Download 3. Extension of Meade`s Model and Endogenous Dynamics

Alternative Approaches to the Building of Economic Dynamics in the
years of High Theory
Collegio S. Chiara
Università di Siena
December 18-19 , 2009
Early Static Keynesian Models and the Emergence of
Endogenous Cycle Theories: an Appraisal of Meade’s
Model (1937)
Michael Assous1
In 1936, J. Meade presented a “Simplified Model of Mr. Keynes’ System” at the Symposium
of the Oxford Econometric Society devoted to Keynes’s General Theory. Three papers were
presented at this occasion; the other two were John Hicks’s famous article “Mr. Keynes and
the Classics” and Roy Harrod’s paper on “Mr. Keynes and traditional theory”. Although
Meade’s paper contains elements that are absent from the traditional IS-LM, it received little
attention over the past half century. Certainly, its most interesting feature is that, bringing
explicitly expectations into the analysis, it directs attention to the stability of the system on
which expectations directly impinge. Stability properties of the system then depend on
whether long-term expectations are treated as exogenous or endogenous. Meade’s argument is
that instability may arise when long-term expectations are overly sensitive to movements in
current profits. This stability issue drew substantial interest from Ragnar Frisch, Jacob
Marshak, Jan Tinbergen and Michal Kalecki who all participated at the Oxford conference2.
1
Address for correspondence: Université de Paris 1 Panthéon-Sorbonne, 106-108 Boulevard
de l’Hôpital, 75013 Paris ; e-mail : [email protected]
2
The discussion on stability was one of the main themes that arose from the IS-LM papers.
As Young (1987) recalls, Meade’s paper was the subject of a lively exchange at the
symposium between Frisch and Meade. In his capacity as editor of Econometrica, Frisch
wrote to Meade about his paper a letter, dated 8 October 1936: “I was much interested in your
paper on the stability questions in connection with Mr. Keynes’ system. If the definition and
the fundamentals of the stability situation could be discussed in more exact terms than those
you use in your paper, I think the paper would be an excellent one and I should be very glad
The stability properties of Meade’s model have been examined later by Peter Rappoport
(1992)3. Meanwhile “Meade’s paper has largely been ignored, and expectations and the
stability issues they raise have had a shadowy existence in the Keynesian arena.” (Rappoport
1992: 356).
Recently, I have re-examined the stability properties of Meade’s model (Assous 2008). Based
on this stability analysis, my aim is to show that, as a result of its focus on the question raised
by endogenous long term expectations, Meade’s model opens new vistas in the field of
endogenous business cycles. More precisely, if one assumes that the sensitivity of long-term
expectations with respect to current profits may vary over time, Meade’s model allows
demonstrating a cyclical relationship between employment and money wages. The basic
features of this dynamics can be stated in a simple way. High employment generates wages
inflation in a Phillips Curve fashion, which reduces profits and decreases investment and
output. This reduction in output reduces in turn labour demand and employment which leads
to lower wage inflation and eventually deflation and severe depression, as soon as
expectations start getting overly sensitive to current changes in profits. After a time, as money
wages decline, and as expected profits to current events get less sensitive, profits start
increasing, which stimulates investment and output. From the time expected profits get more
sensitive to current profits, employment increase strongly, despite the rise in money wages and the rest of the cycle then repeats itself. One thus obtains a Keynesian explanation resting
on both money wage flexibility and expectations which bear some resemblance to MarxianKeynesian such as the one elaborated by Goodwin (1967).
Section 1 of the paper first recall how the emergence of the first endogenous business cycle
model was related to the stability properties of Keynesian macroeconomic equilibrium.
Section 2 describes a two-sector version of Meade’s model and suggests a diagram allowing
addressing graphically the stability issue raised by expectations, once it is allowed that the
expectation elasticity can assume different values. Section 3 reproduces Meade’s local
stability conditions and explores its dynamical structure when money wages are flexible and
expectation elasticity is allowed to vary. The conditions under which endogenous business
cycles occur are derived by utilizing the Poincaré-Bendixon theorem. The final remarks
conclude this paper by comparing this analysis to Keynes’s argument (1936) related to the
business cycle and money wages dynamics.
to see it appear in Eonometrica. I feel rather definitely, however, that before it appears, a
thorough-going overhauling ought to be made, particularly with a view to bringing out the
exact meaning of stability and the assumptions under which your conclusions hold good. May
I suggest that you discuss this matter with Marshak?” (reprinted in Young 1987: 36). Meade
did not however follow Frisch’s instructions. The paper was then not published in
Econometrica.
3
Meade’s paper is also discussed by Darity and Cotrell (1987) and Darity and Young (1995),
who detect strong similarities between Meade’s model and Keynes’s General Theory. They
focus on the fact that Meade’s system contains an account of the nexus between financial
markets and investment. Passing references to Meade’s model are found in MacKay and
Waud (1975), while Young (1987) documents the absence of references to Meade’s work.
These papers do not however centre on the conceptual validity of Meade’s stability analysis.
1. Endogenous Cycles and instability
At the end of the thirties, authors tried to renew endogenous business cycle by reformulating
the General Theory in dynamic terms4. Discussions naturally centred on the stability
properties of Keynes’s macroeconomic equilibrium. It was quite clear that assuming stable
stationary equilibrium implied to admit, because the economy has a natural tendency to reach
