Macroeconomics of Keynesian and Marxian

inspirations: Toward a synthesis

Gerard DUMENIL
and Dominique LEVY
CNRS

Address all mail to: PSE-CNRS, 48 bd Jourdan, 75014 Paris, France.

Tel: 33 1 43 13 62 62, Fax: 33 1 43 13 62 59
E-mail: dominique.levy@ens.fr, gerard.dumenil@u-paris10.fr
Web Site: http://www.jourdan.ens.fr/levy/

1 - Main results and outline

In the wake of World War II, Keynes analysis contributed to
the definition of the new social compromise (including its macro and
welfare components). Within this favorable political context, a Keynesian school, in the broad sense, prospered. After the establishment
of neoliberalism in the 1980s, despite the overall repression of economists critical of the new social order, the Keynesian perspective is
still the object of much interest among the minority of economists
politically leaning to the Left. The relationship between Keynesian
and Marxian economics has always been ambiguous, but there is a
lot in common concerning the macroeconomics.
During the last few decades, one of the fields of our research has
been the analysis of business-cycle fluctuations, theory and empirics.
We understood that Marxs perspective on this issue was macroeconomic as expressed in the phrase crises of general overproduction
general being here the term to be emphasized as opposed to
disproportions among industries. On such grounds, the encounter
with Keynes was unescapable. This is one of the numerous interests we share, we believe, with Duncan Foley to which this volume is
dedicated.
The object of the present study is the investigation of these common Marxian-Keynesian grounds. The Keynesian inspiration refers
to a broad set of approaches, major figures of the past such as Michal
Kalecki, Joan Robinson, or Hyman Minsky, and the contemporary
postKeynesian school. Conversely, by a Marxian inspiration, we
only mean here our own understanding of Marxs analysis, to be
later specified, not the entire set of readings of Marxs crisis theory
by contemporary Marxist economists or in the past. The final ambition is factual analysis: business-cycle fluctuations within sophisticated capitalist economies after the establishment of central banks
and the conduct of macro policies. (This later feature, obviously,
creates some distance with Marxs original approach.)
The purpose of section 2 is the introduction of the overall framework of analysis, but abstracting from money and finance. A broad
typology of theoretical fields is established depending on time frames
and the choice of a multi-industry or macro perspective. Monetary
and financial mechanisms open a second field in section 3. We introduce the framework of co-determination, in which the behaviors

Macroeconomics of Keynesian and Marxian inspirations

of nonfinancial and financial agents concerning money and credit are

jointly considered within single functions. The contrast is sharp between the postKeynesian view of accommodative money and this
new perspective in which the action of central monetary authorities
plays a crucial role in the taming of an otherwise unstable macroeconomy. Section 5 harks back to the perspective of a multi-industry
economy. The section introduces the important thesis that capitalism
is rather efficient in the allocation of capital among industries and,
correspondingly, the production of goods or services in proportions
conform to the pattern of demand but always on the verge of macro
instability (departure toward overheating or recession). These distinct properties provide the theoretical foundations for the adoption
of a macro viewpoint in the modeling of business-cycle fluctuations.
The perspective in section 6 is historical. A link is, first, established
between profit rates and macro stability. Then, our thesis concerning
the tendential instability in capitalism is introduced. Last, section 7 discusses the components of such fluctuations in various time
frames in the U.S. manufacturing sector.
Due to space limitation, two sections of an earlier version of
this study are located in appendices. Appendix A is devoted to the
modeling of a monetary macroeconomy in the short and long terms1 .
Appendix B discusses Minskys financial instability.2

2 - Common and distinct grounds

We first recall our framework of disequilibrium microeconomics
and general disequilibrium models. A typology of alternative theoretical fields is, then, introduced and a link established with the
Marxian and (post)Keynesian perspectives.

2.1 Disequilibrium microeconomics

Heterodox economists reject mainstream economics on account
of their methodology and politics. Optimality properties are used
1. http://www.jourdan.ens.fr/levy/dle2011ga.pdf.
2. http://www.jourdan.ens.fr/levy/dle2011gb.pdf.

Macroeconomics of Keynesian and Marxian inspirations

by the mainstream as an alleged justification of market economies,

another name for capitalism. The problem is not, however, the fact
of accounting for individual behaviors, but what we call equilibrium
microeconomics, that is, maximizing under the assumption of ex
ante equilibrium. To the contrary, we believe decisions are made in
situations of disequilibrium and radical uncertainty (as opposed
to objective probabilities with known distributions). For example,
enterprises are not sure of selling the goods produced and do not
know the probability distribution of demand. They adapt to the
observation of disequilibrium, as in the accumulation of inventories
of unsold commodities:




Modification of
Evidence of

disequilibrium

behavior

This reaction contributes to a correction of the disequilibrium

observed, but it does not result in the immediate return to equilibrium. This outcome can only be achieved progressively as a result of a
gradual and conditional process of adjustment. Smiths metaphor of
the invisible hand points to the gradual collective outcomes of such
individual behaviors not to each movement. This approach to decision making must be extended to all aspects of economic activity
outputs, prices, investments but also decisions to borrow and lend,
including the action of lending institutions and the central bank.
These processes must be studied within dynamic models. As
in other disciplines, such as physics, an equilibrium is the possible
outcome of an actual dynamic process, it is the fixed point of such
a process. Conversely, within mainstream economics, equilibria are
defined independently of any dynamic process. Disequilibrium dynamics are, at best, viewed as inessential developments. When equilibrium is approached as the fixed point of a dynamic process, the
issue of (in)stability comes to the fore. An equilibrium can be stable or unstable; stability is always conditional. In our opinion, the
consideration of an equilibrium locally unstable can be economically
relevant, and the stability conditions are susceptible of economic interpretation.
An institutional framework must be defined in which a list of
agents is established as well as the types of mechanisms through which
they interact. Such models can be called general disequilibrium models, though the ambition must be limited if an analytical treatment is
sought. A common simplifying assumption is the consideration of a
single commodity, as in a macro model. Another option, also typical

Macroeconomics of Keynesian and Marxian inspirations

of Keynesian models, is to assume that enterprises produce exactly

what is demanded. Such assumptions are very helpful, but their consequences must be carefully assessed, and the investigation of more
complex models remains necessary.

