Race, Class and the Changing Division of Labour Under Apartheid

Owen Crankshaw

• Publisher: Routledge
• Year: 2002
• Print ISBN:
  9780415146135
• eBook ISBN:
  9781134757992
• Topic: Social Sciences
• Subtopic: Sociology
• Index: Social Sciences

As the only comprehensive empirical analysis of the changing racial and occupational structure of the urban workforce in South Africa under apartheid, this study will make an invaluable contribution to our understanding of the complex inter-relations of past and present racial inequality and economic development in South Africa.

1
RACE, CLASS AND THE DIVISION OF LABOUR

INTRODUCTION

This book is a study of the changing relationship between racial and class divisions in the urban population in South Africa during the apartheid period. As an analysis of the relationship between racial and class divisions, this study does not entail a new approach to racial inequality in South Africa. A substantial body of research, conducted over the last three decades, has enriched our understanding of how the class dynamics of South Africa’s expanding capitalist economy have complemented and contradicted the policies and practices of the apartheid state. However, with the exception of some earlier studies by Wolpe, and by Simkins and Hindson, there has been no systematic study of the changing relationship between racial divisions and the class structure of the South African population.¹ Where scholars have applied theories of class to racial divisions in South Africa, they have done so in order to understand the political behaviour of the state and of particular social groups. For example, there is a large literature which examines the class basis of changes in the relationships between the apartheid state, organised business and trade unions.² Similarly, a number of scholars have been concerned to understand the class basis for the political behaviour of the white working class and the African middle class.³ So, although these studies have provided us with an understanding of how the changing class interests of organised business and organised labour have influenced their political relationship to apartheid policies and practices, we still lack a systematic understanding of the racial and class structure of the South African population and how this was shaped by capitalist economic growth and apartheid policies. The aim of this study is to fill this gap in our knowledge by documenting and explaining the changing patterns of racial and class inequality among the urban workforce during the apartheid period.

SOUTH AFRICAN LABOUR STATISTICS, CLASS STRUCTURE AND THE STUDY OF RACIAL INEQUALITY

The question of the relationship between racial inequality and class structure in South Africa has been subsumed in a broader debate over the effect of capitalist economic growth on racial inequality and, more generally, apartheid. This debate, which was invigorated by growing international support for trade and investment sanctions against South Africa in the early 1970s, revolved around the central question of whether or not the expansion and development of capitalist relations of production would undermine apartheid. Up until the late 1960s, the interpretation of conventional political economists held sway. In essence, these ‘liberal’ political economists argued that apartheid laws and institutions retarded and distorted economic growth. In contrast, scholars in the emerging ‘revisionist’ school argued that, far from undermining capitalist economic growth, apartheid policies delivered the cheap labour necessary for the survival and expansion of South African capitalism. Because this debate was concerned with the then current political issue of whether or not foreign investment in South Africa would undermine or bolster apartheid, the exchange between liberals and revisionists became extremely polarised and important points of agreement were ignored.⁴ For example, in an unusual reversal of theoretical positions, it was liberal scholars who emphasised the dynamics of capital accumulation which would lead to increasingly capital-intensive production and therefore an increased demand for more skilled black labour.⁵ Similarly, the revisionists, in a surprisingly un- Marxist manner, emphasised the influence of political and ideological relationships which would ensure that white supremacy and apartheid would prove resistant to reform.⁶ The politicised nature of this debate no doubt contributed to the fact that what began as an emphasis on how apartheid policies promoted capitalist interests solidified into a revisionist paradigm which excluded the possibility of contradiction between capitalist interests and apartheid.⁷

Since the 1970s, the terms of this debate have shifted. Although Wolpe was the first to emphasise the importance of class divisions within the African population, his latest theoretical formulation marks a break with his own and other earlier revisionist studies. Wolpe now proposes that the relationship between capitalism and white domination is ‘historically contingent’ rather than necessary. Correspondingly, he eschews the attempt to collapse racial divisions into class divisions, arguing that ‘fissures along class lines may occur within racially defined groups’.⁸ However, although Wolpe does draw attention to the important changes in the occupational structure of the South African population, he does not provide new empirical material.

