A partnership with academia

Building knowledge for trade and development

Vi Digital Library - Text Preview

Structural Change and Economic Development: is Brazil Catching Up or Falling Behind?

Discussion paper by Nassif, André; Feijó, Carmem; Araújo, Eliane, 2013

Download original document (English)

This paper presents a Kaldor-Thirlwall theoretical and empirical framework on the basic driving forces of the behaviour of productivity and economic development in the long-run. By calculating the so- called Thirlwall equation, the main contribution of this research is to examine whether Brazil has been catching up or falling behind. The authors show some empirical evidence based on both descriptive statistics and econometric regressions for Brazil between 1970 and 2010. Some important indicators of descriptive statistics reveal that Brazil has entered into a process of early de-industrialization. In addition, since the econometric estimates also show that there was a dramatic increase in the income elasticity of demand for imports between 1980–1998 and 1999–2010 (from 1.97 to 3.36) and a small decrease in the income elasticity for exports during the same periods (from 1.36 to 1.33), the authors conclude that Brazil not only has already embarked on a trajectory of falling-behind relative to the world economy and the international economic frontier, but also that it might show, in the absence of appropriate policies, lower growth rates in the long run. However, if the opposite occurs, it would face major long-term external constraints to growth.

No. 211


October 2013


STRUCTURAL CHANGE AND ECONOMIC DEVELOPMENT:


IS BRAZIL CATCHING UP OR FALLING BEHIND?






STRUCTURAL CHANGE AND ECONOMIC DEVELOPMENT:
IS BRAZIL CATCHING UP OR FALLING BEHIND?




André Nassif, Carmem Feijó and Eliane Araújo


No. 211
October 2013


Acknowledgements: The authors thank Luis Carlos Bresser Pereira, José Luis Oreiro, David Kupfer, Fernando
Puga, Erika Amorim, Claudio Leal, Jorge Arbache, Samuel Pessôa and the two anonymous UNCTAD referees
for their comments and suggestions. Any remaining failures are the authors’ responsibility.


UNCTAD/OSG/DP/2013/1




ii


The opinions expressed in this paper are those of the authors and are not to be taken as the official views
of the UNCTAD secretariat or its Member States. The designations and terminology employed are also
those of the authors.


UNCTAD Discussion Papers are read anonymously by at least one referee, whose comments are taken
into account before publication.


Comments on this paper are invited and may be addressed to the author, c/o the Publications Assistant,
Macroeconomic and Development Policies Branch (MDPB), Division on Globalization and Development
Strategies (DGDS), United Nations Conference on Trade and Development (UNCTAD), Palais des Nations,
CH-1211 Geneva 10, Switzerland; e-mail: gdsinfo@unctad.org; fax no: +41 22 917 0274. Copies of
Discussion Papers may also be obtained from this address.


UNCTAD Discussion Papers are available at http://unctad.org.


JEL classification: O11; O40; O47




iii


Contents


Page


Abstract ....................................................................................................................................................... 1


I. INTRODUCTION ................................................................................................................................ 1


II. STRUCTURAL CHANGE AND ECONOMIC DEVELOPMENT:
THE THEORETICAL FRAMEWORK ............................................................................................ 3
A. Towards a structuralist theory of economic development: a Kaldorian-Thirlwallian approach ....... 3
B. Stylized facts and a basic hypothesis on structural change and economic development .................. 6


III. STRUCTURAL CHANGE IN THE BRAZILIAN ECONOMY: INDUSTRIALIZATION
(OR DE-INDUSTRIALIZATION), CATCHING UP (OR FALLING BEHIND)
SINCE THE 1970s? .............................................................................................................................. 8
A. An overview of the Brazilian industrialization process .................................................................... 8
B. The connections between the stylized facts on structural change and economic development
and the empirical evidence for the Brazilian economy ................................................................... 10


IV. ECONOMETRIC EVIDENCE: THE KALDOR-VERDOORN’S
AND THIRLWALL’S LAWS ............................................................................................................. 18
A. Estimating the Kaldor-Verdoorn coefficient for the Brazilian economy (1970–2010) ................... 18
B. Estimating the income elasticity of demand for Brazilian exports and imports (1980–2010) ....... 19


V. CONCLUDING REMARKS ....................................................................................................... 22


ANNEX A: Manufacturing industry according to technological intensity ................................................ 24
ANNEX B: Data sources ............................................................................................................................ 25


REFERENCES ........................................................................................................................................... 26


List of tables
1 The main sectors of activity as a share of total real value added in Brazil, selected years .................... 9
2 Participation of employment in Brazil by sector of economic activity, selected years ....................... 13
3 Estimate of the Kaldor-Verdoorn coefficient for the Brazilian economy, 1970–2010,
econometric estimate of equation (2) ................................................................................................... 19
4 Explanatory factors and income elasticity of demand for imports in Brazil ....................................... 20
5 Explanatory factors and income elasticity of demand for exports in Brazil ....................................... 21
6 Thirlwall’s Law .................................................................................................................................... 21


List of figures


1 Value added in the Brazilian manufacturing sector according to technological intensity,
selected years ........................................................................................................................................ 11
2 Percentage of total employment in the Brazilian manufacturing sector according to
technological intensity, selected years .................................................................................................. 12
3 Labour productivity growth of the Brazilian manufacturing industry, 1970–2008 ............................ 15
4 The Brazilian technological gap: relative labour productivity in the Brazilian
manufacturing sector compared with that of the United States, 1970–2008 ........................................ 15
5 Brazilian total exports classified by intensity of value added, 1970–2010 .......................................... 16
6 Trade balance of the Brazilian manufacturing industries classified by
technological intensity, 1970–2008 ...................................................................................................... 17






STRUCTURAL CHANGE AND ECONOMIC DEVELOPMENT:
IS BRAZIL CATCHING UP OR FALLING BEHIND?


André Nassif,* Carmem Feijó** and Eliane Araújo***


Abstract


We present a Kaldor-Thirlwall theoretical and empirical framework on the basic driving forces of
the behaviour of productivity and economic development in the long-run. By calculating the so-
called Thirlwall equation, the main contribution of our research is to examine whether Brazil has
been catching up or falling behind. We show some empirical evidence based on both descriptive
statistics and econometric regressions for Brazil between 1970 and 2010. Some important indicators
of descriptive statistics reveal that Brazil has entered into a process of early de-industrialization.
In addition, since our econometric estimates also show that there was a dramatic increase in the
income elasticity of demand for imports between 1980–1998 and 1999–2010 (from 1.97 to 3.36) and
a small decrease in the income elasticity for exports during the same periods (from 1.36 to 1.33), we
conclude that Brazil not only has already embarked on a trajectory of falling-behind relative to the
world economy and the international economic frontier, but also that it might show, in the absence
of appropriate policies, lower growth rates in the long run. However, if the opposite occurs, it would
face major long-term external constraints to growth.


I. INTRODUCTION


Since classical political economy, questions related to how to accelerate and sustain economic development
in order to achieve both high real income per capita and social well-being levels have always been in the
focus of economics. Many economists believe that economic development depends on mostly economic
factors, but it is also, in fact, influenced by a set of non-economic phenomena, such as history, geography,
anthropology, sociology, culture, among others.1 However, the isolation of any other non-economic or
even economic factor – note that when theoretical economists do this, they use the common expression
“considering all else equal” or what is known as the “ceteris paribus” hypothesis – is not a disadvantage
per se when we are interested in capturing the main economic factors which influence most economic
phenomena. For instance, in the case of economic development, nowadays there is practically a consensus
that it is mostly determined by growth in productivity over time, which is, in turn, concretized by physical


1 See the provocative book by McCloskey (2010).


* Fluminense Federal University and The Brazilian Development Bank (BNDES), Brazil; e-mail: andrenassif27@
gmail.com.


** Fluminense Federal University, Brazil; e-mail: cfeijo@terra.com.br.
*** State University of Maringá, Brazil; e-mail: elianedearaujo@gmail.com.




2


and human capital accumulation as well as technological progress.2 In practical terms, these latter factors
are responsible for sustaining high rates of growth in productivity in a particular country, and therefore,
considering the average growth in the world economy as a given, for accelerating the process of catching up.


Yet, the central point of divergence among economists is concerned with the most important sources for
boosting growth rates in productivity both in absolute and relative terms in a developing country in order
to reduce the technological gap with respect to developed countries and, therefore, to ease the process of
catching up in the long run.3 The focus of this paper has two characteristics: first, it is structuralist and
Kaldorian in the sense that the behaviour of productivity in the long run is mainly determined by factors
that affect aggregate demand, which, in turn, is strongly conditioned by both monetary, fiscal and exchange
rate policies; and second, that behaviour, when successful, is reflected, firstly, in a strong shift of resources
from traditional sectors like the primary one to the manufacturing one, and then, after the country has
achieved high levels of real income per capita, to the service sector. Several authors have argued that,
together with other facts, when the rapid and large loss of participation of the manufacturing sector in
total value added and mainly in total employment occurs before the country has reached high levels of
per capita income, this phenomenon characterizes an early de-industrialization4 and can strongly retard
the process of catching up or even put the economy into a trajectory of falling behind.5, 6 It is important
to remark that the significant loss in participation of the manufacturing sector in total value added is a
natural phenomenon in developed countries, because at this stage of development, the domestic income
elasticity of demand for services is a little higher than that for manufactured goods (see Clark, 1940).


The theoretical framework of this paper is based on Kaldor and Thirlwall theories on the basic driving
forces of the behaviour of productivity and economic growth in the long run. From the Kaldorian view, we
will base our work on the hypothesis that the main sources of the behaviour of the aggregate productivity
come from the manufacturing sector. The main empirical justification is that this sector, compared to
the primary and service sectors, operates under significant static and dynamic economies of scale, such
that it has the highest capacity to disseminate its gains from productivity to the economy as a whole.7
By anchoring on this Kaldorian hypothesis, we also show that the more a country is able to construct a
large and diversified manufacturing sector during its catching up process, the more will be its capacity
to sustain high rates of economic growth in the long run. In the literature on economic development,


2 Within the neoclassical literature, see Lucas (1988), Romer (1986, 1990), and Grossman and Helpman (1991), among
others. For a structuralist-evolutionary view, see Nelson and Winter (1982), Fagerberg (1988), and Dosi, Pavitt and Soete
(1990), among others.
3 In economics, the long run refers to economic time, rather than calendar time. However, even in terms of calendar
time, the period during which a country becomes successful in catching up depends on several factors, such as the size
of the country, the quality of the industrial, technological and institutional policies implemented, and a fine coordination
of these latter policies with macroeconomic policies. For instance, the Republic of Korea could catch up in a period of
around 25 years, while China, if it will actually succeed in doing that, will take much more than 30 years. See Amsden
(1989) for the Republic of Korea, and Amsden (2001), for comparing the hard efforts of several developing countries to
search for adequate strategies to promote structural change and catching up.
4 For other details on the recent literature about early de-industrialization, see Rowthorn and Wells (1987), Rowthorn
and Ramaswany (1999) and Palma (2005).
5 In the definition of de-industrialization, Rowthorn and Wells (1987) gives special importance to the loss of participation
of the manufacturing sector in total employment, rather than in total value added.
6 Then, the significant loss of participation of the manufacturing sector in total value added is a natural phenomenon in
developed countries, because at this stage of development, the domestic income elasticity of demand for services is a
little higher than that for manufactured goods. For other details on de-industrialization, see Rowthorn and Wells (1987),
and Rowthorn and Ramaswany (1999).
7 Static economies of scale occur when a firm doubles its total investment and the amount of the new production more
than doubles. Therefore, considering that the factor prices used in that investment are kept constant, the long-term unit cost
reduces. Dynamic economies of scale occur when a firm is able to reduce long-term unit costs by implementing successful
innovations over time and, therefore, tends to accumulate learning-by-doing, knowledge and major technological capacity.