its equilibrium state, damped cycles. So, authors were well aware that for obtaining undamped cycle, one must assume unstable stationary equilibrium. The problem was that since
Frisch’s critic of Kalecki’s 1933 model, they knew that in that case, linear model give only
rise to explosive cycles, except for a unique value of the parameter.
In 1940, Kaldor suggested to bypass this difficulty by admitting the economy could be
globally stable and locally unstable. So, even if Keynes’s stationary equilibrium is unstable;
the economy will not necessarily grows or decreases ceaselessly. For reaching this result, it is
enough to assume a multiplicity of equilibrium: one unstable and two stable.
« Hence the economic system can reach stability either at a certain high rate of activity or at a
certain low rate of activity. There will be a certain depression level and a certain prosperity
level at which it offers resistance to further changes in either direction. The key to the
explanation of the Trade Cycle is to be found in the fact that each of these two positions is
stable only in the short period: that as activity continues at either one of these levels, forces
gradually accumulate which sooner or later will render that particular position unstable.”
(Kaldor 1940: 83)
It is by reformulating Kalecki’s 1939 non-linear model Kaldor succeeded in developing a new
business cycle model along these new lines. It is worth here presenting Kaldor’s
reformulation of Kalecki’s model. Kalecki’s main argument was that entrepreneur is cycleconscious, more cautious in his investment decisions after a prolonged boom than at the
beginning of it.
“When things are improving, entrepreneurs become more optimistic about their future, and
the rate of investment decisions increases strongly; but after a certain point doubts begin to
arise as to the stability of this development, optimism ceases to keep pace with boom, and the
rate of investment decisions tends to increase less rapidly. In the slump a symmetrical
development is likely to occur” (ibid., p. 310)
The point was that when activity is low, realized profits do not influence strongly
entrepreneurs’ predictions about their future. Entrepreneurs wait for a significant
improvement before taking new investment decisions. Moreover, if entrepreneurs start
initially with important excess capacity, a long time can elapse before they modify their
expectations. This tends to be reflected in very inelastic expectations. After a certain time,
improvement in activity will strongly ameliorate entrepreneur’s predictions about their future
profits. This glowing optimism will be reflected in very elastic expectations. But after a
certain point, disappointments and doubts begin. Optimism in the sustained strength of
4
Harrod (1936) and Samuelson (1939) started directly from Keynes while Kalecki’s (1937, 1939)
developed a new version of its own business cycle model by only introducing elements of
Keynes’s expectation analysis.
prosperity abates. The rate of investment decisions tends to slow down. Waves of economic
optimism are followed by waves of pessimism so that the situation is now reflected by
inelastic expectations. In the slump, the process is more or less reversed. Current profit’s
weighs on forming businessmen expectation about future profits tend thus to vary during the
cycle.
Graphically, this led Kalecki to describe the dynamics with a non-linear function of
investment decisions. Assuming that the expected profits of each investment are estimated on
the basis of current profits being in a fixed relation with the current national income, he
suggested expressing investment decisions of entrepreneurs, I , for a given stock of capital K
, as a S -shaped function of the current national income Y : I  I Y , K  with IY'  0 and
I K'  0 .
Given the investment function is non-linear with respect to the level of activity Y , three
positions of equilibrium can be defined5. More precisely, this implies 1) IY'  0 when
Y   , 2) IY''  0 when Y    K  and 3) IY''  0 when Y    K  and IY''  0 when
Y    K  ; where  is the depreciation rate and   K  is the level of production for which
the capital stock remains unchanged, that is to say the solution of the equation I Y , K    K
(when Y    K  , K  0 and when Y    K  , K  0 ).
Furthermore, Kalecki neglected the variations in the multiplier. Indicating total real income
of capitalists by  , capitalists’ consumption is related to gross real profits by means of the
linear function : C    C where C is a positive constant (fixed volume of capitalists’
consumption) and  is a positive constant, smaller than 1. Ignoring workers’ saving, workers’
consumption is equal to the total real wages received: Cw  wn (where n is the level of
employment). Finally, assuming that the relative share of wages in real national income is
fixed - determined by the degree of competition of the goods market - global consumption can
be related in a linear form to global output. For since wn Y   where  is a fixed
coefficient, we get:
Yd  Cw  C  I  Y   1    Y  C  I Y , K 
IY'   K  , K      1      1 is a necessary condition for the existence of a multiplicity
of equilibriums.
When Y   , dY d dY     1      1.
When Y  0 , dY d dY     1      1.
Kaldor based its own analysis on different arguments. While there is a “normal” level of the
investment propensity IY' in the midrange of real income, IY' will be relatively small for low
as well as for high values of income compared to the normal level. The decreasing slope of
the function for decreasing levels of income is explained by missing profit opportunities in
times of low economic activity levels relative to the ‘normal’ midrange level. When income is
relatively high, decreasing economies of scale as well as rising financing costs lead to a
smaller propensity to invest out of real income.
5
Yd
Yd Y
C
Yd  I C
B
A
 K
Y
Figure 1 Multiplicity of equilibrium and endogenous cycles
The essence of Kalecki’s model is as follows. At the two extreme equilibrium points, A and
C , dY d dY  1 , and at the equilibrium point B , dY d dY  1 , so that A and C are points of
