Love

The Origins of Dependency Analysis
Author(s): Joseph L. Love

Source: Journal of Latin American Studies , Feb., 1990, Vol. 22, No. 1 (Feb., 1990), pp.
143-168
Published by: Cambridge University Press
Stable URL: https://www.jstor.org/stable/157170
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access
to Journal of Latin American Studies
This content downloaded from

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
All use subject to https://about.jstor.org/terms
The Origins of Dependency Analysis
JOSEPH L. LOVE
It is widely recognised that dependency analysis developed out of two

traditions of economic thought, Marxism and Latin American
structuralism, associated with the UN Economic Commission for Latin
America (ECLA). Although structuralism is acknowledged as a pro-
genitor, Marxism is usually viewed, implicitly or explicitly, as the
tradition from which dependency arose. This is perhaps because
dependency per se is so widely perceived as having begun with two books
for which Marxist antecedents were claimed. Dependenciay desarrollo en
AmJrica Latina (1969), by Fernando Henrique Cardoso and Enzo Faletto,
and Capitalism and Underdevelopment in Latin America (i967), by Andre
Gunder Frank, 'stood out as the leading theoretical and systematic efforts
to construct a dependency perspective for Latin America', and remain
'the landmarks to which assessment of dependency perspectives inevitably
return .l
While Frank's Marxism was heterodox (he borrowed the key notion of
surplus' from Paul Baran) and, for many critics, superficial, his allegiance
to the Cuban revolution was an implicit Marxist credential. Few apparently
noticed that Frank himself, writing a decade after the publication of
Capitalism and Underdevelopment, saw the dependency tradition in its
initial, non-radical form as growing out of ECLA-derived structuralism.
Cardoso, for his part, retrospectively made frequent references to his
allegiance to Marxism, and asserted in the same year as Frank's 'Answer
to Critics ' (I1977) that from the beginning of his own work on dependency
'dialectical analysis was the point of departure. What was significant was
the 'movement', the class struggles ... The structures were regarded as
relations of contradiction, and therefore dynamic '.2
1 Steve J. Stern, 'Feudalism, Capitalism, and the World-System in the Perspective of

Latin America and the Caribbean', American Historical Review, vol. 93, no. 4 (Oct.
I988), p. 836. See the similar judgement in William B. Taylor, 'Early Latin American
Social History', in Olivier Zunz (ed.), Reliving the Past: The Worlds of Social History
(Chapel Hill, N.C., I985), p. I28.
2 Andre Gunder Frank, 'Dependence is Dead, Long Live Dependence and the Class
Struggle: An Answer to Critics', World Development, vol. 5, no. 4 (I977), p
Joseph Love is Professor of History, University of Illinois, Urbana-Champaign.
J. Lat. Amer. Stud. 22, 143-168 Printed in Great Britain 143

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
144 Joseph L. Love
Jose Gabriel Palma, whose lucid discussion of dependency in 1978 was

an analytical milestone, considered both Marxist and structuralist
antecedents, but devoted almost six times the space to Marxism as to
structuralism.3 Publishing the same year, Heraldo Munioz considered
structuralism itself the 'old' dependency, the writings of Celso Furtado
and Osvaldo Sunkel as transitional, and Cardoso, Frank, Theotonio dos
Santos, Anlbal Quijano, and Rui Mauro Marini as writers in the 'new'
dependency tradition, which, Mufioz agreed with Cardoso, was based on
dialectical materialism.4 Of course, many 'orthodox' Marxists denied that
dependency could occupy a legitimate space in Marx's mansion, and the
debate continued into the i98os.5 Looking back in 'After Dependency'
(i985), Peter Evans, who was committed to the view that dependency was
compatible with Marxism, emphasised dependency's contribution to
'Marxist theorizing about the nature of capitalism in the Periphery '.6 Thus,
if dependency's contribution to Marxism - though still a matter of dispute
- is what mattered, it is even easier to neglect the structuralist heritage of
dependency. It is the purpose of this article to delineate and assess the
relative importance of these two intellectual traditions in the initial
articulation of dependency in the mid- and latter I96os.
'Dependent' countries and 'dependency' as undefined or ill-defined
terms appeared in a variety of contexts prior to the I96os. In I928 the
Ecuadorian Ricardo Paredes, speaking at the VI Congress of the
Communist International, spoke of 'dependent', 'semicolonial', and
Fernando Henrique Cardoso, 'The Consumption of Dependency Theory in the United

States', Latin American Research Review, vol. I2, no. 3 (I977), p. 14. Sheldon B. Liss see
dependency as traceable to Lenin; see his Marxist Thought in Latin America (Berkeley,
Calif., I984), p. 25.
See Jose Gabriel Palma, 'Dependency: a Formal Theory of Underdevelopment or a
Methodology for the Analysis of Concrete Situations of Underdevelopment?', World
Development, vol. 6, nos. 7-8 (1978). Palma treats Marxism on pp. 882-98, and
structuralism, i.e. the work of ECLA, on pp. 906-8.
Heraldo Mufioz, 'Cambio y continuidad en el debate sobre la dependencia y el
imperialismo', Estudios Internacionales, vol. i i. no. 44 (Oct.-Dec. I978), pp. 90-I.
Mufioz attributes the characterisations 'old' and 'new' to A. G. Frank.
For example see Ronald H. Chilcote, (ed.), Dependency and Marxism: Toward a Resolution
of the Debate (Boulder, Colo., I982). Beyond dependency itself, Marxist elements in
dependency were given further salience in Immanuel Wallerstein's World-System
Theory - a 'neo-Smithian Marxism', in the pejorative phrase of Robert Brenner,
because of Wallerstein's emphasis on the market as the driving force in history. In a
well-known survey, Daniel Chirot and Thomas D. Hall remark that World-System
Theory 'is in most ways merely a North American adaptation of dependency
theory...'. Chirot and Hall, 'World-System Theory', Annual Review of Sociology, vol. 8
(I982), p. 90.
6 Evans, 'After Dependency: Recent Studies of Class, State, and Industrialization', Latin
American Research Review, vol. 20, no. 2 (I985), p. I59.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependency Analysis 145
'colonial' countries in increasing degrees of subjugation to imperialist

powers. In the same year, the one-time Marxist economic historian
Werner Sombart used 'dependent' to describe regions that were economic
satellites of the capitalist West, and explicitly described (but did not
model) a Centre-Periphery relationship. A candidate for a more immediate
precursor of dependency analysis is Celso Furtado's Uma economia
dependente (i95 6), referring to Brazil, though the analysis is stru
Another Brazilian, the Marxist historian Caio Prado Jr., wrote in
underdeveloped countries were 'peripheral and complementary ... in a
subordinate and dependent situation ... '. Other examples could be cited.
But such terms, without specific defining properties, are ambiguous. The
essential elements would seem to be a characterisation of modern
capitalism as a Centre-Periphery relationship between the developed,
industrial West and the underdeveloped, agricultural Third World; the
adoption of a system-wide historical approach, and the consequent
rejection of Boekean dualism and Parsonian modernisation theory; the
hypothesis of unequal exchange, as well as assymetrical power relations
between Centre and Periphery; and the assertion of the relative or
absolute nonviability of a capitalist path to development, based on the
leadership of the national bourgeoisies of the Latin American nations.8
These propositions only cohered in a single argument in the mid-sixties.
Almost a quarter-century after the formulation of dependency analysis,

it is possible to see more clearly how the twin crises in structuralism
(largely associated with ECLA) and in orthodox Marxism resulted in
dependency. We shall begin with the former.
By the early I96os it was clear that the initial analysis and prescriptions
of ECLA, principally formulated in I949 by its director Raul Prebisch,
had failed to bring about the rapid economic development of the region.
Prebisch had introduced the notion of Centre and Periphery and the
hypothesis that the two elements were related by a process of unequal
exchange. Assuming a greater rate of technological innovation in
industrial countries, he had argued that there were different responses to
recessions by primary exporters and those exporting manufactures,
primarily because of the power of organised labour to maintain high
7 Ricardo Paredes, 'Informe de la delegaci6n latinoamericana', Internacional Comunista,

VI congreso, II: Informes y discusiones (Mexico, 1978), p. 179; Werner Sombart, Der
moderne Kapitalismus, III: Das Wirtschaftsleben im Zeitalter des Hochkapitalismus (Munich,
I928), [2 vols. bound as one] vol. i, xiv-xv; Vol. 2, IOI9; Celso Furtado, Uma economia
dependente (Rio de Janeiro, i956); Caio Prado Jr., Esbofo dos fundamentos de teoria
econdmica (Sao Paulo, 1957), p. 190.
8 See Thomas Angotti, 'The Political Implications of Dependency Theory', in Chilcote,
Dependengy and Marxism, pp. 126-7.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
146 Joseph L. Love
wages, and therefore high export prices, in the latter, and secondarily
because of oligopoly in markets for manufactured goods. ECLA's analysis
of the Latin American economies in the late forties and fifties had pointed
to negative features inherent in the Periphery's economy: structural
unemployment, external disequilibrium, and deteriorating terms of trade.
The terms-of-trade thesis became a point of departure for a 'structuralist'
school of development studies that would emphasise not only macro-
economics and the importance of the foreign exchange constraint, but also
long-term (transcyclical) evolution and the role of institutions, especially
the state.
Another structuralist thesis, originating in the ig5os, addressed the

perennial problem of inflation in Latin America. Structuralists held that
under(ying inflationary pressures derive from bottlenecks produced by
retarded sectors of the economy, especially agriculture, whose backward
state yields an inelastic supply, in the face of rapidly rising demand by the
burgeoning urban masses. Other causes identified were the stagnation of
the traditional export sector, owing in part to deteriorating terms of trade,
and a regressive fiscal system that favoured the import-hungry upper
classes.
The foremost policy recommendation of ECLA in its early years had
been state-assisted import-substitution industrialisation (ISI). The dyna-
mism of the manufacturing sector would raise the overall rate of economic
growth, while making the national economy less dependent on the sharp
swings of international commodity prices and the alleged tendency for
those prices to deteriorate in the long run. Furthermore, industry would
also provide additional employment in countries already experiencing
rapid urbanisation, as well as reducing the demand for imports, thus
easing inflationary pressures. Yet already in the latter fifties, and more
obviously by the early sixties, ISI had created major new problems,
including the exacerbation of inflation. ECLA thus began to consider new
strategies beyond the sphere of formal economic analysis.
The reformist views implicit in structuralism were part of an increasing
concern with social issues by the ECLA staff, a concern that quickened
with the growing radicalism of the Cuban Revolution after i959. More
dramatic than ECLA's contribution to the Alliance for Progress in I96I
- it helped win the Kennedy government's acceptance of the goals of
agrarian reform, commodity price stabilisation, and economic integration
- was Prebisch's call for social reform in his essay, Hacia una dindmica del
desarrollo latinoamericano (I963). Here he appealed for specific reforms in
agrarian structure and income distribution.9 Beyond this, he wrote that
9 Rauil Prebisch, Hacia una dindmica del desarrollo latinoamericano (Montevideo, i967
[originally published in I963]), pp. 41, 52. Prebisch had called for land reform in the