3


this relationship between the real output of the manufacturing industry and its positive effects on the
productivity of the economy as a whole is known as the Kaldor-Verdoorn Law.


This Kaldorian theoretical framework will additionally be complemented by the Thirlwall hypothesis
on the importance of a country to have an income elasticity of demand for exports above the income
elasticity of demand for imports if the country intends to sustain economic growth without facing balance
of payments constraints (Thirlwall’s Law). This hypothesis makes it clear why it is important for a
developing country to have not only a large and diversified export composition, but also the majority of
its net exports (exports minus imports) basket composed of goods of high income elasticity of demand
in the long run.


The relevance of the paper for an emerging country like Brazil is that, despite the fact that this country
has been successful in building a relatively large and diversified manufacturing sector, it is still not an
industrialized country in the Kaldorian point of view. That is to say, it has not yet reached the same
stage of maturity as developed countries.8 So, since responding to part of the title of this paper (“is
Brazil catching up or falling behind?”) has normative implications, the paper can assist policymakers
in evaluating whether or not the actual short and long-term economic policies lato sensu (industrial and
technological policy, macroeconomic policy and so on) are in tune with each other in order to sustain
the long-term economic growth of the Brazilian economy and promote the process of catching up with
developed countries.


Besides this Introduction and the Conclusion, the paper is organized as follows. Section II presents
the main hypothesis and stylized facts on structural change and economic development based on the
Kaldor-Thirlwall theories. Section III, by connecting the previous stylized facts and a basic hypothesis
on the same theme with empirical evidence on the Brazilian economy based on descriptive statistics
for the 1970–2010 period, provides a preliminary answer of whether or not Brazil has already entered
into early de-industrialization.9 Section IV presents two important pieces of econometric evidence: first,
by estimating the so-called Kaldor-Verdoorn coefficient for the 1970–2010 period, we will show if the
Brazilian manufacturing sector operates under dynamic economies of scale, a necessary (but not sufficient)
condition for sustaining both high rates of productivity and growth in the long run; and second, we will
estimate the long-term income elasticities of demand for imports and exports for the Brazilian economy
for the 1980–2010 period. This is essential evidence for showing if Brazil has either faced major external
constraints to long-term economic growth or kept more distant from the international technological frontier.


II. STRUCTURAL CHANGE AND ECONOMIC DEVELOPMENT:
THE THEORETICAL FRAMEWORK


A. Towards a structuralist theory of economic development: a Kaldorian-Thirlwallian approach


Most economists today practically agree with the hypothesis that both innovation and technological
spillovers are the main engine for explaining productivity growth. However, while neoclassical economists
tend to give all sectors of the economy equal weight for explaining the productivity behaviour of the
economy as a whole, structuralist economists, by identifying the manufacturing industry as the main
creator and disseminator source of technical progress as well as the principal source of significant static and


8 For a discussion about the stages of development of the Brazilian manufacturing industry, see Feijó and Lamonica (2012).
9 Due to unavailability of data, descriptive statistics on the Brazilian manufacturing sector will only cover the period
between 1970 and 2008.




4


dynamic increasing returns to scale, argue that the manufacturing sector is the main force for explaining
the aggregate productivity.10


The importance given to the maximization of static and dynamic increasing returns to scale as the main
explanatory factor for boosting both aggregate productivity and (therefore) economic development is
relatively old in economics. The general idea was presented not only by authors like Adam Smith (1776)
and Allyn Young (1928), but also by authors like Paul Rosenstein-Rodan (1943), Albert Hirschman (1958)
and Gunnar Myrdal (1957), among others. However, it was not until the publication of a set of Kaldor’s
theoretical and empirical essays in the mid-1960s and the 1970s that the justification of the manufacturing
sector as the location for most industries subject to increasing returns to scale was so clearly and precisely
demonstrated (see Kaldor, 1966, 1967, 1968, 1970 and 1975).


Before summarizing Kaldor’s main hypothesis on the relationship between structural change and economic
development, two points must be stressed: first, as Clark (1940) had already pointed out, Kaldor (1966)
also recognized that as soon as economic development reaches maturity – that is to say, the stage in
which countries, by having already caught up, are able to exhibit high levels of income per capita and
well being –, a relatively significant loss of participation of the manufacturing industry in total real
GDP is accompanied by a major participation of the service sector; and second, one could argue that,
since a lot (but not the majority) of the new ideas, knowledge and dynamic economies of scale are now
being generated in the tradable service sector, the effects of the microelectronic and telecommunication
revolution (for most, the third industrial revolution) on the representative role of the manufacturing sector
for economic development, as supported by Kaldor (1966), is becoming passé.


As to that latter point, the criticism is misleading for several reasons: first, as Kaldor presented his
principal arguments when the service sector was composed basically of nontradables, he certainly would
have recognized the role of the tradable services today as additional forces to those coming from the
manufacturing sector in accelerating and sustaining the development process; second, even if we take into
account the increasing participation of several important tradable services (e.g. software) as being subject
to static and dynamic increasing returns to scale in the entire service sector, the fact is that the majority of
this latter phenomenon occurs and could continue to occur in the process of manufacturing production;11
and third, and perhaps more importantly, following the insights pioneered by Young (1928), Kaldor (1966:
106) stresses that both static and dynamic “increasing returns (to scale) are a “macro-phenomenon” – just
because so much of the economies of scale emerge as a result of increased differentiation, the emergence
of new process and new subsidiary industries, they cannot be discerned adequately by observing the
effects of variations in the size of an individual firm or a particular industry”.


The interpretation of the static and dynamic economies of scale as a “macro-phenomenon” is essential
for understanding Kaldor’s hypothesis on the importance for developing countries to have a strong and
diversified manufacturing industry, especially during the time when their manufacturing sector is in a
trajectory departing from the immaturity to maturity stage. In addition, it is important to stress that, in
a Kaldorian framework, the more a country has its manufacturing industry formed by segments which
operate under static and dynamic increasing returns to scale, the more rapid is its catching up process.
Then, for Kaldor (1966), economic development is a process through which structural change happens.
That is to say, the productive resources are strongly reallocated from the traditional sector (especially
agriculture) to the manufacturing sector (mainly those segments of more technological sophistication,


10 Karl Marx was, perhaps, the first author to emphasize the importance of technical progress in the development process
(Marx, 1887, especially Vol. I, chapter XV, “Machinery and Large-Scale Industry”). Yet, the incorporation of this force
as endogenous to economic development was only first emphasized by Schumpeter (1942).
11 To give an example, various activities generated in the conception of goods produced by Apple, such as the creative
ideas, knowledge and engineering of projects may be developed in the service sector. However, the majority of Apple
family goods (i-pods, i-phones, i-pads and so on) are produced by segments of the manufacturing sector which operate
under conditions of significant increasing returns to scale.




5


namely those that are engineering-, science- and knowledge-based).12 Only when a country has already
reached the maturity stage (in other words, a developed country with an income per capita significantly
above the world average) could the loss of the participation of the manufacturing industry in total GDP
be accepted as natural.


In his seminal paper, Kaldor (1966) enumerated the following hypotheses on structural change and
economic development as well as giving the econometric evidence which supports them:13


(i) Long-term economic growth of a country is largely associated with the relative size and diversification
of its manufacturing industry. Not only due to its high capacity for generating innovation and
disseminating technological spillovers throughout the economy as a whole, but also by virtue of
its above mentioned presence in static and dynamic increasing returns to scale, the manufacturing
sector dictates the dynamism of aggregate productivity growth.


(ii) Insofar as the static and dynamic economies of scale presented in the manufacturing sector are
understood as a “macro-phenomenon”, once economic development is sustained, the synergies
between the increase in productivity in the manufacturing industry and the positive changes in
productivity in the economy as a whole make aggregate productivity become largely associated
with the increase in total output. This result, which is known as the Kaldor-Verdoorn Law, is largely
explained by dynamic economies of scale.14, 15 By associating the different levels of per capita income
reached by a country with a minor or major propensity to consume manufacturing goods, Kaldor
was perhaps the first author to call attention to the importance of preventing a country from early
de-industrialization, especially at the stage during which it has reached a level of income per capita
close to the world average. In fact, since at this stage societies tend to have a high propensity to
consume manufactured goods – that is to say, their income elasticity of demand is significantly above
one –, a high and sustained growth of the manufacturing sector contributes to boosting economic
growth in the long run.


(iii) As if anticipating Thirlwall’s (1979) model of the balance-of-payment constraint to growth, Kaldor
(1966) suggested that, mainly in either intermediate or relatively advanced stages of development,
net exports must increase at a faster rate in order to finance the high need of imported capital
goods. The important question that could be raised as to this point is as follows: if it is a high
demand increase in the manufacturing sector which governs the pace of growth in the economy as
a whole, why should a weak foreign demand for exports constrain economic development even in
large economies like the United States, China or Brazil? The answer is far from being associated
with supporting an export-led growth strategy for these countries, insofar as the large size of their
domestic market is perhaps more important than exports for boosting the advantages of economies
of scale in the manufacturing sector. The main reason is that the more a country can augment and
diversify its exports through a major composition of goods with high income elasticity of demand,
the less will be the external constraint to economic growth in the long run. In fact, if one observes


12 In our empirical analysis ahead, we will break down the manufacturing sector into three groups: labour intensive;
natural resource-based; and engineering-, science- and knowledge-based.
13 McCombie and Thirlwall (1994) call these hypotheses Kaldor’s Laws.
14 Verdoorn (1949) was the first author to suggest this association.
15 Kaldor (1966) showed econometric evidence on this relationship (see table 2: 107). As McCombie and Thirlwall (1994,
chap. 2) summarized, despite a variety of empirical works attempted to validate or not the Kaldor-Verdoorn Law for a
group of countries in the long run, the results, in principle, seemed not to be conclusive. However, since Kaldor (1970)
later argued that it is more appropriate to estimate the “differences in “regional” growth rates [taking into account the]
different areas within the same country […], for there are few, if any, economic barriers to the interregional mobility of
capital and to interstate migration”, McCombie and Thirlwall (1994: 209) also summarized the estimates of the Kaldor-
Verdoorn Law respecting Kaldor’s suggestion and concluded that “the assumption of output growth is fundamentally
demand, rather than supply, determined more plausible at both the international and the regional level”.




