The sexual division of labour at its simplest is an allocation of particular types of work to particular categories of people. It is a social structure to the extent that this allocation becomes a constraint on further practice. This happens in several interrelated ways. First, the prior division of work among people becomes a social rule allocating people to work. An employee entering a firm is given job X if a woman, job Y if a man. The working of such rules is found in almost every study of paid employment that has considered the issue of gender, and this is no primitive hangover found only in low-technology industries. The superb ethnography of a British motor vehicle component assembly plant by Ruth Cavendish, Women on the Line, shows an almost absolute separation between jobs allocated to women and to men. ‘It was obvious’, the author remarks, ‘that the only qualification you needed for a better job was to be a man’. Motor vehicles are not the cutting edge of technology any more, but computers certainly are. Game and Pringle in Gender at Work show that the growth of computing has resulted in job markets as segregated as any in the work-force. Women are hired as data entry key-punchers; men predominate as operators, salesmen, systems analysts and managers.

A segregation rule in operation becomes the basis of new forms of constraint on practice, such as differential skilling. Where women and men have been skilled or trained differentially, discriminatory employment becomes rational from the employer’s point of view. As Carol O’Donnell shows in The Basis of the Bargain, differential skilling and training by sex is a very general feature of the interface between the education system and labour markets. Through such mechanisms the sexual division of labour is transformed into an apparently technical division of labour, resistant to the more obvious antidiscrimination strategies. Where men are usually better prepared or trained than women for a given job, choosing ‘the best applicant’ will normally mean choosing a man. The almost complete dominance of the upper echelons of universities by men is a striking example of this indirect discrimination.

Skilling and training is one of the mechanisms by which the sexual division of labour is made a powerful system of social constraint. Just how powerful is revealed when a conscious attempt is made to change it. The experience of antidiscrimination and affirmative action programs is relevant here, and the slow progress they have made is familiar. Less well known, but equally important, is the evidence that has begun to accumulate about attempts to alter the sexual division of labour in unpaid work, notably housework and childcare. This has become, as Lynne Segal notes, an important focus of the personal politics that grew out of the New Left in the 1970s in Britain. Paul Amato, reflecting on two years experience house-husbanding in Melbourne, notes that the reversal was never accepted by most men he dealt with, who felt that men should work (i. e., housework was not work) and men should not be economically dependent on women. One solution to the dilemma was to regard him as successfully exploiting his wife. Clearly the conventional sexual division of labour has strong cultural supports. A recent survey in southern England by R. E. Pahl shows no tendency for unemployed men to take on more housework; and the Australian study of domestic ‘role reversal’ by Graeme Russell shows shared childcare arrangements between husband and wife to be highly unstable.

Nevertheless the fact that there are some attempts to reverse it, together with the creation of a new sexual division of labour in industries like computing, indicate that the structure is not only a constraint but also an object of practice. The research that documents division in workplaces also documents the activity, the

social labour, that sustains it. In Ruth Cavendish’s factory, management was in the hands of men, so was the union, and it is no accident that both hierarchies resisted the attempt of women workers (eventually by a wildcat strike) to alter their situation. Another British workplace study, by David Collinson and David Knights, vividly shows the methods that sustain the sexual division of labour in a white-collar workplace, in this case an insurance firm. Women who look for promotion above routine clerical work are talked out of it by men in management, appealing to anxieties that the men themselves have done much to create. What amounts to a complete miniature ideology of women’s psychological unsuitability for insurance work is sustained; and the real prejudice of managers against women is rationalized by appeal to the presumed prejudices of customers.

The gender regimes of these two workplaces are, however, significantly different, and it is possible that the mechanisms by which the sexual division of labour is sustained are quite specific. Michael Korda’s Male Chauvinism, an early American account of sexual politics in the workplace, presents the oppression of women mainly as a consequence of individual discrimination by bosses. This cannot be true for a situation of collective exploitation like Cavendish’s factory. But in the New York corporate office world from which Korda writes, where hiring and firing, salary determi­nation and promotion are all highly individualized, the exclusion of women perhaps does mainly operate through the individual sexism of corporate executives.

Beyond the individual workplace there is a wider social process that constitutes the sexual division of labour in terms of whole categories of workers. Margaret Power speaks of ‘the making of a women’s occupation’ as a historical process in which new categories of work and workers are formed. It is by no means a matter of an existing division being mechanically reproduced. We now have case studies of individual occupations that document this. One of the best is Eve Gamarnikow’s study which shows how the modern occupation of trained nurse was constructed by entrepreneurs such as Florence Nightingale. Its basis was a trade-off that accepted the control of medical practice by men (i. e., doctors) in order to open a semi-professional career path to middle-class women.

