Structural

Change and Economic Dynamics 7 (1996) 147-162

STRUCTURAL
CHANGEAND
ECONOMIC
DYNAMICS

Natural dynamics, endogenous structural change and the

theory of demand: A comment on Pasinetti
Davide Gualerzi,

a Istituto di Studi per la Programmazione

Economica (ISPE), Corso V. Emanuele 282/284, 00186

Roma, Italy
b Department of Economics, University of Pisa, 56100 Pisa, Italy

Abstract
This paper examines the most recent contribution
by Luigi Pasinetti to the investigation of
structural economic dynamics by considering in particular the role of demand theory. The
author suggests that a number of critical features of demand (and demand evolution) in modern
capitalist economies cannot easily be reconciled with the theoretical framework underlying
Pasinettis analysis. It is argued that such a framework rests on a conception of growth as a
process driven by exogenous forces and analysed with respect to a path of equilibrium
growth,
which is at variance with the endogenous determination
of consumption
patterns and new
modes of life in actual economic dynamics.
The learning principle at the root of Pasinettis recent contributions
suggests an interesting
bridge between the dynamics of production
and consumption coefficients. However, the most
fruitful way of investigating such a connection would be to consider the income-driven
changes
in the composition
of consumption
expenditure within a specific historical stage of development. It is argued that the objective character of Engel-type demand theory cannot easily
be reconciled with a consideration
of the specific, commodity-based
forms of satisfaction on
broad categories of needs.
Keywords: Structural dynamics of capitalist economies; Learning
tion activities; Compositional
dynamics of consumption
JEL classijkation:

041;

in production

and consump-

P16; D12

1. Introduction
The theoretical

ana.lysis

of structural

change

last ten years in the economic literature

Elsevier Science B.V.
PII SO954-349X(96)00049-5

has received

(Baranzini

growing

attention

in the

and Scazzieri, 1990). Luigi

148

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

Pasinetti has helped to define some of the fundamental questions in this field, along
with a rigorous approach that combines dynamic analysis and structural change.
He points out that his latest contribution (Pasinetti, 1993) is the result of a long
process of refinement of a scheme already formulated in the early 1960s. More
specifically, it grows out of his 1981 essay (Pasinetti, 1981) which already covered
much of the same ground. Some analytical concepts, such as vertical integration,
have been the subject of other contributions in the last 20 years (Pasinetti, 1973, 1988).
In the latest book, Pasinetti presents his argument in the most essential form. The
main results of the theoretical model emerge, in fact, independently of any analytical
treatment of capital goods. They also derive from the new focus on the learning
principle, which is now more explicitly regarded as the fundamental force behind
structural dynamics. As a consequence, the nature of the theoretical approach and
the critical points emerge more clearly.
This paper examines the analytical structure and the results of Pasinettis approach
in its most recent form. Focusing on the theory of demand, it raises questions which
cannot be easily reconciled with the theoretical framework presented by Pasinetti
and calls into question its meaning for the analysis of structural change within
advanced industrial economies.
The paper is organized as follows. Section 2 examines in some detail the pure labour
production model and the derivation of the natural relations that emerge from it.
This leads, in Section 3, to a critical appraisal of some of the fundamental features
of the model. Section 4 focuses on the inadequacy of a conception of the growth
process driven by exogenous forces and argues that an endogenous determination of
consumption patterns would require a change of theoretical perspective. Section 5
concludes.

2. Structural dynamics and natural relations

2.1.

The pure labour

production

model

The revival of the classical political economy approach is the fundamental premise
of Pasinettis work. In the Preface to the 1981 essay he identifies this continuity with
the level of his investigation. It is a distinctive feature of the present theoretical
scheme to begin by carrying out the whole analysis at the level of investigation which
the Classical economists called natural, that is to say at a level of investigation
which is so fundamental as to be independent of the institutional set-up of society
(p. xii). That implies a sharp discrimination between those economic problems that
have to be solved on the ground of logic alone . . and those economic problems
that arise in connection with particular institutions, or with particular groups and
individuals behaviour (p. xiii). He later motivates his choice to develop first of all
a theory which remains neutral with respect to the institutional organisation of
society (p. 25) with the preoccupation
of singling out, to resume Ricardos
terminology, the primary and natural features of a pure production system . . . these
. . will simply emerge as necessary requirements for equilibrium growth (p. 25).

D. GualerriiStructural

Change

and Economic

Dynamics

7 (1996)