6


experiences of economic development in countries like Japan after World War II, the Republic of
Korea from 1960 to1985 and China from 1979 on, it could be concluded that all of them have been
successful in shifting the composition of their exports from small kinds of traditional goods (labour
intensive and natural resource-based) to very diversified goods, especially science-, engineering- and
knowledge-based goods which make up the majority of total exports.16


(iv) The behaviour of economic growth is not supply-constrained neither in the short nor in the long run.
As to this point, if, on the one hand, Kaldor supports Keynes’s (1936) hypothesis that the behaviour of
aggregate demand in the short run explains the business cycles and that the insufficiency of effective
demand explains recessions and depressions, on the other hand, since the Kaldor-Verdoorn Law is in
operation and under the assumption that there are no external constraints to growth, the supply of the
economy can elastically respond to the increase in demand in the long run within limits. As a matter
of fact, Kaldor does not completely discard the role of some supply constraint to long-term growth.
In fact, he observes at least two main forces from the supply side that could constrain long-term
growth: first, the slow capacity of supply to respond to the increase in demand, especially of inputs
and raw materials; and second, the shortage of labour supply. In the first case, Kaldor (1966: 115)
argued that, unless there is a balance of payment constraint, there is no reason to believe in a supply
constraint of this kind because it would only arise if “a particular rate of growth generates a rate of
growth of imports which exceeds the rate of growth of exports”. In the second case, the answer is a
little more complex. Kaldor (1966) accepts that as the country is in a trajectory of catching up and
structural change, sooner or later, the unlimited labour supply (see Lewis, 1954) provided by the
excess of manpower coming from agriculture will be over. While the manufacturing sector is not able
to completely absorb the labour force growth as the economy shifts from immaturity to a developed
stage, since the service sector is less sensitive to demand fluctuations, it could employ the excess
labour supply to demand. This explains why de-industrialization in mature economies tends to reduce
the participation of both manufacturing (expressed by the value added) and employment in total real
GDP. However, by implicitly adopting a Schumpeterian view, Kaldor (1966: 117–121) argues that
technical progress is the dramatic engine that tends to “so radically reduce the labour requirements
in industry as to make it possible to combine growth with falling industrial employment” in such a
way that there is nothing but a “reservoir of surplus labour, or disguised unemployment, instead of
a shortage of labour in the long run”. This conclusion reinforces Kaldor’s main hypothesis that, by
reducing external constraint, it is the dynamic economies of scale associated with the manufacturing
sector, governed, in turn, by the demand side, which explains the long-term growth and the catching
up process.


B. Stylized facts and a basic hypothesis on structural change and economic development


In this subsection, we will present some stylized facts and a basic hypothesis on structural change and
economic development which, together with the above discussed theoretical framework, will serve as
the analytical basis for the empirical evidence that will be shown in sections III and IV.17


Stylized fact 1: Economic development is a process of deep structural change of the economy


As Prebisch (1950) has emphasized a long time ago, although economic development is basically
determined by technical progress, the issue is that technology is neither generated nor equally distributed


16 As will be shown ahead, in the empirical evidence, we will break down the manufacturing industry into three
groups classified according to their technological sophistication: labour intensive; natural resource-based; and science-,
engineering- and knowledge-based. The respective segments of each group are described in annex A.
17 Most of these stylized facts are based on Cimoli and Porcile (2011).




7


between developed and developing countries. Cimoli, Porcile and Rovira (2010), Cimoli and Porcile
(2010a; 2010b) and Cimoli and Porcile (2011) have reactivated some of Prebisch’s and Kaldor’s original
ideas in order to support empirical evidence on how, by dominating the technical progress of the world,
the productive structure of developed countries continues to be much more complex and diversified than
that of developing countries. This means that economic development can be understood as a process
through which a deep structural change occurs in the economy, in such a way that there is a reallocation
of resources from the primary sector to the manufacturing sector, and, then, as soon as countries have
achieved high levels of income per capita, from that latter sector to the service sector. So, economic
development with deep structural change means that both the productive structure of the economy and
the composition of net exports are mostly dominated by the presence of manufacturing segments that
produce science-, engineering- and knowledge-based goods.


Stylized fact 2: Countries characterized by a productive structure and pattern of specialization
concentrated in primary goods or natural resource-based manufactured commodities tend to direct
most of the employment toward these sectors


In countries that basically specialize in primary products and natural resource-based manufactured
commodities, the majority of the employment generated tends to be absorbed by these sectors. Even during
the boom phases of the business cycle, the presence of a significant part of the informal employment
can be dissembled and masked by a “paradox of full employment” (see Cimoli and Porcile, 2010b).
From a structuralist perspective, given the dramatic level of the informality of manpower, however, the
unemployment rate is usually relatively high. Not by chance, these countries tend to weakly sustain the
international competitiveness of these goods based on low wages. However, whenever either the domestic
currencies of these countries are overvalued in real terms or the terms of trade are temporarily favourable,
wages tend to artificially increase and destroy that “spurious” kind of competitiveness (Fajnzylber, 1988).
The main implication is that, by being basically determined by low relative wages, rather than by higher
relative labour productivity, this pattern of static comparative advantage is not favourable to economic
development (see Cimoli and Porcile, 2010a).


Stylized fact 3: The more rapidly a country is able to build and sustain a large and diversified
manufacturing sector with a significant participation in total GDP and total net exports, the more
rapidly it will catch up.


When several factors (lack of appropriate short-term and long-term economic policies, inadequate
institutions, lack of political consensus on how to accelerate economic development, etc.) prevent
countries from developing a large and diversified manufacturing industry, they tend to specialize in goods
with static comparative advantage, especially in labour intensive and natural resource-based goods.18
Since these sectors have low capacity of generating and diffusing technical progress as well as of taking
advantage of dynamic economies of scale, the technological gap increases to such a magnitude that the
development process tends to become locked-in in a regressive technological path (see Arthur, 1989).
So, the catching up process is more rapid when countries are successful in producing structural change
and redirecting most of their exports towards science-, engineering- and knowledge-based industries.


18 One could correctly argue that some natural resource-based manufacturing segments (e.g., oil refining, natural gas, paper
and cellulose, and others) have a high capacity to generate technological spillovers and produce backward and forward
linkages with industries of medium or high technology. For instance, investments and appropriate industrial policies in
the oil industry (e.g. government purchasing policy and tax stimulus) could boost some industries of medium or high
technological sophistication (through mainly backward linkages), such as the shipbuilding industry, oil and petrochemical
equipment and some others. The issue is that a country will be hardly able to produce and sustain significant dynamic
economies of scale in the Kaldorian sense in the long run if it becomes highly dependent on a small group of natural
resource-based industries.




8


Hausmann, Hwang and Rodrik (2007) showed that “what you export matters”. That is to say, countries
whose export basket has a significant part composed of products of high technological sophistication tend
to show higher rates of economic growth in the long run. By contrast, if their productive structure and
pattern of specialization is dominated by labour intensive and natural resource-based goods, countries
tend to experience a falling behind path.


Basic hypothesis on structural change and economic development: The more a country directs its
productive structure and net export basket to very diversified goods, with a major presence of the science-,
engineering- and knowledge-based sectors, the larger will be its degree of intensity of structural change.


This above hypothesis depends on the so-called Thirlwall’s Law (see Thirlwall, 1979), as follows:



M


Xi


Y
Y


p
e


=
*


(1)


Where iY is the rate of economic growth in the domestic country; *Y is the rate of world economic growth;
Xe is the income elasticity of demand for exports; and Mp is the income elasticity of demand for imports.


Thirlwall’s Law suggests that the convergence of the rate of economic growth in a particular country
(say, a developing country) to the world economic growth depends on the ratio between the income
elasticity of demand for exports and income elasticity of demand for imports. In other words, if the
income elasticity of demand for exports increases above the income elasticity of demand for imports,
economic development is sustained because economic growth will not be constrained by balance of
payments constraints in the long run.


III. STRUCTURAL CHANGE IN THE BRAZILIAN ECONOMY: INDUSTRIALIZATION
(OR DE-INDUSTRIALIZATION), CATCHING UP (OR FALLING BEHIND)


SINCE THE 1970s?


This section shows descriptive statistics on the evolution of the structural change and economic
development in Brazil. The statistics on foreign trade cover the period 1970–2010. By contrast, due to
a lack of data, most of the indicators related to Brazil’s manufacturing performance cover the period
1970–2008.


A. An overview of the Brazilian industrialization process


From 1947 to 1980, the Brazilian economy grew (6.4 per cent per year) above the average of both developed
and underdeveloped economies, showing a strong vigour under the leadership of the manufacturing sector.
This sector exhibited an average growth of 8.6 per cent per year in the same period.19 During the 1970s,
particularly, the industrial sector showed average yearly growth rates close to 7.4 per cent. These results
can be explained to a great extent by the strategy of import substitution to complete the industrialization
process of the country. Indeed, between 1947 and 1980, the manufacturing sector increased its participation
in total GDP from 19.3 per cent to 31.3 per cent (table 1). This increase can be, at least partially, credited
to the relative success of the implementation of three national development plans: Targets Plan (Plano de
Metas, 1956–1961), the First National Development Plan (I PND, 1968–1973) and the Second National


19 This expansion was not continuous over time, exposing the period 1963–1967 as the least dynamic (average economic
growth rates of 2.7 per cent per annum).




9


Development Plan (II PND, 1974–1979), which contributed to promoting significant changes in the
productive structure of the country (Castro, 1985).


Although many criticisms received by the actual implementation of the strategy of the import substitution
regime in Brazil are more than valid (e.g. very high rates of effective protection compared with those
imposed by some successful Asian countries, inefficiency in policy implementation, negligence in carrying
out policy evaluation, lack of high investments in education, among others), that strategy was responsible
for the building of a large and diversified manufacturing sector in Brazil. Except for a few short periods,
the extraordinary pace of economic growth was only interrupted, at the end of the 1970s, by the external
debt crisis in 1980–1982, which revealed the unsustainability of the chosen model to finance development
policies with external savings (basically through long-term foreign bank lendings subjected to flexible
interest rates) adopted especially during the 1970s. As a result of the relative loss in dynamism of the
manufacturing sector, its relative weight in total GDP fell. In addition, from 1980 to 2010, since the
general priorities of the economic policy – particularly between 1980 and the mid-1990s – concentrated
in stabilizing high inflation and promoting the adjustment of severe problems of balance of payments
and the external debt, the average growth rate of the real GDP was around only 2.7 per cent per annum.


Although a new growth cycle started in 2004, the 2008 financial global crisis suddenly interrupted this
trend. The average growth rate of the real GDP between 2004 and 2008 increased to 4.7 per cent, though
this performance has been much worse than that of the set of all developing countries.20 Table 1 breaks
down the three sectors of the Brazilian economy and shows how Brazil’s productive structure has been
changing over the last 60 years.21 In the recent debate about the reasons for the relatively slow growth of
the Brazilian economy in comparison with similar economies, the dramatic loss of the relative importance
of the manufacturing sector from the 1980s on has been pointed out as one of the main causes to be
investigated.22


20 To refer to a larger period, the poor performance of the Brazilian economy between 1999 and 2012 can be revealed
by comparing the annual average growth rates of real GDP in Brazil (3.2 per cent) with all developing and emerging
countries (5.7 per cent). Data for Brazil is from the Brazilian Institute of Geography and Statistics and for developing
and emerging countries from the World Bank database. For details, see Nassif (2011).
21 It should be observed that from 2000 onwards the Brazilian System of National Accounts changed the reference base
for its estimates, using data from the annual surveys more extensively. If, on the one hand, this improvement implies
more accuracy in the measure of the contribution of each sector to total value added and employment, on the other hand,
it has also obviously become more difficult to make comparisons of data with those of previous years. See IBGE (2007).
22 For a long-term analysis of the Brazilian experience of growth and a review of the hot debate between opposing
interpretations, see Bresser-Pereira (2006).


Table 1
The main sectors of activity as a share of total real value added in Brazil, selected years


(Per cent)


1947 1960 1970 1980 1990 2000 2010


Agriculture 20.7 17.8 11.6 10.1 6.3 5.6 5.5


Total industry 25.2 32.2 35.8 40.9 30.1 27.7 27.5
Manufacturing 19.3 25.6 27.4 31.3 20.7 17.2 14.6


Services 54.1 50.0 52.6 49.0 63.6 66.7 67.0


Source: System of National Accounts, Brazilian Institute of Geography and Statistics (IBGE).




