Capitalism, Industrialisation and development in Latin America: The dependency paradigm revisited

Capital & Class, Spring 1998 by Ayres, Ron, Clark, David

1 Introduction

THE ROOTS OF DEPENDENCY THEORY go back to Marx and Lenin but the notion that there is a fundamental contradiction between capitalist development in the centre and industrialisation and development in the periphery was first expounded by Paul Baran (1957) and later elaborated by the Dependency School (see Palma, 1978,1994; Larrain, 1989; Brewer, 1990; Clark, 1996).1

The most well known and extreme version of a `dependency theory' was put forward by Andre Gunder Frank in the late 1960s. Frank ( 1967) attempted to show that capitalist development in the periphery, and Latin America (LAC) in particular, is impossible. He contends that the incorporation of LAC into the world capitalist system (dating back to the colonial era) set in motion a metropolis-satellite relationship, or power structure, that permits the former to develop by expropriating economic surplus from the latter.2 This process inevitably leads to the `development of underdevelopment' in the periphery. Furthermore, the only way the LACs and other dependent economies can escape this fate is to break with the existing capitalist system.3 In short, Frank advocates nothing less than (socialist) revolution for development.

While Frank's thesis suffers from a number of theoretical weaknesses4 some socialists have taken issue with its empirical validity. The most notable is undoubtedly the late Bill Warren ( 1973; 1980). Writing in the late 1970s, Warren argued that at the statistical level growth, together with industrialisation, has taken place in LAC, along with other Third World countries. In support of this contention, Warren (1980) presents three sets of data. First, he shows that the growth of manufacturing was faster in the Third World than in the First World between the 1950s and 1970s. Second, he points out that at the same time the share of manufacturing, in terms of GDP and in terms of employment, increased for many Less Developed Countries (LDCs). Finally, Warren presents evidence to show that much of this industrial growth was not orientated solely towards wealthy elites (as some critics have claimed). In fact, highly diversified consumer durable markets not only reach `well into the lower income groups', but are continuing to expand.

Warren's empirical observations are consistent with Classical Marxism but they are also compatible with more orthodox ways of conceptualising economic development and structural change. Kuznets (1973), Chenery et al (1986), Syrquin (1988) and Clark (forthcoming) have all provided useful summaries of the `stylised facts' of structural development. Briefly, structural transformation typically takes the form of:

1. Economic growth, i.e., rising GNP and GNP per capita.

2. High rates of physical and human capital accumulation.

3. Industrialisation.

4. The transformation of manufacturing characterised by a shift from light to heavy industry.

5. The transformation of international trade involving a rise in imports and exports, increasing specialisation, and a move towards the export of industrial products.

6. High rates of technology transfer, followed by R.& D. and expanded indigenous technological capability and a move to high technology production.

7. Rapid and sustained increases in (factor) productivity.

It is worth emphasising that the stylised facts referred to above are derived from, and have been confirmed by, a wide range of time series and cross section studies of countries at different stages of development (levels of per capita income). These studies add momentum to Warren's case.

On the basis of this evidence it is reasonable to conclude with Warren that autonomous capitalist development is taking place in the Third World.5 The idea of stagnation and the notion of impossibility are overthrown. Warren (1973: 4) argues the previous structures of dependency are weakening and that the obstacles to Third World development are internal-not external.

These findings and conclusions, however, were controversial in Marxist circles. Initially, Warren's critics took issue with his methodology and statistical techniques. However by the late 1970s it became increasingly difficult to dismiss his statistics on these grounds (cf. Corbridge, 1986; Clark, 1995). This forced his critics to change tactics. Some Neo-Marxist writers attempted to portray Warren as a `Neo Liberal', others questioned the structure and nature of growth and industrialisation in the Third World. Firstly, it was suggested that any industrial growth that might have taken place in the Third World only incorporates a very primitive, simplistic and labour intensive technology (cf. Corbridge, 1986). Secondly, it was argued that industrialisation should not be dependent on external social and economic forces (Sutcliffe, 1972). Thirdly, the wisdom of equating growth and industrialisation with development was challenged (e.g. Emmanuel, 1974). In a classic paper Seers (1972) argues that economic growth is not sufficient for development unless it is accompanied by reductions in (absolute) poverty, income inequality and unemployment.b