147-162

149

However, in the 1981 essay, the study of the natural relations associated with
structural dynamics was carried out in stages. First they were examined in connection
with a pure iabour production economy and later within a more complex scheme
involving capital goods. Not so in the latest contribution. The introduction of capital
goods is left for a later investigation. Natural relations are analysed with respect to
a pure labour production economy, where labour is the only input and consumption
the only use of output.
Pasinetti argues that, in the same way that the pure exchange model is the essential
core of the exchange paradigm on which neoclassical economics rests, the labour
production model contains those analytical features, and only those features, which
classical and Keynesian theory cannot do without (Pasinetti, 1993, p. xiv). This
choice also emphasizes that technical progress, not capital accumulation, is the key
concept in the explanation of the long term dynamics of industrial economies.
However, his argument must be seen in light of the explicitly reformulated aim of
the theoretical approach: to study the economic consequences of human learning,
as the subtitle of the 1993 book reads.
The learning principle, which is at the root of the pure production (labour) model, goes
down to a more profound level of investigation. The classicaleconomists intuitively perceived
the importance of moving down to this deeper level: they call it natural. In the present
work, the natural relations emerge as having pre-institutional characteristics, and thus as
being even more fundamental than in classical analysis (p. xv).
This is the most important change with respect to the 1981 essay. The fundamental
features of structural dynamics can be adequately conveyed by a simple model of
production with labour alone (p. xiv) and yet such an abstraction is aimed at
grasping basic features of the industrial economies of our time (p. xiii) because of
the role assigned the learning principle, which in turn implies a further development
of the notion of natural relations, beyond the meaning they had in classical analysis.
The investigation is based on a Leontief closed model with two sets of coefficients,
labour input coefficients, Ii, and final demand (consumption) coefficients, Ci. The
scheme is meant to represent an economy governed by a Smithian pure labour theory
of value and by the Keynesian principle of effective demand. Prices reflect labour
content and quantities the level of effective demand, i.e. the level of consumption
demand. In this context the basket of consumption
multiplied by the labour
coefficients determines the level of employment in the economy.
The condition for macroeconomic stability, I;= 1 c,l, = 1, says more: employment
is equal to total labour available in the economy. This relation emerges directly from
the structure of the model and, mathematically, as the condition for economically
meaningful solutions. It implies that all available labour is utilized, i.e. the existence
of equilibrium solutions. Conversely, those solutions imply that employment and the
wage rate are at their natural level.
Pasinetti argues that it would make no sense to consider the natural level of

I Quotations are from Pasinetti(1993) unlessotherwiseindicated

150

D. GualerziiStructural

Change

and Economic

Dynamics

7 (1996)

147-162

employment as anything different from full employment. The total quantity of labour
available in the economy is a magnitude of national relevance. . . And it is clearly
a matter of general concern that is should entirely be employed (p. 23). Similarly,
the natural wage rate emerges as a concept of macroeconomic relevance. The natural
prices, being proportional to labour quantities, channel back to each unit of physical
labour an equal amount of purchasing power (p. 24). The natural wage is their
necessary counterpart and represents each labourers equal share into the net
national product.
When the equilibrium macro condition is undersatisfied we are in a situation of
unemployment, the familiar case examined by Keynes; when it tends to be oversatisfied,
inflation pressures arise.
2.2. Proportional

growth and structural change

This static scheme is the foundation for the study of dynamics. The latter requires
an adequate treatment of the process of change, on the basis of the evolution of the
technical and demand coefficients. This is a fundamental difference with growth
theory models, which can at best accommodate some unrealistic patterns of growth
that have in common the characteristic of maintaining unaltered sectoral proportions.
Proportional
growth has no effect on the macroeconomic condition. Once an
equilibrium is established, it is maintained. Consequently, these models cannot take
into account dynamics in any meaningful way.
A truly dynamic theory must start from the fact that technical change proceeds
unevenly between sectors and therefore labour coefficients decrease at different rates
among sectors. In general, they will not be compensated by a symmetrical and
opposite growth in demand. Indeed, sectoral demands evolve following their own
path. A theory of demand, adequate for the purpose of an inquiry into the evolution
of industrial systems over time (p. 39), can be grounded on two foundations: the
first is the evidence that the structure of consumption always varies as income
changes (p. 38), an empirical regularity known as Engels Law; the second is that
consumers demand follows well defined paths, before slowing down and eventually
reaching saturation (p. 38). Such a theory is therefore based on the recognition that
essential needs are satisfied first, following a hierarchical pattern, and that, in turn,
as income increases, saturation sets a limit to the consumption of any good.
Consequently, in a dynamic setting there is an inherent tendency towards the
non-proportional
growth of demand.
Taking into account these few hardly disputable facts, the labour and consumption
coefficients become time functions of their rates of change, determining the time paths
of prices and quantities, i.e. the dynamic counterparts of the static models unknowns.
2.3. Structural udjjustment and natural

dynamics

On the quantities side, outputs grow at a rate that depends on a scale component,
due to population growth, and on a structural component, due to the rate of change
of per capita demand in each sector. Since technological progress and changes in

