Robert M. Blackburn, Jude Browne, Bradley Brooks and Jennifer Jar man
Explaining gender segregation
ABSTRACT
Occupational gender segregation - the tendency for women and men to work in different occupations - is an important feature of all societies, and particularly the wealthy industrialized ones. To understand this segregation, and to explain its significance, we need to distinguish between vertical segregation entailing inequality and horizontal segregation representing difference without inequality, with overall segregation being the resultant of these components. Three major theoretical approaches to understanding occupational gender segregation are examined: human capital/rational choice, patriarchy, and preference theories. All are found to be inadequate; they tend to confuse overall segregation with its vertical component, and each entails a number of other faults. It is generally assumed or implied that greater empowerment of women would reduce gender segregation. This is the reverse of what actually happens; in countries where the degree of women's empowerment is greater, the level of gender segregation is also greater. An alternative theoretical approach based on processes of social reproduction is shown to be more useful.
KEYWORDS:   Gender segregation; patriarchy; human capital; rational choice; preference theory; social reproduction
The existence of occupational gender segregation, the tendency for men and women to work in different occupations, is well known. In countries with low levels of Gross Domestic Product, such as many of the African states, the degree of segregation is often quite small (Blackburn, Brooks andjarman 2001a). However, in the wealthy industrialized countries of the world the levels of segregation are all at least moderately high. We shall, therefore, confine our attention to these more segregated countries. There are a number of theoretical explanations that have been put forward to explain this segregation, all of which have some explanatory value, focusing attention on some aspect of social reality. Nevertheless, all are unsatisfactory. In this paper we explain why the existing popular explanations are inadequate, and we offer a different sort of explanation.
British Journal of Sociology Vol. No. 53 Issue No. 4 (December 2002) pp. 513-536
© 2002 London School of Economics and Political Science      ISSN 0007-1315 print/1468-4446 online Published by Routledge Journals, Taylor & Francis Ltd on behalf of the LSE DOI: 10.1080/0007131022000021461
514        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
CONCEPTUALIZING SEGREGATION
First of all we need to clarify the relation between gender segregation and inequality. There has been a common tendency to regard the level of segregation as a measure of gender inequality, or at least a strong indicator of inequality, but this is not necessarily correct. As early as 1979 Hakim introduced the idea of vertical segregation to represent inequality, but this was in a limited situs-specific sense; for instance hospital consultants were seen as vertically above nurses and head teachers above teachers in schools, but with no way of placing head teachers above or below nurses (Hakim 1979). Consequently the application of the idea of vertical segregation has been limited. While others have used the idea of vertical segregation, sometimes with a wider meaning, in qualitative research, only recently has there been any attempt to measure vertical segregation as a dimension of segregation. Overall segregation may be seen as the resultant of vertical and horizontal components (Blackburn and Jarman 1997; Blackburn, Brooks and Jarman 1999; Blackburn, Jarman and Brooks 2000; Blackburn, Brooks and Jarman 2001a and b).
As Figure I indicates, we are using the term 'overall segregation' to refer to what is normally referred to simply as 'segregation'. Then 'vertical segregation', and only vertical segregation, measures the component of inequality.1 It measures inequality among all occupations, giving a single vertical dimension rather than a set of separate parallel dimensions2 with one for each situs (as in Hakim's 1979 original formulation). The 'horizontal' component is orthogonal to the vertical one, and so measures the extent of occupational difference without any element of inequality.3
In looking for explanations of gender segregation, therefore, we need to be absolutely clear about the dimensions and resultant segregation. It is not a simple matter of accounting for overall segregation. We need to understand why there is difference and why there is inequality in the occupations of women and men; in other words we need to explain the existence of substantial horizontal and vertical degrees of segregation. On the whole, in
A
Vertical segregation
Horizontal segregation
FIGURE I:    The components of segregation
Explaining gender segregation
515
the countries we have examined (Britain, the USA and Greece (Blackburn, Brooks and Jarman 2001a)), the horizontal component tends to be appreciably greater than the vertical, but both need to be explained. Since the two components are, necessarily, combined in overall segregation, there is a sense in which we are concerned with this resultant, but both components are important. Explanations focused solely on inequality must be, at best, incomplete.
EXPLANATIONS' OF GENDER SEGREGATION
We turn now to the theoretical explanations that have been proposed to explain gender segregation. There have been a vast number of 'theories', and it is not our intention to review them all - in any case space would not permit. Instead we shall consider two groupings: (1) neo-classical economic theories of human capital and rational choice, and (2) feminist and sociological theories of patriarchy. Each theory emphasizes one aspect of potential explanation, and it is rare to consider two in combination. Hence, they tend to appear more as rival theories than as elements that might be combined. We then look at Hakim's (1996, 1998a) 'preference theory' which seeks to integrate these approaches.4 However, our conclusion is that each approach has its problems, whether taken separately or in combination.
HUMAN CAPITAL AND RATIONAL CHOICE
The idea of 'human capital' and its extension in 'rational choice' present a particularly popular and influential approach to occupational differences and inequalities. Arguments in terms of human capital and rational choice theory are developments from within economics. However, the idea of human capital is, in turn, a development from two well-established sociological arguments, indeed so well established that they might be regarded as simply common sense. The two elements of these arguments are qualifications, or the marketable skills the qualifications represent, and experience. In combination these have been treated by economists as a form of capital, notably by Becker (1964), Mincer (1966) and Polacheck (1981), with more explicit reference to gender introduced in Mincer and Polacheck (1974) and Polacheck (1975).5 As we shall see, this combination was a theoretically unsound move. However, the dominance of economics within the social sciences has led to wide-spread adoption of the concept, even among sociologists. There has been some decline in enthusiasm for the approach among sociologists, but without a replacement, and the popularity in economics remains.
Rational choice theory is, in one sense, tautological, asserting that people choose to act in the ways that appear to best serve their interests. It
516        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
would be odd indeed to have a theory that suggested people did not choose to pursue the ends they wanted to pursue. However, the theory is based on the idea of economic choice and interests, and is thus open to rejection.6 It may be seen as closely related to human capital theory in that rational choices depend on what is attainable, which in turn depends on the level of human capital and on the nature of jobs available, the conditions and practices in the labour market, the health of the economy etc.
How do these theories seek to explain gender inequality in the labour market? Firstly we need to recognize that usually they are only concerned with explaining income inequality, as though that is the only aspect of stratification that matters.7 If this were the case we would be faced with a general situation of alienated instrumentalism, where men and women sacrifice the part of their lives while at work in order to gain the money to live the rest of their lives. It means that intrinsic rewards of work, like interest and companionship, are irrelevant, since only the extrinsic reward of money is important. That is, the only benefit from working is seen as the means to purchase significant items for life outside the work situation. It is recognized that women may settle for poorer jobs to fit in with domestic priorities, but this is seen as a distortion of the relation between human capital and income. Nevertheless, money wan essential reward of work, and is well related to other aspects of social stratification.