10


B. The connections between the stylized facts on structural change and economic development and
the empirical evidence for the Brazilian economy


In order to deepen the discussion about structural change in Brazil, this section will focus on the Brazilian
manufacturing industry following the Kaldorian view about its strategic role in the catching up process
of an economy. We will connect the stylized facts already discussed in section II with the indicators of
descriptive statistics on the Brazilian economy in the period under study. Except for a few, most of the
indicators which will be presented were based on the PADI (Analysis Program of Industrial Dynamics)
database of the Economic Commission for Latin American and the Caribbean (ECLAC) that exploits
structural data from the manufacturing sector.23


As seen in table 1, the augmented participation of the manufacturing industry in the total value added
of the Brazilian economy until the 1980s was also accompanied by a broad diversification of its
structure. The 1970s marked the establishment of a diversified productive structure of durable, non-
durable, capital goods and intermediary goods, which represented the conclusion of the last stage of the
process of import substitution (see Castro, 1985).24 As it will be shown below, the diversification in the
industrial production also led to the diversification of exports: in 1964 the share of manufactured and
semi-manufactured goods reached 14.0 per cent and in 1980 this percentage increased to 57.0 per cent.25
The consolidation of a diversified industrial structure occurred in the mid-1980s with the maturation
of investments implemented in the context of II PND (1974–1979). From the early-1990s on, a set of
liberalizing reforms was implemented in the Brazilian economy (e.g. the trade liberalization, external
financial openness, privatization of state enterprises, among others) within a context in which high inflation
and serious problems in the external sector of the economy prevailed. The rapid loss of participation of
the value added of the Brazilian manufacturing sector in total real GDP, which actually had begun in the
mid-1980s, as well as a growing international specialization in exports of primary products and natural
resource-based manufactured commodities, opened an intense debate in the 2000s on whether or not
Brazil suffered from early de-industrialization.26


Stylized facts 1 and 2: structural change in the Brazilian manufacturing industry: resource allocations
and employment


According to table 1 above, the dramatic drop of participation of the manufacturing industry in total
value added since the 1980s could be, in principle, taken as a sign of early de-industrialization in Brazil.
However, according to the old and new literature on the theme, the diagnosis of early de-industrialization
must also take into account other indicators such as the relative growth rates of the manufacturing sector,
the reallocations of resources in terms of value added, employment and exports within the manufacturing
industry and throughout the economy as a whole, as well as the changes in both productivity and the


23 ECLAC-PADI is an internationally harmonized database, which collects statistical information from national statistical
offices. In the Brazilian case, the main source of information comes from the Industrial Censuses and the Annual Industrial
Surveys (Pesquisa Industrial Annual – PIA) carried out by the Brazilian Institute of Geography and Statistics (IBGE).
Data for missing years from IBGE’s surveys have been estimated, following ECLAC’s methodological procedures.
Whenever applicable, all information provided in ECLAC-PADI has been converted from each country’s currency to
1985 constant dollar prices.
24 According to the Industrial Census of the Brazilian Institute of Geography and Statistics (IBGE), in 1980 more than
50 per cent of the value added of manufacturing and mining industries came from segments classified as scale-intensive
(38.6 per cent), differentiated (11.7 per cent) and science-based (2.8 per cent).
25 It is worth noting that even with the debt crisis in the 1980s, considering only manufactured exports, these amounted
to more than 50 per cent of total exports in 1981, and this percentage remained at this level or above until 2008, when it
fell to 47.0 per cent (Brazilian Ministry of Development, Industry and Commerce -MDIC).
26 On the debate about early de-industrialization in Brazil, see Palma (2005), IEDI (2005), Jank et al. (2008), Nassif
(2008), Oreiro and Feijo (2010), and Bacha and Fishlow (2011), among others.




11


technological gap over time. In fact, while Rowthorn and Wells (1987) and Rowthorn and Ramaswany
(1999) prefer to diagnose early de-industrialization looking at the evolution of economic indicators of
a country in absolute terms (like the participation of the manufacturing industry in total value added,
employment, exports and so on), Palma (2005) and Bresser-Pereira (2008, 2010), following structuralist
and Kaldor’s traditions (1966), give special attention to the relative position of a country’s manufacturing
industry in the world economy.


Initially, we will be interested in evaluating what happened only within the Brazilian manufacturing
industry between 1970 and 2008, and not in the economy as a whole. Figure 1 shows how the Brazilian
industrial structure has been changing since the 1970s, considering a taxonomy of the manufacturing
sector according to the technological intensity applied in its productive process. This classification helps
to evaluate whether or not the structural change within the manufacturing sector was directed towards
the set of segments with higher technological intensity. As this taxonomy breaks down the manufacturing
sector into only three groups, a structural change that has been oriented towards the science-, engineering-
and knowledge-based segments might signal major potential for spillover gains from productivity over
the economy as whole, since those segments are more subject to static and dynamic economies of scale,
as Kaldor stressed a long time ago.27


The distribution of the manufacturing value added among the three groups of industries has clearly
moved towards an increase in the weight of the science-, engineering- and knowledge-based segment,
which accounted for 28.1 per cent of the total value added of the Brazilian manufacturing sector in 1970
and reached 45.8 per cent in 2008. This gain has dramatically been offset by the loss in importance of
labour-intensive industries, which reduced their share from 32.0 per cent in 1970 to 16.7 per cent in


27 As our taxonomy breaks down the Brazilian manufacturing industry into only three groups, one could correctly argue
that the high level of aggregation might compromise the analysis. However, unfortunately, in virtue of not having sufficient
data in the ECLAC-PADI database in the long period under investigation, we could not analyse the data of the Brazilian
manufacturing industry at a more disaggregate level.


Figure 1
Value added in the Brazilian manufacturing sector according to technological intensity,


selected years


0


10


20


30


40


50


1970 1980 1990 2000 2008


Science, engineering and knowledge based Natural resource based Labour intensive


Source: Economic Commission for Latin American and the Caribbean (ECLAC)-Analysis Program of Industrial Dynamics (PADI)
database.




12


2008. The natural resource-based industries showed a small loss, although roughly maintained their
participation at over 30 per cent. Looking closely inside each group, we realized that the strong increase
in the participation of the science-, engineering- and knowledge-based segments was due to a growing
importance of machinery and equipment, electrical and transport industries. While in the labour intensive
group all industries reduced their importance, in the natural resource-based one, while the food industry
is still the most important, the oil refining industry gained weight in the period under consideration.


Figure 2 shows the distribution of employment over each group of industries. The same movement is
observed in the labour intensive industries which, differently from the other ones with higher employment
participation, were the only ones to lose importance in job creation over the time period: their contribution
falls from 41.0 per cent in 1970 to 33.8 per cent in 2008. The percentage of employment in the natural
resource-based industries changed little from 1970 (34.6 per cent) to 2008 (35.0 per cent). The science-,
engineering- and knowledge-based group is the one that shows the greatest increase in employment
participation, moving from 24.4 per cent in 1970 to 31.2 per cent in 2008. In 2008 the distribution of jobs
in the Brazilian manufacturing industry is about equal among the three groups of industries.


It is also important to analyse the evolution of the participation of employment in the Brazilian economy
as a whole. Table 2 shows indicators on this subject in Brazil only between 2000 and 2009, the period in
which data are available. Table 2 makes it clear that there was a shift in the workforce from agriculture, but
not from total industry (including manufacturing), to the service sector in the period under consideration.
In the case of the manufacturing industry, there was a small increase in the participation of employment
in total employment from 12.0 per cent to 12.7 per cent between 2000 and 2009. Except for construction,
which had a little increase from 6.7 per cent to 7.1 per cent between 2000 and 2009, the other subsectors
of Brazilian industry pretty much kept their relative participation.


Those combined above mentioned indicators are important for understanding the debate of whether or
not the Brazilian economy actually suffers from early de-industrialization. Given the complexity of this
process, it is not enough to look at only the loss of participation of the manufacturing sector in total value
added, which, in the Brazilian case, was relatively rapid and strong from the early 1990s on, as already


Figure 2
Percentage of total employment in the Brazilian manufacturing sector according to


technological intensity, selected years


0


10


20


30


40


50


1970 1980 1990 2000 2008


Science, engineering and knowledge based Natural resource based Labour intensive


Source: ECLAC-PADI database.




13


shown in table 1.28 Actually, as already discussed, if it is looked at from within its structure, the Brazilian
manufacturing industry not only still keeps a relatively well diversified structure, with more than 40 per
cent of the value added being generated in more sophisticated manufacturing segments, but is also able
to absorb a significant part of the workforce (figures 1 and 2, respectively). However, if we look at the
economy as a whole, the service sector was already responsible for around 62 per cent of the Brazilian
workforce in 2009 (table 2). In addition, as will be shown ahead, the Brazilian manufacturing industry is
getting more distant from the technological frontier. How to reconcile all this evidence?


In principle, the existence of a relatively diversified manufacturing sector and the keeping of a balanced
distribution of employment in manufacturing between old and more technologically sophisticated
industries – by not violating both stylized facts 1 and 2, as discussed in section II – may not be signalling
early de-industrialization in Brazil. However, since the Kaldorian tradition stresses the importance of the
relative position of the domestic manufacturing industry comparatively to the world economy, if a strong
augmentation of the technological gap and a poor performance of the evolution of both exports and trade
balance is verified, these indicators would be signalling a tendency of both early de-industrialization
together with falling behind, insofar as these results would be violating stylized fact 3 and the basic
hypothesis on structural change and economic development, according the Kaldor-Thirlwall theoretical
analysis of section II.


So, thus far, a partial conclusion on the structural change in the Brazilian manufacturing sector is that there
has been more diversification and it has moved towards segments of more technological sophistication.
As discussed in section II, an economy with a more diversified manufacturing structure tends to show
relatively higher aggregate productivity than countries strongly specialized in a few groups of traditional
industries.


28 Rigorously speaking, the loss of participation of the manufacturing sector in the Brazilian real GDP began in the
middle of the 1980s. So, as already argued by Nassif (2008), if we take into account just this indicator, this phenomenon
had begun in Brazil before the introduction of its liberalizing reforms in the early 1990s.


Table 2
Participation of employment in Brazil by sector of economic activity, selected years


(Per cent)


Sectors 2000 2003 2005 2007 2009


Agriculture 22.3 21.0 20.9 18.6 17.4


Total industry 19.5 19.0 20.0 20.1 20.5
Mining 0.3 0.3 0.3 0.3 0.3
Manufacturing 12.0 11.9 12.8 12.8 12.7
Utilities 0.4 0.4 0.4 0.4 0.4
Construction 6.7 6.4 6.5 6.6 7.1


Service 58.2 59.9 59.1 61.4 62.1


Total 100 100 100 100 100


Source: System of National Accounts, Brazilian Institute of Geography and Statistics (IBGE).




14


Stylized fact 3: the diversification of the Brazilian manufacturing industry and its sustainability –
evolution of the labour productivity, technological gap and pattern of international specialization


So far, the Brazilian manufacturing sector has shown a pattern of industrial change that could suggest that
the economy is sustaining a process of catching up. Yet, from a structuralist perspective, another piece of
information should be added to evaluate the profile of the competitiveness of the economy. This will be
done through the analysis of the behaviour of the Brazilian labour productivity as well as its technological
gap in relation to the international technological frontier of the three groups of industries. Figure 3
shows that, since the 1970s, the science-, engineering- and knowledge-based group has been leading the
growth of the Brazilian labour productivity. While the labour-intensive manufacturing segment has been
clearly behind the other ones, the natural resource-based group has followed close to the average of the
manufacturing sector as a whole. Figure 3 also shows that the labour productivity improved significantly
in the early-1990s, following Brazil’s trade liberalization, as has been documented in several studies on
the theme.29 However, around 1997 productivity gains began to slow down and a clear declining trend
is observed in all three groups of industries.