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

151

demand patterns proceed at a different pace, maintaining full employment requires

the continuous process of reproportioning
sectors employment. Indeed, keeping the
macroeconomic condition satisfied in a dynamic setting is a truly complex problem
(p. 50). More precisely, each sector expands, maintains constant, or contracts its
share of total employment over time according to whether the corresponding rate of
growth of per capita demand is greater than, equal to, or smaller than, the
corresponding rate of growth of productivity (p. 51).
Population growth and the rate of retirement in each sector add an element of
flexibility, to the extent that they allow for the redistribution of employment without
labour displacement. But such a non-disruptive way of adjusting sectoral employment
may not be sufficient in the face of the inevitable decline of certain sectors and the
emergence of technological unemployment. In general the continuous decrease in
technical coefficients and the limits set on per capita demand by saturation result in
a tendency for the macro condition to be undersatisfied. There are a few remedies
to this potential situation of stagnation. On the supply side, the diminution of the
activity rate and/or of labour time; on the demand side, external sources of demand,
such as exports, or consumer learning, which speeds up the growth of per capita
demand in the new areas of spending.
The structural dynamics of prices presents more analytical difficulties. In the
physical quantities system the total quantity of labour and its rate of growth can
reasonably be taken as exogenously given. In the price system the model closure
depends on the choice of a numeraire and it is not obvious which price should be
taken as given. Moreover, the expression for the general (average) price level would
become dependent on the rate of productivity growth of such a numeraire.
The question is: Can we define a measure that ensures the constancy of the price
level in the face of the continuous process of price adjustment due to productivity
growth differentials among sectors? In this respect it must be realized that in dynamic
analysis there are two degrees of freedom: one is used to close the system of equations
that determine relative prices, and one is used to close the equation system that
determines growth rates. Using a physical commodity or the wage rate (labour
commanded) as numeraire obscures this fact since, by definition, their rates of change
are equal to 0. The existence of the two degrees of freedom is instead fundamental
for the main objective of the analysis of price level stability: isolating a structural
(real) component from a purely nominal monetary component. This takes a
considerable amount of analytical effort.
Experimenting with the wage rate, expressed in terms of an arbitrary physical
commodity, as numeraire shows that the constancey of the average, or . general
level of prices . . . is not at all a characteristic of a numiraire in physical terms
(p. 66). It is therefore necessary to make reference to a composite commodity the
composition of which reflects the weighted average of the rates of growth of
productivity
of the entire economic system (p. 70), or the standard rate of
productivity growth. When a dynamic standard commodity so defined is used as
numeraire, prices will change according to the difference between their rate of
productivity growth and the standard rate, but the general price level will remain
constant. It is then sufficient to use such a composite commodity as the unit of

152

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

account of the wage rate and setting its growth rate equal to the standard rate to
ensure the stability of the price level through time.
Pasinetti calls the difference between the rate of productivity of any commodity
chosen as numeraire and the standard rate the rate of inflation of the general price
level, due to the structural dynamics of prices (p. 75).
For a perfectly general analysis of price dynamics we need to consider the case in
which the unit of account is a nominal measure, such as any currency. Conceptually
this simply adds the possibility of deviations from the price stability norm, as defined
above, due to the presence of a unit of account fully divorced from the dynamics of
productivity. Expressing the wage rate and its rate of growth in nominal terms
highlights the complete independence of the two degrees of freedom of a dynamic
system and the existence of two components of the sectoral rates of change of prices.
The first, pure nominal, or inflationary, is the result of the deviation of the rate of
change of the nominal wage rate from the ideal situation of dynamic stability in
which the latter is equal to the standard rate of productivity growth (p. 78); the
second is the difference between the standard rate and the sectoral rate of productivity
growth.
Consequently price stability depends not only on the structural component tied
up with problems of attainment of efficiency in each single branch of production
(p. 81) but also on the policy of a specific institution, the central bank, to which the
macro goal of price level stability is entrusted.
2.4. Saoing,$nance

and the natural

rate

of interest

The central bank must also maintain the rate of interest at its natural level, i.e.
at the level that keeps unaltered through time the purchasing power of loans in terms
of labour, i.e. the equality between labour embodied and labour commanded
(P. 92).
Although, by definition, in a pure labour economy the system as a whole does not
carry positive savings from one period to another, a rate of interest arises from within
the operation of the financial sector. Indeed, even in the absence of saving in the
aggregate, it is still possible to have saving and dissaving for single units, or between
the private and the public sectors, provided they cancel out one another. The
purchasing power of the flows originating in these debit-credit relationships will
remain constant with respect to the commodity in which they are denominated, but
will change with respect to all commodities.
Consequently, even without any explicit rate of interest, a set of own rates arises
simply as a consequence of productivity growth differentials; when an actual rate of
interest is stipulated on loans, in terms of a numiraire, say the wage rate, it must be
added to the own rate of interest. If the standard commodity is used as numeraire,
then such rates will be determined by the difference between the standard rate of
productivity and the commodity rate, plus the actual rate stipulated on the loan. For
labour commanded the own rate is simply the actual rate of interest, reckoned in
terms of the standard commodity, minus the standard rate of productivity.
Once we introduce a rate of interest stipulated in a nominal unit of account, we

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

153

need to consider also changes in the price level. In particular, the natural rate of
interest must take into account the average productivity growth and changes in the
price level. The derivation of the natural rate, however, turns out to be remarkably
simple. Pasinetti shows that such a rate is equal to the rate of change in the wage,
independently of any numeraire. Consequently, the rate of change in the wage rate
becomes the reference point for setting the rate of interest at the level consistent with
the internal logic of the natural economy.