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependency Analysis 147
Latin American industrialisation was based on the technology appropriate

to the labour-saving needs of the developed countries, and that the
consumption patterns of the upper strata exacerbated the problem
through their preferences for capital-intensive consumer goods. It was
'absolutely necessary [ineludibe] for the state to deliberately compress the
consumption of the upper strata...'. Given the sharply skewed pattern of
income distribution and the upper classes' high propensity to consume,
deteriorating domestic terms of trade between agriculture and industry
had their explanation 'in the insufficient dynamism of development,
which does not facilitate the absorption of the labour force [because such
absorption is] not required by the slow growth of demand [for agricultural
products] and the increase of productivity in primary activities. This
insufficient dynamism prevents a rise of wages in agriculture parallel to the
increase in productivity, and ... [thus] primary production loses in part or
in whole the gains from its technological progress'. Prebisch further
denounced the actual pattern of industrialisation in Latin America,
pointing out that the exaggerated pattern of protection had allowed
grossly inefficient industries to arise. Latin America had, on the average,
the highest tariffs in the world, depriving it of economies of scale and
opportunities to specialise for export.'0 Thus in retrospect this I963
statement anticipates the sombre Prebisch of Capitalismo periferico (198I).
In the same period Prebisch acknowledged that 'social' as well as
economic forces had to be 'influenced' if reforms were to be achieved."
To this end, in I962 ECLA established an annex called the Latin American
Institute for Economic and Social Planning (ILPES).'2 ECLA's reformism
had definite limits, however, since it was an international agency whose
constituent members were Western Hemisphere states and disengaging
European colonial powers; at a theoretical level, it always assumed that
the state was an exogenous factor in the economic and social system. In
the latter i 960s it reformist efforts were still focused on pressuring
early I950S, though in unspecific terms. Prebisch, 'The Soviet Challenge to American
Leadership: America's Role in Helping Underdeveloped Countries', Prebisch file,
ECLA, Santiago, I95 2(?), p. 6.
0 Prebisch, Hacia una dinaimica, pp. 21, 41, 90,, 99.
Rauil Prebisch, 'Economic Development or Monetary Stability: the False Dilemma',
Economic Bulletinfor Latin America, vol. 6, no. i (March I96I), p. 24. (The Spanish
version was published simultaneously.)
12 ECLA's self-imposed responsibility for directing attention to restructuring the
international trading system was largely shifted in I963 to a new agency, the United
Nations Conference on Trade and Development (UNCTAD), which, under Prebisch's
leadership as its first executive secretary, was clearly the international body most
appropriate for such efforts.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
148 Joseph L. Love
governments of developed countries to liberalise their trade policies, thus

assisting the new UNCTAD, and persuading Latin American states to
accept a greater degree of regional integration. Furthermore, ECLA had
always welcomed foreign investment in Latin America, at least under
certain conditions, and this attitude patently ruled out a number of radical
strategies. 13
The reformism of the sixties was conditioned, and for an increasing
number of structuralists, made irrelevant, by a long evolution of ECLA's
views on its initial key policy recommendation - import-substitution
industrialisation. By 1957 the organisation was distinguishing between
two types of ISI. (In the sixties these would be seen as phases.) The first
involved the relatively easy substitution of simple domestically-produced
consumer goods for previously imported items. The second, more
difficult, type involved the production of intermediate goods and
consumer durables, a shift from 'horizontal' to 'vertical' ISI - so called
because of the substitution of simple goods on a broad front in the first
phase, and in the second, an integrated line of production of fewer final
goods and their inputs. A third phase, the production of capital goods,
would ensue at a later date."4
In i956 ECLA still assumed the existence of a threshold in structural
changes in the economy, beyond which 'dependence on external
contingencies' would diminish. Yet the following year the agency first
suggested that dependence on 'events overseas ' might even increase as ISI
advanced; all the same, it still held that 'import substitution ' consisted of
diminishing 'the import content of supplies for the home market'. 5

Argentina was Latin America's most industrialised country, and despite
the unique political phenomenon of Peronism, ECLA tended to view it in
1957 as a trend-setter for other Latin America nations. Argentina, ECLA
noted, had reduced its imports of finished goods to one-third of total
consumption. Yet its declining capacity to import had meant that
reducing the importation of consumer goods was not sufficient to contain
balance-of-payments difficulties; capital goods and fuels also had to be
reduced, and this fact was reducing the rate of growth. Chile was seen as
13 Norman Girvan, 'The Development of Dependency Economics in the Caribbean an

Latin America: Review and Comparison', Social and Economic Studies, vol. 22, no. I
(March 1973), p. 8; ECLA, International Cooperation in a Latin American Development
Policy (New York, 1954), p. I5.
14 ECLA, Economic Survey of Latin America: ipy6 (New York, 19 5 7), p. I I 6; ECLA, The
Process of Industrial Development in Latin America (New York, I966), pp. I9-20; Octavio
Rodriguez, La teorzia del subdesarrollo de la CEPAL (Mexico, I980), pp. 202-3.
15 'The Situation in Argentina and the New Economic Policy', Economic Bulletin, vol. i,
no. I (Jan., I956), p. 30; ECLA, 'Preliminary Study of the Effects of Postwar
Industrialisation on Import Structures and External Vulnerability in Latin America',
in Economic Survey ipy6, p. II 5.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependency Analysis I49
facing similar though less dire problems. ECLA seemed to wonder aloud
whether the Argentine experience was the future of Latin America. Two
conclusions followed: that primary exports and food production for
domestic consumption had to be increased, and that a region-wide
common market must be developed to assure the future development of
efficient manufacturing industries. 6
Why should industrialisation bring rising import requirements in its
train? Using a simple two-sector model, Furtado in i958 explained the
problem as one in which, by assumption, the advanced sector, A, had a
larger import coefficient than the backward sector, B. As the economy
developed, A's coefficient grew ever larger as a share of the whole
economy's coefficient, and paripassu the average import coefficient tended
to rise."7 If the terms of trade were deteriorating, the pressures on the
balance of payments became even more acute.
Thus, for ECLA economists in the latter fifties, the import requirements
in the later stages of ISI, unless offset by capital inflows or rising exports,
could cause 'strangulation' - a favourite ECLA metaphor for stagnation
caused by insufficient imports of capital goods and other industrial inputs.
As a partial solution to stuttering ISI, ECLA in 95 7-8 formally appealed
to its sponsoring states for a Latin American common market, which,
ECLA held, would provide incentives (through economies of scale) for
the production of capital and intermediate goods.18
Yet in its first decade the Latin American Free Trade Area, established
in I960, was only an expression of hope for alleviating the ills associated
with ISI. In I96I Prebisch wrote, 'It remains a paradox that
industrialisation, instead of helping greatly to soften the internal impact
of external fluctuations, is bringing us a new and unknown type of
external vulnerability '.i9 In I964 came the agonising reappraisal of ISI: In
that year an ECLA study, though blaming Latin America's declining rates
of growth on deteriorating terms of trade in the I 9 5 os, also noted that 8o %
of regional imports now consisted of fuels, intermediate goods, and
capital equipment. Consequently, there was little left to 'squeeze' in the
region's import profile to favour manufacturing.20 Meanwhile, two
monographs, highly critical of ISI, appeared in the agency's Economic
1 ECLA, 'Preliminary Study', pp. I 28, 1 50, 1 5 1.