Most empirical studies of the changing relationship between racial inequality were undertaken by liberal economists who conducted extensive research into racial wage differences during the 1970s and 1980s.⁹ However, important as these studies are, the analysis of wage trends is no substitute for a detailed examination of the racial division of labour. The rise to dominance of the revisionist paradigm in academic circles in the 1970s had important consequences for the study of changes in the racial division of labour in South Africa. Since the racial division of labour was automatically assumed to be congruent with apartheid policies, the assumption of unchanged and even deepening racial inequality was accepted as a basic premise rather than as a subject of enquiry. Where scholars did address the erosion of the racial division of labour, they stressed its slow rate of change and limited prospects for future deracialisation without more than a cursory examination of the evidence.¹⁰ The one exception to this pattern is the work of Wolpe which broke important ground by identifying the emerging contradictions between apartheid policies and the racial division of labour during the 1960s.¹¹ However, Wolpe’s preliminary study did not extend beyond a superficial analysis of Population Census data and, in any event, dealt with employment data only up to 1970.

It was not until the end of the 1970s that new interest arose in the changing patterns of racial inequality in employment. This time, however, most of the interest was generated by official concern about the shortage of skilled white labour and its attendant policy implications. Most of this state-contracted work was undertaken by statisticians and economists who relied on the Manpower Surveys conducted by the Department of Manpower. Because these surveys are a more accurate source than the Population Census for the analysis of occupational structure, these studies were a considerable empirical advance on Wolpe’s earlier contributions and have updated the earlier analysis by Simkins and Hindson. However, without exception, these studies are not suitable as secondary sources for a study of change in the racial and occupational division of labour. The first reason for this is that most of these analyses focus on only one aspect of the division of labour. For example, Roukens de Lange examines only the occupational structure and provides no breakdown of the racial composition of occupations.¹² Similarly, certain reports of the National Manpower Commission and the Human Sciences Research Council examine data only for ‘high-level’ and ‘middle-level’ occupations and exclude all semi-skilled and unskilled manual occupations.¹³ The second reason is that some studies are restricted to specific periods or are now seriously out of date. For example, the study by Ngidi and Zulu covers only the period from 1975 to 1985.¹⁴ Similarly, the work by Terblanche covers data only up to 1979.¹⁵ The third reason is that the authors have employed occupational classifications which are not completely amenable for a study of the impact of economic growth on the racial and occupational division of labour. For example, Terblanche et al. categorise unskilled manual labourers and semi-skilled machine operatives in a single occupational group.¹⁶

CLASS THEORY AND OCCUPATIONAL STRUCTURE

The above review of earlier contributions to the study of the occupational and racial division of labour points to the need for some understanding of the criteria according to which South African scholars have demarcated their occupational categories. Generally speaking, revisionists were strongly influenced by the debate among Marxist scholars over the class determination of people employed in middle class occupations. The most influential authors in this debate were Carchedi (1975), Poulantzas (1975) and Wright (1976) who were concerned to explain why large sections of the populations of advanced capitalist countries did not support working class political parties.¹⁷ As such, the models proposed by these authors place particular emphasis on the boundary between the working class and what they termed, respectively, the ‘new middle class’, the ‘petty bourgeoisie’ and ‘semi-autonomous employees’. Specifically, these models of class determination proposed a tighter definition of the working class which excluded, variously, managerial, non-manual, non-productive and semiautonomous occupations from the working class. These models were therefore appropriate to studies concerned to understand the political implications of the emergence of an African ‘petty bourgeoisie’.¹⁸ However, my own study has a very different purpose. My concern is not with the political implications of the breakdown between the correspondence of racial and class differences. Instead, I wish to understand the forces which have shaped the occupational and racial division of labour during the apartheid period. Since most of the international literature on the class structure of society is concerned quite specifically with the relationship between consciousness, collective action and class divisions, these studies are not directly pertinent to this analysis. The reason for this is that studies of class structure are concerned essentially with the division of society into the bourgeoisie, the middle class and the working class. This concern with the boundaries between these classes has, quite legitimately, led scholars to ignore the social divisions within the middle and working classes. Although the models of Carchedi, Poulantzas and Wright are based on quite different theoretical assumptions, their schemes produce a similar result. For Carchedi, wage earners such as managers and professionals who perform both the ‘global function of capital’ and the ‘function of the collective worker’ are not working class, but belong to the ‘new middle class’.¹⁹ Wright classifies front-line supervisors along with middle and top managers. He also places professionals, semi-professionals and middle-level administrators within his middle class category of ‘semi-autonomous employees’. Similarly, artisans, semi-skilled operatives and unskilled labourers are all classed as proletarian.²⁰ Even Wright’s new theory of class determination, although it relies on criteria of exploitation instead of domination and control, still produces the same result.²¹ For his part, Poulantzas places all productive manual workers in the working class. Conversely, all unproductive service workers, non-manual and managerial employees are classed as ‘petty bourgeois’.²²