3. Structural dynamics, equilibrium

analysis and institutions

In his review of the 1981 volume, Scazzieri noted that Pasinetti had offered a
fresh treatment to many vexed questions in political economy (Scazzieri, 1983). This
is not the purpose of the latest evolution of Pasinettis analysis, which focuses instead
on the complexities of structural dynamics, quite independently of other theoretical
problems. From this point of view the pure labour model may well help to highlight
most cogently the fundamental questions Pasinetti wants to address.
In this respect the sharp separation of structural dynamics from institutional and
historical analysis is also noticeable. In his latest book, institutions are considered
in a separate chapter, mostly to highlight their function of bringing into being and
keeping at their natural level the key macro variables. Similarly, the exposition is
cleared of any descriptive passage on historical dynamics, except for a birds eye
view of the evolution of the structure of production of industrial economies, discussed
in one of the last chapters of the book.
More explicitly than before the distinction between separate spheres of analysis
seems to be a condition for concentrating on the investigation of the natural relations
that lie at the core of structural dynamics.
Pursuing such a research program, however, comes at a price. Deprived of any
reference to historical or institutional dynamics, the system is held together exclusively
by the requirements of full employment, i.e. structural change is analysed in the
framework of an equilibrium growth path.
This is an element of continuity of Pasinettis theoretical approach, and a
troublesome one.
It can be legitimate in order to call attention to the permanent process of
redistribution
of labour imposed by structural change, a highly complex and
unsettling phenomenon. Pasinetti argues that it would be of no interest to consider
a natural system which is not a full-employment system. However, it is one thing to
assume effective demand as a constraint, but quite another to assume full employment
as a constraint. This transforms the full-employment condition from a methodological

Pasinetti points out that the scant attention

given in the theoretical
literature
to the issue of structural
change contrasts
with the numerous
important
contributions
coming
mostly
from the development
literature.
These, however,
never achieved,
nor even pursued,
any status of theory.
For this reason the
analysis
of structural
change should be based on a theoretical
foundation,
beyond the empirical
search
of historical
regularities.

154

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

tool into an attribute of the natural system. What slips into the argument is that a
natural system would be one of full employment, a statement which has little in
common with the ideas of the classics.
A similar criticism was levelled against Pasinetti previously. Shapiro (1984-85) has
argued that the notion of equilibrium imbedded in the model is a fundamental flaw:
the employment effects of technical progress on a full employment growth path are
quite different from what they are outside of it (Shapiro, 1984-85, p. 243).
Consequently, the results of the model hold only for a natural economy, not for a
capitalist economy.
Pasinetti, like other equilibrium
theorists, assumes that the equilibrium
relations are
the necessary ones, and thus the equilibrium
implications
of technical change are its
economic implications.
They have no more significance than the equilibrium
state which
they support and do not necessarily hold outside the conditions which produce them. .
For the employment effects of technical change to be specified, the pattern of technological
change has to be determined. And this can only be done by denaturalizing the innovation
process
(Shapiro, 1984685, pp. 2446245).
Harris (1982) in his review of the 1981 book observed that Pasinetti does not
depart from equilibrium analysis. His entire analysis based on the factors that
systematically upset equilibrium positions. If we further consider that factors of
change are independent from the workings of the economy, the model presents a
continuity with a theoretical framework which would be no cause for discomfort
for the sophisticated neoclassical theorist (p. 37). This is what Pasinetti described is
a super golden age, in which, although everything changes, there still is full
employment. Calling it a natural economic system appears just one way to get around
the problem.
The meaning of equilibrium positions needs to be clarified with respect to the
general purpose they have in the theoretical scheme. Pasinetti has stressed several
times that the model is intended to highlight the necessary decisions that confront
any progressive society because of the existence of technical change, independently
of any institutional set-up. The pure structural and institution-free character of the
model is also underlined by the lack of any gravitational force. Consequently, it seems
fair to say that the scheme of the natural economy should be used to define what
ought to be done. The natural magnitudes then become the targets that institutions
should have as a guiding star to devise policy. Observing that full employment of
labour is clearly a matter of general concern (p. 25) stresses that full employment
must be pursued as an object of policy: there is nothing in the structural evolution
of technical coefficients . and of per-capita demand . . . that will ensure the
maintenance of full employment . . it will have to be actively pursued as an explicit
aim of economic policy (Pasinetti, 1981, p. 90).
Certainly Pasinetti cannot be mistaken for an advocate of laissez-faire. It does
not follow, however, that equilibrium positions are the necessary or best guide to
economic policy. Because of the separation between distinct spheres of analysis, the
burden of explanation falls entirely on the scheme of the natural economy. Its internal
logic defines the economic problem which institutions should govern. Indeed, this

