Put otherwise, the interest rate is
exogenous whilst money supply is endogenous. Tthe conduct of modern
central banks is usually described by the Taylor rule: it = (rt* +
gpt*) + a (gpt - gpt*)+ay (Yt-Yt*) where the asterisked variables
represent desired/target values; the non-asterisked variables
represent observed/expected values; the coefficients a and ay
represent given parameters with positive values.
In the previous equation: the first term on the right-hand side
(rt*+gpt*) represents the nominal interest rate desired by the
central bank i*, given by the sum of the desired real interest rate
rt* plus the desired inflation rate gpt*. the last two terms on the
right hand side (gpt - gpt*) and (Yt-Yt*) tell us that, whenever
inflation gpt or output Y are higher than desired, the central bank
restricts the economy by raising the interest rate i above its
target level i*. The following graph compares the actual Fed funds
rate with the value generated by the Taylors rule. The Taylor rule
seems to be a reasonably good representation of the U.S. monetary
policy!! The same holds for many Central Banks of other
industrialized countries (EMU, UK..).
Thus, if we consider the experience of nowadays, Tobin was right
in highlighting the endogenous nature of money supply.THE DANGEROUS
LIAISON BETWEEN MONEY AND INCOME
As we have seen, the demand for money function is the starting
point of Monetarism. This function is analyzed in more detail in
two famous articles: Friedman 1970, A theoretical framework for
monetary analysis Friedman 1971, A monetary theory of nominal
income. Following Fisher, Friedman starts with the distinction
between the nominal interest rate (i) and the real interest rate
(r). As known, the relationship between the two is: i = r + gep By
holding money, we have two kinds of costs: the real interest rate
on alternative assets (r); the expected fall in moneys purchasing
power due to expected inflation (gep) It is thus the nominal
interest rate (i = r + gep) which is relevant for the demand for
money (L). In a not-inflationary environment in which gep=0,
nominal and real interest rates coincide. The money market
equilibrium condition can thus be written in the usual way, where
r=i is the average interest rate. M=L(i=r) PY In an inflationary
environment in which gep>0, however, nominal and real interest
rates diverge. Since the demand for money depends on the nominal
interest rate,
the money market equilibrium condition becomes: M = L(i=r + gep)
PY As an example, let us start with an initial equilibrium where:
-money supply M grows at a rate of 6% -nominal income PY grows at a
rate of 6% -prices P grow (for instance) at a rate of 3% -real
income Y grows at a rate of 3% -expected inflation gep is 3% (in
equilibrium expectations are correct) 6% 3% 3% 3% M = L( r + gep) P
Y Let us now assume that -the growth rate of money supply raises to
12%. -the real interest rate r does not change We shall reach a new
equilibrium where: -money M grows at a rate of 12% -nominal income
PY grows at a rate of 12% -prices P grow (for instance) at a rate
of 6% -real income Y grows at a rate of 6% -expected inflation (as
current inflation) gep is 6% (in equilibrium expectations are
correct) 12% 6% 6% 6% M = L( r + gep) P Y In each of the two
equilibrium situations, money M and nominal income PY grow at the
same rate: first by 6% and then by 12%. In the transition phase,
however: -expected inflation gep rises from 3% to 6%. -the nominal
interest rate i=r+gep rises -the speculative component of the
demand for money L(..) falls -nominal income grows at a rate higher
than 12% M = L(r + geP) P Y Friedmans analysis can be represented
in the following figure. At time T , the growth rate of money
supply exogenously rises from 6% to 12%. In the old and in the new
equilibrium, money M and nominal income PY grow at the same rate
(firstly 6% and then 12%). The two series, however, do not
coincide. In the transition period, as inflationary expectations
rise, PY grows faster than M.0
logM, logPY
logPY
Transition geP i L PY
log M12%
6%Old equilibrium
T0 Transition
New equilibrium
Time t
The example considered above implies a regular transition to the
new equilibrium. In Friedmans view, however, the monetary sector
may also be unstable. The change in the growth rate of M may then
imply an explosive reaction in PY.