17 Celso Furtado, 'The External Disequilibrium in the Underdeveloped Economi
Indian Journal of Economics, vol. 38, no. I5I (April i95 8), p. 406.
18 'Bases for the Formation of the Latin American Regional Market', Economic Bulletin
vol. 3, no. i (March I958), p. 4. The seventh session of ECLA in I957 had adopted
resolution calling for steps towards the creation of a region-wide common market.
19 Prebisch, 'Economic Development or Monetary Stability', p. 5.
20 ECLA, The Economic Development of Latin America in the Postwar Period (New York,
I964), pp. 14, 21.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
I 5 0 Joseph L. Love
Bulletin - one on the Brazilian experience in particular, and the other on

Latin America in general. These articles pointed to problems that by the
I960s were beginning to affect other parts of the Third World as well.2'
Examining the Brazilian case in the fifties and early sixties, Maria da
Conceicdo Tavares argued that ISI had failed because of the lack of
dynamism of the export sector, coupled with the fact that ISI had not
diminished capital and fuel import requirements. Other problems were
apparent ceilings on the domestic market, owing in part to highly-skewed
income distribution, which also determined the structure of demand; the
constellation of productive resources (e.g. the lack of skilled labour); and
the capital-intensive nature of industrialisation in more advanced phases
of ISI, which implied little labour absorption. In the advanced stages of
ISI, Tavares contended, the low labour absorption of manufacturing
tended to exaggerate rather than to terminate the dualism of Brazil's
economy. Among other things, she argued that bottlenecks in the food
supply, partly due to the antiquated agrarian structure, put unsustainable
pressures on the import bill. Tavares recommended agrarian reform as a
partial solution.22
In the same number of the Bulletin, Santiago Macario wrote a blistering
critique of the way in which ISI had actually been practised in Latin
America, following up Prebisch's observation the previous year (I963)
that the region currently had the highest tariffs in the world. Macario
observed that the governments of the four most industrialised countries
- Argentina, Brazil, Mexico, and Chile - had used ISI as a deliberate
strategy to counteract a persistent lack of foreign exchange, and to create
employment for rapidly expanding populations. But in those four
countries, and in most of the others of the region, protectionism
(primarily in the form of tariff and exchange policies) had been irrational,
in that there was no consistent policy to develop the most viable and
efficient manufacturing industries. On the contrary, the most inefficient
industries had received the greatest protection; there had been over-
diversification of manufacturing in small markets in the 'horizontal'
phase; and these factors had contributed, in some instances, to real
dissavings.23
Nor did Latin American manufactures hold their own in international
markets, continued Macario, at a time when exchange earnings had
21 Maria da Conceisao Tavares, 'The Growth and Decline of Import Substitution in

Brazil', and Santiago Macario, 'Protectionism and Industrialization in Latin America',
Economic Bulletin for Latin America, vol. 9 (I964), pp. I-59, and 6i-ioi, respectively.
For citations of critical studies of ISI strategies in Asia in later years, see H. W. Arndt,
Economic Development: The History of an Idea (Chicago, I987), pp. 82-4.
22 Tavares, 'Growth and Decline', pp. 7-8, Ii, I2, 55.
23 Macario, 'Protectionism', pp. 65-7, 77, 8i.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependency Analysis I 5 I
become critical for the future of industrialisation. On the positive side,

there were tendencies in the early sixties to abolish exchange controls,
quantitative restrictions, and multiple exchange rates, and a related
tendency to begin tariff reduction; yet Macario asserted that Latin
American tariffs were still being built on a makeshift basis, resulting in a
gross misallocation of scarce resources.24
Rational criteria were needed to develop industries - such as the use of
factors in greatest abundance (e.g. labour), or the promotion of industries
that could earn foreign exchange. Equally important, thought Macario,
was the establishment for each country of a 'uniform level of net
protection'. Overall, his thesis was less that ECLA's policy prescriptions
had initially been wrong - which Tavares's analysis did in some ways
imply - than that the region's governments had flagrantly ignored
ECLA's technical advice, pursuing, in Macario's words, 'import
substitution at any cost'.25 Though Albert Hirschman suggested four
years later than Tavares and Macario had been somewhat premature in
issuing the death certificate of ISI, other scholars soon added new charges,
such as that ISI increased the concentration of income with regard both
to social class and to region (within countries).26
In 1956 ECLA had voiced its first doubts whether industry could
absorb surplus labour from agriculture in the world region with the fastest
growing population; nine years later its survey of ISI showed that non-
agricultural employment in Latin America had increased from 13 to 36
million persons between 192 5 and I 960, but that only five of the 23 million
additional employees were absorbed in industrial work.27 Furtado,
writing over his own signature, noted in I966 that while Latin America's
industrial output in the 195os had risen 6.2 % a year, industrial
employment had risen only I.6 % annually, about half Latin America's
average population growth rate. The problem in part was the labour-
saving technology which the Periphery had imported from the Centre.28
24 Ibid., pp. 67, 78, 8i.
25 Ibid., pp. 67 (quotation), 84 (formula for 'uniform level of net protection'), 87.
26 Inter alia, Hirschman argued that the failure of ISI was not inevitable, but depended on
the interaction of social and political factors with economic elements. See his 'The
Political Economy of Import Substituting Industrialization in Latin America' [orig.,
I968], in A. Hirschman, A Biasfor Hope (New Haven, 1971), pp. 85-123, esp. p. 103.
On ISI failures, see the discussion of Sunkel and Furtado in the following section, and
the literature surveyed in Werner Baer, 'Import Substitution and Industrialization in
Latin America: Experiences and Interpretations, Latin American Research Review, vol.
7, no. i (spring 1972), pp. 95-122, esp. p. 107.
27 ECLA, 'The Situation in Argentina', p. 42; ECLA, The Process of Industrial Development
(New York, I966 [Spanish orig., I965]), p. 38.
28 Celso Furtado, Subdesenvolvimento e estagnafo na Am(rica Latina (Rio, 2nd edn,
[orig., I966]), pp. 9-IO. As noted above, Prebisch had previously made this point in
Hacia una dinamica, p. 38.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
Furtado explained that ISI was fundamentally different from European

industrialisation in the eighteenth and nineteenth centuries. In the classic
phase, technology continually cheapened the relative cost of capital goods,
creating the possibility of solutions to social problems. In twentieth-
century Latin America, unlike nineteenth-century Europe, technology
was exogenous to the regional economy, and was specifically designed for
the requirements of the developed countries. Factor absorption, therefore,
did not depend on the relative availability of factors, but on the type of
technology used, and over this matter Latin Americans could exercise
little choice.29 Among other things, they had to compete in their own
national markets with high-technology, multinational corporations.
This tendency by the mid-sixties to take the longer view and seek
lessons in history was partly the result of the fact that, from ECLA's
perspective, Latin America now had 3 5 years of import-substituting
experience, but it was also rooted in the long-term perspective that was
part of ECLA's original style. Prebisch's I949 Survey had tried to view
sweep of economic history from the I88os to the mid-twentieth century
for the region, and in more detail for the four most industrialised nations,
Argentina, Brazil, Chile, and Mexico. In some ways this volume was a
model for the case studies on those countries that were to be carried out
between I959 and I963 - Furtado on Brazil, Pinto on Chile, Aldo Ferrer
on Argentina; and later, Osvaldo Sunkel and Pedro Paz on the whole
region. Furtado's study, however, derived principally from h

ECLA interests in the defining features of colonial Brazil. In fact, more
than anyone else at ECLA, Furtado was responsible for 'historicising' the
analysis, thus departing from cyclical concerns.3"
29 Furtado, Subdesenvolvimento, pp. 9-I 1. The perspective of the I98os permits a sligh
more sanguine view of industrial employment than that adopted by ECLA and Furtado
in the mid-sixties: although industry in Argentina and Chile absorbed a smaller
percentage of the labour force in I98I than in I965, in I98I it employed a larger share
in Brazil, Mexico, Colombia, and Venezuela. But in no Latin American country in I98 I
did industry account for as much as a third of the labour force. International Bank for
Reconstruction and Development, World Development Report: 198y (New York, I985),
pp. 214- I 5.
30 Celso Furtado, Formafdo econrmica do Brasil (Rio de Janeiro, I959); Anibal Pinto Santa
Cruz, Chile, un caso de desarrollo frustrado (Santiago, I95 9); Aldo Ferrer, La economia
argentina: las etapas de su desarrolloj problemas actuales (Mexico, I963); Osvaldo Sun
and Pedro Paz, El subdesarrollo latinoamericano y la teorla del desarrollo (M6xico, 1970).
Subsequently a more specialised structuralist work appeared on Mexico: Rene
Villareal's El desequilibrio externo en la industriali.acidn de Mexico (1929-7y): Un enfoque
estructuralista (Mexico, 1976). Villareal argues, however, that structuralism accounts
more adequately for Mexico's external disequilibrium in the period 1939-58 than in
195 9-70.
31 Furtado's pre-ECLA dissertation does not contain much formal economic analysis of
any kind. See Furtado, 'L'economie coloniale bresilienne (XVIe et XVIIe siecles):
Elements d'Histoire Economique Appliques', PhD thesis, Faculte de Droit, Universite

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependengy Analysis I 5 3
Explicit in the I949 Survey was the thesis that industrialisation had
historically occurred in periods of crisis; or, as a critic of ECLA put it,
ECLA viewed development as occurring through the agency of 'external
shocks'.32 In Brazil, Furtado pointed to rapid industrial growth in the
Depression, partly due to 'the socialization of losses' through exchange
devaluation, which none the less helped maintain domestic demand - a
point made in the I949 Surv6y in different language for the region as a
whole, contemporaneously with Furtado's first use of the idea.33 Furtado
viewed expansionary fiscal and monetary policies as a form of unwitting
Keynesianism, and his views on Brazilian industrialisation in the
Depression touched off a long debate. Yet it now seems clear for Brazil,
as for the other most-industrialised countries, that the World Wars and the
Depression were less important in producing 'inward-directed growth'
than was believed by some contemporaries to these events, and by ECLA
economists later.34 A now widely-held view is that investment in industry
(capacity) grew in line with export earnings for the period 1900-45, while
output (but not capacity) tended to rise during the 'shocks', when imports
had to be curtailed. Capacity could not grow appreciably during the
Depression for lack of exchange credits to buy capital goods and inputs,
nor during the World Wars because of the unavailability of capital goods
and fuels from the belligerent powers.35
de Paris, 1948. But A economia brasileira (Rio de Janeiro, 1954) is a structuralist analysis
of Brazil's economic history. Furtado sees as one of his major contributions to
structuralism a complete break with a cyclical framework for a fully historical
approach. Furtado to author, Paris, 22 Dec. i982.
32 Carlos Manuel Pelaez, Hist6ria de industrialiZafdo brasileira: critica a teoria estruct
no Brasil (Rio de Janeiro, 1972), ('external shocks'). This view is stated concisely in the
Economic Survey 1949 7), where it is asserted that, in Argentina, the crisis of I 890 began
industrialisation, and that World War I produced new industries, 'which developed
most vigorously during the Great Depression and the following war'. (See citation in
note 33, below.)
3 For the best summary of Furtado's arguments and the subsequent debate in Brazil, see
Wilson Suzigan, Induistria brasileira: Origem e densenvolvimento (Sao Paulo, i986), pp.
2 1-73. For a similar thesis about the socialisation of losses through exchange
depreciation and government maintenance of aggregate demand, see ECLA, Economic
Survge of Latin America: J949 (New York, I 9 5 I [Spanish orig., I9 5 0]), pp. 6o, 171-
Furtado's argument appeared in print the same year; see Furtado, 'Caracterfsticas
gerais da economia brasileira', Revista Brasileira de Economia, vol. 4, no. I (March 195o),
pp. 1-37. As early as 1932 Prebisch at Geneva had held that '...currency depreciation
has lessened the effect on the home market [in Argentina] of the world drop in prices'.
Prebisch, 'Suggestions Relating to the International Wheat Problem', League of
Nations: Commission Preparatoire/Conference Monetaire et Economique/E8, i i
Dec. 1932 (mimeo), p. 3. (Located at the Palais des Nations, Geneva).
34 See essays in Rosemary Thorp (ed.), Latin America in the ipjos: The Role of the Pe
in World Crisis (New York, I984).
3 During World War II, Brazil's growth was perhaps less hampered because of the