2.2 Toward the distinction between theoretical fields

Due to the complexity of the mechanisms under investigation,
economic analysis is conducted within distinct theoretical fields
sets of notions and mechanisms defined by implicit or explicit abstraction. It is assumed that the quantitative relationships between
a number of variables are weak and can be neglected. Implicitly or
explicitly, two basic distinctions play a central role within economic
theory:
1. Proportions and dimension. In our terminology, proportions
refer the relative values of variables among industries or enterprises,
for example, the relative stocks of capital or prices. Dimension
refers to the general or average values of the variables, for example,
total output as in a macro model.3
2. Time frames. The distinction between various time frames is based
on the hypothesis that one set of variables is involved within movements faster than another set of variables. For example, production
may be considered a fast variable, and the stock of fixed capital, a
slow variable. In a similar manner, a given variable can be broken
down into a short-term component (accounting for its fluctuations)
and a long-term component (that is, a trend, around which the fluctuations are observed).
In the investigation of properties related to a long term, one
can assume that the short-term dynamics have converged, that is,
assume that short-term equilibrium prevails. This is the method of
temporary equilibrium. Long-term dynamics are approached as the
succession of temporary short-term equilibria. Thus, the dynamic
properties of the model can be investigated in two successive steps:
3. Keynes is very conscious of this distinction and criticizes the almost
exclusive emphasis on proportions in what he calls classical economics:
Most treatises on the theory of Value and Production are primarily concerned with the distribution of a given volume of employed resources between various uses and with the conditions which, assuming the employment of this quantity of resources, determine their relative rewards and the
relative values of their products. (J.M. Keynes, The General Theory of
Employment, Interest and Money, London: Macmillan (1936), p. 4).

Macroeconomics of Keynesian and Marxian inspirations

(1) short-term equilibrium and its stability; and (2) long-term equilibrium and its stability.
We call historical term (or term of historical tendencies) a
further time horizon in which the historical trends of distribution and
technology, as well as the historical transformation of institutions, are
considered.

2.3 A taxonomy: Knitting Marxian and Keynesian

perspectives
The distinctions between, on the one hand, proportions and dimension, and, on the other hand, short, long, and historical terms
allow for a taxonomy of economic theories.
Short term Long term
(production) (growth)
Dimension
(macro)
Proportions
(relative values)

[1]

[3]

[2]

[4]

Historical term
(historical dynamics)
[5]

Five basic configurations are discussed here:

- Configuration [1] in the table defines dimension in the short term.
The economy is considered globally, and the stock of capital is
constant. Thus, abstraction is made of accumulation and growth.
Proportions are set aside. This is the field of Keynesian macroeconomics and the short-term equilibrium can, consequently, be
called a Keynesian equilibrium.
- Configuration [2] refers to proportions in the short term. It can
be understood as the extension to various industries of configuration [1]. A given pattern of relative inter-industry capacity
utilization rates allows for the equality between outputs and demands.
- Configuration [3], dimension in the long term, is the field of
postKeynesian economics. In a postKeynesian trajectory (or
traverse)4 , a sequence of Keynesian short-term equilibria con4. M. Lavoie, P. Ramirez-Gaston, Traverse in a Two-Sector Kaleckian
Model of Growth with Target Return Pricing, University of Ottawa, Ottawa (1993); A. Dutt, Convergence and Equilibrium in Two-Sector Models of Growth, Distribution and Prices, Zeitschrift f
ur National
okonomie,
Journal of Economics, 48 (1988), p. 135 -158.

Macroeconomics of Keynesian and Marxian inspirations

verges toward a steady state in which resources are not fully

used.
- Configuration [4], proportions in the long term, defines the Classical-Marxian field of the formation of production prices within
competition. Considering the average values of the variables, one
can also define a Classical-Marxian long term macro configuration, as in [3], evocative of the postKeynesian long-term equilibrium, with the important difference that the Classical economists and Marx assume that capacity utilization rates converge
to normal (or target) values. (More rigorously, they abstract from business-cycle fluctuations.)
- In the third column, only problems of dimension [5] remain,
as proportions have been determined within shorter terms. In
this context, dimension refers to the historical tendencies of the
growth rate and the profit rate, and the trajectory of technical
change (notably, the ratios between inputs such as the capitallabor ratio, or ratios between outputs and inputs such as productivities).
A given school may refer one particular or various configurations. The field of Marxs analysis of competition is [4]. Involved
are only the relative values of prices, capital stocks, outputs. The
problem of dimension the mechanisms by which the rate at which
productive capacities are used converge to normal values is assumed away, and the business cycle is not discussed. Marxs analysis
of the business cycle and crisis of general overproduction is clearly
macro as in [1], that is, abstracting from proportions. Although
Marx refers to partial crises within specific industries5 , he strongly
rejects analyses of capitalist crises in terms of disproportions. Marxs
study of historical tendencies belongs to [5]. Keynesian economists
do not distinguish between the long term and the historical term, and
abstraction is made of proportions. Thus, only cases [1] and [3] are
considered. Keynes work belongs to [1], dimension in the short term.
The field of postKeynesian economists combines [1] and [3]. Temporary equilibria are defined in [1] and the sequence of such equilibria
in the long term belongs to [3].
A priori, each configuration in the table can be considered from
the viewpoints of equilibrium and disequilibrium dynamics (including
stability conditions). Keynes analysis in [1] focuses on equilibrium,
5. K. Marx, Theories of Surplus Value, Moscow: Progress Publisher (1862),
p. 620-621.

Macroeconomics of Keynesian and Marxian inspirations

while Kaleckis analysis belongs to [1] and [3], with an emphasis on

dynamics. Concerning postKeynesian economists, only equilibrium
is considered in the short term, as in [1], while both equilibrium and
disequilibrium dynamics are considered in the long term, as in [3].
The interest for the disequilibrium dynamics around short-term
equilibrium defines an important difference between our approach
and the traditional (post)Keynesian perspective. This difficulty harks
back to a basic ambiguity within the Keynesian paradigm, beginning
with Keynes himself. It is not clear whether Keynesian equilibria are
supposed to explain durable shifts or sudden collapses in the general
level of activity (the 1920s in England or the Great Depression in the
U.S., for example). The core Keynesian analytical device seems to
account for durable shifts, whereas Keynes analysis of the business
cycle6 emphasizes the extreme volatility of the marginal efficiency of
capital, which determines investment and the level of output.