stable equilibrium and B is a point of unstable equilibrium. The aggregate demand curve is a
short-period function and will shift as capital stock shifts. A rise in capital stock will shift the
aggregate demand curve down and a fall in capital stock will shift it up. Thus if the system is
in equilibrium at C ( Yc    K  so K  0 ) and capital is being accumulated ( I K'  0 ), C is
shifted to the left and A to the right. When B and C coincide the aggregate demand curve
and the 45° line are tangent to each other and the system is unstable in a downward direction
and output falls to A . There the movements in the aggregate demand schedule are in the
opposite direction provided capital is being decumulated ( YA    K  so K  0 ), thus A and
B move together until they coincide and the system is unstable in an upward direction and
employment rises until C is reached again, upon which the process repeats itself. Here the
economy never reach the point Y    K  for which K  0 .
Reviewing Kalecki’s model, Goodwin wrote:
“Particularly important is his introduction of variable coefficients into the system. To the
earlier model, Frisch offered the objection that it was not convincing to assume that the
coefficients had exactly those values which would give undamped cycles. Now, Kalecki
meets the objection by the use of variable coefficients, which make his system necessarily
undamped. The quantitative nature of the variation in the coefficients is not stated, hence,
unfortunately, his proof of no damping is plausible but not rigorous.” (Goodwin: 710)
Although Kalecki did not develop a fully specified mathematical model, his analysis based on
a non linear investment function constitutes a major step in the development of endogenous
dynamics approach. By focussing on the concept of expectation elasticity and centring on the
stability issue of Keynes’s macroeconomic equilibrium, one can show Meade’s model make
early static Keynesian model very close to the first Keynesian endogenous business cycles
models.
2. Stability and expectations in Meade’s Keynesian Model
Meade’s two-sector model with a consumption and an investment good sector sees aggregate
current profit as forming the basis of long-term profit expectations. Like Hicks (1937), Meade
treats wages as a fixed variable in the relevant period. The model also shares with Hicks’s
(1937) model the assumption that the flow of new capital goods is negligible relative to the
existing stock as well as the hypothesis that each sector’s output price is determined by
marginal product factor pricing. Meade’s model departs however from Hicks’s in that it is
based on the equilibrium condition that the marginal efficiency of capital equals the interest
rate (see Keynes, 1936, Chapter 1). Meade’s model can be summarized by the following
eleven equations6:
i  f i ni 
(1)
c  f c nc 
(2)
W  p f i ni 
s
i
'
W  pc f c' nc 
(3)
(4)
Y  pc c  p i
Y    Wn
n  ni  nc
(5)
pid i  sY
E  
pid 
r
pid K
pd K
Lr  
 M  vY  i , L'r  0
Lr 
M  vY
d
s
pi  pi
(8)
s
i
(6)
(7)
(9)
(10)
(11)
The exogenous variables in the model are the money wage rate ( W ), savings rate ( s ), the
capital stock ( K ) and the transactions demand for money coefficient ( v ).
The first two equations indicate that output in each sector depends on the volume of
employment in each sector where i is the output of investment (or capital) goods and c is the
output of wage (or consumer) goods . ni is employment in the investment-goods sector and nc
6
Meade original model contains ten equations in which equation (11) defining the equality
between the investment supply price and the investment demand price is not explicitly
written.
is employment in the wage goods sector. The third and fourth equation are the supply price
equations, establishing that the real wage equals the marginal product of labour where pis is
the supply price of the capital good and pc is the price of the wage good; W is the common
nominal wage rate; and f i ' and f c' are the sectoral marginal products. The fifth and sixth
equation are the equation for nominal national income, the sum of the money value in each
sector, where Y represents nominal GDP with no output produced for the public sector, as
well as the sum of the money wage bill Wn and nominal profits  . The seventh equation
says total employment n is the sum of employment in each of the two sectors.
Meade closes his model with the four remaining equations. The first of these is the
investment-equals-saving equation condition, where s is the marginal propensity to save. The
second of these sets the investment demand price, pid , equal to the present value of the
expected stream of net profits associated with ownership of a capital asset, where future
expected net profits are discounted at the current rate of interest. The expression E 
denotes expected future profits as a function of current profits. In that way Meade’s brings
endogenous expected profits explicitly into the analysis. Equation ten is Meade’s asset market
clearing conditions, where the right hand side is the ratio of the money value of the existing
stock of capital goods to cash balances held for speculative purposes. Agents decide to hold
their wealth in money, financial or real assets. They decide to hold a quantity of money
M Y in their portfolio. The ratio between the value of real assets they hold, p id K , and the
quantity of money held as cash balances is an increasing function of the interest rate where v
denotes the transactions demand for money coefficient. Equation (11) is an equilibrium
condition requiring that the price of the good viewed as output equals the price it can
command as an addition to the capital stock.
The endogenous variables are : i , c , pis , pid , pc , Y , nc , ni , n ,  and r . The exogenous
variables are: M , s , K and  .
Meade assumes that the labour share is constant and equal in both sectors of the economy. For
the sake of clarity, we propose to express both sectoral production functions as Cobb-Douglas
functions i  Ani and c  Bn c where  represents the labour share and A and B are fixed
coefficients.
From (1), we get :
1
 i 
ni   
 A
From (3), it is evident that the investment supply price is equal to:
1
1
1 
s
pi  A Wi 

