Latin American Sociology - Dependency and Its Legacy

Latin America produced the only body of social theory from outside Europe and North America that reshaped the discipline globally.

It did so by explaining why its own societies had not developed as the reigning theory said they should — and one of its principal authors later became President of Brazil and implemented policies his own theory had been read as forbidding.

A puzzle in a set of trade statistics.

Raúl Prebisch is an Argentine economist. In the 1940s he is running the UN Economic Commission for Latin America in Santiago, and he is looking at a long run of trade data.

The reigning theory of international trade — Ricardo's comparative advantage, taught everywhere — says that if Argentina is good at producing beef and wheat and Britain is good at producing manufactures, both gain by specialising and trading. Everyone should get richer, and the gains should be shared.

Prebisch looks at the numbers over sixty years and finds something the theory does not predict.

The prices of primary commodities — the things Latin America exports — decline relative to the prices of manufactured goods. Not in a single year, but as a long-run trend.

Which means: the same quantity of coffee, copper or beef buys progressively less machinery over time. A country specialising in primary exports must run faster to stay in the same place, and the gains from trade are systematically distributed towards the industrial economies.

Hans Singer, working independently at the UN in New York, reaches a similar finding in the same period. The Prebisch–Singer hypothesis enters the literature.

And the explanation Prebisch offers is sociological rather than technical.

In the industrial centre, productivity gains are captured by workers through organised labour and by firms through market power, so they show up as higher wages and profits rather than as lower prices.

In the periphery, labour is unorganised, there is surplus labour from the countryside, and producers compete, so productivity gains show up as lower prices.

Same technological progress, opposite price effects — because of the institutional structure on each side.

And therefore: the international division of labour is not a neutral allocation of tasks. It is a structure that transfers the gains from productivity in one direction, and it will do so as long as it operates.

The theoretical problem Latin American social science faced in the 1950s was the same one this course has met twice already, and its answer was different.

Modernisation theory (5.1.3) was the reigning account of development: societies progress through stages from traditional to modern; the obstacles are internal — traditional values, extended kinship, low achievement motivation, insufficient entrepreneurship, dual economies; and the remedy is the diffusion of capital, technology and modern attitudes from the developed world.

Its clearest statement is W. W. Rostow's The Stages of Economic Growth (1960), subtitled "A Non-Communist Manifesto" — five stages from traditional society through the preconditions for take-off, take-off, the drive to maturity, and the age of high mass-consumption.

The framework is functionalist in structure (5.1.1), stadial in the way 3.3.1 warned about, and it locates the cause of underdevelopment inside the underdeveloped country.

Latin American economists and sociologists looking at their own countries found this did not fit.

Their societies were not "traditional." They had been integrated into the world economy for four centuries — since silver from Potosí financed European commerce in the sixteenth century. They had cities, universities, railways, industry and republics. They had been "modernising" for a very long time and the gap was widening.

And the causal direction was wrong. Modernisation theory treated underdevelopment as an original condition to be escaped. The Latin American argument was that it was a produced outcome — that development and underdevelopment were not two stages but two aspects of one process.

André Gunder Frank's formulation is the sharpest: "the development of underdevelopment."

Dependency theory

The core proposition: underdevelopment is not a stage prior to development. It is a condition produced by a country's mode of integration into the world economy.

The structure: a centre (or core) of industrial economies and a periphery of primary exporters, connected by trade, investment and finance in relations that transfer surplus from the second to the first.

The mechanisms:

Declining terms of trade — Prebisch–Singer, above.

Profit repatriation — foreign investment produces returns that leave, frequently exceeding new inflows.

Technological dependence — capital goods and technology must be imported, on terms set elsewhere, and licensing keeps the rents at the source.

Debt — borrowing to finance imports, with servicing costs that constrain policy.

And internal alliance. This is the sociologically important one: the arrangement is sustained by an alliance between external interests and a domestic elite whose position depends on the export sector and on imported consumption. The periphery is not simply acted upon; it has a class structure that reproduces its position.

There is no single dependency theory, and the differences matter enormously — including politically.

The stagnationist wing. André Gunder Frank (a German-American economist working in Latin America) and, in a different register, Ruy Mauro Marini argued that dependency makes autonomous development impossible; that the periphery's "national bourgeoisie" is not a progressive force; and that the only exit is revolutionary rupture from the world system. Frank's slogan: "underdevelopment is not original or traditional... it is generated by the same historical process which also generated economic development."