2.4 The components of the general level

of activity
The framework in the previous sections has important implications concerning the analysis of the fluctuations of the general level
of activity. Various time frames must be distinguished. This is illustrated in Figure 1. The variable is the capacity utilization rate within
the U.S. Manufacturing sector, broken down into three components,
uHT , uLT , and uCT , which sum up to u:
) defines a first
- The segment from zero to the continuous line (
component, uHT , with a slow downward trend since the 1960s.
Its interpretation is difficult (section 7). For brevity, it can be
denoted as the historical component, but something else than
historical tendencies in Marxs sense is implied, maybe a mere
figment of the data.
- The the long-term component, uLT , is the difference between the
) of u and the above first component
long-term fluctuation (
(
). It reveals broad fluctuations during significant periods of
time, about 10 years long. It can be approached as a sequence
of temporary equilibria.
- A short-term component is manifest in the distance between
). The sharp deu ( ......... ) and the long-term fluctuation (
partures typical of this component are the expression of the
(in)stability of short-term equilibrium.
6. J.M. Keynes, The General Theory, op. cit. note 3,

ch. 22.

Macroeconomics of Keynesian and Marxian inspirations

Figure 1

92
88
84
80
76
72
68
64

The capacity utilization rate in the U.S. Manufacturing

sector

..
..
.
.
.
... ..... . ...u.... ....
.. .
.. ...... ... .. .... u . . . ...... ...... ........
. . ... ..... .. ..... . .. ... ... ........ . ...... ..
..
. ..... .
. .. .... .. .. .
..
.
.
.. .. ..
.
... . u ... ... .
..
..... . ...
. ....
..
.
.
.
.
and
: 48.1 07.2 .
.....: 48.1 11.4
.....
....
..............
...
.. ..
..
.... .. ... ST
... .... .
.. .. ..
.
... .. .. ..
.... ... ..............
...... ... ... .........
.......
. . .... . .
.. ..
.. .. .. .. . ...
.. ...
......................... ....
.
... ... ... ... .... ............
....
.. .. . ... . ..
... .
.....
.... . .. LT..... ... ..
....
.
.. ...... . .. .. .. . .
...
... . ...................................................................................................................... . . . ................. . ..... . . ............ . . . . . . . . . . .......... . . . . . . ........ ....... . . . . . . . . . . . . . . . . . .
. ...
..
....................... .. ... ..... ......... .... .. ...................................... ... ..... ...
.
. ..... .................... ....
.. .. ................ .. . .. .
............... .. ..
.
.. .....
.
.
.
.
.
.
.
.. .
.. .. .. ....... ...
....
.. .. ...........................
... .
.... .. ........ ..
...
.. .. .. ..... ......................... ............... ....... ..................
....... ....
....
.. .. . .. ..
..
............. ... .
.. .. .. ... .... ..
..
.. ..
....
.. .. .. ..
.
.. ..
..... ......... ................................................... ....
..
.. .. ........
.. ..
.
.
.
.
.
.
.
............... ................
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.. ..
......... .. ........
.
....
.. ...
.. ... .. ..
.
.
.
..... ............... ..
.
.
.
.. ..
.
.
..
.
.
.
.
.. ..
.. .. .. ..
......
..... ..........
.... ..
.
.
.
.
.
.. ..
.
.
.
.
.
.
.
.
.
.
... .. ..
.. .......... ..
.. ...
.. ... ....
.
.. ..
.. .. ..
...
.
..
.. ..
.... ....
..
.. ..
..
.. ...
.. ....
.. ..
.... ....
.. ...
.. ..
.. ..
.. ..
....
.... ...
..
.. ..
.. ..
.. ..
.. .....
....
....
.. ..
.
......
.. ..
.. ....
....
..
.
.. ...
..
.
.. ..
....
...
.
.. ..
...
.. .
...
.
.. ..
.. ..
..
.. ..
....
.. ..
..
....
.. ..
..
...
.. ..
..
.
.. ..
..
.. ..
..
.. ..
..
....
....
..
...
..
.
..

1950

1960
LT

1970
ST

1980

1990

u (= u + u + u ):
Sequence of short term equilibria (= u + uLT ):
Sequence of long-term equilibria (= u): (
)
Average value of u before 1970: ( ......... )

2000

2010

( .... .... )

In 2009, Manufacturing industries only accounted for 10 percent of total

U.S. income. The fluctuations of the manufacturing capacity utilization
rate, u, are, however, tightly correlated to those of the value added of
the Nonfinancial corporate sector. The purpose of the figure is merely
illustrative. The two lines (
and
) have been determined using
Whittaker filter. Both have been calculated to the second quarter of 2007,
that is, prior to the current crisis. More time will be necessary to determine
new tendencies.

In this paper, we denote as business-cycle fluctuations the two latter categories of fluctuations above, the short-term and long-term
components. Interpretations are suggested in section 7.

3 - Money and credit in the macroeconomy

The present section is devoted monetary and credit mechanisms,
of which abstraction is made in the previous section. Financial
mechanisms proper, in a broader sense, are treated in appendix B.