(12)
(13)
The investment supply price schedule slopes upward in p i - i space. It is shifted upward by an
increase in the nominal wage7.
7
Similarly, the supply price of consumer goods increases proportionately when money wages
increase. It also increases when output expands. Meade defined by  the short run supply
Equation (5) contains an implicit multiplier relationship, so that any rise in nominal
investment expenditure must be associated with a rise in nominal aggregate income, given a
fixed saving rate:
pdi
(14)
Y i
s
Knowing that global nominal profits are equal to:
  1   Y
we get :
pdi
  1    i
s
(15)
(16)
Equation (9) defines the rate of interest as the ratio of expected profits to the investment
demand price:
E  
(17)
r d
pi
When E      , where  
dE   
, denotes the expectation elasticity, we get:
E   d

p id i 


1




s 
r
p id

(18)
or:

i
r  1    p
(19)
 
s
If one assumes that the quantity of money is given, equilibrium in the money market implies:
pdi
pid k
(20)
M  i 

s





1

i

L 1    pid
  
 s  

The expression of the investment demand price results from this equation. When expected
profits are related to current profits, one obtains an investment demand price schedule which
slopes either downward or upward in p i - i space, depending on the value of the expectation
elasticity.
Equilibrium is represented at the intersection of the pis and pid lines (see Figure 1). When the