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

155

suggests that, without ill-conceived policies, full employment would be within reach.
The approach bears little resemblance to the natural system of Smith and Ricardo,
where the attainment of natural long-run positions is intrinsically linked to the work
of competition and is distinct from the question of full employment.
To sum up: keeping the analysis strictly institutions-free seems to create more
problems than those it intends to solve. Recently, Baranzini and Scazzieri (1990) have
argued that institutional mechanisms and structural dynamics are two distinct levels
of analysis which should be kept together: In particular, it emerges that the analysis
of actual economic dynamics requires the integration of structural and institutional
types of analysis (Baranzin and Scazzieri, 1990, p. 257). In Pasinettis scheme the
trouble arises with the very notion of a natural economy. While it can hardly justify
the reference to equilibrium positions, it suggests that institutions come into being
from some process distinct from that of economic growth and then can be added to
the scheme of necessary structural relationships. This way of proceeding may make
the analysis of the long-term evolution of industrial economies more elusive than
substantial.
In fact, it is not clear how institutions could prevent the tendency for the
macroeconomic condition to become undersatisfied, unless, as noted by Harris, they
could stop the course of technological progress (Harris, 1982, p. 35). The very
fundamental mechanism contrasting the lapse of the system into depression and
underemployment,
the learning by consumers of new preferences, conceivably
combined with the introduction of new products, brings to the fore the question of
the institutional factors that shape consumption in advanced capitalistic economies.
From this point of view the natural economy only works to move the real questions
of development and accumulation a step away, displacing them into the domain of
policy. It seems, quite frankly, a far too extreme conclusion.
It is also not immediately clear what can come out of efforts to improve the model
by adding new parameters to the fundamental relationships; or model historical or
institutional processes into the scheme.
As an example, one may recall a recent attempt to insert the long wave into
Pasinettis model. Reati and Raganelli (1994) show that, as the innovation process
unfolds along the path of the long wave, prices converge towards natural prices.
However, this only shows that there are way to model the unsteady character of
technical change which are consistent with natural relations, much less that they
capture the essence of the actual dynamics of structural transformation of industrial
economies. As Harris has noted, the consistency of the model and, more generally,
the logic of structural models, is not a matter of dispute. The problems arise with
the theoretical practices, purposes, and uses to which the model is put (Harris, 1982,
p. 28).

4. Endogenous growth, exogenous forces of change and the theory of demand

In light of these criticisms it may be useful to reconsider the development of
Pasinettis approach in the 1993 book. As pointed out in Section 2, Pasinetti argues

156

D. Gualerzi/Structural

Change

and Economic

Dynamics

7 (1996)

147-162

that the essential features of structural dynamics may be analysed independently of

the process of capital accumulation; from this fundamental change of perspective the
prominent role of the learning principle takes its full relevance. On the basis of these
premises, Pasinetti can claim that the pure production model, and the natural
relations that emerge from it, can capture the essential aspects of the long-term
dynamics of structural transformation of industrial economies.
This development of the approach raises new interesting questions. In particular,
the relevance accorded to the learning principle shows an important similarity with
the fundamental theme common to the new models of endogenous growth (Romer,
1986; Lucas, 1988). In these models a long-term growth trend is obtained without
any reference to exogenous technical change, but rather by taking into account the
effects of the accumulation and diffusion of knowledge, either in the form of new
technical knowledge, which becomes widely available (Romer, 1986) or spreads as
a consequence of specialization of production (Romer, 1988) or as a result of the
improvement of the general level of education that affects human capital resources
(Barro, 1989). Although such a similarity should not be stretched too far,3 the analogy
between this view of the growth process and the role of learning in Pasinettis scheme
is quite clear.
In particular, the new growth theory, which directs attention to the endogenous
and cumulative nature of the growth mechanism that originates in the accumulation
of social knowledge, helps to formulate the question of endogenous growth within
Pasinettis model.
Indeed, the fundamental aspect that has eluded previous criticisms is its reliance
on exogenous forces of change, which remains despite the centrality of the learning
principle. An exogenous process of technical progress creates the conditions for
income growth. As income grows, the consumption structure evolves, following its
own independent path, modelled after Engels Law. The potential for more output
in each sector must be validated by a symmetrical increase in demand, bringing to the
fore a conflict which is at the core of structural adjustment. However, beyond the
non-proportional
pattern of expansion of Engels Law, final demand composition is
not specified and the rates of productivity growth are fully exogenous. Income growth
appears, from this point of view, as something totally unexplained and leading to
stagnation. Uneven demand expansion, before indicating the possibility of economic
expansion, results in a tendency for markets to become saturated.
The notion that the forces of change can be taken as exogenous rests on the
questionable argument that they fall mostly outside the domain of economic analysis.
While there is in principle no reason to disagree with the fact that a theory of technical
change would pertain to a much wider field than economics (Pasinetti, 1981, p. 67)
it is equally true that economic theory must be able to explain how these forces
work their way into the growth process.
Endogenous
growth
models develop out of the effort to rescue growth theory from the contradiction
between capital accumulation
and decreasing
returns typical of the neoclassical
production
function.
This
is clearly irrelevant for Pasinettis scheme, which does not make reference to the neoclassical
theory of
production.