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
Thus the perceived failure of ISI as a historical process - and perhaps

the growing suspicion that industrial growth had varied directly and not
inversely with export earnings - was a leading cause of pessimism among
structuralists in the mid- to late sixties. In a variety of other areas there
were also reasons for pessimism, both immediate and long-term. ECLA's
efforts at 'programming' - calculating required savings and inputs to
meet government-specified development targets - had borne little fruit by
the middle of the decade. ECLA publications were now criticising the
agency's own earlier efforts in the field for failing to take into account
'social and political viability' as an essential criterion when attempting to
set and meet realistic development targets. Another problem perceived in
I966 which had an obvious social dimension was the rising degree of
income inequality among social classes in Argentina, usually viewed as the
pacesetter.36 Increasing income concentration was already observable in
Mexico, and would be apparent in Brazil after the census of 1970.
Meanwhile Sunkel, writing in his own name in I966, criticised
agricultural performance, documenting the fact that while food crop
yields in the United States had continually risen from the late nineteenth
century through the early I96os, those in Latin America had stagnated.
This was an unsettling discovery for those who hoped to save ISI by
reducing food imports in a region where urban population was growing
even more rapidly than general population.37 (Others argued that ISI,
with its high-cost manufactured goods, had turned the domestic terms of
trade against agriculture.) For Sunkel, other problems without solution
were the apparent fall in the overall rate of economic growth from I950
to I965; the persistent rather than diminishing dependency on the foreign
trade sector as the key to growth; the unemployment or underemployment
in both urban and rural sectors, as population climbed higher; and the
apparent concentration of income in the upper income brackets. In
existence of a small capital goods sector. For discussions of the revisionist literature,
see Suzigan in note 3 3; Joseph L. Love, Sao Paulo in the Bra.Zilian Federation, J889
(Stanford, Calif., I980), pp. 57-9; Laura Randall, An Economic Historj of Argentina in
the Twentieth Centurj (New York, 1978), p. 125; Ciro Flamarion S. Cardoso and Hector
Perez Brignoli, Historia econdmica de Amjrica Latina (Barcelona, 1979), vol. II, pp. 197
(summarising a literature on Mexico, Argentina, Brazil, and Chile) and I99; Perez
Brignoli, 'The Economic Cycle in Latin America: Agricultural Export Economies
(I880-I930)', Latin American Research Review, vol. i5, no. 2 (I980), pp. 31-2 (on
Argentina).
36 Rodrfguez, Teoria, p. 223 (on planning); 'Income Distribution in Argentina',
Economic Bulletin for Latin America, vol. ii, no. I (I966), pp. I06-3 I.
3 Osvaldo Sunkel et al., Inflacidn y estructura econdmica (Buenos Aires, I967), cited in
Cardoso and Perez Brignoli, Historia econdmica, pp. I 87-9; Baer, 'Import Substitution',
p. I05.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependengy Analysis 55
Sunkel's view, the long-term prospects for Latin America were

'frightening .38
Other structuralist criticisms of the Latin American economies in the
latter sixties included the perception of a sustained balance-of-payments
disequilibrium and a 'debt spiral' for the most industrialised countries,
and of high rates of inflation, with consequent social tensions and political
instability. For ECLA, these problems were chiefly structural problems -
the agrarian pattern of latifundia and minifundia; industrial structure
(indivisibilities of scale, tending to produce underutilisation of capital, and
high capital density, with the consequence of low labour absorption); the
rigidly stratified social structure; and the consequent maldistribution of
39
income.
On the political front, another source of pessimism was the end of the
'developmentalist' experiment in Brazil, where the populist President
Joao Goulart was ousted by a military coup in i964; two years later, a
coup in Argentina installed a conservative and authoritarian government.
Furthermore, the diminished interest of the USA under Lyndon Johnson
in the reform and development goals of the Alliance for Progress, coupled
with the invasion of the Dominican Republic in I965 (the first such action
in Latin America since the twenties), was another blow to reformism.
More broadly, the intellectual and political climate in which dependency
analysis would be received was radicalised by the international resistance
to the US war in Vietnam, of which the Dominican intervention was a
consequence (to prevent a 'second Cuba'). Resistance to the Vietnam War
interacted with anti-establishment protest in a variety of countries, often
led by students, and reached a peak in the demonstrations and repressions
of I968-70.
Yet there was a tendency, if still not a dominant one, for official
economic policy in Latin America to move in a contrary direction from
the radicalism of the streets. Anti-ECLA orthodoxy made its reappearance
in the programmes of 'monetarists', in the anti-populist and authoritarian
regimes of Brazil after I 964 and Argentina after I 966. In Brazil, at the end
of the decade, it was a professedly neoclassical economist, Antonio Delfim
Neto, who implemented ECLA's earlier calls for the export of
manufactures.40 These shifts presaged the dominant position that liberal
38 Osvaldo Sunkel, 'The Structural Background of Development Problems in Latin

America', [orig., I966] in Charles T. Nisbet (ed.), Latin America: Problems in EconomiC
Development (New York, I969), pp. 7, II, 13, 23.
3 Rodriguez, Teoria, pp. I87-8, 214-7.
40 In 1957 ECLA has still denied that Latin America could compete in manufactured
exports on the world market, but in the i 960s the agency viewed the export of
manufactures as a requirement for continued development. ECLA, Economic Survey
J9y6, p. I5I; Rodriguez, Teoria, p. 222.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
orthodoxy would enjoy at the policy level under military regimes in the
1970s. The ascending star of monetarism and associated policies in the
latter sixties (even though not necessarily incompatible with structuralist
policy in actual practice, as in the Brazilian case), probably contributed to
a frustration in intellectual circles which prepared the way for the
acceptance of dependency analysis.
In this regard, Chile's experience under its only Christian Democratic
administration was probably even more important than that of Brazil,
even though Chile's military coup was still several years away. The host
country for ECLA, Chile was the site of a major reformist experiment
under the administration of Eduardo Frei for the six years beginning in
November, I964. Frei's chief economic adviser, Jorge Ahumada, a former
ECLA analyst, sought in the economic plan of I965 concurrently to
achieve faster growth, a redistribution of income (in part through agrarian
reform), and the elimination of the structural causes of inflation.
After an initial success in stimulating the economy through expansion-
ary policies, the Frei government saw the growth rate slow from an
average of 6.o % in i965-6 to 3.20% in I967-70. The rate of investment
also dipped in the Frei years, as employers became leary of the
government's social experiments, and unemployment rose. To contain
inflation, Frei turned increasingly to monetarism after I967, without
notable success: the rate of inflation, held to 17.0% in I966, averaged
z8.0 % in I967-70. The foreign sector, so important in structuralist
analysis, was not responsible for these difficulties, as copper prices
remained high, along with exports and imports in general.4"
The Frei government's agrarian reform programme produced more
impressive results, but aspirations of the rural population rose faster than
land redistribution, resulting in a tendency toward political polarisation in
the countryside and beyond. In addition, the degree of social trans-
formation achieved under the Christian Democratic government was
modest compared with that of the Unidad Popular administration that
succeeded it.42 Consequently, the Frei experience revealed the enormous
Sergio Bitar, Transif,co, socialismo e democracia: Chile com Allende, tr. by Rita Braga (Sa
Paulo, I980), pp. 49-5o; Anibal Pinto, 'Desarrollo econ6mica y relaciones sociales', in
Pinto et al., Chile, hoy (Mexico, 1970), p. 47; Enrique Sierra, Tres ensayos de estabil
en Chile: Las politicas aplicadas en el decenio J9y6-J966 (Santiago, 1970), pp. 91-4, I83-5.
Vittorio Corbo Lioi argues that most of the alleged structural causes of inflation were
eliminated during this period, and notes that the Frei government attributed the failure
of its stabilisation programme 'to the behaviour of wages', not to the inelasticity and
instability of the demand for exports, deficiencies of the fiscal system, or the inelasticity
of the supply of agricultural goods based on traditional patterns of land tenure. Corbo,
Inflation in Developing Countries: An Econometric Study of Chilean Inflation (Amsterdam
1974), p. I5.
42 The amount of land distributed and the rural workers unionised in the six years