As this case shows, the construction of the sexual division of labour is not just a matter of the allocation of work to people. It also involves the design of the work. The argument about


‘appropriate technology’ – machinery or techniques that are less environmentally destructive, or cheaper, or particularly suited to Third World needs – has made it obvious that there is never just one technique for doing a job. There are even alternative ways of building a nuclear bomb. A socio-technical system, such as industrial manufacturing or domestic work, can be set up in various ways. The particular designs and practices current at a given time therefore represent a social choice of some kind. The current labour practice gets embodied in technology designed with given social arrangements in mind – among them the sexual division of labour. For instance the waterframes in the cotton mills of north-western England and southern Scotland in the early Industrial Revolution were designed to be operated by women and children. As T. C. Smout notes, women and children were preferred as workers for the new factories because they were presumed more docile in the face of unprecedented demands for labour discipline. The factory owners faced the interesting problem of what to do with the husbands in order to attract their preferred labour force.

A more complex example is the technology of housework. Machines like vacuum cleaners or washing machines are equally suited to be operated by women or by men. The models widely sold are however designed to be used by just one household and presuppose one permanent domestic worker per household. This arrangement happens to be provided by the conventional sexual division of labour and not by any other likely or practicable arrangement. Accordingly they are always advertised with pictures of smiling women operating them, not smiling men. It is quite possible to design equipment for doing the same work under other social arrangements. For instance the communal laundries operated by British local government provided a practical alternative way of mechanizing laundry in the 1930s and 1940s. They were squeezed out by post-war privatization. The consumer-goods boom of the 1950s was, among other things, a reconstitution of the domestic sexual division of labour by new means.

These considerations imply that the idea of a ‘division of labour’ is itself too narrow. We are dealing not just with the allocation of work, but with the nature and organization of that work. It is impossible to separate either fact from the distribution of the products of work – that is, the distribution of services and income. Coming back to Cavendish’s factory, the fierce insistence on the sexual division of labour by the men makes more sense when we learn that the men’s jobs were much better paid. Some of the men in the plant were paid twice what the women got, for doing easier jobs. The women got ‘married women’s wages’ whether married or not. This is far from being an isolated case. As shown in chapter 1 for a country which has theoretically had ‘equal pay’ for men and women for fifteen years, the actual average income of women in Australia is still less than half the average income of men.

So the ‘sexual division of labour’ can no longer be seen as a structure in its own right. It must be seen as part of a larger pattern, a gender-structured system of production, consumption and distribution.

To think of a gender structuring of production, not just a sexual division, allows a clearer recognition of differentiations within the work-force that have to do with sexual politics but which operate within the broad categories of sex. Some have to do with the marketing of sex itself, in prostitution or the entertainment industry. But the issue is much broader. Cases include the making of jobs like receptionist, air hostess and secretary as a combination of particular technical skills with a particular femininity. Particular industries, notably clothing and theatre, are associated with homosexual masculinity. On the other hand business management is integrated with forms of masculinity organized around interper­sonal dominance: ‘tough’ stances towards labour are admired in business, and phrases like ‘aggressive marketing’ are terms of approval in management jargon.

This extension of the concept, however, runs up against the concept of the ‘mode of production’, or its near equivalent the ‘social division of labour’, in Marxist theory. There is a good deal of literature attempting to link the position of women to these concepts; as noted in chapter 3, this has been the most important extrinsic theory of gender. The barrenness of this theoretical enterprise is largely because the theorizing has not been bold enough. Almost all this literature has accepted the traditional definition of the ‘capitalist mode of production’ in Marxism as a production system fundamentally defined by class relations. Even the attempt to define a ‘domestic mode of production’ located in the household has left the class analysis of capitalism undisturbed.

It must now be accepted that gender divisions are not an ideological addendum to a class-structured mode of production. They are a deep-seated feature of production itself. They are not confined to domestic work, or even to the division between unpaid domestic work and paid work in industry. They are a central feature of industrial organization too. They are not a hangover from pre-capitalist modes of production. As the cases of computing and world-market factories show, they are being vigorously created in the most advanced sectors of the capitalist world economy.

A series of arguments lead firmly to the conclusion proposed by Game and Pringle. Gender divisions are a fundamental and essential feature of the capitalist system; arguably as fundamental as class divisions. Socialist theory cannot any longer evade the fact that capitalism is run by, and mainly to the advantage of, men.