D. Gualerzi~Structural

Change

and Economic

Dynamics

7 (lYY6)

147-162

157

It is precisely on this issue that Pasinettis model contains a promising new insight.
A characteristic of his approach is the integration of the demand side into the analysis
of growth and the link established between the production
and consumption
coefficients. The reason for this insight is fully brought to light by the most recent
development of the model: at the roots of economic progress is indeed the general
principle of learning. Once disentangled from the notion of improvement of the
production methods, which is dominant in economic theory, it is clear that the same
principle applies also to demand, i.e. to consumers.
Having gone this far, however, Pasinetti does not follow through with the
implications of his own reasoning. Learning may indeed be the core of the
endogenous self-determination of the growth process. But then the latter cannot be
modelled in a way that leaves exogenous the forces of change. Pasinetti seems to
have overlooked the fact that precisely the evolution of his argument about structural
dynamics exposes the inadequacy of the stylized facts on which the model is based.
Both technological innovation and taste evolution attain their specific connotations
in so far as they are endogenous to the process of economic development; however,
their reciprocal determination is the source of expansion of the market economy and
consequently also where the process of structural change originates. Accordingly. as
opposed to the general principle of learning, there exist a number of specific learning
processes that are part and parcel of the process through which scientific and
technological progress, as well as taste and preferences, are organized to become
markets. Alternatively, the entire model is pulled by exogenous factors. It has no
dynamics of its own.
The reference to the forces of change as exogenous is not therefore justified
conceptually, whereas theoretically it leads to a model in which structural dynamics
and the composition of industrial output are entrusted to processes independent of
economic development. In fact, Pasinettis structural dynamics ends up in a paradox:
it rests on consumer sovereignty. Learning new preferences is the main force
counteracting the tendency to market saturation, but the rates of change of per capita
demand for each good cannot be determined without reference to an exgenous process
of the formation of consumer preferences.
Ultimately, the integration of the demand side into the analysis of growth, which
is potentially the most fruitful step forward, does not lead to an analysis of the
endogenous growth mechanisms because of a fully inadequate theory of demand.
In the 1981 essay Pasinetti argued that preferences ultimately
depend on
human nature. which represents, in the same way as the technical conditions of
production do. a fundamental external datum for any meaningful economic investigation (p. 68). As questionable as it is, it was an argument for asserting the exogeneity
of taste. In the latest book the notion of human nature has disappeared and we have
instead a careful drawing of all the implications of Engels Law and numerous hints
that go well beyond it.
Pasinetti observes that technical progress implies the possibility of obtaining . . .
entirely new goods and services and that this aspect of technical change opens the
way to further, autonomous decisions distinct from the pure availability of more
income. It thus becomes necessary to discuss the relation that exists between

158

D. Guulerri~Strucrural

Champ

and Economic

Dynamics

7 (1996)

147-162

increases in real income and expansion of demand for consumption goods (p. 37).
In this respect the traditional theory of consumer demand is completely useless.
However, he observes that for the purpose of developing a reasonable theory of
consumption decisions in a dynamic context, we already have many important
fragments (p. 107). Among these the inevitable asymmetry between consumption
decision concerning already known goods and consumption decision concerning
goods that are consumed for the first time and the fact that an increasing
proportion of consumption goods require a process of preliminary training and
learning, in order to be enjoyed (p. 108). Finally: It may in fact become necessary
to investigate in some detail the inherent characteristics of human needs . about
which to pretend to remain agnostic would be simply self-defeating (p. 109).
All these observations, however, do not affect the analytical structure of the model,
nor can be considered a more adequate theoretical treatment of the relationship
between demand, structural change and growth.
In the first place it must be recognized that the dynamic theory of consumption,
based on the objective ground (p. 37) of Engels Law, cannot say anything about
the specific, commodity-based forms of satisfaction of such broad categories of needs
as those considered by the income-driven changes in the composition of consumption
expenditure. These are specific to a social and production structure at the historical
stage of development. The technology content, product complementarities
and
investment strategies associated with it are the key to the determination of the pace
and pattern of growth. This line of reasoning implies that much attention must be
directed to new products. Pasinetti seems to attribute to them some autonomous
role in the determination
of the patterns of consumption. The variation in the
composition of consumption may well occur independently of the increase in income
and of the changes in prices, as a consequence of the appearance on the market of
newly invented goods and services (p, 40). However, although essential to keep the
macroeconomic condition satisfied, they are not analytically treated in any way
distinguishable from the exogenous increases in productivity, nor, he elaborates, on
their influence on taste formation. Finally, this approach to demand even contradicts
the notion of learning. Despite the role of prime mover within the theoretical scheme,
learning receives scant attention and is hardly analysed. According to Pasinetti, the
saturation of certain needs imposes the speeding up of consumer learning. The
analysis of this process would presumably indicate that new preferences are not
discovered by consumers, but rather developed within an adaptive, socially conditioned process, where contact with products, and especially new products, as well
as the social meanings of consumption, are essential aspects.
We are left to wonder why the analysis of demand remains instead rooted in a
premise. the exogeneity of taste, which is at odds even with the fragments of analysis
that Pasinetti mentions. Pasinetti notes that Fortunately, at least with regard to the
present analysis, we do not need to develop a complete theory of demand (p. 37).
The question raised here. however, is a different one. It concerns the very premise
on which demand is analysed and consequently the analytical foundation of the
theoretical approach, not its completeness. The only plausible reason to obtain
exogenous taste formation seems to be the logic on which the analysis of structural

D. Gualerzi/Struclural

Change

and Economic

Dynamics

7 (1996)