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependengy Analysis I 5 7
political difficulties of implementing ECLA-inspired reforms as envisioned

in Prebisch's Hacia una dinaumica and other structuralist writings of the
I 960s.
Two positive developments in the late sixties seemed to have had little
effect on the emerging vision of dependency, namely, a cyclical
improvement of the terms of trade for primary producers, and the efforts
by Argentina and Brazil to decrease their tariffs in order to stimulate
industrial efficiency.43 Perhaps the former process was not perceivable in
the short run, and the latter was carried out by governments having little
sympathy for ECLA.
Ironically enough, the perspective of the late I98os allows an even more
favourable view of the economic climate in which dependency analysis
emerged. The rate of population growth in Latin America, highest among
the world's major regions, had peaked in the early I96os, and had begun
a long-term decline (something Sunkel could not have known in the
middle of the decade). The international economy was more dynamic in
the period I960-73 (ending with the OPEC oil price shock) than in any
other period in the postwar era, permitting unparalleled diversification of
Latin American exports, including manufactures. Perhaps most ironic
was the fact that the postwar economic growth rate for the region reached
a peak in I965-73, and in the same years, industrial output in the region
averaged 8.o % annually, a higher rate than in any other period before or
since.44 But analysts of the mid-sixties, in the midst or on the cusp of these
developments, had no way of knowing Latin America was beginning its
most successful period of economic growth and diversification.
In any case ECLA's theories and policy prescriptions were not only
challenged by the neoclassical Right, but also by a heterodox Left, some
of whose exegetes had been leading figures in ECLA itself, notably
Furtado and Sunkel. This new Left would quickly make 'dependency
theory' famous. Although ECLA itself had produced nothing if not a
kind of dependency analysis, the new variety was set off by its more
clearcut 'historicising' and 'sociologising' tendencies in both its reformist
and radical versions.
Not only had Furtado and Sunkel adopted in the mid-sixties an
Frei government were only half the numbers attained in the three years of Allende's
presidency. For details, see Sergio G6mez, Instituciones y procesos agrarios en Chile
(Santiago, 1982), pp. 24-33.
4 Baer, 'Import Substitution', p. IIO (on Argentina and Brazil); Palma, 'Dependency',
p. 908 (on terms of trade).
44 Ricardo Ffrench-Davis and Oscar Munioz, 'Latin America in the International
Economy: 195o-86' (manuscript for Cambridge History of Latin America, revised), pp.
i, 6, 1I, 26, 36.
6 LAS 22

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
explicitly historical view of development, noted above, but Furtado now

elaborated on an earlier contention that development and underdevelop-
ment were linked. As early as I 9 5 9 he had written that there was a
'tendency for industrial economies, as a result of their form of growth, to
inhibit the growth of primary economies', and he expanded on the idea
in Development and Underdevelopment two years later. In I966 he again
argued that because the two processes were historically associated,
underdevelopment could not be a phase in the passage to development.45
In his analysis of economic history, Furtado also introduced the
element of social class. He argued in I964 that class struggle had
historically been the engine of economic growth in the West: workers
'attack' through organisation to raise their share of the national product,
and capitalists 'counterattack' by introducing labour-saving technology.
Since labour is unorganised in the Periphery, above all in the rural sector,
he asserted, the process fails to work there. These marxisant propositions
are perhaps less surprising than they appear at first glance, since they are
an extrapolation of Prebisch's initial explanation of declining terms of
trade.4" Furtado's several essays in the latter sixties, written largely in hi
Paris exile, pointed to the need for an analysis of the whole capitalist
system, Centre and Periphery together, a problem which Samir Amin,
Immanuel Wallerstein, Johan Galtung, Arghiri Emmanuel, Andre
Gunder Frank, and Furtado himself began to address in the seventies.
Celso Furtado, A operafao nordeste (Rio de Janeiro, I 9 5 9), p. I 3; Furtado, Desenvolvimento
e subdesenvolvimento (Rio de Janeiro, I96 i) p. I8o and passim; Furtado, Subdesenvolvimento,
pp. 3-4. These statements lend credence to a claim of priority for Furtado as the first
theorist of dependency, but I believe H. W. Arndt has exaggerated in tracing Furtado's
dependency position back to The Economic Growth of Bra.zil (Berkeley, Calif., I963
[Portuguese orig., 1957]), which in my view is more correctly described as the full
historicisation of structuralism. At all events, in a recent retrospective Furtado dates his
central contributions to dependency analysis in books and articles published between
1970 and 1978. In these works, Furtado views as a central feature of underdevelopment
the adoption of the consumption patterns of the developed West by the upper strata
of underdeveloped areas, as these regions entered the international division of labour.
This process was the 'result of the surplus generated through static comparative
advantages in foreign trade. It is the highly dynamic nature of the modernized
component of consumption that brings dependence into the technological realm and
makes it part of the production structure'. Novel items of consumption require
increasingly sophisticated techniques and increasing amounts of capital. But capital
accumulation is associated with income concentration, so industrialisation 'advances
simultaneously with the concentration of income'. Celso Furtado, 'Underdevelopment:
To Conform or Reform', in Gerald Meier (ed.), Pioneers in Development, second series
(New York, I987), pp. 2I10-II.
Thus the evolution of Furtado's thought on this matter was similar to Prebisch's
between Hacia una dindmica (I963) and Capitalismo perje'rico (I98I). For Furtado's
statement of this position, see Underdevelopment and Dependence: The Fundamental
Connections (Cambridge, Centre of Latin American Studies, Cambridge University,
I973 [offset]), i8 pp.
46 Furtado, Subdesenvolvimento, p. 7; Furtado, Diagnosis of the Brazilian Crisis, tr. Suzett
Macedo (Berkeley, I965 [Portuguese orig., I964]), pp. 48-5 i.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependengy Analysis Is 9
Another Brazilian, Fernando Henrique Cardoso, during his association

with ECLA in the latter sixties (through its sociological annex, ILPES),
played a major role in moving the dependency perspective towards an
analysis of social relations. Indeed, Santiago, with its various research
institutions through which structuralists and their leftist critics moved,
was the crucible of dependency analysis.47 Independently of ECLA,
Cardoso had arrived at a pessimistic view of the 'national bourgeoisie',
through his empirical studies of industrialists in Brazil and Argentina; and
his view that Latin America lacked what Charles Moraz6 has called a
' conquering bourgeoisie' was shared by other sociologists.48 Cardoso had
reached his position before the presence of multinational corporations
became so prominent (and native industrialists relatively inconspicuous)
in the 'open economies' of the neoclassical policymakers of Argentina and
Brazil in the late I96os and 1970s.
While Sunkel spoke of the international capitalist system as 'a
determining influence on local processes', and one which was 'internal' to
the Periphery's own structure, Cardoso and his Chilean collaborator,
Enzo Faletto, preferred to speak of two subsystems, the internal and the
external, and emphasised that the international capitalist system was not
solely determining. There was a complex internal dynamic to the system,
they asserted.49
Beyond this, Cardoso and Faletto stressed the mutual interests among
social classes across the Centre-Periphery system. The interests of the
bourgeoisie in the Centre (and by implication, those of its proletariat)
overlapped those of the bourgeoisie of the Periphery; these links became
all the more intimate as multinational corporations loomed ever larger in
Latin America. (In I965 the Brazilian political scientist H6lio Jaguaribe
had already specified a 'consular' bourgeoisie, with distinct interests from
E.g. note the Brazilian contributors to dependency who were in Santiago in the years
following the I964 coup in their country: Furtado, Cardoso, Theot6nio dos Santos,
Rui Mauro Marini, Vania Bambirra, and Jose Serra (still a student). Europeans in
Santiago in the ig6os influenced by the emerging dependency perspective included
Johan Galtung and Keith Griffin.
48 Fernando Henrique Cardoso, Empresdrio industrial e desenvolvimento econ6mico no Brasil
edn (Sao Paulo, I97z [orig., i9641), pp. i8i, I9I, i95; Cardoso, 'The Entrepreneurial
Elites of Latin America', Studies in Comparative International Development, vol. z (I966),
p. I47; Cardoso, Ideolog4as de la burguesia industrial en sociedades dependientes (Argentina y
Brasil) (Mexico, 1971) [data collected in 1963, 1965 -6, pp. 1, 103, 146, 158, 2i5; Dardo
Cuineo, Comportamientoy crisis de la clase empresarial (Buenos Aires, 1967), pp. I 29, I 72,
I92; Claudio Veliz, 'Introduction', in V6liz (ed.), Obstacles to Change in Latin America
(London, I965), pp. 2 , 7-8; and in a more popular vein, Rodolfo Stavenhagen, 'Seven
Fallacies about Latin America' [orig., I965], in James Petras and Maurice Zeitlin (eds.),
Latin America: Reform or Revolution?: A Reader (Greenwich, Conn., I968), pp. 20, 22.
9 Osvaldo Sunkel, 'The Pattern of Latin American Dependence', in Victor L. Urquidi
and Rosemary Thorp (eds.), Latin America in the International Economy (London, 1973),
p. 6; Fernando Henrique Cardoso and Enzo Faletto, Dependenciay desarrollo en Am(rica
Latina (Mexico, i969), pp. I7, 28, 38.
6-2