Macroeconomics of Keynesian and Marxian inspirations

3.1 The credit demand channel

Monetary mechanisms are approached here in a simple framework, in which banks make loans to the nonfinancial agents
households, the government, and enterprises from which demand
emanates, assuming the traditional confrontation between the lender
and the borrower (households and enterprises), as in the past, still
typical of procedures in many countries. (Money and credit must be
treated jointly, as the two facets of a same coin, since money is issued
when loans are granted, and destroyed when loans are paid back.) An
important assumption is that demand is financed out of previously
garnered income and borrowing:
 
 


Borrowing
Demand
Income
+
=
in period t + 1
in period t
in period t + 1
Other monetary actions are also involved with equivalent impacts
on demand. They must be jointly considered with borrowing in the
determination of financing:
Borrowing (1)
Paying back loans (2)

Depositing on bank accounts (3)

Using these deposits for demand (4)

A household, for example, can finance more than its income by drawing on its bank account; it can also use the money deposited at the
bank to pay back loans. And the converse will happen if demand is
inferior to income. The outcome of such monetary-credit flows is the
variation of the net debt (borrowing for short).
From the viewpoint of real flows, the variation of the net debt
of an economic agent is the difference between its total spendings
(consumption and investment) and its income. Equivalently, the opposite of the variation of the net debt is equal to financial savings,
as opposed to savings, the difference between income and spendings.
With the notation D for demand, Y for income, and N for the
net debt, one has:
Dt+1 = Yt + Nt+1

(1)

This approach to credit mechanisms can be used to account for the

formation of the demand of specific agents within models in which
several goods are considered, as well as within a macro model in
which a single nonfinancial agent is at the origin of demand. We
consider this equation the cornerstone of the modeling of a monetary
economy.

10

Macroeconomics of Keynesian and Marxian inspirations

A first remark must be made concerning the aggregate net debt.

If nonfinancial and financial agents are simultaneously considered,
the variation of the aggregate net debt is null since the loans of financial agents are identically equal to the borrowing of nonfinancial
agents, and the deposits of nonfinancial agents are held by financial
agents. Conversely, considering only nonfinancial agents, the variation of their aggregate net debt is a priori different from zero.
A second remark concerns the distinction between various nonfinancial agents. Borrowing (the variation of the net debt) may affect
to different degrees the demands of distinct agents. For example,
the demand of capitalist investors could be larger than their income,
while wage-earners would consume less than their income. We call
credit macro demand channel the overall impact of credit mechanisms on aggregate demand independently of its distribution among
agents.

3.2 Built-in instability in a credit economy

The financing of demand via the credit channel opens a field of
analysis of major import. Like Marx and Kalecki, we believe that
the propensity to borrow of economic agents in a capitalist economy
is procyclical. When the economy is booming, economic agents are
prone to borrow more, and conversely when the economy is depressed.
If the aggregate net borrowing of nonfinancial agents is procyclical, short-term equilibrium is necessarily unstable, with cumulative
movements upward or downward. Notably, the tendency on the part
of enterprises to borrow when the economy is growing comparatively
fast in order to expand their investment has destabilizing effects. The
same is true of households borrowing for residential investment.7 The
converse is true in a recession, with cumulative movements downward. Thus, the unrefrained propensity to borrow (or the unchecked
propensity to diminish expenses and pay back debts) would result in
a built-in instability inherent in capitalism.8
This built-in instability renders necessary the stabilizing countercyclical action of a segment of financial institutions beyond the
narrow profit motive. Private or government central institutions are
7. Consumption might increase less than output, with stabilizing effects.
Empirical investigation is required here.
8. G. Dumenil, D. Levy, Modeling monetary macroeconomics: Kalecki
reconsidered, Metroeconomica, 63 (2012), p. 170 -199.

Macroeconomics of Keynesian and Marxian inspirations

11

involved in such practices. In the United States, under the National

banking system, such functions were performed by the large banks
of New York in the context of the gold standard.9 Since the creation of the Federal Reserve in 1913, and its more active action after
World War II, this management of the macroeconomy is the object
of monetary policy, supplemented by fiscal policy.

3.3 Co-determination
A central aspect of the postKeynesian analysis is the endogenous or accommodative character of the issuance of money. Lenders
(commercial banks for short) and central banks could only accommodate the demands for loans. Symmetrically, the opposite assumption concerning the exogenous character of money, in which monetary authorities determine the amount of credit and money, is harshly
criticized. We reject this dilemma. Money is neither endogenous nor
exogenous, but is co-determined a third viewpoint.
We call co-determination the general twofold principle stating that: (1) in the analysis of the formation of demand, real and
monetary variables must be simultaneously determined; and, (2) less
trivially, the actions of nonfinancial and financial agents are jointly
involved. The behavior of households responds to their eagerness to
buy, the level of their income, their monetary holdings, and the cost
at which funds can be borrowed. Lenders typically assess risks; they
can set an upper limit to the amounts borrowed or straightforwardly
deny lending; they also take account of their own capability to lend
and obtain refinancing on the interbank market, given the policies
of the central bank and the opportunity to refinance opened by the
securitization of loans; they are constrained by existing regulations;
but the profit motive stimulates their propensity to lend. The central
bank has multiple objectives such as managing the macroeconomy, in
particular taming inflationary pressures, ensuring the smooth functioning of financial institutions, and limiting the degrees of indebtedness and the risks of financial crisis. One lever in the conduct of
such policies is the modification of the interest rate at which banks
are refinanced, but various balance-sheet ratios and regulations are
also involved. The variations of net debts are the outcomes of such
mechanisms.
9. G. Dumenil, D. Levy, Capital Resurgent. Roots of the Neoliberal Revolution, Harvard: Harvard University Press (2004), Box 18.1.

12

Macroeconomics of Keynesian and Marxian inspirations

The determination of the flows of new loans is traditionally approached as the confrontation of two functions, a (potentially vertical) supply curve for loans and a demand curve for borrowing. Conversely, we model the variation of the net debt as in equations 1 or 3
by a single function accounting for all aspects of monetary and credit
mechanisms, the co-determined monetary function, N = F . A
simple expression for F is, for example:
Ft+1 = + Yt Nt jt

(2)

Besides the constant , the term Y simultaneously accounts for the

determination of borrowers, given their income and the tendency of
commercial banks to accommodate this demand when the economy is
booming (and symmetrically within opposite situations). The term
N accounts for the concerns vis-`a-vis the levels of indebtedness on
the part of both borrowers and lenders. The last term, j, describes
the aversion of the central bank for inflation and the consequences
of its policy concerning the refinancing of banks. A similar function
can be written making explicit the role of the interest rate, in turn
manipulated by the central bank.
To the countercyclical action of central monetary authorities, one
must add fiscal policy on the part of the government. As any other
agent, the government may borrow to spend more than its revenue
(and symmetrically, may pay back its debt if its revenue is larger
than its spending). Contrary to other agents, however, this behavior
can be countercyclical, as part of a deliberate action to stabilize the
macroeconomy or due to the stickiness of expenses (as in built-in
stabilizers).
Concerning the capability of central monetary institutions and
the government to influence money, credit, and spendings, two aspects must be combined: (1) These institutions strongly impact the
macroeconomy and, in the absence of this action, the macroeconomy
would be unstable; (2) The efficacy of these actions is conditional, not
always and unambiguously ensured. Government demand policy is
specifically important when the channels of monetary policy become
inefficient (for example, in a credit crunch during a crisis of financial
institutions).