147-162

159

dynamics is based. Specifically, it is necessary to develop the analysis of the demand

side within an equilibrium growth path. Indeed, if we were to carry out the analysis
of consumption focusing on the relationship between innovation, new commodities
and endogenous taste formation, then it would become apparent that it conflicts with
the very structure of the approach.
Only this perspective can give full relevance to the notion of learning. The latter
should be considered as the process that endogenizes the interaction between
technological innovation and the evolution of consumer taste. As such it needs to
be integrated into an adequate approach to demand centred on the relationship
between production and consumption. It must be stressed that such a relationship
in no way implies a fully endogenous theory of taste, nor that demand theory must
be exclusively an economic theory of demand. It indicates instead that economic
theory should develop a theory of demand consistent with the analysis of growth.
Such a theory may be developed starting from two premises. First, in order to
consider human needs within economic analysis we cannot abstract from their social
nature and from the systems of commodities that are designed to satisfy them.
Specifically, the analysis of endogenous growth mechanism must focus on the
relationship between technological conditions of production and commodity-specific
forms of satisfaction of socially defined needs that determine the evolution of taste.
Otherwise we abstract from the potential for novel consumption, which is a
characteristic of the production of commodities. Second, it seems safe to say that
firms influence taste both directly, by means of product innovation, and indirectly,
through the impact of their strategies of expansion on the social environment of consumption. This criticism of exogenous taste is more fundamental than that implied
by consumer manipulation by means of advertising and other marketing efforts.
In a similar vein, several years ago Leon discussed the relationship between
production and consumption, the active role of entrepreneurs and the limits of Engels
Law for the analysis of consumption in a capitalist economy (Leon, 1967).5 He first
observes that the inclinations and tastes of consumers in fact depend on production,
when they are viewed in a dynamic setting, because they can only be manifested
when commodities are already present in the market (p. 124). He then argues that
Engels Law, like other phychological laws, although operative in all economic
systems, is not capable of explaining why certain tendencies rather than others are
present in the evolution of the consumer budget in a capitalist society (p. 12.5). Only
through the action of entrepreneurs does the pattern of growth of consumption
becomes determinate.
4 For a discussion
of the theoretical
framework
that makes reference
to the notion
of the social
env,ironment
of consumption.
set Levine (1981). In that framework
the reciprocal
determination
of firms
market
expansion
strategies
and the constitution
of individuals
within a mode of life determines
the
evolution
of the social environment
of consumption,
The latter is the source of new needs and the potential
for new commodities.
at least to the extent that needs can be satisfied
by commodities.
Within
his model the dynamics
of demand,
based on Engels Law. is the basis for a permanent
differentiation
of protit rates. This in turn leads to a possible conflict
with the aims of the entrepreneurial
class. Resolution
of this conflict
requires
an analysis
of consumption
and is the key to the theory of its
dcvclopment
within capitalism.

160

D. GuulerziiStructurtl

Change

and

Economic

Dynamics

7 (1996)

147-162

Another research path is the more articulated analysis of consumption spending

and of the household sector.
Eichner (1987) has shown how the hierarchical pattern of the Engel curve can be
combined with a disaggregation of consumption to construct a model of household
behaviour. The latter distinguishes between needs and products and elaborates on
the role of habit formation and non-routine expenditure in determinjng the expenditure patterns. In a recent article Appelbaum (1922) developed some of Eichners ideas
about the modelling of the household sector. Following the idea that household types
react differently to the process of economic development, the model IS combined with
a taxonomy of households. It is then possible to link technical change, consumption
activities and labour supply and to analyse their interdependence within an inputoutput model.
In yet another recent contribution, Landesmann and Scazzieri (1994) go in a new
direction. They elaborate a methodology for the analysis of structural dynamics based
on the notion of the production process. The latter is specified in three dimensions:
tasks, capabilities and fabrication stages. From the interaction of these three sets of
elements. which characterize the organization of the produclion process in any
historical phase, endogenous structural change results. This scheme can capture
qualitative change which eludes simpler representations. We may conclude that the
intrinsically simple decomposition technique used by Pasinetti may limit the kind of
questions that the model can answer and consequently cannot capture the underlying
logic of structural change.

5. Concluding remarks
With his latest contribution
Pasmettis approach to the analysis of growth and
structural change has attamed a new degree of theoretical clarity. Starting from the
most essential model that embodies classical and Keynesian features he derives the
necessary relationships that hold in a natural economy driven by the non-uniformity
of productivity growth and demand structure evolution in order to maintain full
employment. This leads to a rigorous analysis of the many complexities of structural
dynamics, and particularly of the structural dynamics of prices, and highlights the
questions faced by a policy pursuing macro stability.
Its results must be evaluated with respect to the type of questions asked. The model
can be used to uncover, in an analytically precise manner, the consequences of
structural change. Pasinetti has extensively explored this issue and, for such a purpose,
the model certainly represents an important theoretical achievement. Its contribution
is most strmulating when it directs attention to the relationship between productivity
and demand. However, the latter is used exclusively to discuss the implications of
natural magnitudes and adjustment processes with respect to macroeconomic
stability,
rather than to develop new insights into the theory of structurnl chanye.
In general, the relevance of the approach for a theory of the structural transformation of advanced industrial economies is very much in doubt because it rests on the
concept of a growth process driven by exogenous forces and analysed with respect