To a large extent these economic arguments are not explicitly about segregation, but rather about the effect of segregation on differences in pay. As Anker (1998: 14) points out 'many theories and explanations treat the determinants of occupational segregation by sex and male-female pay inequality as if they are synonymous, [which] is unfortunate' since segregation is only one cause among many of pay differentials. We would add that pay differentials are not even a measure of overall segregation but only of vertical segregation, a distinction regularly overlooked in discussions of human capital and segregation. Furthermore, pay is only one way of conceptualizing a vertical dimension, giving rather different results from a measure of general social advantage (Blackburn, Brooks and Jarman 2001b).
A key element in the economic (and other) explanations is the impact of domestic work on earnings. It is argued that spending time and energy on domestic work prevents women from investing in human capital. (Even the prospect of future domestic work associated with child-rearing influences many younger women to make decisions not to develop their human capital.) The rational choice, therefore, is for the person with more human capital, the man, to be the principal earner, while the woman takes primary responsibility for domestic work.
The basic logic of this argument has long been familiar; it is economically rational to prioritize the employment of the higher earner of a partnership. In so far as most couples pool their money or place it in the care of the woman (Pahl 1983, 1989) such an approach is logically sound for both partners, assuming that they stay together over the long-term and
Explaining gender segregation
517
that there is basic agreement about how their finances should be handled. Even where the partnership is less egalitarian, most of the income usually goes on items that are beneficial to the whole household, such as food and shelter (Blackburn 1999), so the approach may still be sound.
However, we should recognize that the argument in no way depends on the concept of human capital. The rational economic choice depends on the fact that men tend to be paid more than women, whatever the reason for the difference in pay. The situation is self-reproducing; because men can earn more than their partners, their market employment is prioritized, and so they continue to earn more. Since the women need to expend considerable energies on domestic work, when they enter the labour market they often do so part-time or in full-time jobs that are less demanding and so less rewarded. In these ways a segregated labour market is sustained.
Human capital takes a central role in the explanation of the gender gap in earnings because it is seen as the basic determinant of occupational pay. It is argued that the greater the amount of human capital that is involved in the satisfactory performance of an occupational role, the more important it is that the occupational incumbent should have the requisite level of human capital. Since there is a cost, at least in time, in acquiring human capital, it has a scarcity value. Therefore market principles ensure that levels of remuneration are directly related to levels of human capital.8
Problems with Human Capital and Rational Choice Explanations
While there is certainly some truth in the argument that human capital has an impact upon pay, and that this accounts for gender differences, there are serious theoretical and empirical flaws with the concept. In the first place we should note that gender differences in pay are greater than would be predicted on human capital arguments (Treiman and Hartmann 1981; World Bank 1994). Then there is an extensive literature criticizing this sort of economic approach (e.g. Anker 1998; Dex 1988; Stewart et al. 1985; Anderson and Shapiro 1996). The usual lines of criticism of human capital theory are that as it assumes perfect labour market information on the part of actors, it does not properly account for discrimination or fluctuations in labour market demand (Dex 1988: 283; Krahn and Lowe 1988: 73; Gallie 1988: 21; Ashton, Maguire and Spilsbury 1990; Harris 1987). However, here we will confine ourselves to the more fundamental faults in the very notion of human capital, that have not been as extensively discussed. Some of the criticisms apply to human capital theory in general, but a fortiori they are relevant to supposed explanations of gender segregation.
It is claimed that childcare and housework do not contribute to the growth of an individual's human capital. Yet they entail and develop skills which have marketable value. Cooking skills can be quite highly rewarded; sewing and cleaning are also occupations paid for in the labour market. Childcare also entails a range of activities such as nursing, teaching and
518        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
general supervision of children which are relevant to waged employment, and running a household involves a number of administrative and managerial tasks. Admittedly most of the experience is not directly relevant to highly paid forms of employment. Nevertheless, neither the fact that it is usually done by women nor that it is not waged work justify the attribution of zero value to the experience.
The notion of human capital is extremely problematic, if not nonsensical (Blackburn 1988).9 The idea of capital is used by analogy with economic capital, typically measured in money value. However, the question of ownership is very strange in this case. If someone invests their money in shares it is quite clear who owns the shares; with human capital it is quite different. The society collectively invests extensively in education and training, so might be expected to own the resultant human capital. However, it could not claim sole ownership. The educators themselves, as direct investors, might be entitled to a share, as might the parents of individual workers, since they contribute in many ways, including financial ones (particularly but not only at higher education levels). Then the employers provide experience for the worker, and they may also provide specific training. Finally, the individuals themselves have to make the effort to contribute to their qualifications and experience. So, on sound theoretical grounds who should be the owner? It seems ownership should be regarded as widely spread, but this is not how it is treated. Contrary to the logic of the concept, ownership is attributed entirely and solely to the individual worker.
A further problem with this concept of 'capital' concerns the way to spend it. The value of economic capital can always be realized in cash, at least in principle, and spent. Rash or irresponsible individuals can 'blow' all their capital on extravagant living. For human capital it is totally different. The way it is 'spent' is by using it to obtain an appropriately good job. However, far from reducing the stock of capital, this sort of spending actually increases the capital. In fact there is no way to use it up. Financial capital can increase in value if left invested to acquire interest. Again it is the opposite for human capital; the way it can lose value is by leaving it 'unspent' so that the skills deteriorate. For instance, unemployed workers do not become more employable as their period of unemployment increases.
Even if we could ignore these problems, we are faced with the fact that human capital explanations are circular. In general the capital is seen as increasing with experience, but more precisely we may say it increases unless it decreases. The use of skills increases the capital, through growth of experience, unless the skills become outdated and so less valuable. The only way we can tell if the value of the capital goes up or down is by observing whether the pay increases or decreases, so the idea of capital adds nothing to our understanding.
A similar circularity exists in the human capital supposed explanation of the lower wages of women. We are told that women's qualifications are
Explaining gender segregation
519
worth less, and therefore are paid less. How do we know they are worth less? Because women are paid less. Such a line of reasoning may have some validity in accounting for pay differences within each gender, but it cannot be applied legitimately to differences between genders (Browne 2000b). Some jobs are said to be more suitable for women because the women have 'nimble fingers'. Packing chocolates from a conveyor belt into boxes certainly fits this notion of manual dexterity but is not well rewarded. A good secretary has to combine many skills: a high standard of fast typing (more than just manual dexterity); an efficient shorthand; an ability to look after her/his boss, keeping track of his/her activities and appointments and keeping the office paperwork under control; being agreeable in person or on the telephone to outside people; and so on. It would be hard in terms of qualifications and experience to justify the unremarkable level of pay for such work, unless we adopt the human capital tautology that the value of the capital corresponds to the level of remuneration. Indeed, comparable worth studies, which may be regarded as providing approximate measures of human capital requirement, show a persistent undervaluing of women's occupations (Anker 1998: 30-33).
A related variant of the qualifications argument, as used for instance by Becker (1991), is that women have lower productivity (whether associated with inferior qualifications, or learnt or natural gendered skills). This is supposed to be a result of their lower commitment to work following from their greater involvement in domestic roles. It may well be that they are less interested in promotion, but it is questionable whether they can really be said to have lower commitment (Purcell 1988). As for productivity, Joshi and Paci (1998) have shown there is no relation between gender differences in pay and levels of productivity. This is particularly evident when we consider the low returns to part-time workers, who are predominantly women.