Although it can be realized that the rate of growth in the Brazilian manufacturing labour productivity has
been higher in segments of major technological sophistication, it is important to check whether or not
the speed of the productivity growth has been enough to keep the economy in a trajectory that reduces
its technological handicap. In the Kaldorian perspective, this point is absolutely important. In fact, as
Kaldor (1966: 104)) strongly stressed, “rather than the level, the differences in growth rates (among
countries) are largely accounted for by differences in the rates of growth of productivity (…) and also
that the incidence of technical progress – as measured by the rate of growth of productivity – is higher
in manufacturing activities than in other fields, so a great concentration of manufacturing increases the
overall rate of advance” (italics from the original).


An estimate of how far productivity gains are from the technological frontier is shown in figure 4, which
measures the rate of growth in the Brazilian manufacturing labour productivity by technological intensity
in comparison with that of the corresponding groups in the United States.30 This indicator can be evaluated
as a proxy for the technological gap.


According to figure 4,31 the productivity gap began to quickly and dramatically widen in all industrial
groups near the end of the 1990s. The science-, engineering- and knowledge-based industries showed
the lowest productivity gap compared with the other groups, until 2006, when from then on the natural
resource-based industries have shown better performance, even though they are still very far from the
technological frontier. It should be remarked that, in 1980, the science-, engineering- and knowledge-
based industries registered the lowest relative distance from the technological frontier (51 per cent).
So, in spite of the movement of the industrial structure towards more diversified and technologically
sophisticated sectors, the international comparison suggests that the Brazilian manufacturing industry
might be lagging behind. Rigorously speaking, independently from what happens with the other sectors


29 See Moreira and Correa (1998), Bonelli (2002), Feijó and Carvalho (2002) and Nassif (2003), among others.
30 Even taking into account that the United States has been losing importance in the global economy and is not in the
vanguard of all segments of the manufacturing sector, most studies consider that, on average, the country is still at the
international technological frontier.
31 This figure is the only one that follows the original classification of the manufacturing sector according to the
technological intensity available in ECLAC-PADI database, which misleadingly considers the chemical industry as natural
resource-based, instead of as science-, engineering- and knowledge-based. While constructing the other indicators, we were
able to take out the chemical industry from the natural resource-based group and include it in the science-, engineering- and
knowledge-based industries. But for the indicator shown in figure 4 this was not possible, since ECLAC-PADI database
does not show the behaviour of the labour productivity of the chemical industry separately for the United States. Even so,
particularly in the Brazilian case, by analysing the rate of growth of the labour productivity by groups, it does not make
much difference if the chemical industry is included in either one or the other group.




15


Figure 3
Labour productivity growth of the Brazilian manufacturing industry, 1970–2008


(Index numbers, 1970 = 100)


50


100


150


200


250


1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008


Science, engineering and knowledge based Natural resource based
Labour intensive Manufacturing industry


Source: ECLAC-PADI database.


Figure 4
The Brazilian technological gap: relative labour productivity in the Brazilian manufacturing


sector compared with that of the United States, 1970–2008
(Per cent)


0


10


20


30


40


50


60


1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008


Science, engineering and knowledge based Natural resource based
Labour intensive Manufacturing industry


Source: ECLAC-PADI database.


of the Brazilian economy (primary and service sectors), in a Kaldorian perspective, the technological gap
of the manufacturing sector as a whole has been increasing at such a high pace since the end of the 1990s
that the level registered in 2008 is enough to draw the conclusion that the Brazilian economy has been
characterized by signs of early de-industrialization and a falling behind long-term path. The important
issue to be answered is whether or not this falling behind path could be reversed in the medium term.




16


Before doing some econometric exercises that could provide us with preliminary answers to the above
question,32 we will investigate the evolution of the Brazilian exports and trade balance. Figure 5 shows the
behaviour of Brazilian exports since 1970. It is a remarkable fact that throughout the 1970s, a period during
which the import substitution regime was most intense, manufactured goods increased their importance
in Brazilian total exports. Starting at less than 20 per cent of total exports in 1970, this participation more
than doubled by 1980. In 1984, manufactured exports represented around 55 per cent of Brazilian total
exports, and this participation kept around this mark until 2006, when exports of manufactured goods
started to lose participation. This loss is being offset by the increase in exports of basic goods.


Two facts could explain the poor performance of the last few years. On the one hand, related to a more
structural source, it could be associated with the enlargement of the technological gap in the manufacturing
sector as a whole, a phenomenon that has been happening since the end-1990s, as already analysed.33 On
the other hand, the loss of participation of manufactured goods in total exports could be associated with a
persistent trend of overvaluation of the Brazilian currency since the 1990s.34 This factor, combined with the
significant growth in world trade in the 2000s, could be responsible for putting the Brazilian economy on
a dangerous path of specializing in goods in which it has static comparative advantage. By looking at the
Brazilian total exports, we realized that this situation was so strongly accelerated between 2006 and 2010
that in this latter year Brazilian exports of basic products (especially primary products and commodities)
surpassed those of manufactured goods (figure 5). This set of poor indicators (high technological gap,
sharp drop in manufactured exports in recent years and a rapid increase of participation of primary and
commodities goods in total exports) are clear preliminary signs that Brazil has actually not only entered
into a process of early de-industrialization, but also embarked on a trajectory of falling behind.


32 This will be carefully analysed in section IV, related to some econometric estimates.
33 As an illustration, the share of the Brazilian manufacturing industry in GDP (around 18 per cent) is close to the value
observed in developed countries, whose per capita income is on average 7 times the Brazilian’s. This shows the premature
character of the de-industrialization process of the Brazilian economy (IEDI, 2005).
34 For details, see Nassif, Feijó and Araújo (2011).


Figure 5
Brazilian total exports classified by intensity of value added, 1970–2010


(Participation in percentage points)


0


10


20


30


40


50


60


70


80


1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010


Basic


Semi-manufactured


Manufactured


Source: BrazilianOfficeofForeignTrade/MinistryofDevelopment,IndustryandCommerce(SECEX/MDIC).




17


Another piece of evidence to support the argument of both early de-industrialization and falling behind
in the past few years could be analysed by the structural characteristics of the Brazilian trade balance.
Figure 6 shows that, while the science-, engineering- and knowledge-based manufacturing industries
showed persistent trade deficits, these negative trade balances not only continued to increase sharply over
the 1990s, but also dramatically accelerated between 2006 and 2008. Labour intensive manufacturing
industries are those characterized by minor importance in Brazilian manufacturing exports and showed
their worst trade performance in both the second half of 1990s and in the final years of the series. In turn,
primary resource-based manufacturing industries are the only group that had trade surplus during the
entire time. The positive trade balance of this latter group has sharply accelerated in the last few years.


A preliminary conclusion could be drawn from this section. If, on the one hand, the Brazilian manufacturing
industry can still be characterized as a relatively large and diversified one, on the other, the enlargement of
the technological gap, combined with a persistent real exchange rate overvaluation, has been responsible
for the loss of international competitiveness in the manufacturing sector (except from the natural resource-
based industries), as well as for putting the Brazilian economy on a dangerous path of falling behind.
The science-, engineering- and knowledge-based manufacturing industries, particularly, have sharply
enlarged both the technological gap and the trade deficits. From a Kaldorian perspective, the combination
of this set of negative factors, by reducing the ability of the manufacturing sector to spillover its gains
from productivity to the rest of the economy, might definitively deepen the Brazilian process of early
de-industrialization and accelerate the actual falling-behind path in the long run. In the next section,
we will carefully investigate not only if the Brazilian economy might be able to reverse this negative
trajectory in the medium term, but also if it is subjected to any external constraint to sustain economic
growth in the long run.


Figure 6
Trade balance of the Brazilian manufacturing industries classified by technological intensity,


1970–2008
(Billions of US dollars)


Science, engineering and knowledge based Natural resource based Labour intensive


-20


-15


-10


-5


0


5


10


15


20


1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008


Source: ECLAC-PADI database.




18


IV. ECONOMETRIC EVIDENCE: THE KALDOR-VERDOORN’S
AND THIRLWALL’S LAWS


Aiming at speculating on the potential capacity of the Brazilian economy to sustain its long-term growth,
we need to implement at least two important econometric exercises: the first one involves estimating the
so-called Kaldor-Verdoorn coefficient in the 1970–2010 period. This indicator reveals whether or not
the Brazilian manufacturing industry operates under static and (mainly) dynamic economies of scale,
as interpreted by Kaldor (1966) and as already discussed in section II;35 the second one is related to the
estimate of the income elasticity of demand for Brazilian exports and imports between 1970 and 2010.
This estimation is essential for drawing some conclusions on whether or not the Brazilian balance of
payments constraints in the long run have been increasing according to the so-called Thirlwall’s Law.


A. Estimating the Kaldor-Verdoorn coefficient for the Brazilian economy (1970–2010)


The Kaldor-Verdoorn Law was originally specified by Verdoorn (1949) and used by Kaldor (1966)
according the following equation (we will maintain the original notation):


p = a + bq (2)


where p is the exponential growth of labour productivity in the manufacturing sector and q is the exponential
growth of manufacturing output. The estimated coefficient, b, is the Kaldor-Verdoorn coefficient.


According to Kaldor’s (1975: 693) interpretation, for having unequivocal evidence of the existence of
static and (mainly) dynamic economies of scale in manufacturing, there must be a “statistically significant
relationship between p and q, with a regression coefficient which is significantly less than 1”. At the same
time, for the existence of constant returns to scale to be rejected, the above mentioned coefficient must
statistically be significantly different from zero. Following McCombie and Thirlwall (1994), we also
estimated the Kaldor-Verdoorn coefficient as specified in equation (2) by ordinary least squares (OLS).36


The estimation covers the 1970–2010 period.37 However, in order to investigate if the post-trade
liberalization period has produced any change in the Kaldor-Verdoorn coefficient in Brazil, we also
implemented the econometric estimation for the subperiods between 1970–1989 and 1990–2010.38 The
results are shown in table 3, whose footnotes also inform on the implemented variables and data sources.


The results show that the Brazilian economy has, at least in principle, potential for growing in the long run.
The estimated Kaldor-Verdoorn coefficients b not only were statistically highly significant and different
from zero, but also were less than 1 in all estimated periods. This reveals that the Brazilian manufacturing
industry operates under dynamic economies of scale, in the sense analysed by Kaldor (1966). The
estimated coefficient b for all estimated periods (1970–2010) was the same for the 1970–1989 period
(0.39). Yet, between 1990 and 2010, the estimated Kaldor-Verdoorn coefficient significantly improved


35 McCombie and Thirlwall (1994, chap. 2, especially pages 163 and 231) present a detailed critical review on the Kaldor
theory of growth in the long run, as well as a lot of econometric evidence (and the several issues related to its estimate)
on the Kaldor-Verdoorn coefficient.
36 For a detailed discussion on problems related to the original Kaldor theoretical and empirical equation (and also other
following proposed theoretical and econometric specifications), see McCombie and Thirlwall (1994: 175–231).
37 For labour productivity in the manufacturing sector (defined as the ratio of value added to total employed workforce),
we used the original source from ECLAC-PADI. However, as its available data only go until 2008, we estimated the value
added and total employment in the Brazilian manufacturing sector in 2009 and 2010 based on the yearly growth rates for
both variables estimated by Brazil’s Ministry of Development, Industry and Commerce (MDIC).
38 If the period from 1970 to 2010 was divided into a larger number of subperiods, the regression methodology would
have little information for estimating the coefficients and making the results reliable.