Log PY Log M, LogPY
Log M
Time t
The cumulative process which leads to the explosive reaction of
income is due to the interdependence between gep and L(..). On the
one hand, the rise in expected inflation (gep) implies an increase
in the nominal interest rate i=r+ gep which reduces the demand for
money L. gep i L On the other hand, the decrease in L implies an
increase in the demand for goods which fuels actual and thus
expected inflation. L gp gep Starting with the monetary nature of
aggregate demand, Friedman consequently comes to the conclusion
that money: is a powerful weapon for the control of nominal
expenditure; but is also so powerful that it can be destabilizing.
Friedmans belief is consequently that central banks: should not
pursue a short-term perspective, using money as a countercyclical
tool for the fine-tuning of economic activity; should instead adopt
a long-run perspective, using money to accommodate the growth of
real income (gy) without leaving any room to inflation (gp).
Friedmans monetary policy rule consequently is: gm=gy Comments
on Friedmans analysis about the nature of AD i) The real interest
rate r To simplify, let us consider a not inflationary environment.
In terms of the IS-LM model with P given that was used by the
Synthesis, the Monetarists framework is the opposite of the
Orthodox Keynesians one: -the interest rate (i=r) is given by the
real sector (a flat IS curve) -real income is given by the monetary
sector (a rigid LM curve)LM
r=i
IS
Y
By assuming that the real interest rate r is given, however,
Friedman too introduces an ad hoc assumption. With this, he can no
longer present Monetarism as THE general theory. Downsizing its
ambitions, he is forced to recognizes that: -Old Classics took real
income as given at its full-employment level -Keynes (orthodox
interpretation) and Orthodox Keynesians took nominal wages and
prices as given -Monetarists take the real interest rate as given.
At this stage, Friedman also comes to admit that his ad hoc
hypothesis of a given interest rate lacks a convincing theoretical
justification. To quote his own words, in the absence of this
justification his monetary theory of nominal income resembles the
Shakespearean play of Hamlet without its Prince. If Friedman
subscribes to the Quantity Theory of Money, however, it is because
he firmly believes that (contrary to the money sector) the real
sector is stable. -Thanks to the price mechanism, the labour market
tends to its equilibrium. -Aggregate supply consequently tends to
its full employment level Yfe. -The goods market (the IS curve)
gives the interest rate (a real variable) which aligns Yd to Yfe.
-The LM curve is passive: it adapts itself to the ASfe-IS
intersection through the fluctuations in P. -Equilibrium money
balances M/P are given by the ASfe-IS intersection (by the real
sector). -Money is neutral: money supply can only affect the price
level.
ASfe
LM(M/P)
r=i
ISYfe
ii) The nominal interest rate According to Orthodox Keynesians,
a monetary expansion implies a fall in the nominal interest rate. M
i According to Friedman, the opposite happens. A monetary expansion
raises actual and thus expected inflation. Given the real interest
rate, it consequently raises the nominal interest rate. M i = r+
geP iii) The velocity of circulation V According to the Old
Classics, the velocity of circulation was an institutional
constant. Real income was at its full employment level. The result
was the neutrality of money.. MV=PY According to Keynes, the
equilibrium condition of the money market is M=L(i) PY The velocity
of circulation is the reciprocal of L(i). 1/L(i)=V(i) This means
that V is a positive function of the interest rate i (instead of
being a constant). i L(i) V=1/L(i) In the presence of a monetary
expansion which lowers the interest rates, the velocity of
circulation falls and this mitigates the impact of M over PY To
conclude, in Keyness view, V performs an anti-cyclical role. M V(i)
= P Y According to Friedman, the equilibrium condition of the money
market is M = L(r+gep) YP. On this basis, a monetary expansion:
-increases actual and expected inflation; -raises the nominal
interest rate i=r+gpe; -implies the fall in L(..) and the increase
in V=1/L(..). M gp gpe i L V=1/L The velocity of circulation V
accentuates the impact of M over PY Friedmans conclusion is that
the role of V is pro-cyclical.