Women's pay disadvantage is not confined to mothers, or even to married/cohabiting women. It has been argued that the main differential is between married women and men (e.g. Blau and Khan 1992), which is so for fairly obvious reasons. Nevertheless, there are many occupations where single women are poorly paid. Rational choice theory10 accounts for this in terms of a general expectation that women will have children. Against this, several studies have pointed out that the evidence indicates greater concern with career and income, and less commitment to having children than has generally been assumed (Browne 2000b; Franks 1999). Nevertheless, having children is a realistic expectation for most young women, and they know this will alter their lives significantly. However, if children rank below income and career in women's expectations, it hardly seems reasonable to suppose they choose poorer jobs because they expect to have children.
While rejecting the fashionable human capital-rational choice approach, we do recognize some value in its traditional sociological form. Qualifications and experience do tend to be reflected in levels of pay. Equal pay
520        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
legislation has led to some tendency for them to have similar relevance for men and women. However, the substantial degree of gender segregation means that it is still possible for 'male' and 'female' jobs to be paid at different rates in relation to the skills and experience entailed. Far from being able to rely on them to explain vertical segregation, it seems that vertical segregation helps to explain the value put on qualifications, skills and experience. The gendered pattern of skills acquired is related to horizontal segregation (Browne 2000b), but in so far as there is a causal direction it is from segregation to the skills that appear relevant to men and women as they prepare to enter the labour market. As we have seen, the introduction of the concept of'human capital' has simply served to obscure the situation and undermine theoretical understanding. It may appear to be a useful single term to cover qualifications, skills and experience, but it is theoretically misleading in general, and counter-productive in relation to gender segregation.
PATRIARCHY
A more social approach to understanding gender segregation has been in terms of patriarchy. Feminist writings extended the original commonsense meaning of the term to apply to a situation of male dominance in a society. This immediately raises the question of whether the term is purely descriptive, or whether it refers to an explanatory theory. It is important to appreciate this difference as many sophisticated feminist arguments use the term in its descriptive sense.
The general idea of patriarchy has spread beyond feminist approaches to more general sociological arguments. At the same time there has been growing awareness of its limitations, but the result has been a range of variations rather than a recognized alternative. The essential ideas cover male power and control, exclusion of women from the best jobs, keeping women dependent on men (because their earnings are low), and generally exploiting women as workers in the labour market and in the home.
As a descriptive term patriarchy certainly has some validity. Traditionally men have dominated the positions of power in a society, and although this has been changing there is no country where women have gained equal access to power. The United Nations measure of Gender Empowerment, GEM, - which is actually a measure of women's empowerment - is less than 1 for all countries measured. Since 1 is the value representing equality, this means the advantage everywhere lies with men. Looking at components of the GEM, we see that everywhere women are under-represented in seats in parliament, in administrative, managerial, professional and technical occupations, with lower levels of income (United Nations 2001).
Explaining gender segregation
521
Problems with Patriarchy Explanations
As an explanation of segregation, however, the concept of patriarchy is, to say the least, far less useful. We understand why Hakim (1996: 210) claims that 'theories' of patriarchy are essentially 'just description'. Nevertheless, several writers have maintained that segregation is a result of male dominance, of patriarchy (e.g. Delphy 1977, 1984; Hartmann 1976; Walby 1986, 1990). Many more have used the concept in discussions of gender inequality. Patriarchy appears to be intended as an explanation of vertical segregation, but in common with most segregation theorizing, overall segregation is taken to be a vertical measure. Thus difference is commonly equated with inequality. This is a significant mistake since only the vertical dimension represents inequality. Most of the difference in segregation is actually located on the horizontal dimension.
One problem with 'explanations' in terms of patriarchy is that the argument appears circular; patriarchy is used to explain a situation of patriarchy - male dominance is 'explained' by the fact that men dominate. Beyond this there are some serious problems with the approach.
In the first place we should note that there is a clear tendency for the GEM and its separate components to be positively related to the level of overall segregation; that is, the higher the level of women's empowerment in a country, the more the women are segregated from men (Blackburn, Jarman and Brooks 2000). This is the opposite of what we would expect according to patriarchy theories.11 If segregation is a result of male dominance, we would expect it to decline as women's empowerment increased. The reverse is what generally happens in reality. Thus, for example, egalitarian Sweden has particularly high segregation while the level in Japan is unusually low.
However, there is also some tendency for vertical segregation to decrease as overall segregation increases. If we focus, not on overall segregation but on its vertical component, which directly measures gender inequality, then it is possible, as we would expect, that vertical segregation decreases as women's empowerment increases (Blackburn, Jarman and Brooks 2000; Blackburn, Brooks and Jarman 2001b). Thus, if we consider vertical segregation instead of overall segregation, the patriarchy approach may retain some plausibility. However, there are other problems with the approach.
A major weakness of theoretical approaches such as that of Walby is that they underestimate women's agency and over-estimate men's agency (Browne 2000a). Women are presented as passive victims in their relationships with men. When Walby claims, for example, 'men have usually been successful in excluding women from the better work' (1986: 248) the men are accredited with too great a degree of collective organization. One may wonder why there are, then, such huge differences among men, with many having less desirable occupations and poorer living conditions than many women.
Patriarchy is often linked with capitalism and class in the analysis of
522        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
women's oppression, as in the argument that men exclude women from the attractive occupations. 'Job segregation by sex ... is the primary mechanism in capitalist society that maintains the superiority of men over women' (Hartmann 1982: 448; see also Hamilton 1978; Middleton 1983; Barratt 1988). There is an unexamined assumption that women are concentrated in lower level occupations. Yet this is not really the way things are. In terms of vertical segregation measured by social stratification, there appears to be little difference between men and women, with a slight tendency for women to be advantaged (Blackburn, Brooks and Jarman 1999, 2001b; Vaiou 1999; Fox and Suschnigg 1989; England 1979). While men may still dominate in the most desirable occupations, they also outnumber women in the low-skilled manual occupations, contrary to 'patriarchy' assumptions. On the other hand, when occupations are ordered according to male pay (to avoid effects of discrimination or length of service) the vertical dimension favours men in Britain (though there is little difference in the USA). Even so, the horizontal dimension of difference (without inequality) is appreciably larger than that of vertical inequality (Blackburn, Brooks and Jarman 2001b). Whether this contrast between social stratification and pay is due to gender choices or historical processes, it is hard to see how it can be attributed to patriarchy.
The arguments of patriarchal exploitation tend to be based on a selective rewriting of history, and an interpretation of the past from a contemporary perspective. As we explain below, the situation was very different from that presented by Walby, with most wives wanting their husbands to be effective 'breadwinners' (Humphries 1977, 1988). The patriarchy arguments also fit badly with recent developments, where changes in the labour market are destroying traditional 'male' occupations. The decline of coal, steel, chemicals, shipbuilding and docking has adversely affected men, while the growth of service industries has favoured women (Browne 2000a). Consequently, in many formerly industrial cities outside the south-east of Britain, women are now a majority of the labour force. By the middle of 1998 the most female labour force was in Middlesborough, with 55.3 per cent women, followed by Liverpool - 54.5 per cent, Glasgow - 51.5 per cent, Manchester - 51.5 per cent, Bristol - 51.3 per cent and Sheffield - 50.8 per cent (Guardian 1998). This development is hardly in keeping with male control of employment.