3.4 Modeling a monetary macroeconomy

Appendix A introduces a number of models of a monetary macroeconomy, built along the lines introduced in the previous sections. The
main results are as follows:

Macroeconomics of Keynesian and Marxian inspirations

13

- Concerning short-term equilibria, we believe that their stability

is always subject to conditions and that these conditions are susceptible of economic interpretation. As no decentralized mechanisms account for the stability of the macroeconomy, otherwise
autonomously unstable, the action of the central bank is crucial.
But this action is also subject to significant limitations. Notably, perturbations on financial markets may recurrently destabilize the macroeconomy. Overheatings and recessions are always around the corner. Overall, the stabilizing action of the
central bank confines the deviations of the general levels of activity within certain limits. More than constantly maintained,
stability is recurrently restored.
- Concerning long-term equilibria, assuming the preservation of
the stability of short-term equilibria, long-term equilibria remain
attractors around which the rather hectic short-term oscillations
are observed. The problem concerning long-term equilibrium
is the management of its level. It is a slow process, in which
institutional transformations are implied. Thus the gravitation
of short-term equilibria around a normal value of the capacity
utilization rate is constantly shifted, as the macroeconomy is
involved in slow long-term fluctuations.

4 - PostKeynesian views concerning money

and credit
A first section briefly recalls Kaleckis emphasis on the use of borrowing to finance investment to which we devoted a specific study.10
The two other sections compare our views to the postKeynesian approach of money as endogenous.

4.1 The financing of investment in Kaleckis

framework
The financing of spending by borrowing, as introduced in section
3.1, is reminiscent of Kalecki statements concerning financing. The
10. G. Dumenil, D. Levy, Modeling monetary macroeconomics, op. cit.
note 8.

14

Macroeconomics of Keynesian and Marxian inspirations

view that the expansion of loans (credit inflation) can be used to

finance investment is repeatedly stated by Kalecki. Loans are made
to capitalists and finance their investment beyond their savings11 :
How can capitalists invest more than remains from their
current profits after spending part of them for personal consumption? This is made possible by the banking system in
various forms of credit inflation.12
Thus, within Kaleckis perspective, our equation 1 translates into:
It+1 = Savingst + Nt+1

(3)

4.2 Exogenous money, endogenous money,

and co-determination
To the eyes of most postKeynesian economists, money is endogenously determined, meaning that the financial system always
provides the loans demanded by nonfinancial agents. Thus, the absence of monetary variables within models is, sometimes, presented
as a deliberate option. In the words of Basil Moore: The financial
system has no choice but to accommodate13 . Moore bases his assessment on the observation that commercial banks are enterprises,
similar to other enterprises, responding to the demand of their customers. They are price setters and quantity takers.14 By lack
of capability or willingness, central banks accommodate the demands
for refinancing.
A few postKeynesian economists contend that, as in our approach of co-determination, the power is shared between nonfinancial
and financial agents. In Mark Lavoies formulation: Other post Keynesians15 prefer to recognize that the stock of money has endogenous
11. The quotation is made by Malcom Sawyer (M. Sawyer, Kalecki and
Finance, European Journal of the History of Economic Thought, 8 (2001),
p. 487 -508).
12. M. Kalecki, Capitalism: Business Cycles and Full Employment, Collected Works of Michal Kalecki, Vol. I (Edited by J. Osiatynski), Oxford:
Clarendon Press (1990), p. 148.
13. B.J. Moore, The Endogenous Money Stock, Journal of Post Keynesian Economics, 2 (1979), p. 49 -70, p. 58.
14. B.J. Moore, The Endogenous Money Supply, Journal of Post Keynesian Economics, X (1988), p. 372 -385, p. 381.
15. With reference to S. Weintraub, Capitalisms Inflation and Unemployment Crisis, Reading: Addison-Wesley (1978) and P. Davidson, The DualFaceted Keynesian Revolution, Journal of Post Keynesian Economics, 2
(1980), p. 291 -307.

Macroeconomics of Keynesian and Marxian inspirations

15

as well as exogenous aspects.16

Money endogeneity can also be questioned in reference to a Keynesian liquidity constraint. Foley emphasizes this mechanism in
relation to the interconnectedness of transactions inherent in market
economies, described in terms of externalities:
Keynes insisted by contrast that in real-world monetary
economies many households and firms are liquidity-constrained, that is, unable to finance spending beyond their
immediate cash inflows by borrowing. In this more realistic
world, spending itself has an important externality. Each
household or firm that spends money not only accomplishes
its own ends (consumption or production), but also relieves
the liquidity constraint of other spending units.17

4.3 Financing channels within postKeynesian models

The treatment of financing channels is complex, as stocks and
flows must be carefully articulated within accounting frameworks.
This is manifest in the works of Marxist economists concerning Marxs
analysis of the circulation of capital in Volume II of Capital.18 A
number of postKeynesian economists use transactions flow matrices and Balance sheet matrices.19
The approach of postKeynesian economists is quite distinct from
ours (as introduced in the previous sections):
1. Consumption and investment functions. Within postKeynesian
models, traditional consumption and investment functions are defined
prior to the consideration of financing.
2. Intermediation, and credit and deposits as residuals. Concerning
enterprises or the government, a first set of external channels is the
16. M. Lavoie, The Endogenous Flow of Credit and the Post Keynesian Theory of Money, Journal of Economic Issues, 18 (1984), p. 771 -797,
p. 776.
17. D. Foley, The Anatomy of Financial and Economic Crisis, New School
University, New York (2009), p. 15.