Why are these neo-Marxist models of class structure too crude for an analysis of the changing relationship between racial and class divisions in South Africa? First, the use of broadly-defined classes simply does not provide the detail that is required for an understanding of the exact ways in which the racial division of labour may be eroding in South Africa. In an earlier study of class theory and the growth of the African middle class, my results revealed that, according to Wright’s class criteria, the size of the African middle class rose from 3.5 per cent of the urban African workforce in 1969 to 8.8 per cent in 1983. According to Poulantzas’ criteria, the proportion rose from 19.0 per cent to 35.0 per cent over the same period.²³ However, as I pointed out at the time, these results could not identify the precise form of the African middle class. Clearly, in a study of racial inequality it would be crucial to know whether or not the increase of the African middle class was due, for example, to the expansion of African employment in teaching and nursing jobs or in middle management jobs. Without such information one could not draw reliable comparisons of the African middle class with its white counterpart, since their occupational composition could differ substantially.

The second problem with using these neo-Marxist models of class structure is that their class categories comprise occupations which are often reproduced by entirely different, and even contradictory, processes. For example, the working class comprises unskilled manual labourers, semiskilled machine operatives and skilled artisans. The fortunes of these different occupational groups within the working class have undergone important shifts with the development of capitalist production from simple manufacture to machinofacture, yet class theories take no account of these changes.

At face value, neo-Weberian class schemes such as that proposed by Goldthorpe would appear to suit the purposes of my study because they use a more detailed occupational breakdown than most neo-Marxist models.²⁴ A further advantage of Goldthorpe’s scheme is that, unlike those of Marxist scholars, it retains occupational descriptions as the basis for allocating individuals to different classes. Since the Manpower Surveys are based on occupational descriptions, this method suits the constraints of my study. However, the disadvantage of Goldthorpe’s scheme is that occupations are grouped together solely on the basis of the similarity or dissimilarity of the market and work situations of their incumbents.²⁵ Although this method is quite acceptable for classifying occupations within an occupational structure, it does not provide any basis for understanding how the occupational structure is itself reproduced and changed. Since one of my aims is to understand how specifically capitalist production relations have shaped the division of labour in South Africa, I still require a theory of the division of labour which incorporates the dynamics of the capitalist labour process and its impact on the division of labour. For these insights I have turned to labour process theory and research.

Unlike neo-Marxist and neo-Weberian models of class structure, labour process theory approaches the problem of class boundaries from the point of view of the labour process and how its dynamics develop and change the division of labour.²⁶ For example, although Braverman’s seminal study of the labour process was concerned with understanding the structure of the working class, his approach was to investigate how the class boundary of the proletariat has been shaped historically by changes in the organisation of work.²⁷ In summary, Braverman argues that, in pursuit of higher productivity, the organisation of production has been transformed by the ever-increasing division of labour whereby tasks are subdivided into increasingly simpler and more routine tasks. The de-skilling of work by this process of fragmentation has been accompanied by both mechanisation and a growing division of labour between the conception, management and execution of work. Braverman’s formulation therefore provides a rationale for identifying divisions between artisans, semi-skilled machine operators and unskilled manual labourers employed in the primary and secondary sectors. Similarly, his scheme also identifies the emergence in the twentieth century of routine white-collar workers as a class distinct from managers and commercial profe...

Table of contents

• COVER PAGE
• TITLE PAGE
• COPYRIGHT PAGE
• FIGURES
• TABLES
• 1. RACE, CLASS AND THE DIVISION OF LABOUR
• 2. THE EXTENT AND PATTERN OF AFRICAN ADVANCEMENT
• 3. CAPITALIST INTERESTS, WHITE LABOUR AND APARTHEID LABOUR POLICY
• 4. RACIAL DIFFERENTIATION, CLASS FORMATION AND THE LABOUR PROCESS IN THE CONSTRUCTION, MANUFACTURING AND MINING SECTORS
• 5. RACIAL SEGREGATION AND DE-RACIALISATION IN THE TERTIARY SECTOR
• 6. ECONOMIC GROWTH AND TRENDS IN THE RACIAL WAGE AND INCOME GAP
• 7. THE INCREASING SIGNIFICANCE OF CLASS
• APPENDIX TO CHAPTER 1
• APPENDIX TO CHAPTER 2
• APPENDIX TO CHAPTER 3
• APPENDIX TO CHAPTER 5
• APPENDIX TO CHAPTER 6
• NOTES
• BIBLIOGRAPHY