A final point to note before leaving our discussion of patriarchy is that gender does not provide a satisfactory basis for distinguishing good and bad jobs. As Siltanen (1994) demonstrates, although 'component' jobs (those not providing a living wage) are mainly held by women they are also held by some men. The crucial factors determining who holds such jobs are not gender persebut the social circumstances of the individuals. This simply is not compatible with universal patriarchal dominance.
Explaining gender segregation
523
PREFERENCE THEORY
The ideas collected under the name of'preference theory' were developed by Hakim (1996, 1998a,b, 2000), who sees the theory as providing the missing link between biological theories12 and accounts of patriarchy, while being a refinement of rational choice and human capital theories. She criticizes Becker's rational choice perspective for treating all women as equivalent in relation to childcare - the same sort of fault we noted in the patriarchy approach. Instead she identifies three basic types of women in relation to their family and work commitments.
Two of Hakim's type groupings are clearly polarized; these are the work-centred and the home-centred women. However, the largest of Hakim's three groupings is made up of 'adaptive' women, which essentially comprises those who do not fit into either of the other two groupings.
The work-centred are, as the term indicates, career-oriented women whose priorities are all focused on the public sphere. They tend to have high commitment to work and low commitment to family life. Many of them are single - appreciably more than in the population as a whole - and even more are childless, though some do successfully combine work and family. Where they are married they are likely to have what is often seen as a more egalitarian life-style, in the sense of more equal sharing of domestic activities.13 The work-centred women choose to, and are able to compete on equal terms with men, and are often employed in gender-mixed occupations, where there are similar numbers of women and men. Occupational gender segregation is low for these women. However, Hakim (1996: 209) argues that because they are competing as equals they are more likely to suffer from discrimination: this would contribute to positive vertical segregation (in our approach to conceptualizing gender segregation, where the sign is positive or negative depending on whether the advantage lies with men or women).
In contrast, the family-centred women regard their families as the central feature of their lives. They hold a more traditional view of women's place, or at least their own place as wives and mothers, being in the home. They prefer to be with their children and not to be in employment away from home. Qualifications do not serve a direct purpose; they are symbolic of education, which in turn contributes to the intellectual life of the home. Segregation only becomes an issue if they do enter the labour force, when we may expect their work to be temporary, part-time and flexible.
While each of these group types is seen as accounting for something in the region of 20 per cent of women, the majority are in the 'adaptive' group. This is a much more diverse category. It includes those women who make a deliberate choice to combine work and family, whether the reasons be financial pressures or interest. It also includes those with no clear strategy for organizing work and family life - the 'drifters'. These women present a clear contradiction of human capital theory in that they do not make economic use of their qualifications. For all the various sorts of
524        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
women in the adaptive group, however, it is the case that whatever the reasons for working, they are not sufficient to dominate choices. Unlike the work-centred, these women are not ambitious for career success and are content to settle for less demanding 'female' occupations. Many combine home and work by working part-time. They thus contribute to the maintenance of both horizontal and vertical segregation.
Hakim criticizes Becker's approach for entailing the assumption that people have stable preferences. Against this she argues, with good justification, that preferences may change through the life-cycle. For instance, mothers may become more work-centred as the children grow up and leave home. She also criticizes his emphasis on the limitations imposed by child-care. While agreeing that such limitations are real, she maintains that women have considerable choice in relation to the impact of children on their lives, including at the extreme the option of remaining childless.
Problems with Preference Theory
Rather than seeking explanation in terms of a single cause, Hakim attempts to take account of the different approaches we have considered together with biological determinism. In one sense this is a definite advance which overcomes the narrowness of the other approaches. However, as the other approaches were each flawed in several ways, the combination does not entirely overcome their weaknesses. Indeed, not only are there some distinct problems with the preference theory approach, but there is also a tendency to accumulate the various problems of the separate approaches (Browne 2000a, b).
We have criticized the dominant approaches for their strong general tendency to treat men and women as two separately homogeneous categories. Hakim's 'preference theory' approach is also critical of such assumptions of homogeneity, particularly with respect to women. However, she then goes on to do the same thing (Crompton and Harris 1998a, b), except that she has expanded the number of categories from one to three - Home-centred, Adaptive, and Work-centred (Browne 2000a). While the typology is useful in differentiating significant aspects of women's experience, it does not begin to capture the variety of people's orientations and experience in large contemporary societies.
At the same time the approach is highly individualistic in its dominant emphasis on choice. At all points it is assumed that people choose their course of action. Of course there is a sense in which this is necessarily true. However, choices are always constrained by social circumstances, and this is insufficiently allowed for. The availability and cost of childcare may be crucial. At one extreme a willing grandmother living close by may make full-time employment possible (Young and Willmott 1957); at the other the cost of paying for childcare may make employment impossible. The choices women, and men, make are not necessarily their true preferences but
Explaining gender segregation
525
rather what is possible. A variety of conflicting considerations - attractiveness of available work, possible pay levels, possible childcare arrangements, financial need, etc. - all combine to restrict and shape choices. Preferences may often have little relation to outcomes.
As Crompton and Harris (1998a) point out, 'preference theory' is really a revival of 'orientations' theory with a new name. Although Hakim (1998a) vigorously denies this, her own account of preference theory looks very like accounts of orientations.14 The range of orientations theory is much greater than Hakim (or her critics Crompton and Harris) acknowledges, going back to Thomas and Thomas (1928) and developed in theoretical terms by Parsons and Shills (1951). Hakim's discussion of the relation between employment and motherhood was largely anticipated in Beynon and Blackburn (1972), and the most thorough, detailed examination of the approach (Blackburn and Mann 1979) concluded that preference choices were very highly constrained.15 Admittedly this last study was of men in manual occupations, but there is little to suggest women have greater occupational choice. A few women who can afford it may choose to be at home, but practical and financial constraints are usually decisive.
One factor that is important here is class or stratification. This is not totally ignored by Hakim, or in the various approaches on which she draws, but tends to be inadequately considered in them all. In the first place we may note that the opportunity to choose varies directly with the stratification level of the household. The higher the level of the household, the greater the income and the less need for a woman to work, and so the more choice about whether to do so or not; the lower the level, the more an extra income is needed. The better qualified the individual, the more attractive are the opportunities to work, as human capital theory asserts in relation to income. In general we see that choice varies directly with stratification level (Browne 2000b).
There is a general assumption that women have poorer occupations, whether it is due to biology, patriarchy, lack of human capital or whatever. However, as we have seen (p. 521), this is not entirely correct. In terms of stratification, and even in terms of pay in the USA but not Britain, those women who are in the labour market are, on average, as well placed as men (Blackburn, Brooks and Jarman 1999, 2001b). Hakim (1998b) is at least aware of this but her account, drawing as it does on the other approaches, takes little note of this situation.