18. G. Dumenil, Marx et Keynes face `

a la crise, Paris: Economica
(1977),
ch. II, Lelaboration des schemas: Flux et stocks, and ch. III, Lelaboration
des schemas: Les mecanismes du credit; D. Foley, Understanding Capital,
Marxs Economic Theory, Cambridge: Harvard University Press (1986),
ch. 5, The reproduction of capital.
19. W. Godley, M. Lavoie, Monetary Economics. An Integrated Approach
to Credit, Money, Income, Production and Wealth, London: Palgrave
(2007).

16

Macroeconomics of Keynesian and Marxian inspirations

collect of funds by the issuance of securities such as stock shares,

bonds, and the like. (For example, a given percentage of investment
by an enterprise is financed by the issuance of shares.) These channels
share the common property of being mechanisms of intermediation in
which existing purchasing powers are transferred, as opposed to the
issuance of money as in bank loans in which purchasing power is created. As in portfolio theory, households determine the proportions of
their monetary and financial holdings. (For example, a given percentage of their financial assets is held as stock shares.) The equilibrium
on financial markets is ensured by the prevalence of prices that clear
each of these markets. Loans and deposits are treated as residual
variables allowing for the consistency of spending plans (otherwise
determined) with the financing made available by the above channels of intermediation. Since the credit system accommodates the
demands of borrowers, and since no decisions are made concerning
deposits, this consistency is guaranteed by construction.
3. The prices of financial assets. Correspondingly, the focus of
postKeynesian frameworks is more on the price of financial assets
rather than the amount of loans. For example, the price of fixed capital is compared to the price of stock shares, as in Tobins q. This is
also an important aspect of Minskys analysis of business cycles.20
Note that we do not question the importance of financing procedures distinct from the credit channel. Their potential destabilizing
impact is obvious. Within sophisticated monetary and financial economies, the multiplicity of financing channels renders the action of the
central bank even more difficult.

5 - Articulating theoretical fields

Vindicating macroeconomics
The present section focuses on general disequilibrium models, in
which several industries are considered and other assumptions lifted.
Only general principles are introduced, and the emphasis is on the
main results. The objective is to show how the four first theoretical
20. H. Minsky, Financial Instability Hypothesis, in P. Arestis, M. Sawyer
(eds.), The Elgar companion to radical political economy, Aldershot: Edward Elgar (1994), p. 153 -157, p. 143.

Macroeconomics of Keynesian and Marxian inspirations

17

fields in the table of configurations in section 2.3 (short and long

terms, and proportions and dimension) can be articulated within a
unique coherent framework. Important conclusions follow, notably
capitalist economies are stable in proportions, and at the limit of
(in)stability in dimension, a thesis to be made explicit in section 5.2.
The third section briefly opens the broad field of the relationship
between the historical term (the term of tendencies) and the business
cycle.

5.1 Production and investment

This section introduces the main mechanisms typical of a general
disequilibrium model:
1. The decision to produce. Production is decided and realized within
enterprises, and this is also where investments are made. Production
is decided before demand is known, and supplies and demands are
confronted on markets. Inventories of unsold commodities may exist
(or rationings). These inventories, St , are transferred to the next
market as part of supply, St + Yt 21 :
St+1 = St + Yt Dt
2. Accumulation and growth. Capital invested during one period
becomes part of the stock of capital, K, in the following period.
Instead of variables such as I, Y , or S, it is more convenient to use
ratios to the stock of fixed capital, K (or to productive capacities,
Y Max ), such as = I/K, u = Y /Y Max , or s = S/Y Max . We define u
and s, the values of such ratios, as targeted by enterprises.
3. Main behaviors. We use adjustment to disequilibrium as in section 2.1. For simplicity, in the equations below, abstraction is made
of the superscript i indicating that one particular agent is considered.
When inventories are smaller than targeted by enterprises (s < s),
they increase their capacity utilization rate. To a smaller extent, they
also raise prices. When capacity utilization rates are large (u > u),
enterprises slowly increase their prices. They also tend to invest more
(that is, = I/K rises). When the profit rate in one enterprise or
industry is larger than the average, the difference stimulates investment in this enterprise or industry. Since enterprises increase their
21. The case of services requires a specific treatment from which we abstract here.

18

Macroeconomics of Keynesian and Marxian inspirations

investment when capacity utilization rates are large, they simultaneously initiate a return toward the normal use of productive capacities.
(This twofold behavior aims at the adjustment of productive capacities in line with demand.) For example, the decision to produce (the
determination of the capacity utilization rate) can be modeled as the
adjustment of ut in reaction to two disequilibria:
ut+1 = ut (st s) (ut u)

(4)

Other behaviors must be modeled in a similar manner.

To obtain a full-fledged general disequilibrium model, the consumption and investment functions must be expressed, in conformity
with the principle of co-determination also accounting for the central
role of monetary and credit mechanisms.

5.2 Stability in proportions instability

in dimension
We presented general disequilibrium models in various earlier
works.22 For each enterprise or industry, four variables must be considered: u, s, K, and p. To this first set of variables, one must add
money stocks (deposits) and debts for all agents.
Usually, the determination of equilibria is rather easy, while the
study of the stability of equilibrium is difficult. The distinction between various theoretical fields as in section 2.2 points, however, to
two possible ways out that may render the treatment of stability
manageable. First, under the assumption that it is possible to sort
out variables whose variations are judged rapid, and others, slow, the
distinction between time frames, short and long terms, suggests the
use of the methodology of temporary equilibria. Second, the distinction between proportions and dimension allows for a factorisation of
the model. New centered or relative variables are defined. For
example, to profit rates in each enterprise or industry, one can substitute the differences between these rates and an average profit rate;
the same can be done for capacity utilization rates and prices (distinguishing between the general level of prices and relative prices).
The average profit rate, the average capacity utilization rate, and the
22. G. Dumenil, D. Levy, The Economics of the Profit Rate: Competition,
Crises, and Historical Tendencies in Capitalism, Aldershot: Edward Elgar
(1993); La dynamique du capital. Un si`ecle deconomie americaine, Paris:
Presses Universitaires de France (1996).