19


to 0.52, indicating that an increase in the growth of
the Brazilian manufacturing output by one percentage
point increases the growth of labour productivity by
a bit more than one-half of a percentage point. This
is proof that, even when taking into account the
severe micro and macroeconomic problems suffered
by the Brazilian economy through part of this latter
subperiod (such as the necessary, but relatively rapid
trade liberalization, high inflation, the persistent trend
of overvaluation of the Brazilian currency in real
terms, among others), the Brazilian manufacturing
industry still operates under substantial dynamic
economies of scale. In other words, in principle, it
has potential for boosting both labour productivity
and, therefore, economic growth in the long run. It is
important to stress that Kaldor implicitly considered
a coefficient near 0.50 as idealistically good for a
developing country sustaining a catching up process
(see Kaldor, 1966). This is because countries like the
United States and Germany, for instance, during the
period in which they were close to catching up with
the United Kingdom, between the last quarter of the
nineteenth century and the early-twentieth century,
showed Verdoorn coefficients of 0.42 and 0.49,
respectively (as originally calculated by Verdoorn
and presented by McCombie and Thirlwall, 1994:
170, table 2.5).


Obviously, the result presented in table 3 must be understood as a preliminary exercise that has some
limitations, such as a linear technical progress function as originally specified by Kaldor (1966), simple
econometric regression, omission of variables, among others. However, after summarizing several
empirical studies from the most simple to highest econometric sophistication, McCombie and Thirlwall
(1994:167) showed that almost all of them have econometric issues. Not by chance, these same authors
concluded that “the debate over the Verdoorn Law would make a good textbook example of the problems
that can beset statistical inference!”


B. Estimating the income elasticity of demand for Brazilian exports and imports (1980–2010)


This section aims at estimating the income elasticity of demand for Brazilian imports and exports. The
estimated model uses quarterly data covering the 1980–2010 period. The econometric model closely
follows Cimoli, Porcile and Rovira (2010), who implemented two regressions for estimating the demand
functions for imports and exports, respectively, according to the following specifications:


mt = c + ψ(rert ) + π(yt ) + et (3)


xt = c + φ(rert ) + ε(y*t ) + et (4)


where m is the growth rate of imports; c is the exogenous constant term; ψ is the price elasticity of
demand for imports; rer is the growth rate of the real exchange rate (expressed as the domestic price of a
foreign currency); π is the income elasticity of demand for imports; y is the growth rate of the domestic
real GDP; e is a white noise error; x is the growth rate of exports; φ is the price elasticity of demand for
exports; ε is the income elasticity of demand for exports; y* is the growth rate of the world economy and


Table 3
Estimate of the Kaldor-Verdoorn coefficient


for the Brazilian economy, 1970–2010,
econometric estimate of equation (2)


(Dependent variable: rate of growth of labour
productivity in manufacturing)


1970–2010 1970–1989 1990–2010


Constant a -0.002
(0.008)


-0.006
(0.014)


0.010
(0.011)


K-V coeff. b 0.392***
(0.129)


0.387**
(0.178)


0.521**
(0.210)


R2 0.19 0.21 0.24


R2 adjusted 0.17 0.20 0.17
DW 1.99 2.07 2.01
n 40 20 21


Source: Brazilian System of National Accounts from the
Brazilian Institute of Geography and Statistics;
and Brazil’s Ministry of Development, Industry and
Commerce.


Note: *** Significantat1percentlevel;
** Significantat5percentlevel;
* Significantat10percentlevel.


Values in parentheses indicate standard deviation;
DW is the Durbin-Watson statistics; n is the number
of observations; p the difference of the logarithm of
the labour productivity in the Brazilian manufacturing
sector; q the difference of the logarithm of the value
added of the Brazilian manufacturing industry,
expressed in millions of US dollars.




20


t is the respective quarterly time. For estimating the
coefficients of equations (3) and (4), we had either
to construct or take data from the several sources
detailed in annex B.


Before implementing the econometric exercise, we
analysed the potential non-stationarity of the time
series through appropriate tests. First, by applying the
Augmented Dickey-Fuller (ADF) and Phillip-Perron
(PP) unit root tests, we realized that all the series are
stationary.39 Thus, since all time series revealed to
be stationary, the estimated coefficients by ordinary
least squares (OLS) are superconsistent, making the
results of our estimations reliable.


Since we were also interested in comparing the more
recent period with the period immediately following
the 1980s, by avoiding an arbitrary year to divide
the series into two subperiods, our first step was
to implement several econometric structural break
tests for both import and export functions. For this
purpose, we implemented the CUSUM, the recursive
residuals and the Chow tests.40 We only identified
one structural break in January 1999 for imports,
but no structural break for exports. Despite this,
since by 1999 the Brazilian economy had already
undergone substantial institutional changes, we think


the division before and after that year is also economically justified to be the divisory line for analysing
both import and export series.41


Tables 4 and 5 present the results of our estimates of the elasticities of demand for Brazilian imports and
exports between 1980 and 2010, respectively.


First of all, all estimated coefficients for both import and export functions revealed to be statistically
significant for the two subperiods. By comparing the estimated coefficient of income elasticity of demand
for imports p for the 1980–1998 subperiod (1.97), see table 4, we noticed that it sharply increased (to
3.36, or around 70 per cent). Yet, the estimated coefficient of the income elasticity of demand for exports
e marginally decreased between the two analysed subperiods (from 1.36 to 1.33, see table 5).


As pointed out by Thirlwall (1979) and McCombie and Thirlwall (1994), the income elasticities of
demand for imports and exports reflect competitive factors associated with the productive structure of the
economy as a whole. These elasticities are determined, in turn, by the content and other characteristics
of both imported and exported goods, such as the degree of technological sophistication, the level of
product differentiation and the domestic capacity to respond to changes in global demand. Countries whose


39 All results of these tests can be made available by the authors upon request.
40 All applications of these tests as well as growth rates of the adopted series implemented in the econometric exercise
can be made available by the authors upon request.
41 Among other institutional liberalizing reforms already adopted by 1999, we can mention the trade liberalization
(1990–1994), the privatization of state enterprises (from 1990 on), the openness of short-term capital account (1992–1993),
an adoption of a floating exchange rate regime (1999), as well as targets for both inflation rates and fiscal surplus (from
1999 on).


Table 4
Explanatory factors and income elasticity of


demand for imports in Brazil
(Third quarter of 1980 to the end of 1998 compared with
the first quarter of 1999 to the second quarter of 2010)


III/1980–
IV/1998


I/1999–
II/2010


C (constant) 0.615
(0.164)


-0.757
(1.798)


ψ (price-elasticity of
demand for imports)


-0.612***
(0.245)


-0.279***
(0.172)


π (income-elasticity of
demand for imports)


1.967**
(0.732)


3.361***
(1.148)


R2 0.327 0.228


R2 adjusted 0.316 0.192


DW 2.367 2.162


n 74 46


Note: ***Significantat1percentlevel;
** Significantat5percentlevel;
* Significantat10percentlevel.
Values in parentheses indicate standard deviation.


DW is the Durbin-Watson statistics; n is the number
of observations; rer (real exchange rate) was used
with a time lag.




21


net import structure are characterized by higher
technological content than the export one have higher
income elasticity of demand for imports than for
exports. These characteristics of imports and exports
structure tend not only to augment the country’s
technological gap with respect to the international
technological frontier, but also to put the country into
a unsustainable economic development trajectory,
insofar as it will face major external constraints to
growth in the long run. This seems to be the case of
Brazil in the last decade, which presented clear signs
of falling behind, as can be seen in table 6.


The last column of table 6 is the empirical calculation
of equation (1) for Brazil – which is related to our
basic hypothesis on structural change and economic
development –, based on the estimated income
elasticity of demand for exports and imports. Since


the
M


X


p
e


ratio between 1999 and 2010 sharply


decreased, compared with the 1980–1998 period, this
results means that Brazil, by having augmented the
technological gap and being notably far from the
average world economic growth in the last decade,
entered into a clear trajectory of falling behind.


However, taking into account that, according to the Kaldor-Verdoorn estimated coefficient, its
manufacturing industry operates under dynamic economies of scale, there is still time to redirect the
economy to a process of catching up. To achieve this goal, the Brazilian government needs to be successful
in adopting a fine coordination between the long-term policies (such as industrial and technological
policies, infrastructure and education policies, among others) and the short-term macroeconomic policies
(especially the monetary, fiscal, financial, credit and exchange rate policies).42


42 For detailed discussion, see Nassif (2011).


Table 5
Explanatory factors and income elasticity of


demand for exports in Brazil
(Third quarter of 1980 to the end of 1998 compared with
the first quarter of 1999 to the second quarter of 2010)


III/1980–
IV/1998


I/1999–
II/2010


C (constant) 0.670
(0.019)


1.782
(2.042)


φ (price-elasticity of
demand for exports)


0.371**
(0.181)


0.374*
(0.244)


ε (income-elasticity of
demand for exports)


1.358**
(0.575)


1.329***
(0.476)


R2 0.218 0.308


R2 adjusted 0.201 0.276


DW 2.09 2.30


n 74 46


Note: ***Significantat1percentlevel;
** Significantat5percentlevel;
* Significantat10percentlevel.
Values in parentheses indicate standard deviation.


DW is the Durbin-Watson statistics; n is the number
of observations; y* (world real GDP) was used with
a time lag.


Table 6
Thirlwall’s Law


Period


Income elasticity of
demand for Brazilian


exports (εx )


Income elasticity of
demand for Brazilian


imports (πM )


Thirlwall’s Law


M


X


Y
Y


p
e


=
*


III/1980–II/2010 1.059 1.993 0.531


III/1980–IV/1998 1.358 1.967 0.690


I/1999–II/2010 1.329 3.361 0.395


Source: Authors’ own elaboration, based on the econometric estimates.




22


One could argue that a high Kaldor-Verdoorn (K-V) coefficient which implies a manufacturing sector
operating under conditions of dynamic economies of scales should be incompatible with a falling behind
trajectory. In Brazil, however, this is not necessarily the case. In fact, it is possible that the high K-V
coefficient is asymmetrically influenced by some segments of the manufacturing industries like the natural
resource-based ones that are characterized, at the same time, by a high capital/output ratio. This has positive
economic policy implications because, with smart long-term technological policies, governments can
use the state’s purchasing policy to boost and construct dynamic comparative advantages in industries
such as shipbuilding, machines and equipment for extracting oil, machines and equipment associated
with paper and cellulose, petrochemicals, pharmaceutical products, and so on.


V. CONCLUDING REMARKS


This paper analysed a theoretical and empirical framework on the relationships between structural change
and economic development based on Kaldorian and Thirlwallian theories. According to Kaldor (1966,
1970), when a country embarks on a sustainable path of catching up with both levels of income per capita
and quality of life close to that of developed countries, this process is accompanied by structural change,
in the sense that there is a strong reallocation of resources from traditional sectors to the manufacturing
sector. At the same time, Kaldor’s theories on long-term growth emphasize the role of the manufacturing
industry to boost and spill over technical progress throughout the economy as a whole. And since that
industry, differently from the traditional ones (primary and nontradable service sectors), operates under
static and dynamic economies of scale, it tends to augment the aggregate productivity. Kaldor was one
of first authors to discuss the dangers to a country entering into a process of early de-industrialization
before it has achieved levels of income per capita comparable to those of developed countries.
However, this is a complex phenomenon which is not only measured by the loss of participation of the
manufacturing sector in total value added, employment, and exports, but also by indicators that capture
the country’s international relative competitiveness, such as the technological gap, the participation of
more technologically sophisticated sectors in the country’s trade balance, among others.