A general weakness of the explanations we have been considering is that they do not take proper account of the power and choices of employers. Of course it is assumed that employers take account of human capital and/or act in patriarchal ways, but this does not include the effects of varying degrees of gender bias or the employers' pursuit of their perceived economic and social interests or needs. A degree of gender bias is inevitable, as people are generally slow to adapt to changing circumstances. This effect may be expected to decline as more and more women demonstrate their ability to hold all sorts of jobs. Employers' interests and needs
526        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
with relation to employees vary with changing social and economic circumstances, as discussed in the next section.
We have seen that the different approaches have problems even for explaining overall segregation. In some cases it is actually vertical segregation they attempt to explain, though generally presented as overall segregation. All work on the assumption that vertical segregation is always positive, that is, working to the disadvantage of women. Even as explanations of positive segregation they are seriously deficient. Since they have not conceptualized a dimension of vertical segregation which can be negative or positive, they do not even recognize the need to explain those situations where vertical segregation is found to be non-existent or negative.
SOCIAL REPRODUCTION AND CHANGING GENDER RELATIONS
The basic problem with all the types of explanation we have considered is that they are external to the social processes concerned. It is very common in social theory to look for some factor to apply to the situation as a supposed explanation, be it patriarchy, human capital or whatever. Such approaches are essentially static, and unable to explain a real situation which is continually changing and developing. The explanation that we outline in this section is based around three important processes that have been ongoing which, in combination, have crucial influences on gender segregation and inequality.
(1) There has been a substantial expansion of education.
(2) The occupational structure has changed considerably.
(3) Adult female participation in the labour force has become easier, (particularly for married women with children).
It would be possible to give explanations of these processes, but for present purposes we simply need to recognize how they provide the basis for a proper theoretical explanation of the current patterns of occupational gender segregation.
We see that employer action and family choices, in changing circumstances not directly related to gender segregation have combined to produce the current high level of horizontal segregation and the smaller, but nevertheless important degree of vertical segregation. Employers have to solve their staffing needs in relation to the changing nature of the potential labour force and relevant ideology. For instance, at one time the dominant ideology was that married women - and especially mothers - should be in the home, whereas the ideology has now changed to extent that the British government policy is that even single mothers should be employed in the labour market.
If we are to understand the pattern of vertical and horizontal segregation, we need to see how the present situation has developed. Social
Explaining gender segregation
527
reproduction is a continuous process of transformation, so that the current position is not simply a reflection of contemporary social influences. Present arrangements emerge as the social processes modify what already exists. It is impossible to give a total explanation of any phenomenon, whether we are dealing with natural or social science, but it is possible to identify key elements which significantly advance our understanding. For our present purposes, using Britain as an example, it will be useful to look back over about a hundred years to the start of the twentieth century. This will cover a period of major changes in the labour market, and in gender relations, both in the home and especially in the workplace.
It is not simply that changes have occurred, but that these changes bring in a wide range of influences on the gendered pattern of employment. Theoretical explanation requires both an adequate understanding of what is being changed - what the situation was previously - and also an understanding of the causes and effects of the ongoing social processes. Useful theory in any field of knowledge must be directly related to actual ongoing processes.
The first point to appreciate is that as a labour market changes it affects the participation patterns of both men and women. Looking back a hundred years we see a much larger manual workforce and far fewer non-manual workers. The British census classification of 1911, where the British Registrar General, Dr. T. H. C. Stevenson, first introduced social classes, included only one non-manual class but seven manual ones. Many contemporary professions did not exist then and the total numbers in the professions were quite small. Clerical work was also very limited. Today manual work accounts for only a minority of jobs, and many of the traditional 'men's' occupations, like mining and shipbuilding, have largely disappeared. There has been a general shift 'upward' into non-manual work, first at the clerical level and then at the professional level. In 1911 only 5.1 per cent of male and 3.3 per cent of female employees were clerks; by 1971 the corresponding figures were 6.7 per cent and 27 per cent, and in 1991 they were 6.8 per cent and 28 per cent respectively.16 Thus there was a huge expansion of clerical work. More recently the major expansion has been in the professions. The newly created jobs usually had to be filled with the people available, and as we shall see those who were available were mainly women. The effect of this may be seen very clearly in the growth in numbers of female clerks, and more recently in the growing number of women in some newer professions.
As the occupations expanded, their status and pay inevitably declined; being a clerk in 1890, for instance, was very different from being one in 1990. Similarly the status and pay of university staff has declined as their number has grown, though in this case low turnover has meant that, despite the growing number of women, the majority are still men.17 At the same time men have continued to compete for the lucrative jobs in business. In the USA there has been a similar pattern, which Reskin and Roos (1990) describe as job and gender queues, though the collective
528        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
power of men to achieve high (or total) vertical segregation is much less than they imply, in the USA or Britain.
In the early years of the twentieth century, young people could start work at the age of 12 (not many years earlier it had been at 8 years old). Domestic service was the main form of employment for women, accounting for a third of all female employment; it was also quite important for men, but to a much lower extent. Since then, employment in domestic service has declined steadily and is no longer a major source of employment.18 This is one aspect of the fundamental change in female employment in the last century.
A hundred years ago women accounted for 30 per cent of the labour force. This remained the proportion, more or less, into the 1960s, apart from wartime. However, in those early days the women workers were mostly teenagers and women in their early twenties; only one in ten were married. Few women had any sort of career in market employment at that time. This does not mean women were not economically active in any sense. Some older women, that is older than 25, were employed in the labour market, notably in textiles, and such employment could be critically important in poorer families. Other women worked from their homes and were not recorded in the employment figures; in Britain, as elsewhere, such activities included taking in laundry and brewing before they (especially brewing) became male-dominated industries (Hollingsworth and Tyyska 1988). There were hardly any careers leading to senior jobs for women, and top jobs were virtually all held by men. Such a situation could only change slowly.19
The expansion of education was not, for the most part, deliberately or directly related to gender. It did, however, have significant consequences for women's employment. In 1918 the school leaving age was raised to 14,20 then in 1945 to 15, and in 1974 to 16. On top of this there has been a growing tendency to stay on in education beyond the minimum leaving age; 37 per cent of 16 year olds and 28 per cent of 17 year olds stay on at school, while a further 34 per cent and 29 per cent respectively are in full-time further education (Department of Education and Employment 2000). There has also been a huge expansion in the numbers engaged in higher education. Even as late as 1938, just before World War II, less than 2 per cent of the relevant age group were in higher education and the proportion for women was less than 0.5 per cent (Blackburn and Jarman 1993). Today the proportion is about one third, with women outnumbering men.21
This expansion has removed from the labour market a very large proportion of women in what had been the main age range for employment. Consequently employers have had to recruit older women. Labour shortages during and after the Second World War led to the abandonment of the bar on employing married women that many firms had operated.22 The initial impact was for women to continue in employment after marriage before starting a family. The next stage was widespread employment of
Explaining gender segregation
529
mothers, often on a part-time basis (United Nations 1956; Stewart 1961). In this way employers maintained the number of working hours contributed by women, while a much larger proportion of women are now involved.23 Also these developments have meant that today the women in employment are generally older than in the past and likely to have family responsibilities.
Gradually women have caught up with and overtaken men at the different levels of education, most recently including graduates - a pattern that may be observed in a number of countries including Britain (Shavit and Blossfeld 1993). As women gained qualifications they were able to obtain good jobs, for instance in the expanding professions, on the basis of formal criteria which countered sexist discrimination, and so encouraged more women to obtain qualifications (Blackburn and Jarman 1993).