Macroeconomics of Keynesian and Marxian inspirations

19

general level of prices account for dimension, and the centered (and
relative) variables, for proportions. The main results are as follows:
1. A Keynesian equilibrium prevails in the short term (with the equality between production and demand in each enterprise or industry),
and a Classical-Marxian equilibrium in the long term (with a normal
use of productive capacities and prices of production).
2. Within rather simple models, the characteristic polynomial of the
Jacobian matrix used in the study of stability conditions can be factorized into four factors corresponding to proportions and dimension,
respectively, in the short and in the long terms. Within more complex models, the factorization is not rigorously possible, only quasi
factorization (given the low values of a number of parameters).
3. An important result is that, in the short term as well as in the
long term, the conditions for stability in proportions appear to be easily met, while the conditions for stability in dimension can be easily
violated. This result points to the basic property of capitalist economies, which we denote as stability in proportions and instability in
dimension:
- Stable in proportions refers to the capability of capitalist markets to allocate capital in the long term, and determine outputs
in the short term, in various industries, in line with demand patterns. Potential buyers generally find what they seek on markets.
- Instability in dimension or, more rigorously, at the limit of
stability in dimension, points to a twofold propensity of capitalist macroeconomies to deviate from normal levels. This is, first,
manifest in the recurrence of overheatings and recessions (the
expression of the instability of short-term equilibrium). Second,
for more lasting periods of time, the average levels of output
around which short-term fluctuations are observed may stray at
a distance from normal levels (as defined by the historical trend).
4. General disequilibrium models allow for the interpretation of these
properties. The same mechanisms simultaneously account for stability in proportions and instability in dimension in the short term.
Involved is the swift capability of enterprises to react to the signals
manifesting a divergence between supply and demand (parameter
in equation 4). This prompt adjustment ensures the availability of
output or the limitation of the growth of inventories of unsold commodities, with favorable effects concerning proportions. But the vigor
of the same adjustment may entail cumulative movements of output,
upward or downward, the expression of instability in dimension.

20

Macroeconomics of Keynesian and Marxian inspirations

These findings straightforwardly echo Marxs analysis in Capital. In the theory of competition in Volume III, Marx accounts for
the mechanisms leading to the allocation of capitals among industries
and the prevalence of prices ensuring in each industry an equalized
average profit rate, with corresponding prices and outputs. (Involved
is, actually, a gravitation process because of recurrent shocks of variegated nature.) Marx believed this mechanism is efficient in capitalism (that is, stability in proportions). Conversely, the general level
of output follows the characteristic pattern of the cycle of industry
(the business cycle), with recurrent departures into overheatings and
recessions (instability in dimension).

6 - A historical perspective
This section is devoted to the analysis of the historical transformations of stability conditions. A first aspect is the impact of
declining profit rates. A second aspect is the effect of the progress of
management and the development of monetary mechanims on macro
stability.

6.1 Profit rates and stability conditions

An interesting hypothesis proper to Marxs analysis is the view
that profit rates their values and trends impact macroeconomic
stability. Recurrent fluctuations downward of profit rates may cause
recessions. For example, in the short term, as in Marxs analysis
of overaccumulation, the growth of employment during a phase of
expansion may push wages upward and temporarily diminish profits,
and provoke a recession.23
This impact may be felt in the longer term, and this suggests an
interesting link between the historical and short terms. If the profit
rate declines and remains low for more durable periods of time, a
structural propensity to instability prevails with more frequent and
deeper recessions, as during the 1970s.
23. K. Marx, Capital, Volume III, New York: First Vintage Book Edition
(1894), ch. 15.

Macroeconomics of Keynesian and Marxian inspirations

21

One mechanism accounting for the impacts of profit rates is their

potential influence on parameter in equation 4 (accounting for the
decision by enterprises to increase or diminish production depending
on the levels of inventories). As stated in the previous section, if
the response to excessive or deficient inventories is too strong, this
mechanism may trigger cumulative movements of output, downward
or upward. Profit rates command cashflows and liquidities and, thus,
affect . (The continuation of production at given levels depletes
liquidities if sales are diminished.)

6.2 The tendential instability thesis

It is a well-known stylized fact that the emergence of the modern pattern of business-cycle fluctuations echoed the development of
money and finance in capitalism. The built-in instability began to
manifest itself when monetary-financial mechanisms reached a sufficient degree of development, as during the first half of the 19th
century in Europe. In combination with the progress of the management of enterprises, notably since the end of the 19th century,
the gradual expansion of financial mechanisms fostered a tendency
toward increasing macro instability, a tendential instability.
To account for such trends, the potential impact of the variations of reaction parameters in the modeling of behaviors can be
considered in a historical perspective. The financial system became
gradually more inclined to accommodate the demand for loans, a rise
of parameter in equation 2, a tendency in which financial innovation
played a prominent role. The progress of the management of enterprises induced a rise in parameters such as in equation 4. These
historical drifts in the value of parameters rendered the satisfaction
of stability conditions more difficult.
The second facet of the tendential instability thesis is, however,
that this movement toward increased instability was checked by the
historical progress of countercyclical central mechanisms, thus becoming gradually more necessary. The Federal Reserve, created in
1913, was substituted for the National banking system, although the
movement toward a genuine management of the macroeconomy was
only gradual. The corresponding stabilizing mechanisms only reached
matured forms after the New Deal and World War II, in the context
of the Keynesian revolution. Neoliberalism did not destroy the domestic framework in which monetary policy is conducted, rather the

22

Macroeconomics of Keynesian and Marxian inspirations

contrary, but altered its objectives.24 The new conditions created by

financial globalization during the 2000s finally unsettled the foundations of the policies conducted by the Federal Reserve.25
Thus, the tendential instability thesis points to an historical process in which resistances and set-backs are recurrently observed. But
the pressure of events, as in recurrent crises, finally has the edge,
though the costs may be huge as in structural crises.