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
i 6o Joseph L. Love
the 'national' or 'industrial' variety.) Cardoso and Faletto analysed the

development of the 'populist' coalition of national and foreign capital
with the working class, corresponding to the successful phase of ISI, and
linked the failure of import substitution with the demise of the populist
political style. In the current phase of capital accumulation, they believed,
authoritarian regimes were required to assure a political demobilisation of
the masses.50
Their treatment of dependency, despite its early appearance, was more
nuanced than others, emphasising contradiction, shifting alliances, and a
range of historical possibility. Cardoso and Faletto distinguished between
simple enclave economies and those controlled by local bourgeoisies. For
the latter, they entertained the possibility of significant manufacturing
sectors. In a scheme they called 'associated development ' or 'development
with marginalisation [marginalidad]' and which Cardoso would later term
'associated-dependent ' development, they noted that contemporary
foreign capital was focusing its investment in manufacturing operations.
Furthermore, the public sector, multinational capital, and the 'national'
capitalist sector were joining hands under authoritarian rule. Like
Furtado, Cardoso and Faletto pointed to the international system as a
whole as the proper unit of analysis; and, like Furtado, they saw
development and underdevelopment not as stages, but as locations within
the international economic system, for which they offered a schematic
historical analysis of the Periphery's class dynamics.5"
Concurrent with the efforts of Cardoso and Faletto, another researcher
in Santiago was producing a radical version of dependency, and no
dependency model was more widely and hotly debated than that of Andre
Gunder Frank. Because he borrowed from a Marxist tradition - probably
without being decisively influenced by it - it is necessary to consider the
history of Latin American Marxism which, like structuralism, was
undergoing a fundamental reassessment in the I96os.
In fact, the crisis of Marxism in Latin America revolved around one of
50 Cardoso and Faletto, Dependencia, pp. 27, I43, I54, I55; Helio Jaguaribe, 'The
Dynamics of Brazilian Nationalism', in Veliz (ed.), Obstacles, p. i8z2 ('consular
bourgeoisie'). Lenin in I9I6 had already noted the mutual interests 'between British
finance capital and the Argentine bourgeoisie'. V. I. Lenin, Imperialism: the Hi-ghest
Stage of Capitalism: A Popular Outline (New York, I939), p. 8.
5 Cardoso and Faletto, Dependencia, pp. 28, 32-3, I35 (quotation), I42, I47, i55. For
Cardoso's views on dependency in the early I970S, see his 'Associated-Dependent
Development: Theoretical and Practical Implications', in Alfred Stepan (ed.),
Authoritarian Brazil: Orgins, Poliies and Future (New Haven, I973), pp. I42-78.
The role of the multinationals became the subject of a major research agenda in the
I970S for students of dependency and for ECLA as such; in the absence of a
'European' bourgeoisie, the multinational corporations would dominate the new
phase of industrialisation in Latin America.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependeny Analysis 6i
the same issues that had troubled structuralists, namely, the role of the
local or national bourgeoisies. From the twenties to the sixties, the
Communist parties of Latin America had vacillated - in truth, oscillated
- between the view that the local bourgeoisie was progressive, and the
view that it was reactionary, depending in part on Soviet instructions.
In I929 the Peruvian Marxist Jose Carlos Mariaitegui had argued that
capitalism had reached Latin America too late for the local bourgeoisies
to emulate the historic role of their European counterparts.52 Yet the
prevailing doctrine in most Latin American Communist Parties, from the
Popular Front period, beginning in 1934, through the early sixties, was
that the local bourgeoisie was a progressive force: contending that their
continent had large feudal residues, Party spokesmen argued that
proletarians and bourgeois must struggle together, at the present phase of
history, to eliminate feudal residues and to contain imperialist penetration.
They agreed with Mariategui that capitalism was relatively recent in Latin
America; for Party theorists, it had made its first appearance there in the
nineteenth century. 5 But unlike Mariategui, they concluded that
capitalism could and must, by grim necessity, be developed in the region.
Support for the national bourgeoisie and capitalist industrialisation was
the position, for example, of the Communist Party of Brazil, and the Party
52 Jose Carlos Mariategui, 'Point de vue antiimperialiste' [i929], in Michael Lowy, Le

Marxisme en Amirique Latine de 1909 a nos jors: Anthologie (Paris, I980), p. II3.
Mariategui is usually regarded as the most original Latin American Marxist writing
before World War II. Part of his originality lay in his heterodoxy, including a
voluntarism reminiscent of Russian populism. The Peruvian's praise of 'Incaic
socialism' (a discovery of Plekhanov, a generation earlier) seemed to a Comintern critic
in I94I a reincarnation of Russian populism. Mariategui hinted that because of its
indigenous ayllu, the surviving Incaic form of agrarian communism, Peru could move
from its present semifeudal stage of development directly to socialism. Thus
Mariategui was a 'stage-skipper', perhaps unwittingly following the tradition of
Alexander Herzen and a long line of Russian populists (narodniki). Lenin had repudiated
the populist contention that Russia could 'skip' capitalism, and argued in I899 that
capitalism had already triumphed in Russia, reducing the peasant commune (mir) to a
relic. Mariategui's view that peasant collectivism could be the basis for passing from
feudalism to socialism was shared by one of the leading Latin American spokesmen at
the sixth congress of the Communist International in 1928, Ricardo Paredes of
Ecuador; his views were echoed by a Uruguayan delegate, Sala. On these matters, see
V. M. Miroshevski, 'El " populismo " en el Peru: Papel de Mariategui en la historia del
pensamiento social latinoamericano', in Jose Aric6 (ed.), Mariciteguiy los origenes del
Marxismo latinoamericano, 2nd edn., rev. (Mexico, I980), pp. 55-70; V. I. Lenin, The
Development of Capitalism in Russia: The Process of the Formation of a Home Market for
Large-scale Industry (Moscow, 1956, [orig., I899]); Internacional Comunista, VI
congreso, II, pp. i8o-i, 367. Also see note 69.
53 E.g., V[olodia] Teitelboim, 'El desarrollo del capitalismo en Chile', in Alexei
Rumiantsev (ed.), El movimiento contemporaneo de liberaci`ny la burguesia nacional (Prague,
I96I), p. I56. On feudal residues in contemporary Latin America, see Rodney
Arismendi, 'Acerca del papel de la burguesfa nacional en la lucha antiimperialista', in
ibid., pp. 1 34, 1 36.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
I62 Joseph L. Love
was in effect part of Brazil's 'developmentalist' and populist coalition for

much of the period 1945-64. In Cuba, as late as August I960, just two
months before Castro's sweeping nationalisation of the economy, the
Communist leader Blas Roca announced that the Cuban Revolution was
not socialist, but 'bourgeois-democratic'.5 Yet the issue that the Cuban
Revolution posed after October 1960 was the viability of 'the
uninterrupted path to socialism', a thesis long defended by Latin
American Trotskyists, but now more audibly proclaimed as viable policy
by Ernesto 'Che' Guevara. After Fidel Castro's public adherence to
Marxist-Leninism in December I96I, the thesis that contemporary Latin
America could only sustain bourgeois-democratic regimes would have to
be reappraised.55
A pre-existing historiography was now discovered - beginning with
Sergio Bagui's La economia de la sociedad colonial in 1949, an essay which
argued that Latin America had never had a feudal past at all, but in broad
outline had evinced fundamental features of capitalism since the sixteenth
century. The Spanish and Portuguese empires in the New World, Bagui
and others alleged, were basically commercial enterprises, for which
'feudal' titles and trappings were but a veil.56 This school, composed
chiefly of nonconformist Marxists, would reject Communist orthodoxy on
the role of the bourgeoisie not by asserting that capitalism had arrived too
late in Latin America, as Mariategui had contended, but that, on the
contrary, capitalism had already prevailed too long in the region:
consequently, there was nothing to hope from the local bourgeoisie. This
class, rather than overseas imperialists, would constitute the 'immediate
enemy', as Frank put it, for Cuba-inspired groups.
In Brazil, for instance, as early as 1947, i.e. before Bagui's Economia de
la sociedad colonial, Caio Prado Jr. defended the thesis that Brazilian
agriculture had been capitalist rather than 'feudal' in the colonial era.57
Between I960 and I966 he elaborated the argument that from the outset
Brazilian agriculture had been capitalist in its essential features, i.e. that
54 Lowy, Marxisme, pp. 223-6 (Brazil), p. 47 (Cuba).
55 [Ernesto] Che Guevara, Guerrilla Warfare (New York, 196i [Spanish orig., 1960]), p.
I 5; Lowy, Marxisme, p. 269.
56 Sergio Bagii, La economz'a de la sociedad colonial: Ensayo de la historia comparada de Am6rica
Latina (Buenos Aires, 1949). Bagii's work was not explicitly Marxist, unlike those of
four other contributors to this thesis: Marcelo Segall and Luis Vitale of Chile,
Milciades Pefia of Argentina, and the above-mentioned Caio Prado Jr. of Brazil, all of
whose writings are on this matter are anthologised in Lowy, Marxisme, pp. 243-5 3,
413-22. From a non-Marxist perspective, Roberto Simonsen has denied that Brazil had
known anything other than capitalism in its colonial history, well before the Marxist
debate. Simonsen, Histdria econ6mica do Brasil (Iyoo-Ic820), 4th edn. (Sao Paulo, I 962 [orig.
1937]), pp. 80-3.
57 [Caio Prado Jr.], Tres etapas do comunismo brasileiro', Cadernos de Nosso Tempo, no.
2 (Jan.-June, 1954), p. 127.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependengy Analysis I 63
the Portuguese colony was a mercantile enterprise in which there existed