d  1
i
slope of the pis line is larger than that of the pid , to the left of i * , pid is greater than pis and
elasticity of the capital goods, that is to say, the ratio between the rate of change of production
to the rate of change of the marginal cost:



di dni di
f i ' ni 
1 


''
i d ni di 
f ni  f i ni 

2
investment output increases by the multiplier dynamic. To the right of point i * , pis is greater
than pid and investment output falls. The equilibrium point i * as drawn in Fig. 1 is thus stable.
pid , pis

pis W

pi *
 
pid M

i*
i
Figure 2 Stable Equilibrium
In Figure 2, we see linear pis and pid functions again, but this time the slope of the pid curve
is greater than that of the pis curve. Now, to the left of the equilibrium i * , pis is greater than
pid , thus investment output will contract and the economy will move away from i * . In
contrast, to the right of i * , pid is greater than pis , so output will increase and the economy
will move further to the right of i * . Thus, equilibrium in Fig. 2 is unstable.
pid , pis
 
pid M


pis W

pi *
i
i*
Figure 3 Unstable Equilibrium
Global stability or instability are implied by linear pis and pid curves. What happens if we
assume that the pid curve is non-linear (see Figure 3).
pis W 
C
 
pid M

B
pi *
A
iA
i*
iC
i
Figure 4 Multiplicity of Equilibrium
The non-linear pid curve shown in Figure 3 can be explained by simply recognizing that the
rate of investment will be quite low at extreme investment output levels. Why? One can here
refer to Kalecki’s analysis. At low investment output levels (e.g. at i A ), realized profits do not
influence strongly entrepreneurs’ predictions about their future. Entrepreneurs wait for a
significant improvement before taking new investment decisions. Moreover, if entrepreneurs
start initially with large excess capacity, a long time can elapse before they modify their
expectations. This tends to be reflected in very inelastic expectations and a constant or even
decreasing investment demand price. After a certain time, improvement in activity will
strongly ameliorate entrepreneur’s predictions about future profits. This glowing optimism is
reflected in very elastic expectations and a strongly increasing investment demand price. But
after a certain point, disappointments and doubts begin. Optimism in the sustained strength of
prosperity abates. Thus, the rate of investment will also be relatively low. At output levels
between i A and iC (e.g. at i * ), the rate of investment is higher. Thus, the non-linearity of the
pid curve is wholly justified by the evolution of expectations.
How can this explain cycles? If we are at A, we stay at A. If we are at C, we stay at C. If we
are at B, we will move either to A or C with a slight displacement of the pis and pid line. No
cycles are however apparent. To examine the endogenous cycle brought by changes in the
elastic expectations, one possibility is to examine the evolution of money wages and the
evolution of investment on which they directly impinge.
3.
Meade’s
Model
and
Endogenous
Dynamics:
an
interpretation
Consider the following adjustment mechanism:

i  h pid  pis
W  g n  n



h'  0
(22)
g '  0 when n  n
(23)
where n is the supply of labor (assumed inelastic), and n is the demand for labor across
sectors.
The first equation says that output increases when the investment demand price exceeds the
investment supply price, and embodies the mechanism described by Keynes in chapter 11 in
the General Theory. The second equation says that money wages decrease in response to
unemployment as Keynes envisages it in chapter 19 of the General Theory. The linear
approximation of this system in the neighbourhood of the equilibrium can be formulated by
utilizing the Jacobian matrix:
1
λ
 dp
dp 

h 


di
di


'
d
i
MJ 
1 i
 
s  A 
s
i
A

i
λ
1 
λ
1 

0
The necessary and sufficient conditions for stability are
 dp d dp s
h '  i  i
di
 di
1 i
 
s  A 
where  is the share of wages in all sectors.
1 


  0 (T<0)