Marini's specific contribution — super-exploitation — holds that peripheral capitalism compensates for unequal exchange by paying labour below the cost of its reproduction, which is an important and undertaught mechanism.

The dependent development wing. Fernando Henrique Cardoso and Enzo Faletto's Dependency and Development in Latin America (1969, English 1979) is the most sophisticated statement and rejects the stagnationist conclusion.

Their argument: dependency is real, and it does not preclude development. Development can occur within dependency — "associated dependent development" — producing industrialisation, growth and a domestic bourgeoisie, while remaining structurally subordinate and generating extreme internal inequality.

And their method is explicitly historical and sociological rather than economic. They insist there is no general law of dependency, that each country's trajectory depends on its specific class structure, state and historical integration, and that the analysis must be of concrete situations rather than of a model.

Cardoso complained repeatedly that his work was read in the United States as a version of Frank's. He was right, and the flattening has persisted.

The world-systems wing. Immanuel Wallerstein generalised the framework into a theory of a single capitalist world-economy from the sixteenth century, with core, semi-periphery and periphery, and cycles of hegemony. He acknowledged the Latin American debt — which much of the field did not.

And the African version: Samir Amin's unequal exchange and delinking; Walter Rodney's How Europe Underdeveloped Africa (6.4.1).

Dependency theory made predictions, and the subsequent fifty years tested them. The results are genuinely mixed and worth stating precisely.

Where the theory did badly:

East Asia. South Korea and Taiwan were peripheral, primary-exporting, aid-dependent and authoritarian in 1960. By 1995 they were industrial economies with high incomes. The stagnationist prediction that development within the world system is impossible was falsified in the clearest possible way.

And the explanation of how they did it is instructive: not delinking but export-oriented industrialisation — precisely the strategy dependency theory regarded as reproducing subordination — combined with land reform, state-directed credit, protection of infant industries, heavy education investment, and geopolitical circumstances (Cold War aid, US market access, and land reforms imposed or accelerated by American occupation and threat) that were not generally available.

China. Integration into global production networks from 1978 produced the largest reduction in absolute poverty in recorded history. Whatever else this is, it is not what the framework predicted.

And the import-substituting industrialisation that dependency analysis supported ran into severe difficulties across Latin America — inefficient protected sectors, chronic balance of payments problems, inflation, and the debt crisis of the 1980s.

Where the theory did well:

Commodity dependence remains a real trap. The countries that remained primary exporters have, on the whole, done badly, and the "resource curse" literature — Sachs and Warner's cross-country findings, and the subsequent institutional refinements — documents a real association between primary export concentration and poor growth.

The terms of trade hypothesis has held up better than it was given credit for. Later econometric work using long series, including Ocampo and Parra's, found real declines in commodity terms of trade over the twentieth century, though with substantial cycles and heterogeneity across commodities.

And the internal alliance argument is well supported. The observation that peripheral development is shaped by an alliance between external capital and a domestic elite, and that this produces growth alongside extreme inequality, describes Latin America's actual trajectory closely. Latin America has the world's highest regional income inequality, and it grew during periods of growth.

And then the author became president.

Cardoso was Brazil's finance minister and then President from 1995 to 2002. He implemented the Plano Real, which ended hyperinflation; he privatised state enterprises; he opened the economy; and he negotiated with the IMF.

He was accused of betraying his own theory. His response — consistent, and consistent with what Dependency and Development actually said — was that the theory had never prescribed autarky, that it was an analysis of constraints rather than a programme, and that governing requires acting within constraints one has correctly identified.

Whether that is a vindication or a capitulation is genuinely arguable. What is not arguable is that it is a rare event: a social theorist obtaining the power to act on his own analysis of a country's structural position, and doing so for eight years. The record is there to be assessed.

Go deeper

Before and beyond dependency

Latin American sociology is older and broader than dependency theory, and reducing it to one export is exactly the error this Part is about.

The positivist period (3.3.1, 4.6.3). Comte's system had more institutional purchase in Latin America than anywhere in Europe — "Order and Progress" on the Brazilian flag, the Mexican científicos, sociology chairs in Colombia (1882) and Argentina (1898). And it was substantially used to justify racial hierarchy: blanqueamiento (whitening) immigration policies, Lombrosian criminology, and the treatment of indigenous and Black populations as obstacles to national progress. This is not a proud inheritance and the tradition knows it.