MV=PY iv) The role of monetary policy The high confidence placed
in the link from M to PY might suggest that Friedman is favorable
to monetary policy. As we have seen, this is absolutely false!!!
Money may have explosive effects on nominal income: we have to use
it the least possible. Money supply should simply accommodate the
real growth of the economy, without giving any room to inflation.
Friedmans rule for monetary policy consequently is: gm=gy v)
Expectations As we have seen, in the transition period inflationary
expectations gradually adapt themselves to their new equilibrium
values. With regard to this, Friedman assumed adaptive
expectations. Todays expectations gpet are equal to yesterdays
expectations gpet-1 adjusted by a fraction P) will imply a
negotiated real wage W*/P which is higher then the desired one
(W/P)*. Firms (workers) will offer (require) a real wage above
their traditional Ld (Ls) curve. W*>(W)* P P An underestimation
of the price level (Pe < P), will imply a negotiated real wage
W*/P which is lower then the desired one (W/P)*. Firms (workers)
will offer (require) a real wage below their traditional Ld (Ls)
curve. W* gep=gw* =0%Ls0
W/P0 W/P0
0
W/P falls below W/Pfe L rises above Lfe. Y rises above Yfe u
falls below ufe.
Lfe
L0
However, Friedmans short-run is really short-lived: price data
are published relatively frequently. With time, workers will then
realize their forecasting errors. As a result, they will gradually
correct them (adaptive expectations). Specifically, they will ask
for higher money (and real) wages W* (W*/P). Their labor supply
curve will go back from the Ls0 to the traditional Ls0 curve. The
system will move back from short-run equilibrium 0 to long-run
equilibrium 1. Employment (unemployment) will go back to its
natural level Lfe (ufe). Current and expected nominal variables
will now grow at a 10% rate.IMPLICATIONS FOR THE
PHILLIPS CURVE
Following Friedman, we shall distinguish between: the short-run
the long-run. FRIEDMANS SHORT-RUN PHILLIPS CURVE As we have seen:
what matters in the labour market are real wages W/P; ceteris
paribus, desired money wages W*=(W/P)* Pe depend on expected prices
Pe; ceteris paribus, the growth rate of money wages (gw) will then
reflect expected inflation (gep) On the basis of these premises,
Friedman claims that the short-run Phillips curve has to be
inflation augmented. Behind the lines, it subtends a given expected
rate of inflation gep. gw=f(u) +gep The negative slope of Friedmans
short-run Phillips Curve At the given expected rate of inflation
gep: the short-run Phillips curve will be downward sloping. there
will be a negative relationship between the growth rate of money
wages gep and the rate of unemployment u. This result derives from
Friedmans analysis of the labour market. As we have seen, when the
growth rate of money and prices rises from 0 to 10%: workers do not
realize the change (their gep=0) they ask for the same money wage
as before (gW=0) they involuntarily ask for lower real wages
(gp=10%)
their Ls curve moves rightwards employment and output rise
unemployment falls. The relevant short-run Phillips curve implies a
0% expected inflation (gep=0). When actual inflation rises to 10%,
there is an inflationary surprise (gp - gep) which stimulates the
economy. The higher rate of inflation thus implies a fall in the
rate of unemployment u. We move from point 0 to point 1 (the new
short-run equilibrium).
gP
SRPC(gep=0) 1 0 ufeu
10%
0
Notice that in Friedman: causality runs from inflation gp to
unemployment u. For Keynes and for the OKS, the direction of
causality was the opposite: the activity level determined marginal
costs and prices. fluctuations in employment and output are supply
led. If unemployment is high, it is because labour supply is low.