1
A
1 


(24)
i

 0 (D>0)
(25)
The determinant is always positive. Analysing the singular point one can conclude that the
stationary equilibrium is stable or unstable accordingly as the slope of the investment demand
price is greater and smaller than the slope of the investment supply curve.
The most interesting point in this dynamic system is thus the ambiguous element
dpid di  dpis di . According to the graphical representation of the preceding section, it is
assumed that this element is positive, which means that the slope of the p id curve is taken to
be greater than the slope of the pis curve - for a normal level of income and second, that it
becomes negative for abnormal high or abnormal low levels of income. The stationary state
equilibrium has a normal level of income. This is illustrated in figure 3 (upper diagram) by the
fact that the normal level of investment output i is unstable, while the extreme values of i are
stable. Mathematically, this means that the trace changes sign during cycles. This is the
negative criterion of Bendixon for limit cycles, i.e. if the trace does not change signs, limit
cycles cannot exist. What is necessary for the existence of cycles is only that the p id curve
has a greater slope than the pis curve at some level of investment output.
To obtain the phase diagram, we must obtain the isoclines di dt  0 and dW dt  0 by
evaluating each differential equation at steady state. The shape of the isocline for di dt  0
depends upon the value of the difference between the slope of the investment demand price
and the investment supply price. As we claimed earlier, for extreme values of i (around i A
and iC ), the investment supply curve is steeper than the investment demand curve, thus
dW di i0  0 , i.e. the isocline is negatively shaped. However, around middle values of i
(around i B ), the investment demand curve is steeper than the investment supply curve, thus
dW di i0  0 , i.e. the isocline is positively shaped. From figure 3, we see in the lower
diagram that at low values of i and high values of i , the isocline is negatively-sloped - this
corresponds to the areas in figure 4 where the investment supply price curve is steeper than
the investment demand price curve (e.g. around i A and iC ). However, around i B , the isocline
is positively-sloped - which corresponds to the areas where the investment demand price
curve is steeper than the investment supply price curve. The off- isocline dynamics are easy,
namely differentiating the differential equation di dt for W : d di dt  dW  0 . Thus, above
the isocline, di dt  0 , so output falls whereas below the isocline, di dt  0 , so output rises.
The directional arrows indicate these tendencies. The isocline for dW dt  0 is vertical.
dW di W 0 is a constant. The off- isocline dynamics are easy, namely differentiating the
differential equation dW dt for i : d dW dt  di  0 . Thus, to the right of the isocline,
dW dt  0 , so money wages increase whereas to the left of the isocline dW dt  0 , so
money wages decrease.
pis W2 
pis W0 
D
C
pis W1 
G
B
E
p id
A
F
iA
i
iC
iB
W
W  0
E
D
W2
B
A
C
W0
W1
F
G
i  0
i
Figure 4. Endogenous cycles
For a given level of W0 , we can find the corresponding equilibria i A , i B and iC
at the
intersection between the investment demand price and the investment supply price p W0  .
Suppose we start at point C so that W begins to rise. Notice that when W0 rises to W2 , the
point iC and i B begin to move together and finally “to merge” at the critical point D ; the
s
i
underlying dynamic (represented by the phase arrows) implies that point D is completely so
there will be a catastrophic jump from D to the lower equilibrium E (which is where A
moved to as W rise from W0 to W2 . Notice that this catastrophic fall in output is driven solely
by the fast multiplier dynamic – the slower - moving dynamic is inoperative as, in moving
from D to E , money wages are constant at W2 . The rest of the story then follows in reverse.
At E , we are at a low output level and thus money wages start decreasing and so money
wages decline from W2 past W0 and then towards W1 . In the meantime, our lower two
equilibria begin to move towards each other and A and B meet and merge at point F at i A .
Now, we must have a strong jump in output from F to G (the high equilibrium to which
point C moved to as money wages fell from W2 to W1 . From G , the process then begin
again as money wages rises at high output levels from W1 to W0 and onto W2 .
4. Final remarks
In this contribution, the contention of the existence of an approach to economic fluctuation in
Meade’s model has been confirmed. Indeed, the non linear model of endogenous fluctuations
we have proposed has shown that in Meade’s approach business cycles can be explained by
expectations (including on the Stock market) of investment. Meade’s model characterization
of investment and the role of expectations capture important features of Keynes’ discussion of