Gilberto Freyre's Casa-Grande e Senzala (1933) broke with race science and argued that Brazilian society was constituted by the interpenetration of African, indigenous and Portuguese cultures — and underwrote the myth of racial democracy which subsequent research demolished (4.6.3).

Florestan Fernandes (1920–1995) is the founder of modern Brazilian sociology and the demolition's author: rigorous empirical studies of race relations in São Paulo, commissioned partly by UNESCO in the 1950s, which documented systematic discrimination and destroyed the racial democracy thesis. His work on Brazilian capitalism as structurally dependent and on the "bourgeois revolution" that never happened is substantial and undertranslated.

Pablo González Casanova's internal colonialism (1965) — the argument that the centre/periphery relation is reproduced inside countries, between metropolitan regions and internal peripheries, and specifically in the treatment of indigenous populations. This is one of the most portable concepts in Part 6 and it applies directly to Adivasi India (6.2.5) and to regional inequality everywhere.

Paulo Freire's Pedagogy of the Oppressed (1968) — conscientização (critical consciousness), the critique of "banking education" in which knowledge is deposited in passive recipients, and dialogical method. One of the most influential books in education worldwide, and a genuine piece of social theory: it is Gramsci's organic intellectual (5.5.1) turned into a pedagogy.

Orlando Fals Borda's participatory action research — the argument that the researched should be co-researchers, that knowledge production should serve those studied, and that the separation of researcher and subject is itself a political arrangement. A methodological tradition (7.4) originating in Colombia.

Liberation theology — Gustavo Gutiérrez, the base ecclesial communities, and the "preferential option for the poor" — which produced a substantial sociology of religion and of social movements, and which was suppressed institutionally in the 1980s and 90s.

And the coloniality group: Aníbal Quijano's coloniality of power (6.7), Walter Mignolo, María Lugones — the argument that colonialism ended and coloniality did not, and that racial classification is the organising principle of the global division of labour. This is Part 6's most active contemporary tradition and it gets its own lesson.

The honest ledger on dependency

What was right:

That the international economy has a structure, that position in it matters enormously, and that the position is not simply a consequence of internal characteristics.

That underdevelopment can be produced rather than being an original condition.

That commodity dependence is a genuine trap with documented consequences.

That peripheral development is shaped by an alliance between external and domestic interests, and generates growth alongside extreme inequality.

And that development is political, requiring a state with capacity and autonomy — which the East Asian cases confirm rather than refute.

What was wrong:

The stagnationist claim that development within the system is impossible. Falsified.

The tendency to treat the periphery as passive and internal factors as derivative. The East Asian states did things; the difference between Korea and the Philippines is not principally a difference in external constraint.

Import-substituting industrialisation as a general prescription.

And a persistent methodological weakness: the theory frequently explained outcomes after the fact and specified few conditions under which it would be wrong.

What replaced it:

World-systems analysis, which generalised it.

The developmental state literature — Chalmers Johnson, Alice Amsden, Robert Wade, Peter Evans — which took the political insight and specified the institutional conditions: state capacity, embedded autonomy, disciplined subsidy, export discipline.

Global value chain analysis, which asks where in a production chain value is captured and by whom — dependency's question, with better data and a firmer empirical base.

And the coloniality tradition, which took the framework in a cultural and epistemological direction (6.7).

"Dependency theory said development is impossible." The stagnationist wing did. Cardoso and Faletto explicitly argued for "associated dependent development" and objected to being read as Frank.

"East Asia refuted dependency theory." It refuted the stagnationist claim decisively. The insight that development is political and requires state capacity was confirmed, and the developmental state literature is dependency's successor as much as its refutation.

"Prebisch–Singer has been disproved." Later long-series econometric work has found real declines with substantial cycles and heterogeneity. It is contested, not disproved.

"Latin American sociology is dependency theory." It includes Fernandes on race, González Casanova on internal colonialism, Freire on pedagogy, Fals Borda on method, liberation theology and the coloniality group.

"Cardoso betrayed his theory." Dependency and Development analysed constraints and prescribed no programme; whether his presidency vindicates or capitulates is arguable, and it is a rare case of a theorist testing his own analysis in office.

"Freyre showed Brazil had no racial problem." He broke with race science and underwrote racial democracy, which Fernandes's UNESCO-era research and everything since has refuted.