Thirlwall, in turn, developing Kaldor’s original ideas, showed the importance for a country to keep
an income elasticity of demand for exports above that for imports in order to avoid persistently facing
external constraints to long-term growth. Combining Kaldor’s and Thirlwall’s ideas, they suggest that the
farther a country is from the international technological frontier, the more the income elasticity of demand
for imports tends to exceed that for exports. If this is the case, economic development is recurrently
constrained by a balance of payments crisis.


By empirically analysing the Brazilian case between 1970 and 2010, we could draw some important
conclusions. First, although there was a dramatic decrease in the participation of the manufacturing
industry in total value added from 31.3 per cent in 1980 to 14.6 per cent in 2010, within the manufacturing
industry per se, however, except for the labour intensive industries, there was a reallocation of resources
from the traditional segments (labour intensive and natural resource-based) to the more technologically
sophisticated ones (science-, engineering- and knowledge-based). Second, the level of employment, despite
remaining relatively well balanced among all groups of manufacturing classified by technological intensity
in the past few years, was not reallocated from the manufacturing industry to the service sector between
2000 and 2008. Third, within the manufacturing industry, the science-, engineering- and knowledge-
based segments represented almost 50 per cent of total Brazilian manufactured exports, followed by
the natural resource-based segments (around 42 per cent) and labour intensive (less than 10 per cent).
Fourth, by analysing total Brazilian exports, basic products surpassed those of manufactured goods
between 2009 and 2010 (boldface ours). Fifth, the sectoral trade deficits of the engineering-, science- and
knowledge-based manufacturing sector has significantly increased in recent years. Sixth, between the
end of the 1990s and 2008, not only did the labour productivity of the Brazilian manufacturing sector
decrease, but also its technological gap dramatically increased, revealing that it is getting farther from
the international technological frontier. Seventh, our econometric estimates show that, since the income




23


elasticity of demand for imports significantly augmented and kept above the income elasticity of demand
for exports, these results not only reflect the above mentioned augmentation of the technological gap, but
also that Brazil sharply increased its external constraints to sustain economic growth in the long run. The
conclusions from the fourth to seventh are sound evidence that Brazil has been in a process of early de-
industrialization and falling behind since the end of the 1990s, comparatively with developed countries
or even other emerging economies.


However, as the estimated Kaldor-Verdoorn coefficient surprisingly revealed that the Brazilian
manufacturing industry operates under dynamic economies of scale, this suggests that it has, in principle,
the potential for sustaining the growth in productivity of the economy as a whole and, therefore, Brazilian
economic growth in the long run. It is important to stress that to still have a large and diversified
manufacturing industry subject to dynamic increasing returns to scale is a necessary, but not a sufficient
condition to ensure economic development in the long run.


In fact, the findings in this paper bring about important long-term and short-term economic policy
implications. The late Brazilian economist Antonio Barros de Castro used to repeat the word “breath”
in his lectures to refer to the great capacity of the Brazilian manufacturing industry to face a diversity of
internal and external shocks (high inflation, trade liberalization, real exchange rate overvaluation, and
so on). Obviously, this capacity is not unlimited. But with appropriate and coordinated long-term (e.g.
industrial and technological policies, infrastructure and educational policies, among others) and short-
term economic policies (coordination among monetary, fiscal, financial, credit, and, especially, exchange
rate policies pro-growth), there is still time to put the Brazilian economy into a process of catching up
again. In 2008, the Brazilian government made an important step in this direction with the adoption of
the Policy for Productive Development (Política de Desenvolvimento Produtivo – PDP), which, among
other instruments, introduced tax and credit stimulus to innovation, infrastructure and exports.


Although suggestions of economic policy instruments escape the scope of this study, we would like to
finish this paper by stressing that any agenda that is oriented to sustaining economic development in
Brazil should be fulfilled by two complementary conditions: (i) policies of supply-side stimuli should
be balanced by demand-side ones; (ii) long-term policies such as industrial and technological policies,
infrastructure, education, and so on, will not have good performance if they are not well coordinated
with other economic institutions, notably short-term macroeconomic policies (mainly monetary, fiscal
and exchange rate policies).


As to the role of the real exchange rate as an important macroeconomic instrument (if not the most)
to sanction the micro and meso economic policies introduced by governments in favour of economic
development, Kaldor (1970: 152) argued a long time ago that, “of these two instruments for counteracting
adverse trends in “efficiency wages” – protection and devaluation – the latter is undoubtedly superior to
the former. Devaluation, as has often been pointed out, is nothing else but a combination of a uniform
ad-valorem duty on all imports and a uniform ad-valorem subsidy on exports”.




24


ANNEX A


Manufacturing industry according to technological intensity


Science-, engineering- and knowledge-based
Metal products
Non electrical machinery
Electrical machinery
Motor vehicles
Scientific instruments
Chemicals


Natural resource-based
Food
Beverages
Tobacco
Wood products
Paper and cellulose
Petroleum refining
Oil and carbon products
Rubber products
Glass
Other non-metallic mineral products
Iron and steel
Nonferrous metals


Labour intensive
Textile
Clothing
Leather manufactures
Footwear
Furniture
Paper printing
Other chemicals
Plastic products
Pottery
Other manufactured products


Source: ECLAC-PADI database.




25


ANNEX B


Data sources


(i) Brazilian imports: imports expressed in United States million dollars CIF (cost,insurance and freight)
according to the International Monetary Fund, International Financial Statistics, browser on CD-
ROM; current values were deflated by the United States Wholesale Price Index (WPI); growth rates
were based on the construction of index numbers (average of 2005 = 100).


(ii) Brazilian real exchange rates: we transform the monthly rates series available at the Brazilian
Institute of Applied Economic Research (IPEA) into quarterly real exchange rates – http://www.
ipea.gov.br; growth rates were based on the construction of index numbers (average of 2005 = 100).


(iii) Brazilian real gross domestic product (GDP) at market prices: adjusted seasonally by the Brazilian
Institute of Geography and Statistics (IBGE/SCN2000-Qtr) – http://www.ibge.gov.br; growth rates
were calculated based on index numbers (average of 2005 = 100).


(iv) Brazilian exports: exports expressed in millions of United States dollars FOB (free on board)
according to Brazil’s Central Bank Bulletin, Balance of Payments Section (BCB Bulletin/BP) –
http://www.bcb.gov.br; current values were deflated by the United States Wholesale Price Index
(WPI); growth rates were based on the construction of index numbers (average of 2005 = 100).


(v) World quarterly real GDP: available at the International Monetary Fund (IMF) http://forums.imf.
org/showthread.php?t=6124, calculated by IMF forum participants, based on official websites of
country-members of the IMF, transformed into United States million dollars and subtracting Brazil’s
quarterly real GDP; growth rates were based on index numbers (average of 2005 = 100).




26


REFERENCES


Amsden AH (1989). Asia’s next giant: South Korea and late industrialization. Oxford: Oxford University Press.
Amsden AH (2001). The rise of “the rest”: challenges to the west from late-industrializing economies. Oxford:


Oxford University Press.
Arthur WB (1989). Competing Technologies, Increasing Returns, and Lock-in by Historical Events. The Economic


Journal, 99: 116–131. March.
Bacha E and Fishlow A (2011). ‘The recent commodity price boom and Latin American growth: More than new


bottles for an old wine? In: Ocampo JA and Ros J, eds. The Oxford Handbook of Latin American Economics.
Oxford: Oxford University Press.


Bonelli R (2002). Labor Productivity in Brazil during the 1990’s. Texto para Discussão nr 906. Rio de Janeiro: IPEA.
Bresser-Pereira LC (2006). Novo desenvolvimentismo e ortodoxia convencional. In: Eli Diniz, ed. Globalização,


Estado e Desenvolvimento. São Paulo: FGV Editora.
Bresser-Pereira LC (2008). The Dutch disease and its neutralization: a Ricardian approach. Revista de Economia


Política, Brazilian Journal of Political Economy, vol. 28, no. 1 January–March.
Bresser-Pereira LC (2010). Globalization and competition. Why some emerging countries succeed while others fall


behind. Cambridge: Cambridge University Press.
Castro AB (1985). Ajustamento versus transformação: a economia brasileira de 1974 a 1984. In: Castro AB and de


Souza FEP, eds. A Economia Brasileira em Marcha Forçada. Rio de Janeiro: Paz e Terra.
Cimoli M and Porcile G (2010a). Specialization, Wage Bargaining and Technology in a Multigoods of Growth


Model. Metroeconomica 61 (1): 219–238.
Cimoli M and Porcile G (2010b). Global Growth and International Cooperation: a Structuralist Perspective.


Cambridge Journal of Economics, Vol. 35: 383–400.
Cimoli M and Porcile G (2011). Tecnología, Heterogeneidad y Crecimiento: Un Caja de Herramientas Estructuralista.


Santiago de Chile: The Economic Commission for Latin America and the Caribbean (ECLAC). Unpublished.
Cimoli M, Porcile G and Rovira S (2010). Structural Change and BOP-Constraint: Why did Latin America Fail to


Converge. Cambridge Journal of Economics 34: 389–411.
Clark C (1940). The conditions of economic progress. London: Macmillan.
Dosi G, Pavitt K and Soete L (1990). The Economics of Technical Change and International Trade. London:


Harvester Wheastsheaf.
Fagerberg J (1988). Why growth rates differ? In: Dosi G, Freeman C, Nelson R, Silverberg G and Soete L, eds.


Technical Change and Economic Growth. London and New York: Pinter Publishers.
Fajnzylber F (1988). Competitividad Internacional: Evolución y lecciones. Revista de la Cepal nr. 36.
Feijó CA and Lamonica MT (2012). Importancia del sector industrial para el desarrollo de la economía brasileña.


Revista de la CEPAL nr.107. (CEPAL Review nr.107, also in English edition; available at: http://www.eclac.org).
Feijó CA and Carvalho PGM (2002). Uma interpretação sobre a evolução da produtividade industrial no Brasil nos


anos noventa e as “leis” de Kaldor. Nova Economia, vol. 12, no. 2: 57–78.
Grossman GM and Helpman E (1991). Innovation and Growth in the Global Economy. Cambridge (Mas.): The


MIT Press.
Hausmann R, Hwang J and Rodrik D (2007). What you Export Matters. Journal of Economic Growth, 12 (1):1–25.
Hirschman AO (1958). The Strategy of Economic Development. New Haven: Yale University Press.
IBGE (2007). Methodological Note nr. 9. Brazilian National Account System: reference 2000. Rio de Janeiro:


Brazilian Institute of Geography and Statistics; available at: http://www.ibge.gov.br.
IEDI (2005) Ocorreu uma desindustrialização no Brasil? São Paulo: Instituto de Estudos para o Desenvolvimento


Industrial. November, paper not published; available at: http://www.iedi.org.br.
Jank M et al. (2008). Exportações: existe uma doença holandesa? In: Barros O and Giambiagi F, eds. Brasil


Globalizado: o Brasil em um Mundo Surpreendente. Rio de Janeiro: Ed. Campus.
Kaldor N (1966). Causes of the slow rate of economic growth of the United Kingdom. An inaugural lecture.


Cambridge: Cambridge University Press.
Kaldor N (1967). Strategic factors in economic development. New York: Ithaca.