long term expectations. In a context of money and price flexibility, expectations can be a
destabilizing force, as a result of which the system fluctuates endlessly without reaching full
employment.
One must of course recognise that this interpretation departs from Keynes’s analysis in
several respects. Certainly, the major difference between this analysis and the one of chapter
19 of the General Theory dealing with money wages flexibility is that instability is local in
Meade’s model and not global. Keynes seems to say that money wages flexibility is likely to
destabilize the economy and even to lead it to its destruction. Mead’s amended model thus
departs from Keynes. Moreover, this model does not deal with several elements present in
chapter 19, as the liquidity trap or the effects of the debt deflation on investment and
consumption. Finally, in Meade’s model, one must keep in mind that recovery is ultimately
due to the fall in money wages.
References
Assous, M. (2003). Kalecki's Contribution to the Emergence of Endogenous Business Cycle
Theory: an Interpretation of his 1939 Essays. History of Economic Ideas, XI: 109-123.
Assous, M. (2008). Le modèle keynésien de James Meade (1937): stabilité et flexibilité des
salaires”. Cahiers d’économie politique, 54 : 7-26.
Cottrell, A. and Darity, W. (1987). Meade's General Theory Model: a Geometric Reprise. Journal of
Money, Credit and Banking. 19: 210-221.
Darity, W. and Young, W. (1995). IS-LM. An Inquest. History of Political Economy, 27: 1-41.
Goodwin, R. M. (1967). A Growth Cycle, in C.H. Feinstein, editor, Socialism, Capitalism and
Economic Growth. Cambridge: Cambridge University Press.
Harrod, R. F. (1937). Mr. Keynes and Traditional Theory. Econometrica, 5: 74-86.
Hicks, J. R. (1937). Mr. Keynes and the 'Classics' : a Suggested Interpretation. Econometrica, 5:14759.
Kalecki, M. (1939) [1990]. Collected Works of Michal Kalecki, vol. 1: "Capitalism, Business Cycles
and Full Employment". Oxford: Clarendon press.
Keynes, J. M. (1936).The General Theory of Employment, Interest and Money. London: Macmillan.
Lopez, J. and Assous, M. (2009). Michal Kalecki: an Intellectual Biography, Palgrave
Macmillan (forthcoming).
MacKay, R. and Waud, R. N. (1975). A Reexamination of Keynesian Monetary and Fiscal Orthodoxy
in a Two-Sector Keynesian Paradigm. Canadian Journal of Economics, 8: 548-73.
Meade, J. E. (1937). A Simplified Model of Mr. Keynes's System. Review of Economic Studies, 4: 98107.
Phelps-Brown, H.-E. (1937). Report of Oxford Meeting. Econometrica, 5:25-29.
Rappoport, P. (1992). Meade's 'General Theory' model: stability and the role of expectations. Journal
of Money, Credit, and Banking, 24: 356-69.
Tobin, J. (1975). Keynesian models of recession and depression. American Economic Review, 65: 195202.
Young, W. (1987). Interpreting Mr. Keynes: the IS-LM enigma, London: Polity Press.
on « Mr. Kaldor, who has shown that the amplitude of the business cycle is stabilised by the
fact that the influence of incomes on investment and of investment on incomes (the multiplier
effect) vary with the phases of the cycle – being weakest at the top of the boom and the
bottom of the depression. Dr Kalecki’s argument in the volume under review is very similar,
except that he neglects the variations in the multiplier and pins his main argument on the
entrepreneur being cycle-conscious, and hence more cautious in his investment decisions after
a prolonged boom than at the beginning of it” (Scitovsky in Kalecki (1999b) , p. 548)
“Citation de Kalecki concernant les anticipations…”
« This theory [Kalecki’s theory] explains the business cycle in terms of fluctuations of the
marginal return on investment resulting from the accumulation and decumulation of capital
and from the effect of investment on income. […] Professor Schumpeter’s criticism that
flexibility of interest rates must stop the Kalecki mechanism (vol I, p. 188) is not justified.
Such flexibility presupposes a monetary system different from the actual one, and even if that
be granted, uncertainty and elastic expectations may foil the effect.” (O. Lange, p. 193)
Kaldor and Kalecki tried to reformulate the General Theory in dynamic terms. Keynes
phrased the problem in a static form; a given level of investment determines effective
demand, output and hence employment. But investment is not given, nor is it solely
determined by the marginal efficiency of capital (MEC) or the interest rate.
2. Stability and expectations in Meade’s Model
Meade pose la question de la stabilité dans un modèle centré sur les interactions entre les
marches des biens et de la monnaie.
3. Extension of Meade’s Model and Endogenous Dynamics:
Concluding remarks
Meade/ Kalecki
Kalecki admet que les anticipations de long terme sont susceptibles de varier au cours du
cycle. C’est le cas de Meade.
La différence entre Meade et Kalecki est que l’élasticité d’anticipation varie au cours du
cycle. Ce cas n’est pas envisagée par Meade.