This is one of the grounds for a wider rethinking of the socialist analysis of capitalism. For the feminist argument has an interesting parallel with the view from Third World radical movements, which have seen capitalism mainly as a system of global inequality and imperialism. Together they suggest a new view of capitalism, as a system for the concentration and regulation of profits extracted by a number of qualitatively different mechanisms of exploitation, rather than the basically homogeneous structure implied by the concept of a ‘mode of production’. If that is broadly correct, we no longer need the kind of sideways skip performed by Eli Zaretsky in Capitalism, the Family and Personal Life, proposing that capitalism took over the existing patriarchal organization of gender or domestic life and used it for its own reproduction. The connection is more direct. Capitalism was partly constituted out of the opportunities for power and profit created by gender relations. It continues to be.

Given this understanding of gender relations in production and consumption, what are their main principles of organization? What kind of ‘system’ are we dealing with? At this point my argument becomes more speculative and uncertain. I offer the following as an organizing hypothesis rather than a firm conclusion. It is, nevertheless, based on the research on domestic and industrial labour already referred to, as well as some practical experiences of reform. Five points in particular seem important:

1 the sheer scale and insistence of demarcations between women’s and men’s work, despite their technical irrationality and the impossibility of making them complete;

2 the connection of many demarcations with issues of profitability

or labour control, or both, in the workplace;

3 the way they function to exclude virtually all women from opportunities for the accumulation of wealth on a scale usable as capital, or from career paths that would lead to the control of significant capitals;

4 the importance of practices promoting the solidarity of men – often across class lines – in maintaining these demarcations;

5 the consistency with which divisions of labour and differences of income operate to allocate childcare to women, especially younger women.

Much of this can be understood in terms of two major principles. One might be called the gendered logic of accumulation. The overall gender organization of labour concentrates economic benefits in one direction, economic losses in another, and on a scale sufficient to produce a dynamic of accumulation in its own right. Christine Delphy identified such a mechanism in her study of French marriages, but by confining the issue to the household and the one-to-one marriage relation, missed the larger possibilities of accumulation in industry. The benefits and costs are not distributed in an all-or-nothing manner between the sexes as blocks. The printing tradesmen studied by Cynthia Cockburn are minor beneficiaries; the media capitalists who employ them are major beneficiaries. Not all women are major losers, a fact of strategic importance to feminism. Overall the benefits, opportunities and costs are large enough to be worth fighting for, and motivate the active practice of demarcation and exclusion by many groups of men.

There are two internal limits to the pace and scale of gender – based accumulation. One is the fact that the division of labour, while marked, is far from absolute. Very few women are sailors and very few men are secretaries, but quite a lot of both are clerks, shopkeepers, sales personnel, programmers and teachers. Much of the labour in peasant agriculture is shared. The second is the fact that marriage is a one-to-one relationship. There is limited scope to extract material advantages from the labour of just one other person. So the scale of economic inequalities resulting from labour organized through marriage is severely limited, when compared with the economic inequalities that can be produced by accumu­lation in industry. In this respect the nuclear family form must be seen as an important restraint on sex inequality. The current commercialization of domestic work, for instance in the growth of ‘fast food’ outlets, is likely to increase the economic inequalities of gender.

The second principle might be called, perhaps less happily, the political economy of masculinity. A number of important practices have to do with the definition of masculinity and its mobilization as an economic resource. Ann Curthoys has argued that childcare is the basis of the sexual division of labour and that the childcare issue is the structural basis of feminism. This overstates the case, but the general importance of the issue is undeniable. Curthoys acutely observes that childcare is not just an issue for women but an issue about men: ‘the notion that caring for young children is not a fit occupation for men is extraordinarily deepseated’. Since men have more control over the division of labour than women, their collective choice not to do childcare, as Margaret Polatnick has argued, reflects the dominant definition of men’s interests, and in fact helps them keep predominant power. The ability of management, in many industrial conflicts, to mobilize men workers and their unions in tacit alliances against women workers confirms the strength of these definitions of interest. How particular definitions of masculinity are interrelated will be discussed later. Here I will simply note that a hegemonic pattern of masculinity, in organizing the solidarity of men, becomes an economic as well as a cultural force.

This force does not operate without resistance. The sexual division of labour itself creates bases for solidarity among women. In industrial employment the widespread exclusion of women from career grades has the effect of giving them shared work experiences and little structural reason to be in competition with each other. The routines of commuting make women the daytime majority in dormitory suburbs; and studies of new suburbs by sociologists such as Lyn Richards emphasize how important, how carefully sought and monitored, are their relationships with each other. Beatrix Campbell, discussing the current recession in Britain, notes how the division of labour in childcare means that young single mothers on state welfare benefits enter a community of women that links very different age-groups. While none of these are blanket effects, they do indicate a potential for self-definition or resistance.