For Keynes and for the OKS, unemployment was due to a low demand
for labour. The position of Friedmans short-run Phillips Curve
depends on the given expected rate of inflation gep. When u=ufe, we
are in general equilibrium. This means that foresight is perfect.
At u=u*, expected and current inflation consequently coincide
(gp=gep). The inflation rate gp corresponding to ufe along the
curve thus gives us the given expected rate of inflation gep.
underlying the whole curve.
gp
SRPC(gep=10%)
gp=10%0
ufe
u
Instead of the traditional unique Phillips curve, we shall have
a whole set of short-run Phillips curves; each of them implies a
given expected rate of inflation gep higher Phillips curves imply
higher expected rates of inflation gep
SRPC(gep=30%) SRPC(gep=15%) SRPC(gep=0%) gp=30% gp=15%
gp=0%ufe
u
LONG-RUN
PHILLIPS CURVE
As we have seen, according to Friedman, in the long-run: the
labour market reaches its full employment equilibrium; unemployment
is at its natural/equilibrium rate ufe; the long-run Phillips curve
is a vertical line at u=ufe. ; there is perfect foresight: current
and expected variables coincide. actual and expected nominal
variables grow at the same rate as money supply;
the growth rate of nominal variables does not affect the real
sector of the economy.gp LRPC
Initial long-run equilibrium 0 (with gm=0) u=ufe and
gm=gp=gep=gw=0 Subsequent long-run equilibrium 1 (with gm=10%) u=
ufe and gm=gp=gep=gw=10% Long-run Phillips curve u=ufe
gm=gp=10% gm=gP=0%
1
0
ufe
In the long-run, there is no trade-off between unemployment and
inflation!FRIEDMANS
PHILLIPS CURVE AND MONETARY POLICY
MONETARY EXPANSION Let us start from an initial not-inflationary
long-run general equilibrium point 0. unemployment is at its
natural rate u* nominal variables are stable (gm=gp=gw=0)
expectations are correct (gep=0) we are both on the LRPC and on the
SRPC corresponding to gep=0
gp
LRPC
SRPC (gep=0%)
0%
0ufe
Let us now assume an expansionary monetary policy. Money supply
and prices start rising at 10%. gm = gp=10% New short-run
equilibrium. Workers do not realize the change; they keep expecting
a stable price level (gep=0%). We shall move leftwards on the
corresponding SRPC (gep=0%)
gp
When actual inflation rises to 10%, there is an inflationary
surprise (gp - gep) which stimulates the economy.SRPC (gep=0%)
The higher rate of inflation thus implies a fall in the rate of
unemployment u. We move from point 0 to point 0(a short-run
equil.). Unexpected inflation stimulates economic activity.
10%
0
0%
0ufe
In the short-run, the monetary expansion is effective However,
it acts through an inflationary surprise which damages workers
inducing them to accept lower wages. Transition to the new long-run
equilibrium short-run equilibrium 0 is inevitably temporary:
expected inflation is 0% whilst actual inflation is 10%; gradually,
workers realize and correct their errors; they consequently ask for
a higher money (and real) wage; the short-run Phillips curve moves
upwards. New long-run equilibrium We end up at point 1, at the same
time: on the LRPC on the new SRPC corresponding to gep=10%. In the
new long-run equilibrium 1, actual and expected nominal variables
grow by 10%: u= ufe and gm=gp=gep=ga =10%
gp
SRPC(gep=10%)
LRPC
SRPC(gep=0%)
10%
0
1
0%
0 ufe
In the long-run, money is neutral and monetary policy has no
real effect. In the short-run, a higher rate of growth in money
supply stimulates economic activity. In the long-run, however, the
only effect is a higher growth rate in current and expected nominal
variables. This is undesirable, since inflation implies the
inflation tax. The negative effects of a monetary expansion are
accentuated by Friedmans accelerationist hypothesis.
If monetary authorities want to keep u