7 - Business-cycle fluctuations
This section provides intepretations of the fluctuations of the
capacity utilization rate as in Figure 1, in line with the analytical
framework introduced in the present study. The notion of fluctuation implies the reference to a given center around which the macroeconomy oscillates, abstraction being made of growth. In our models,
this center of gravitation is denoted as u and assumed constant over
time. The examination of the series of capacity utilization rates over
the more than seventy years in Figure 1 suggests fluctuations around
a constant value to 1970 (83.1% for the average 1948-1970) and, in the
subsequent decades, around a slowly declining trend. Thus, between
2003 and 2007, the figure estimates uHT at 77.7%. We do not believe
enterprises went on building capacities during several decades, while
they were not able to reach their target utilization rates. Thus, in the
downward trend of uHT , we see a decline of u not a growing negative
gap between such a target and the rates actually achieved.26
The following interpretations can be given of the observed patterns of fluctuations:
24. As in the Deregulation and Monetary Control Act of 1980. See G.
Dumenil, D. Levy, Crise et sortie de crise. Ordre et desordres neoliberaux,
Paris: Presses Universitaires de France (2000), box 18.5.
25. G. Dumenil, D. Levy, The Crisis of Neoliberalism, Cambridge: Harvard
University Press (2011), ch. 14.
26. The difficult assessment of capacity utilization rates over such long periods of time is clearly expressed in the definition of the variable itself: The
Federal Reserve Boards capacity indexes attempt to capture the concept
of sustainable maximum output the greatest level of output a plant can
maintain within the framework of a realistic work schedule, after factoring
in normal downtime and assuming sufficient availability of inputs to operate the capital in place (Capacity Utilization Explanatory Notes, Federal
Reserve, http://www.federalreserve.gov/releases/g17/CapNotes.htm).

Macroeconomics of Keynesian and Marxian inspirations

23

1. The long-term component. The broad fluctuations in uLT (

)
reveal the weakness of the centripetal forces generated by the action of
the central bank, intending to confine output in a vicinity of levels
considered normal or appropriate. To this end, the institutional
setting and the rules in which monetary policy is conducted were
recurrently modified:
- The awareness of the partial recovery of output in the recession
of the late 1950s induced the bold fiscal policy of the 1960s on the
part of President Kennedys advisers, aiming at the stimulation
of the macroeconomy. In combination with the financing of the
Vietnam war, this policy contributed to the rise of output to
unusual levels.
- Then, the decline of profit rates altered the basic conditions underlying the trade-off between the fight against inflation and the
preservation of growth in the direction of increased inflation. In
the context of the rather accommodative policy of the central
bank after 197327 , an upward trend of prices was observed, paralleling the decline of the long-term component in what has been
called stagflation.
- The shift of the long-term component to lower levels during the
structural crisis of the 1970s was prolonged into the early 1980s,
when the sudden rise of interest rates, the 1979 coup, further
bent the long-term component (and caused the recession of the
early 1980s, whose analysis belongs to the short-term component).
- During the 1980s and 1990s, the preservation of decent levels became gradually more problematic in the context of the
growing deficits of foreign trade and globalization.28 The severe
character of the situation was temporarily hidden by the boom
of information technologies during the 1990s, with a significant
stimulation of investment. When this unexpected bonanza came
to an end around 2000, a new decline in the long-term component occurred, whose severity was only partially checked by the
upward trends in residential investment.
In these movements the recurrent actions of the central bank and
government are manifest, in particular in the upward movement from
27. The fact that the theory of accommodative money was developed during those years is certainly not coincidental.
28. See G. Dumenil, D. Levy, The Crisis of Neoliberalism, op. cit. note
25.

24

Macroeconomics of Keynesian and Marxian inspirations

the late 1950s to the 1960s, the fight against inflation culminating in
the early 1980s, and the attempt to boost the U.S. macro trajectory
by a bold mortgage policy (and deregulation) prior to the current
crisis.
2. The short-term component. The sudden deviations of uCT around
the long-term fluctuation suggest quite specific patterns of variations:
- The 1950s and 1960s manifest the typical pattern of the stop
and go, in the context of still immature stabilizing procedures.
- One can, then, observe the occurrence of sharp departures upward and downward, with rapid a few quarters long transitions (for example, the dramatic two-quarter fall in 1974.)
- A closer examination reveals clusters of observations, when the
capacity utilization rate stabilizes (gravitates in a vicinity of
given positions): (1) close to the long-term component (as in
1963, 1984, and during the long boom); (2) for comparatively
higher positions, as in overheatings (as in 1979); and (3) for
comparatively lower positions as in the lasting recession of 20012002.
- The three broad cycles during the 1970s and early 1980s (the
period of declining profitability) are spectacular.
3. The circumstances (deregulation and the relaxation of lending
practices) that led to the subprime crisis in the United States. Concerning these latter years, the circumstances created by the duration
of the current structural crisis have a considerable impact on macro
mechanisms and disturb usual patterns. More time will be necessary
to provide reliable assessments. After the plateau of the capacity
utilization rate above the long-term component during the second
half of the 2000s prior to the recession, a new sharp plunge occurred,
followed by a still quite limited recovery.

Macroeconomics of Keynesian and Marxian inspirations

25

Contents
1 - Main results and outline . . . . . . . . . . . . . . 1
2 - Common and distinct grounds . . . . . . . . . . . 2
2.1 Disequilibrium microeconomics . . . . . . . . . . . 2
2.2 Toward the distinction between theoretical fields . . . 4
2.3 A taxonomy: Knitting Marxian and Keynesian perspectives 5
2.4 The components of the general level of activity . . . . 7
3 - Money and credit in the macroeconomy . . . . . . 8
3.1 The credit demand channel . . . . . . . . . . . . . 9
3.2 Built-in instability in a credit economy . . . . . . . 10
3.3 Co-determination . . . . . . . . . . . . . . . . 11
3.4 Modeling a monetary macroeconomy
. . . . . . . 12
4 - PostKeynesian views concerning money and credit
13
4.1 The financing of investment in Kaleckis framework . 13
4.2 Exogenous money, endogenous money, co-determination 14
4.3 Financing channels within postKeynesian models . . 15
5 - Articulating theoretical fields Vindicating...
. . 16
5.1 Production and investment . . . . . . . . . . . . 17
5.2 Stability in proportions instability in dimension . . 18
6 - A historical perspective
. . . . . . . . . . . . . 20
6.1 Profit rates and stability conditions . . . . . . . . 20
6.2 The tendential instability thesis . . . . . . . . . . 21
7 - Business-cycle fluctuations . . . . . . . . . . . . 22