In a practical sense, it is now much easier for women to participate in the labour market. They are generally healthier than a hundred years ago, and they have fewer children, with major developments in contraception giving men and women greater control over the timing and number of children. Furthermore, women live longer and until quite recently increasing life-expectancy has meant they had a longer life of working age. At the same time there are many developments, from washing machines to fast food to disposable nappies, to make domestic work easier. As housework became easier, expectations and standards became higher and initially there was little or no saving in time. More recently, however, there has been a reduction in the time devoted to domestic labour (Bowden and Offer 1994).
Where women spent most of their adult working-age life in child-bearing and childcare, men had to be the main earners. Thus most women needed and expected their men to be the 'breadwinners'. Since hours in employment were long and the work was often arduous and unpleasant in the nineteenth and earlier years of the twentieth century, there were also those who considered such work was not suitable for women (Humphries 1977). However, there were women who did work, often out of necessity, and others who would have welcomed the opportunity. Many of these women were interested in challenging the dominant situation with its associated 'justification' of lower wages for women and preference given to men for available jobs. However, for those women who were in employment the position was not always straightforward, and there have been periods when there was disagreement about whether they should, on trade union principles, work for the same wage rates as men, or whether they should undercut men's rates to secure jobs.
Work in the home, as well as in the labour market, has changed considerably and family dependence on the earnings of the man (with possible additions from older children) has greatly declined. As the dominance of the 'male breadwinner' model was increasingly challenged, during the latter part of the twentieth century, women gained steadily in employment. However, even without significant opposition, careers take time to develop. For instance, few new graduates, male or female, become managing
530        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
directors overnight, and men in senior positions have not rushed to step aside for the younger generation. Nevertheless, the extent and speed of changes that have taken place are quite remarkable.
Looking at changes in education, family life-styles and health, in conjunction with the transformation of the occupational structure, it is easy to see why there is substantial horizontal segregation. The near disappearance of teenage employment, compared to early in the last century in Britain, has certainly affected male employment patterns, but it has transformed female patterns. The large decline in domestic service has added to the effect. The family and social circumstances of the women in employment are now quite different to their counterparts a hundred years ago. As different parts of the occupational structure have declined or expanded, women have generally been more readily available in the growth areas, and horizontal segregation has continued.
The explanation of vertical gender segregation is somewhat less obvious. Here we need to recall that in terms of stratification there is no general male advantage. In both manual and non-manual employment men have been able to hold on to the better jobs, at least in comparison with women working part-time. However, holding on to old advantages has meant that men have been less successful in entering the expanding non-manual areas. There is a well-known tendency for young men to enter occupations that are newer and more desirable - at least in terms of career prospects - than the occupations of their elders. A similar pattern may be observed for the young women (now as well as or better educated than their male counterparts), though it remains to be seen whether this will eventually lead to similar career achievements.
In terms of pay we would expect men to have an advantage. Even when we characterize occupations by male pay levels, to control for age, length of service, hours worked and gender discrimination effects, we find a substantial differential between 'male' and 'female' occupations. Because women have largely entered the expanding areas, where declining ex-clusiveness has meant pay levels have declined, their pay is generally below that in occupations where men still predominate.24 Furthermore, the gender difference in pay is not a new phenomenon; women have been socialized to expect and to accept lower levels of pay, and equal pay legislation only provides for equality within their (and comparable) occupations. There may also be a gender element of choice; at least at higher occupational levels with men tending to prioritize pecuniary rewards while women may prefer socially worthwhile professions. Thus we see women have comparable levels of occupations in terms of general attractiveness (Cambridge Scale), yet inferior occupations in terms of pecuniary rewards. However, it is not clear how far this reflects genuine choice or simply constrained opportunities where professional qualifications give protection against discrimination.
In this rather brief discussion of the processes of the labour market over time, we have attempted to indicate how and why gender segregation in
Explaining gender segregation
531
employment exists today. Rather than applying some external theory (such as rational choice or patriarchy), in the traditional style of metaphysics, we have looked for scientific explanation in the ongoing developments of societies. We argue that segregation is best understood through the examination of social reproduction and changing gender relations. It is vital to recognize that all aspects of social and technical change contribute to the changing pattern of gender in employment.
(Date accepted: July 2002) Robert M. Blackburn
Social Science Research Group Cambridge University
Jude Browne Centre for Business Research Cambridge University
Bradley Brooks Statistics Canada Ottawa
JenniferJarman Department of Sociology and Social Anthropology
Dalhousie University
NOTES
1. Since the vertical dimension may entail advantage to either men or women, we adopt the convention of giving a positive value for male advantage - the usual and widely assumed situation, and a negative value for female advantage (Blackburn, Jarman and Brooks 2000; Blackburn, Brooks and Jarman 2001b).
2. Strictly speaking there can only be one vertical dimension. The Hakim scheme uses a set of parallel measures.
3. It will be seen that we are using the concepts of 'vertical' and 'horizontal' with their mathematical and commonsense meanings. This is a break with a common convention in segregation research, where 'horizontal' has been used to refer to overall segregation. Because overall segregation contains a vertical component, what has been termed 'horizontal' has often been used as a measure of (vertical) inequality.  We  suggest  this  usage is
extremely confusing, and recommend our usage, as illustrated in Figure I.
4. She combines them with biological theories of human nature. However, such theories have little relevance to present concerns (see note 12) and are not discussed here.
5. There is an extensive literature by these authors, and others. For an excellent overview statement see the entry for 'human capital' in The New Palgrave (Eatwell et al. 1987).
6. Some economists are critical of the narrow economistic conception of rational choice (e.g. Sen 1999) but, aswe have seen, to widen it makes it tautologically useless.
7. There has been some criticism that the conception of human capital is too narrow (Sen 1999) and a tendency to broaden the concept (e.g. Becker 1996). Indeed some economists have followed the sociological approach of taking account of
532        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
other variables besides money, human capital, along with personal preferences, may then be taken to determine whatever rewards are received. In effect this is a rejection of human capital theory as an explanation of pay differences. However, this approach does not appear to be usual, and despite our criticisms we consider human capital is quite well related to pay differences.
8. The argument is reminiscent of the sociological 'functional theory of stratification' (Davis and Moore 1945), which has been damningly criticized, notably by Tumin (1953, 1967), and is no longer accepted in sociology. However, variants of it are still to be found in popular accounts and in economics.
9. Bourdieu's (1986) notion of social capital, or capital that is made up of social obligations, networks and connections, is more satisfactory in that it explicitly recognizes that the existing structure of social and economic privilege itself contributes to lower human capital formation, but to some extent suffers from the same limitations.
10. A particular problem with the idea of rational choice is that the choices of employers and employees may well be at odds. This has been discussed as an economic version of the 'prisoner's dilemma', but with no adequate 'economic' solution (see Smith 2000 for a thorough discussion). The arguments appear particularly relevant to the employment of women.
11. It also contradicts the usual forms of human capital and rational choice theories, where segregation is seen as resulting from women's choice of less 'empowering' occupations.