(at a theoretical level), legal equality among the settlers. In A revoluftio
brasileira (i966), a work which influenced a generation of urban guerrillas
and was the unstated reason for his incarceration in I969, Prado reviewed
the course of Brazilian history to show that his country, as fully
'capitalist', was ripe for revolution, contrary to the official position of the
nation's Communist Party (of which he had long been a member). Prado's
thesis doubtless had a major influence on Andre Gunder Frank, who
expanded Prado's argument on capitalism in Brazilian agriculture in
Prado's journal, Revista Brasiliense, in i963.58 Frank's work forms the most
obvious point of tangency between the revisionist Marxists who
emphasised relations of exchange and the structuralists. In i965-7 he
would in fact rework ECLA's theses to yield a radical conclusion.59 Frank
was explicit about at least some of his sources: from Sergio Bagui - whose
work was not cast in Marxist terms - he borrowed the proposition that
Latin America's economy had been essentially capitalist from the colonial
era. From Paul Baran, whose Political Economy of Growth was widely read
in Spanish and Portuguese editions, Frank took the proposition that
capitalism simultaneously produces underdevelopment in some areas as it
produces development in others.60 Frank briefly worked on commission
for ECLA, from which he presumably took his thesis of deteriorating
terms of trade, though he was to emphasise monopoly elements in the
process much more than ECLA had. Frank's antinomy 'metropolis-
satellite' is surely derived from 'Centre-Periphery', and his notion of
58 Caio Prado Jr., A revolufdo brasileira (Sao Paulo, I966); Fernando Gabeira, 0 que e isso
companheiro? (Rio de Janeiro, 1979), pp. 31-2 (on Prado's influence on urban guerrillas).
Prado's previous articles on capitalism in early Brazilian agriculture were 'Contribuisao
para a analise da questao agraria no Brasil', Revista Brasiliense, no. 28 (March-April,
I960), pp. I 65-23 8, and 'Nova contribuisao para a analise da questao agraria no Brasil',
in ibid., no. 43 (Sept.-Oct. I962), pp. 11-52. These and other articles of the early I96os
are collected in Prado, A questao agrdria (Sao Paulo, 1979). Andre Gunder Frank's essay
reflecting Prado's influence was 'A agricultura brasileira: capitalismo e o mito do
feudalismo', in Revista Brasiliense, no. 5I (Jan.-Feb., I964), published in English in
Frank, Capitalism, pp. 2 I9-77.
Fernando Henrique Cardoso in 1977 saw Prado as one of a group of Brazilian
scholars trying to identify a colonial mode of production. I do not believe Prado saw
the problem in that way in the early I960s: His category was capitalism. See Cardoso,
'Consumption of Dependency Theory', pp. I I-I2.
5 Frank's model has the force and crudity of W. W. Rostow's 'stages of growth ' model,
to which it has been compared. See Aidan Foster-Carter, 'From Rostow to Gunder
Frank: Conflicting Paradigms in the Analysis of Development', WUorld Development, vol.
4, no. 3 (March 1976), pp. I67-80, esp. p. I75; and W. W. Rostow, The Stages of
Economic Growth: A Non-Communist Manifesto (Cambridge, 1960).
60 Despite Frank's attribution of Baran, H. W. Arndt has pointed out that Frank went
beyond Baran, who had held that capitalism was an obstacle to the underdeveloped
world's progress, to argue that underdevelopment was caused by capitalism. Arndt,
Economic Development, p. I27.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
I 64 Joseph L. Love
'involution' - the development of the satellite in periods of crisis in the

metropolis - is directly analogous to ECLA's historical analysis of
'inward-directed' growth.
From Pablo Gonzalez Casanova, the Mexican political scientist, Frank
borrowed the thesis of 'internal colonialism', whereby industrial and
political Centres within the satellite exploit their dependent regions
through fiscal and exchange policies, and by draining off capital and
talent. 61
Frank linked transnational exploitation with internal colonialism,
hypothesising a concatenation of metropolis-satellite relations from Wall
Street down to the smallest Latin American village, in which only the end
points of the continuum would not stand in both relationships.62 Frank,
perhaps more than any other writer except the Mexican sociologist
Rodolfo Stavenhagen,63 hammered away at the theme that dualism did
not exist in Latin America: all areas were linked by an unequal exchange
of goods and services, consequent on underdeveloped capitalism. Thus
Frank attacked the traditional Communist positions on 'feudal residues'
and non-Marxist dualism as well.64 Frank's polemic paralleled the
61 Andre Gunder Frank, On Capitalist Underdevelopment (Bombay, 1975),

Prebisch), z6 (on Bagui), 68 (terms of trade), 73 (Gonzalez Casanova); Frank, Capitalism
and Underdevelopment in Latin America (New York, I967), pp. xi, xviii (on Baran), xii
(association with ECLA). On Frank's professional development, see David Booth,
'Andre Gunder Frank: an Introduction and Appreciation', in Ivar Oxaal, Tony
Bennett and David Booth (eds.), Beyond the Sociology of Development (London,
5o-8 5.
62 This was an independent 'rediscovery', I believe, of a model the Rumanian trade
theorist Mihail Manoilescu had developed a generation earlier. See Manoilescu, 'Le
triangle economique et social des pays agricoles: La ville, le village, l'etranger',
Internationale Agrarrundschau, no. 6 (June I940), pp. I6-26. Manoilescu's scheme, like
Frank's, specifically linked the transnational extraction of surplus with internal
colonialism, and was not limited to an analysis of the latter phenomenon. On the
history of the concept of internal colonialism, including the models developed by Hans
W. Singer and Celso Furtado in a structuralist discourse during the I950s, see Joseph
L. Love, 'Modeling Internal Colonialism: History and Prospect', World Development,
vol. 17, no. 6 (June I989), pp. 905-22.
63 Stavenhagen, 'Seven Fallacies', pp. I5-I8.
64 The non-Marxist dualism Frank and Stavenhagen attacked was less that of EC
that of J. H. Boeke, who had first developed the concept of virtually unrelated
modern and peasant sectors in the Indonesian economy. Although ECLA had
hypothesised the existence of a dual economy in the 1949 Survey, in its model the
surplus labour force passed from subsistence to modern sectors, assuming a highly
wage-elastic labour supply. (ECLA preferred the term 'heterogeneous' to 'dualist.') In
I963 the ECLA sociologist Jose Medina Echavarria noted that the differences between
traditional and modern sectors of Latin American societies were due to 'internal
development processes' more than to the imposition of a European society on an
indigenous one, and that interpenetration was a basic aspect of their existence. Medina
EchavarrIa, 'A Sociologist's View', in Medina and Benjamin Higgins, Social Aspects of
Economic Development in Latin America (Paris, 1963), p. 29.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
The Origins of Dependengy Analysis i 6 5
simultaneous but less didactic and less explicit efforts of Furtado and
Sunkel to find causal links between development and underdevelopment.
Frank the synthesiser was also an effective wordsmith, and he termed
the plight of Latin America, and by extension, that of the Third World,
the 'development of underdevelopment'. For Frank, Latin America had
been 'underdeveloping' for more than four centuries, a process which he
divided into four phases, each defined by the principal form of monopoly
exercised by the metropolis: commercial monopoly (in the age of
mercantilism); industrial monopoly (during the age of classical liberalism);
monopoly of capital goods (I900-50); and monopoly of technological
innovation (1950 to the present).65
It is notable that the stages developed by Frank, Theotonio dos Santos,
and Johan Galtung (who had extensive contacts in Santiago)66 were stages
in the development of the entire capitalist system, not stages a la Rostow,
in which the underdeveloped countries would repeat the unilinear
trajectory of the advanced capitalist nations.67 A similar emphasis on the
whole system was also implicit in the work of Furtado and Cardoso and
Faletto.
For Frank, exit from the system in a revolutionary struggle, following
the Cuban example, was the path to development. Only in that manner
could 'involution', a partial and temporary exit, be transformed into
continuous development. There was an urgency in Frank's voluntarist
view that the continued underdevelopment inherent in capitalism would
make the breakthrough all the more difficult.68 He argued that the gap
between metropolis - the United States - and satellite - Chile, a case
study - was widening 'in power, wealth, and income'; and that the
'relative and absolute' income of the poorest classes in Chile was
decreasing.69 Frank clearly disagreed with Marx and Prebisch, who,
65 Frank, Capitalism, p. 2 II.

66 Galtung was in Santiago in 1962-3, and maintained contact with Santiago-based
personnel later. His model of imperialism cites the dependency literature, and Sunkel
criticised the essay in manuscript. See Johan Galtung, 'A Structural Theory of
Imperialism', Journal of Peace Research, no. 2 (197 1), pp. 8 I- 1 17.
67 Theot6nio dos Santos, 'The Structure of Dependence', American Economic Review:
Papers and Procedings, vol. 6o, no. 2 (May, 1970), pp. 23I-26, esp. p. 232; Galtung (in
preceding note).
68 Revolution would be more costly in human terms because of the strengthening
capitalist institutions in the development/underdevelopment process. Frank, On
Capitalist Underdevelopment, p. I I o.
69 Frank, Capitalism, pp. 47-8. In 1974 Fernando Henrique Cardoso argued that Frank
and Rui Mauro Marini, two 'left' members of the dependency school, were repeating
an error of the narodniki in denying the possibility of capitalist development in the
Periphery. Cardoso, Autoritarismo e democratiZafeo (Rio de Janeiro, 1975 [orig., '9741),
pp. 27-30. One might add that Frank's sense of urgency seems analogous to the hic et
nunc attitude of the People's Will, which in the X87os and early 188os held that