Why it matters

It is the one theory the periphery exported to the centre. Dependency reshaped development studies, sociology and political economy globally, and it did so from Santiago, Rio and Mexico City. After Part 6's four previous lessons on exclusion, this is the case where it did not happen — and understanding why is instructive: the theory addressed a question the centre cared about, in a vocabulary the centre used, at a moment when its own theory was failing.

It supplies internal colonialism, which travels to any country with a metropolitan core and an internal periphery, and which describes Adivasi India, the Andean highlands, the American South and the north of England.

It produced two methodological traditions that are now standard: Freire's dialogical pedagogy and Fals Borda's participatory action research, both of which start from the claim that the separation of researcher from researched is a political arrangement rather than a scientific necessity (7.4).

And it is the clearest case in this course of a theory being graded honestly. The stagnationist prediction failed; the structural insight held; the political insight was confirmed by the cases that falsified the prediction; and the tradition's own author tested it in office. That is what a research programme looks like when it is taken seriously by the people who built it.

Prebisch and Singer: the long-run decline in commodity terms of trade — explained sociologically, by productivity gains being captured as wages and profits in an organised centre and as lower prices in an unorganised periphery. Same technology, opposite price effects, because of institutional structure.

Against modernisation theory (Rostow's five stages), which located underdevelopment inside the underdeveloped country: Frank's "development of underdevelopment" — development and underdevelopment as two aspects of one process.

Mechanisms: declining terms of trade, profit repatriation, technological dependence, debt, and — sociologically the most important — an internal alliance between external interests and a domestic elite.

Three wings, and the differences matter: stagnationist (Frank; Marini's super-exploitation), holding development within the system impossible; dependent development (Cardoso and Faletto), holding that development occurs within dependency, producing industrialisation alongside extreme inequality, and insisting on concrete situations rather than general laws; and world-systems (Wallerstein).

The record: East Asia falsified the stagnationist claim — through export orientation, land reform, state-directed credit and geopolitical circumstance — and confirmed the political insight that development requires state capacity. Commodity dependence remains a documented trap; terms of trade decline has held up with cycles; and the internal alliance argument describes Latin America's growth-with-extreme-inequality trajectory closely.

And Cardoso became President of Brazil (1995–2002) and governed within the constraints his own theory had identified — a rare and assessable case.

Beyond dependency: the positivist inheritance and its racial uses; Fernandes demolishing Freyre's racial democracy; González Casanova's internal colonialism; Freire's conscientização; Fals Borda's participatory action research; liberation theology; and the coloniality group.

The successors: world-systems analysis, the developmental state literature (Johnson, Amsden, Wade, Evans), global value chain analysis — dependency's question with better data — and coloniality.

Prebisch–Singer hypothesis — the long-run decline in the terms of trade of primary commodities against manufactures.

Centre and periphery — the structural positions in the international division of labour.

Development of underdevelopment — Frank's claim that the two are produced by one process.

Associated dependent development — Cardoso and Faletto's finding that development occurs within dependency, with extreme inequality.

Super-exploitation — Marini's mechanism: peripheral labour paid below its cost of reproduction.

Internal colonialism — González Casanova's centre/periphery relation reproduced within a country.

Import-substituting industrialisation — the protectionist development strategy dependency analysis supported.

Developmental state — the state capacity and disciplined industrial policy identified in the East Asian cases.

Conscientização — Freire's critical consciousness, developed through dialogical rather than "banking" education.

Participatory action research — Fals Borda's method treating the researched as co-researchers.

Resource curse — the documented association between primary export concentration and poor growth.

One — check the terms of trade. Find a long series for a commodity your country exports or imports against manufactured goods prices. The trend and the cycles are both visible.

Two — apply internal colonialism. Take a country you know and identify its metropolitan core and internal periphery. Ask where the resources flow and where the decisions are made.

Three — test the internal alliance argument. For any commodity-exporting country, ask whose position depends on the export sector, and what they would lose from diversification.

Four — grade the theory yourself. Take East Asia and one Latin American country in 1960 and now. Ask what dependency theory predicted for each and what happened. Being able to state where a theory failed and where it held is the exercise.

Five — read Freire's second chapter. The critique of banking education is about fifteen pages, has been applied in every continent, and is a piece of social theory disguised as an education text.