27


Kaldor N (1968). Productivity and growth in manufacturing industry: a reply. Economica. November.
Kaldor N (1970). The case for regional policies. Scottish Journal of Political Economy. November.
Kaldor N (1975). Economic Growth and the Verdoorn Law: A Comment on Mr. Rowthorn’s Article. Economic


Journal, Vol. 85, Nr. 340. December.
Keynes JM (1936). The General Theory of Employment, Interest and Money. Cambridge, UK: Cambridge University


Press.
Lewis WA (1954). Economic Development with Unlimited Supplies of Labor. Manchester School of Economic


and Social Studies, 22 (2): 139–191.
Lucas Jr RE (1988). On the Mechanics of Economic Development. Journal of Monetary Economics 22: 3–42. July.
Marx K (1887). The Capital. A critical analysis of capitalist production. Vol. I. First English edition, edited by


Frederick Engels. London: Penguin Books, 1990.
McCloskey DN (2010). Bourgeois dignity: why economics can’t explain the modern world. Chicago: The University


of Chicago Press.
McCombie JSL and Thirlwall AP (1994). Economic growth and the balance-of-payments constraint. London: St.


Martin’s Press.
Moreira MM. and Correa PG. (1998). A first look at the impacts of trade liberalization on the Brazilian manufacturing


industry. World Development, Vol. 26 (10): 1859–1874. October.
Myrdal G (1957). Economic theory and underdeveloped regions. London: Duckworth.
Nassif A (2003). Liberalização comercial e eficiência econômica: a experiência brasileira. PhD Dissertation. Rio


de Janeiro: Federal University of Rio de Janeiro (Universidade Federal do Rio de Janeiro – UFRJ). March.
Nassif A (2008). Há evidência de desindustrialização no Brasil? Brazilian Journal of Political Economy, Vol. 28,


No. 1 (109), January–March.


Nassif A (2011). Overcoming the “impossible trinity”: towards a mix of macroeconomic policy instruments for
sustaining economic development in Brazil. Brazilian Journal of Political Economy, Vol. 31, No. 5 (125):
912–927, Special edition. December.


Nassif A, Feijó C and Araújo E (2011). The long-term “optimal” real exchange rate and the currency overvaluation
trend in open emerging economies: The case of Brazil. UNCTAD Discussion Paper Nr. 206. Geneva. December.


Nelson RR and Winter SG (1982). An Evolutionary Theory of Economic Change. Cambridge, Mass.: Harvard
University Press.


Oreiro JL and Feijó C (2010). Desindustrialização: Conceituação, Causas, Efeitos e o Caso Brasileiro. Brazilian
Journal of Political Economy, Vol. 30, No. 2: 100–115. January–March.


Palma G (2005). Four sources of de-industrialisation and a new concept of the Dutch Disease. In: Ocampo JA, ed.
Beyond reforms. Palo Alto, CA: Stanford University Press.


Prebisch R (1950). The Economic Development of Latin America and its Principal Problems. New York: United
Nations.


Romer PM (1986). Increasing Returns and Long-Run Growth. Journal of Political Economy, 94: 1001–1037.
Romer PM (1990). Endogenous Technological Change. Journal of Political Economy, 98, Supplement to no. 5:


71–102.


Rosenstein-Rodan PN (1943). Problems of Industrialization of Eastern and South-Eastern Europe. Economic
Journal, 53: 202–211. June–September.


Rowthorn R and Ramaswamy R (1999). Growth, trade and de-industrialazition. IMF Staff Papers no. 46, 1.
Washington, DC: International Monetary Fund.


Rowthorn R and Wells JR (1987). De-industrialization and Foreign Trade. Cambridge: Cambridge University Press.
Schumpeter J (1942). Capitalism, Socialism and Democracy. London and New York: Routledge (1994 edition).
Smith A (1776). An inquiry into the nature and causes of the wealth of nations. New York: Modern Library, 1937.
Thirlwall AP (1979). The balance of payments constraint as an explanation of international growth rates differences.


Banca Nazionale del Lavoro Quarterly Review, Vol. 128.
Verdoorn PJ (1949). Fattori che regolano lo sviluppo della produttivitá del lavoro. L’Industria.
Young AA (1928). Increasing Returns and Economic Progress. The Economic Journal, Vol. 38, No. 152: 527–542.


December.




28


No. Date Author(s) Title


210 December 2012 Giovanni Andrea
Cornia and Bruno
Martorano


Development policies and income inequality in selected
developing regions, 1980–2010


209 November 2012 Alessandro Missale
and Emanuele
Bacchiocchi


Multilateral indexed loans and debt sustainability


208 October 2012 David Bicchetti and
Nicolas Maystre


The synchronized and long-lasting structural change on
commodity markets: Evidence from high frequency data


207 July 2012 Amelia U. Santos-
Paulino


Trade, income distribution and poverty in developing
countries: A survey


206 December 2011 André Nassif,
Carmem Feijó
and Eliane Araújo


The long-term “optimal” real exchange rate and
the currency overvaluation trend in open emerging
economies: The case of Brazil


205 December 2011 Ulrich Hoffmann Some reflections on climate change, green growth
illusions and development space


204 October 2011 Peter Bofinger The scope for foreign exchange market interventions
203 September 2011 Javier Lindenboim,


Damián Kennedy and
Juan M. Graña


Share of labour compensation and aggregate demand
discussions towards a growth strategy


202 June 2011 Pilar Fajarnes An overview of major sources of data and analyses
relating to physical fundamentals in international
commodity markets


201 February 2011 Ulrich Hoffmann Assuring food security in developing countries under the
challenges of climate change: Key trade and development
issues of a fundamental transformation of agriculture


200 September 2010 Jörg Mayer Global rebalancing: Effects on trade flows and employment
199 June 2010 Ugo Panizza,


Federico Sturzenegger
and Jeromin Zettelmeyer


International government debt


198 April 2010 Lee C. Buchheit and
G. Mitu Gulati


Responsible sovereign lending and borrowing


197 March 2010 Christopher L. Gilbert Speculative influences on commodity futures prices
2006–2008


196 November 2009 Michael Herrmann Food security and agricultural development in times of
high commodity prices


195 October 2009 Jörg Mayer The growing interdependence between financial and
commodity markets


194 June 2009 Andrew Cornford Statistics for international trade in banking services:
Requirements, availability and prospects


193 January 2009 Sebastian Dullien Central banking, financial institutions and credit creation
in developing countries


192 November 2008 Enrique Cosio-Pascal The emerging of a multilateral forum for debt
restructuring: The Paris Club


191 October 2008 Jörg Mayer Policy space: What, for what, and where?
190 October 2008 Martin Knoll Budget support: A reformed approach or old wine in new


skins?
189 September 2008 Martina Metzger Regional cooperation and integration in sub-Saharan Africa


UNCTAD Discussion PaPers




29


No. Date Author(s) Title


188 March 2008 Ugo Panizza Domestic and external public debt in developing
countries


187 February 2008 Michael Geiger Instruments of monetary policy in China and their
effectiveness: 1994–2006


186 January 2008 Marwan Elkhoury Credit rating agencies and their potential impact on
developing countries


185 July 2007 Robert Howse The concept of odious debt in public international law
184 May 2007 André Nassif National innovation system and macroeconomic policies:


Brazil and India in comparative perspective
183 April 2007 Irfan ul Haque Rethinking industrial policy
182 October 2006 Robert Rowthorn The renaissance of China and India: implications for the


advanced economies
181 October 2005 Michael Sakbani A re-examination of the architecture of the international


economic system in a global setting: Issues and proposals
180 October 2005 Jörg Mayer and


Pilar Fajarnes
Tripling Africa’s Primary Exports: What? How? Where?


179 April 2005 S.M. Shafaeddin Trade liberalization and economic reform in developing
countries: structural change or de-industrialization?


178 April 2005 Andrew Cornford Basel II: The revised framework of June 2004
177 April 2005 Benu Schneider Do global standards and codes prevent financial crises?


Some proposals on modifying the standards-based approach
176 December 2004 Jörg Mayer Not totally naked: textiles and clothing trade in a quota


free environment
175 August 2004 S.M. Shafaeddin Who is the master? Who is the servant? Market or


Government?
174 August 2004 Jörg Mayer Industrialization in developing countries: some evidence


from a new economic geography perspective
173 June 2004 Irfan ul Haque Globalization, neoliberalism and labour
172 June 2004 Andrew J. Cornford The WTO negotiations on financial services: current


issues and future directions
171 May 2004 Andrew J. Cornford Variable geometry for the WTO: concepts and precedents
170 May 2004 Robert Rowthorn and


Ken Coutts
De-industrialization and the balance of payments in
advanced economies


169 April 2004 Shigehisa Kasahara The flying geese paradigm: a critical study of its
application to East Asian regional development


168 February 2004 Alberto Gabriele Policy alternatives in reforming power utilities in
developing countries: a critical survey


167 January 2004 Richard Kozul-Wright
and Paul Rayment


Globalization reloaded: an UNCTAD Perspective


166 February 2003 Jörg Mayer The fallacy of composition: a review of the literature
165 November 2002 Yuefen Li China’s accession to WTO: exaggerated fears?
164 November 2002 Lucas Assuncao and


ZhongXiang Zhang
Domestic climate change policies and the WTO


163 November 2002 A.S. Bhalla and S. Qiu China’s WTO accession. Its impact on Chinese
employment


162 July 2002 Peter Nolan and
Jin Zhang


The challenge of globalization for large Chinese firms




30


No. Date Author(s) Title


161 June 2002 Zheng Zhihai and
Zhao Yumin


China’s terms of trade in manufactures, 1993–2000


160 June 2002 S.M. Shafaeddin The impact of China’s accession to WTO on exports of
developing countries


159 May 2002 Jörg Mayer,
Arunas Butkevicius
and Ali Kadri


Dynamic products in world exports


158 April 2002 Yılmaz Akyüz and
Korkut Boratav


The making of the Turkish financial crisis


157 September 2001 Heiner Flassbeck The exchange rate: Economic policy tool or market price?
156 August 2001 Andrew J. Cornford The Basel Committee’s proposals for revised capital


standards: Mark 2 and the state of play
155 August 2001 Alberto Gabriele Science and technology policies, industrial reform and


technical progress in China: Can socialist property rights
be compatible with technological catching up?


154 June 2001 Jörg Mayer Technology diffusion, human capital and economic
growth in developing countries


153 December 2000 Mehdi Shafaeddin Free trade or fair trade? Fallacies surrounding the theories
of trade liberalization and protection and contradictions in
international trade rules


152 December 2000 Dilip K. Das Asian crisis: Distilling critical lessons
151 October 2000 Bernard Shull Financial modernization legislation in the United States –


Background and implications
150 August 2000 Jörg Mayer Globalization, technology transfer and skill accumulation


in low-income countries
149 July 2000 Mehdi Shafaeddin What did Frederick List actually say? Some clarifications


on the infant industry argument
148 April 2000 Yılmaz Akyüz The debate on the international financial architecture:


Reforming the reformers
147 April 2000 Martin Khor Globalization and the South: Some critical issues
146 February 2000 Manuel R. Agosin and


Ricardo Mayer
Foreign investment in developing countries: Does it
crowd in domestic investment?


145 January 2000 B. Andersen,
Z. Kozul-Wright and
R. Kozul-Wright


Copyrights, competition and development: The case of
the music industry


Copies of UNCTAD Discussion Papers may be obtained from the Publications Assistant, Macroeconomic and
Development Policies Branch (MDPB), Division on Globalization and Development Strategies (DGDS), United
Nations Conference on Trade and Development (UNCTAD), Palais des Nations, CH-1211 Geneva 10, Switzerland;
e-mail: gdsinfo@unctad.org; fax no: +41 22 917 0274.


UNCTAD Discussion Papers are accessible on the website at http://unctad.org.




Login