12. Hakim draws on the work of Goldberg (1973, 1979, 1993), describing his theory as 'unassailable' (1996: 212) and 'cannot be refuted on the evidence' (2000: 282). However, although biology has some obvious relevance, it is a simple category error to try to explain the social in terms of biology. For instance, we could hardly take seriously a purely biological (or psychological) account of events on a football field, let alone the technological collaboration in launching a space capsule.
13. It is difficult to determine what lifestyles are more egalitarian. It may be argued that where the home-centred do actually stay at home they are enjoying equality (or advantage if the man does not like his work). There is a common error of confusing equality with identity - which would require men and women to do the same things. Furthermore, feminist writers, who tend to be in professional occupations, are apt to see equality from their perspective where equality coincides with identity.
14. There are also similarities with Gouldner's (1959) theory of'latent roles', relating to age, sex, etc., which are drawn on in interpreting work situations.
15. Even the name 'preference' theory is not really new; Blackburn and Mann (1979) used the name 'persistent preferences' to refer to the components of orientations theory.
16. These and other figures are taken from the relevant British census data. The 1911 figures were estimated by Bain (1970) in a way that makes them comparable with the 1971 figures.
17. When the major university expansion began in Britain, following the Robbins Report, there were far more qualified men than women to fill the new posts, and most of them are still there. As they retire we may expect women to fill many of the vacancies, though women graduates are still a minority in some subjects.
18. At the present time employment in domestic service is increasing again, though still much less than in the past and in a quite different form, with a lower proportion of live-in servants (Anderson 2001; Gregson and Lowe 1994).
19. The transformation of women's labour market participation was even greater in the LSA. In 1900 only 20.6% of women over 14, and only 5.6% of married women were in the labour market (Reskin and Roos 1990: 10), while the proportion of women, aged 16 and over, had risen to 58% by 1992 and 60% in 2002. Of the women in employment in 1992 and 2002 over half were married - 56% and 54% respectively (Bureau of Labor Statistics 2002).
20. The 1921 census nevertheless recorded occupational data for children of
Explaining gender segregation
533
12 and over. (The previous census had started at age 10.) However, the numbers were relatively small below the official leaving age.
21. The estimate is based on figures from Department of Education and Employment (2000) for the UK. It is confined to those studying full-time and so not normally available for regular employment; most are attending universities. Since students enter full-time higher education at different ages, and remain for varying periods (some drop out, others continue to higher degrees) it is not possible to give a precise figure for the impact on availability for employment.
22. For instance, previous to the abandonment of the marriage bar, the Cadbury company gave three presents to women who got married: a bible, a white carnation, and the sack (dismissal). All three 'gifts' were in keeping with their policy of caring paternalism, as understood at the time. Men who married might risk dismissal for rather different reasons. Thus the clearing banks would not allow male staff to marry until they were deemed to earn enough to keep a wife, and so free from the temptation to steal from their employer.
23. Women's share of the labour force has increased from 30% at the beginning of the last century to roughly 45% by the end of the century.
24. For details of male and female pay by occupation see the New Earnings Survey (2002).
BIBLIOGRAPHY
Anderson, B. 2001 'Why Madam Has So Many Bathrobes: Demand for Migrant Domestic Workers in the EU', Tijdschrift voor Economiche en Soziale Geografie 92(1): 18-26.
Anderson, D. and Shapiro, D. 1996 'Racial Differences in Access to High-Paying Jobs and the Wage Gap between Black and White Women', Industrial and Labor Relations Review 49 (2): 273-86. Anker, R. 1998 Gender and Jobs, Sex Segregation of Occupations in the World, Geneva: International Labour Office.
Ashton, D., Maguire, M. and Spilsbury, M.
1990 Restructuring the Labour Market: The Implications for Youth, London: Macmillan. Bain, G. S. 1970 The Growth of White-Collar Unionism, Oxford: Clarendon Press. Barrett, M. 1988 Women's Oppression Today, London and New York: Verso [first published 1980].
Becker, G. 1964 Human Capital: A Theoretical and Empirical Analysis with Special Reference to Education, New York: National Bureau of Economic Research. --1991 A Treatise on the Family, Cambridge Mass.: Harvard University Press.
--1996 Accounting for Tastes, Cambridge
Mass.: Harvard University Press. Beynon, H. and Blackburn, R. M. 1972 Perceptions of Work, London: Cambridge University Press.
Blackburn, R. M. 1988 'Understanding unemployment: the need for a social perspective', The International Journal of Sociology and Social Policy 8(1): 1-22.
-- 1999 'Is Housework Unpaid Work?',
International Journal of Sociology and Social Policy 19(7/8): 1-20. ' Blackburn, R. M. and Jarman, J. 1993 'Changing Inequalities in Access to British Universities', Oxford Review of Education 19(2): 197-215. '
Blackburn, R. M. and Jarman, J. 1997 'Occupational Gender Segregation', Social Research Update 16(Spring) [http:// www.soc.surrey.ac.uk/ sru/SRU16/SRU16. html].
Blackburn, R. M. and Mann, M. 1979 The Working Class in the Labour Market, London: Macmillan.
Blackburn, R. M., Brooks, B. and Jarman, J.
1999 Gender Inequality in the Labour Market: The Vertical Dimension of Occupational Segregation, Cambridge Studies in Social Research 3, Cambridge: Cambridge University, SRG Publications.
--2001 a The Gendering of Work Around the
World: Occupational Gender Segregation and Inequality (Report to the United Nations on International Occupational Gender Segregation, 1999), Cambridge Studies in Social Research 9, Cambridge: Cambridge University, SRG Publications.
-- 2001b 'The Vertical Dimension of
Occupational Segregation', Work, Employment and Society 15(3): 511-38. Blackburn, R. M., Jarman, J. and Brooks, B.
534        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
2000 'The Puzzle of Gender Segregation and Inequality: A Cross-National Analysis', European Sociological Review 16(2): 119-35.
Blau, F. and Khan, L. 1992 'The gender earnings gap: learning from international comparisons', American Economic Review 82(2): 533-8.
Bourdieu, P. 1986 'The Forms of Capital' in J. G. Richardson (ed.) Handbook of Theory and Research for the Sociology of Education, New York: Westport, London: Greenwood.
Bowden, S. and Offer, A. 1994 Household Appliances and the Lse of Lime; the United States and Britain Since the 1920s', Economic History Review 47(4): 725-48. Browne, J. M. 2000a Vertical Occupational Gender Segregation in the British Labour Market,   Cambridge  Studies  in Social Research 7, Cambridge: Cambridge University, SRG Publications. -2000b Vertical Occupational Gender Segregation in the British Labour Market, PhD, Cambridge University. Bureau of Labor Statistics 2002 Statistics of Employment, Washington: US Department of Labor.
Crompton, R. and Harris, F. 1998a 'Explaining women's employment patterns: "orientations to work" revisited', British Journal of Sociology 49(1): 118-36.
-- 1998b 'A reply to Hakim', British
Journal of Sociology 49 (1): 144-9. Davis, S. K. and Moore, W. E. 1945 Some Principles of Stratification', American Sociological Review X(2): 242-9. Delphy, C.  1977  The Main Enemy: A Materialist Analysis of Women's Oppression, London: WRRC Publications. -1984 Close to Home: A Materialist Analysis of Women's Oppression, London: Hutchinson.