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
i 66 Joseph L. Love
holding that rising productivity was the essence of capitalist development,

did not believe that the development of the Centre had to be primarily at
the expense of the Periphery.
In fact, Frank's use of Marxist analysis was soon challenged. Ernesto
Laclau charged him with 'circulationism' - putting primary emphasis on
relations of exchange rather than relations of production, whereas, for
Marx, capitalism was defined by a (free) wage labour market :70 only when
labour power had become a commodity could relative surplus value be
maximised, and this process occurred in the Centre long before it appeared
in the Periphery.
If Frank's sources are evident because of his own acknowledgments,
Cardoso's are less so. This is important, because, as noted at the outset,
Marxism has so frequently been cited as the main tradition out of which
capitalism had to be smashed in Russia before it destroyed the primitive communism

of the peasant mir. (Of course, Frank argued that Latin America had never known any
other mode of production than capitalism).
70 Ernesto Laclau, 'Feudalismo y capitalismo en America Latina' [orig., I97I], in Carlos
Sempat Assadourian (ed.), Modos de produccidn en Ame'rica Latina (Mexico, I973), pp
28-37, 43.
The degree to which the emerging modes-of-production debate influenced early

dependency analysis depends in part on how the 'modes' debate is characterised. By
definition the dependentista attack on economic dualism was a challenge to Communist
orthodoxy on the transition of feudalism to capitalism in Latin America, and in I966
Frank had engaged in an explicit 'modes' debate with Rodolfo Puiggros, a former
leader of the Argentine Communist Party. The debate occurred in the Mexico City
newspaper El Dz'a, where Stavenhagen had denied the existence of dualism in his
'Seven Fallacies' in the previous year.
Puiggr6s argued that Latin America's past had originated in a feudal regime, and
Frank held that it had been capitalist from the beginning; Frank, however, nuanced his
view by conceding that capitalism could accommodate a variety of subordinate modes,
thereby anticipating the 'articulation' thesis of writers influenced by Louis Althusser,
Etienne Balibar, and Pierre-Philippe Rey in the early I970s. See Puiggr6s and Frank,
'Polemica sobre los modos de producci6n en Iberoam6rica', A UN [Cuadernos
Universitarios, serie hist6rica, ficha 2], Buenos Aires, n.d. [orig. in El Dz'a (Mexico
City, I966)], especially p. 4I (Frank's views on articulation).
Although Balibar's essay on the modes issue ('Elements for a Theory of Transition')
was published in Althusser's Lire le Capital in I96 5, and a Spanish edition of Althusser's
Pour Marx had been published in Mexico in I967, structuralist Marxism would have
its greatest impact in Latin America in the early I970S, after a readership had been
prepared by Marta Harnecker's Althusserian primer Los conceptos fundamentales del
Marxismo (I969). In short, the appearance of Althusserian Marxism in the region
overlapped the formation of dependency analysis in the mid- and late Ig6os, but
debates between the two schools and mutual adjustments were largely a phenomenon
of the I970S. Sempat's Modos (I973), in which Laclau's 'Feudalismo' of I97I appeared,
was the first major statement of the (highly heterogenous) 'modes' group. The French
impact was broader in fact than that of the Althusserians, however, because as early as
I969, Laclau's thought was influenced by the debate in the Marxist journal La Pens6e
(involving Maurice Godelier, Catherine Coquery-Vidrovich, and Jean Suret-Canale) in
I963-7. (Interview with Ernesto Laclau, Urbana, Ill., I2 Nov. I984).

147.96.221.252 on Tue, 08 Dec 202u, 01 Jan 1976 12:34:56 UTC
The Origins of Dependengj Analysis I67
dependency arose. The priority of Marxism has been reinforced by the
English edition of Dependengy and Development (I979) by Cardoso

Faletto, in which preface, postscripts, and parts of the text show a strong
Marxist orientation. By contrast, the first Spanish edition (I969) is far less
obviously influenced by Marxism, and the unpublished original draft
(i965) is recognisably an ECLA product.7" In this first version the
challenge the Parsonian categories of modernisation theory, and are
pessimistic about the reformism of local bourgeoisies, but from an eclectic
perspective. No Marxist studies were cited in the draft, and Marxist
categories are almost completely lacking. The theme receiving most
attention in the I965 version was the inadequacy of the bourgeois-directed
project of development, partly resulting from increasing market domi-
nation by multinational corporations.72
The issue is clouded, however, by elements in Cardoso's I964 study on
Brazilian entrepreneurs. That work adumbrates one of his most important
contributions to the dependency tradition - namely, his denial of the
adequacy of the modernisation paradigm, although in the limited context
of the role of entrepreneurs.73 In that work, Cardoso, though eclectic in
methodology, cast his major conclusions within a Marxist paradigm.74
Thus the sources of Cardoso's contribution were various, and a safe
conclusion would seem to be that he could make his statement in either
a structuralist or a Marxist idiom. Yet it was initially made in the former,
as dependency emerged in Santiago.
As the sixties yielded to the seventies, many writers on dependency
adopted an exclusively Marxist perspective, and dependency analysis for
this group matured as a 'region' of Marxism: It offered a perspective on
imperialism which the classical Marxist theorists of the subject had
ignored, namely, the view from the Periphery. A respectable Marxist
pedigree was apparently required to validate the dependency perspective
after its radicalisation, and a fortiori after it was challenged by those
claiming to represent an orthodox Marxist tradition.
71 See the review of the changes in the editions from I969 to I979 by Robert A.
Packenham, 'Plus sa change .. .: The English edition of Cardoso and Faletto's
Dependenciay desarrollo en Am6rica Latina', Latin American Research Review
(i982), pp. I3I-5I.
72 See Cardoso and Faletto, Dependencia y desarrollo en AmJrica Latina: Ensayo de
interpretacidn sociologica (Mexico, I969); and the original draft, 'Estancamiento y
desarrollo econ6mico en America Latina: Condiciones sociales y poifticas
(Consideraciones para un programa de estudio)', mimeo., late I965. (Adol
kindly supplied this document and other drafts of Dependenciay desarrollo from the
ILPES files in ECLA's Santiago headquarters).
7 Cardoso denied that the roles played by Europe's historical bourgeoisies in economic
development could be replicated by Brazilian entrepreneurs in the ig6os. Cardoso,
Empresdrio, pp. 46, 49, I94. 7 Ibid., pp. I92-98.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC
i68 Joseph L. Love
Yet three of the four defining elements of dependency analysis - the

historical Centre-Periphery perspective, unequal exchange, and the denial
of dualism - derived more directly from Latin American structuralism
than from Marxist theories of imperialism. The first two theses flowed
directly from structuralism.75 The third was compatible with both
Marxism and neoclassical economics,76 and when adopted by Marxists, set
them against Marxist orthodoxy as defined by national Communist
parties. However, the fourth thesis - the alleged nonviability of the
national bourgeoisies - was incompatible with Prebisch's structuralism,
even if some who perceived it did not cast their findings in a Marxist
framework.
Arghiri Emmanuel's Marxist theory of unequal exchange was published in I969, but
owed nothing to Latin American Marxists - who worked within the tradition of
historical materialism rather than formal Marxist economics. But it did owe something
to Prebisch and Hans Singer. Meanwhile, by I969 dependency analysis had been born.
See Emmanuel, Unequal Exchange: A Study of the Imperialism of Trade (New York, I 972
[French orig. I969]). See especially pp. 80-7 on the Prebisch-Singer thesis on
deteriorating terms of trade.
76 For a neoclassical approach denying the existence of dualism in a Peruvian case study,
see Adolfo Figueroa, Capitalist Development and Peasant Economy in Peru (Cambridge,
I984), especially p. I20.

147.96.221.252 on Tue, 08 Dec 2020 18:24:39 UTC

Love

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Love

Uploaded by

Copyright:

Available Formats

The Origins of Dependency Analysis

Author(s): Joseph L. Love

Stable URL: https://www.jstor.org/stable/157170

This content downloaded from

It is widely recognised that dependency analysis developed out of two

1 Steve J. Stern, 'Feudalism, Capitalism, and the World-System in the Perspective of

Joseph Love is Professor of History, University of Illinois, Urbana-Champaign.

J. Lat. Amer. Stud. 22, 143-168 Printed in Great Britain 143

This content downloaded from

Jose Gabriel Palma, whose lucid discussion of dependency in 1978 was

Fernando Henrique Cardoso, 'The Consumption of Dependency Theory in the United

This content downloaded from

'colonial' countries in increasing degrees of subjugation to imperialist

Almost a quarter-century after the formulation of dependency analysis,

7 Ricardo Paredes, 'Informe de la delegaci6n latinoamericana', Internacional Comunista,

This content downloaded from

Another structuralist thesis, originating in the ig5os, addressed the

This content downloaded from

Latin American industrialisation was based on the technology appropriate

This content downloaded from

governments of developed countries to liberalise their trade policies, thus

diminishing 'the import content of supplies for the home market'. 5

13 Norman Girvan, 'The Development of Dependency Economics in the Caribbean an

This content downloaded from

1 ECLA, 'Preliminary Study', pp. I 28, 1 50, 1 5 1.

This content downloaded from

Bulletin - one on the Brazilian experience in particular, and the other on

21 Maria da Conceisao Tavares, 'The Growth and Decline of Import Substitution in

This content downloaded from

become critical for the future of industrialisation. On the positive side,

This content downloaded from

Furtado explained that ISI was fundamentally different from European

region. Furtado's study, however, derived principally from h

This content downloaded from

This content downloaded from

Thus the perceived failure of ISI as a historical process - and perhaps

This content downloaded from

Sunkel's view, the long-term prospects for Latin America were

38 Osvaldo Sunkel, 'The Structural Background of Development Problems in Latin

This content downloaded from

This content downloaded from

political difficulties of implementing ECLA-inspired reforms as envisioned

This content downloaded from

explicitly historical view of development, noted above, but Furtado now

This content downloaded from

Another Brazilian, Fernando Henrique Cardoso, during his association

This content downloaded from

the 'national' or 'industrial' variety.) Cardoso and Faletto analysed the

This content downloaded from

52 Jose Carlos Mariategui, 'Point de vue antiimperialiste' [i929], in Michael Lowy, Le

This content downloaded from

was in effect part of Brazil's 'developmentalist' and populist coalition for

This content downloaded from

the Portuguese colony was a mercantile enterprise in which there existed

This content downloaded from

'involution' - the development of the satellite in periods of crisis in the

61 Andre Gunder Frank, On Capitalist Underdevelopment (Bombay, 1975),

This content downloaded from

65 Frank, Capitalism, p. 2 II.

This content downloaded from

holding that rising productivity was the essence of capitalist development,

capitalism had to be smashed in Russia before it destroyed the primitive communism

The degree to which the emerging modes-of-production debate influenced early

This content downloaded from