D. Gualerzi!Structural

Change

and Economic

Dynamics

7 (1996)

147-162

161

to an equilibrium growth path. There are therefore reasons to be dissatisfied with

the approach which are independent of the exclusion of capital goods from the
analysis.(j Even the question of learning, which emerges as the key concept in the
scheme of a naturally growing economy, remains largely marginal because of this
underlying concept of the growth process. For Pasinetti the latter is adequately
represented by the inherent uneven dynamics of the labour and consumption
coefficients.
This view contrasts with theoretical abstraction which is based on the endogenous
mechanism that connects income growth and structural change, i.e. the emergence
of new industries and new modes of life. In turn that requires an appropriate
treatment of demand. The question is not how much of an exogenous component
must be considered, but rather how to develop a theory of demand consistent with
the dynamics of growth. It should centre on the issue of the reproduction of social
life in a system dominated by the production and consumption of commodities and
account for the internal drive for expansion which results from that.
Admittedly it is not clear how to model the interaction between firms orientation
toward market development and endogenous taste formation. The remarks formulated
above and the most recent contributions to the analysis of the household sector,
together with new efforts to develop an adequate methodology for the analysis of
structural change, need further development to establish how product innovation
and need development may be satisfactorily integrated into the analysis. This
constructive task can be undertaken only with respect to a theoretical scheme
centred on the making of modern consumption and its fundamental relationship with
investment directed toward the creation of new commodities markets. There is
nothing natural about such an evolution, which is indeed social in its essence.

Acknowledgements
An earlier version of this paper was presented at the URPE session Nonneoclassical Approaches to Macroeconomics at the ASAA meetings in New Orleans,
3 January 1992, and at the XIa Riunione Scientifica de1 Gruppo di Studio delle
Teorie e delle Politiche Economiche, held at the Universita della Calabria, Cosenza,
13114 March 1992. I wish to thank Prof. E. J. Nell, the editors of this Journal and
two anonymous referees for their comments.

6 This is indeed a different

problem.
One may wonder,
could affect the derivation
of the natural
rate of interest.

for example,

how

the presence

of capital

goods

References
Appelbaum.
E., 1992. The integration
of household
structure
and industrial
input-output
model. in: Milberg,
ed.. The megacorp
and macrodynamics:
Eichner
(M.E. Sharp. Armonk.
New York).

structure:
Essays

An extension
of the
in memory
of Alfred

162

D. Gualerii,Structural

Change

and Economic

Dynamics

7 (1996)

147-162

Baranzini,
M. and R. Scazzieri.
1990, Economic
structure:
Analytical
perspectives,
in: M. Baranzini
and
R. Scazzieri,
eds., The economic
theory
of structure
and change
(Cambridge
University
Press,
Cambridge,
UK).
Barro, R., 1989, Economic
growth
in a cross section of countries,
NBER Working
Paper, February.
Eichner, AS., 1987, The macrodynamics
of advanced
market economies
(M.E. Sharp, Armonk,
New York).
Harris.
D.J., 1982, Structural
change and economic
growth:
A review article, Contributions
to Political
Economy
1.
Landesmann.
M. and R. Scazzieri,
1994, Growth,
economic
structure
and decomposition
techniques,
in:
M. Landesmann
and R. Scazzieri,
eds., Production
and economic
dynamics
(Cambridge
University
Press, Cambridge,
UK).
Leon, P., 1967, Structural
change
and growth
in capitalism
(The Johns Hopkins
University
Press,
Baltimore,
MD).
Levine, D.P., 1981, Economic
theory. vol. II (Routledge
and Kegan Paul, London).
Lucas, R., 1988. On the mechanics
of economic
development,
Journal
of Monetary
Economics
22.
Pasinetti,
L.L., 1973, The notion of vertical
integration
in economic
analysis,
Metroeconomica
25.
Pasinetti,
L.L., 1981, Structural
change and economic
growth
(Cambridge
University
Press, Cambridge,
UK).
Pasinetti,
L.L.. 1988, Growing
sub-systems,
vertically
hyperintegrated
sectors and the labour
theory of
value, Cambridge
Journal
of Economics
12,
Pasinetti,
L.L., 1993, Structural
economic
dynamics
(Cambridge
University
Press, Cambridge,
UK).
Reati. A. and V. Raganelli,
1994, Introducing
long waves into Pasinettis
model of structural
change: The
price movements,
Economic
Appliqut,
in press.
Romer, P., 1986, Increasing
returns and long run growth,
Journal
of Political
Economy
94.
Romer. P.. 1988. Growth
based on increasing
returns due to specialization,
American
Economic
Review,
Papers and Proceedings,
May.
Scazzieri,
R., 1983. Economic
dynamics
and structural
change:
A comment
on Pasinetti,
Rivista
Internazionale
30.
Shapiro, N.. 1984-85. Involuntary
unemployment
in the long run: Pasinettis
formulation
of the Keynesian
argument.
A review article, Journal
of Post-Keynesian
Economics
7, no. 2.