Department of Education and Employment 2000 Education and Training Statistics for the United Kingdom, 2000, London: Stationery Office.
Dex, S. 1988 'Gender and the Labour Market', in D. Gallie (ed.) Employment In Britain, Oxford: Basil Blackwell. Eatwell, J., Milgate, M. and Newman, P.
1987 The New Palgrave: A Dictionary of Economics, London: Macmillan.
England, P. 1979 'Women and Occupational Prestige; A Case of Vacuous Sex Equality', Signs 5: 252-65. Fox, J. and Suschnigg, C. 1989 A Note on Gender and the Prestige of Occupations', Canadian Journal of Sociology 14(3): 353-60. Franks, S. 1999 Having None of It, London: Granta.
Gallie, D. (ed.) 1988 Employment in Britain, Oxford: Basil Blackwell. Goldberg, S. 1973. The Inevitability of Patriarchy, New York: William Morrow.
-- 1979 Male Dominance: The inevitability
of patriarchy, London: Abacus Sphere Books.
--1993 Why Men Rule: The Theory of Male
Dominance, Chicago: Open Court. Gouldner, A. 1959 'Organisation Analysis', in R. K. Merton, L. Broom and L. S. Cotrell Jr. (eds), Sociology Today: Problems and Prospects, Vol. II, New York: Harper and Row.
Gregson, N. and Lowe, M. 1994 Servicing the Middle Classes; Class, Gender and Waged Domestic Labour, London: Routledge. Guardian, The 8 August 1998. Hakim, C. 1979 Occupational Segregation: A Study of the Separation of Men and Women's Work in Britain, the United States and Other Countries, Research Paper 9, London: Department of Employment.
--1996 Key Issues in Women's Work:Female
Heterogeneity and the Polarisation of Women's Employment, London: Athlone.
--1998a 'Developing a sociology for the
twenty-first century: Preference Theory', British Journal of Sociology 49(1): 137-43.
-- 1998b Social Change and Innovation in
the Labour Market, Oxford: Oxford University Press.
-- 2000 Work-Lifestyle Choices in the 21st
Century: Preference Theory, Oxford: Oxford University Press.
Hamilton, R. 1978 The Liberation of Women: A Study of Patriarchy and Capitalism, London: Allen and Unwin. Harris, C. 1987 Redundancy and Recession in South Wales, Oxford: Blackwell. Hartmann, H. 1976 'Capitalism, Patriarchy and Job Segregation by Sex', Signs 1: 137-69.
Hollingsworth, L. andTyyska, V. 1988 'The hidden producers; Women's household
Explaining gender segregation
535
production during the great depression', Critical Sociology 15(3): 3-29. Humphries, J. 1977 'Class Struggle and the Persistence of the Working Class Family', Cambridge Journal of Economics 1 (3): 241-58. - 1988 'Protective legislation, the capitalist state and working class men: the case of the 1842 Mines Regulation Act', in R. Pähl (ed.) On Work: Historical, Comparative and Theoretical Approaches, Oxford: Blackwell.
Joshi, H. and Paci, P. 1998 Unequal Pay for Women and Men: Evidence from the British Birth Cohort Studies, London: MIT Press. Krahn, H. and Lowe, G. 1988 Work, Industry and Canadian Society, Scarborough: Nelson Canada.
Middleton, C. 1983 'Patriarchal exploitation and the rise of English Capitalism', in E. Gamarnikow, D. Morgan, J. Purvis and D. Taylorson (eds) Gender, Class and Work, London: Heinemann. Mincer, J. 1966 'Labor Force Participation and Unemployment: A Review of Recent Evidence', in R. Gordon and M. Gordon (eds) Prosperity and Unemployment, New York: Wiley.
Mincer, J. and Polacheck, S. 1974 'Family Investment in Human Capital: Earnings of Women', Journal of Political Economy 82: Part 2, S76-S108.
New Earnings Survey 2002, London: ONS. Pähl, J. 1983 'The allocation of money and the structuring of inequality within marriage',   The  Sociological  Review 13(2): 237-62.
-- 1989 Money and Marriage, Basingstoke: Macmillan.
Parsons, T. and Shils E. A. (eds) 1951 Towards a General Theory of Action, New York: Harper and Row. Polachek, S. 1975 'Discontinuities in Labor Force Participation and its Effects on Women's Market Earnings', in C. B. Lloyd (ed.) Sex Discrimination and the Division of Labor, New York: Columbia University Press.
-1981 'Occupational Self-Selection: A
Human Capital Approach to Sex Differences in Occupational Structure', Review of Economics and Statistics (Feb.): 60-9. Purcell,   K.   1988   'Gender   and the Experience of Employment', in D. Gallie
(ed.) Employment in Britain, Oxford: Black-well.
Reskin, B. F. and Roos, P. A. 1990 Job Queues, Gender Queues, Philadelphia: Temple University Press. Sen, A. 1999 Development as Freedom, Oxford: Oxford University Press. Shavit, Y. and Blossfeld, H. P. 1993 Persistent Inequality. Changing Educational Attainment in Thirteen Countries, Boulder: Westview Press.
Siltanen, J. 1994 Locating Gender, London: UCL Press.
Smith, M. 2000 'On the Use of the Prisoner's Dilemma to Analyze the Relations Between Employment Security, Trust, And Effort', Review of Social Economy LVIII(2): 153-75.
Stewart, A., Blackburn, R. M. and Prandy,
K. 1985 'Gender and Earnings: The Failure of Market Explanations', in R. Finnegan, D. Gallie, and B. Roberts (eds) New Approaches to Economic Life - Economic Restructuring: Unemployment and the Social Division of Labour, Manchester: Manchester University Press. Stewart, C. M. 1961 'Future Trends in the Employment of Married Women', British Journal of Sociology 12(1): 1-11. Thomas, W. I. and Thomas, D. S. 1928 The Child in America, New York: Alfred A. Knopf.
Treiman, D. and Hartmann, H. (eds) 1981 Women, Work and Wages, Washington, DC: National Academy Press. Tumin, M. W. 1953 'Some Principles of Stratification: a critical Analysis', American Sociological Review 18 (4): 387-94.
-- 1967 Social Stratification, Englewood
Cliffs: Prentice-Hall.
United Nations 1956 Demographic Aspects of Manpoweer, Report 1, New York: United Nations.
--2001 Human Development Report 2001,
Oxford and New York: Oxford University Press, for the United Nations Development Programme.
Vaiou, L. 1999 Women's employment in Greece: Their participation in the labour market, occupational gender segregation and the role of the family, MPhil dissertation, Social and Political Science Faculty, Cambridge University.
536        Robert M. Blackburn, Jude Browne, Bradley Brooks and JenniferJarman
Walby, S. 1986 Patriarchy at Work, Cambridge: Polity Press.
-- 1990 Theorizing Patriarchy, Oxford:
Basil Blackwell.
World Bank 1994 Enhancing women's participation in economic development, Washington.
Young, M. and Willmott, P. 1957 Family and Kinship in East London, London: Roudedge and Kegan Paul.
Copyright of British Journal of Sociology is the property of Routledge, Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.