LSE the London School of Economics and Political Science ■ Explaining Women's Employment Patterns: 'Orientations to Work' Revisited Author(s): Rosemary Crompton and Fiona Harris Source: The British Journal of Sociology, Vol. 49, No. 1 (Mar., 1998), pp. 118-136 Published by: Blackwell Publishing on behalf of The London School of Economics and Political Science Stable URL: http://www.jstor.org/stall in ^^^^^^^^^^^^^^^^^^^^^^^^^^M Accessed: 22/03/2011 15:19 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/action/showPublisher?publisherCode=black. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Blackwell Publishing and The London School of Economics and Political Science are collaborating with JSTOR to digitize, preserve and extend access to The British Journal of Sociology. STOR http://www.jstor.org Rosemary Crompton and Fiona Harris Explaining women's employment patterns: 'orientations to work' revisited* ABSTRACT Explanations of the persisting differences in the structure of men's and women's employment have long been debated in the social sciences. Sociological explanations have tended to stress the continuing significance of structural constraints on women's employment opportunities, which persist despite the removal of formal barriers. Neo-classical economists, in contrast, have emphasized the significance of individual choice, an argument which has been recently endorsed by Hakim who suggests that patterns of occupational segregation reflect the outcome of the choices made by different 'types' of women. In this paper, a previous debate relating to the explanatory utility of men's 'orientations to work' is used to argue that employment structures are the outcome of both choice and constraint, and that this is the case for women, as well as men. The argument is illustrated with evidence from cross-nationally comparative biographical interviews carried out in five countries. KEYWORDS: women; employment; occupational segregation; cross-national; work orientations INTRODUCTION It has recently been claimed that women's relatively disadvantaged position in the labour market is not a consequence of the institutional and/or structural disadvantages they suffered. Rather, women's position reflects the outcome of their varying choices. Hakim (1991; 1995; 1996), argues that there are two 4qualitatively different' types of working woman, the 'committed' and the 'uncommitted', the former giving priority to their employment careers, the latter to their domestic responsibilities. 'Committed' women work full-time, 'uncommitted' women work part-time. The existence of these different orientations to employment, Hakim argues, explains the apparently contradictory finding that part-time workers express themselves as highly 'satisfied' with their low-level, poorly paid, employment. Furthermore, Hakim argues that 'feminists' have deliberately perpetrated Brit.Jnl of Sociology Volume no. 49 Issue no. 1 March 1998 ISSN 0007-1315 © London School of Economics 1998 Explaining women's employment patterns 119 'fashionable but untrue' ideas concerning women's employment preferences (and related employment patterns) for overtly 'political' reasons, that is, in order to '. . .effectively dictate ... a narrow range of acceptable conclusions . . . that women are victims who have little or no responsibility for their situation' (1995: 448). In this paper, we will not seek to deny that women make choices, and that these choices are reflected in aggregate patterns of employment amongst women. We will, however, dispute the assertion that variations in women's 'orientations to work' (or 'choice') is the major independent variable explaining women's employment patterns. Rather, we shall argue that women's employment behaviour is a reflection of the way in which women actively construct their work-life biographies in terms of their historically available opportunities and constraints. This argument will be developed using evidence from a cross-national study which has gathered biographical interviews from women in the same occupations in five different countries. A previous variant of the 'orientations to work' debate, which was developed in respect of men's employment, will be examined in order to make some more general points relating to the deployment of individualist, voluntarist explanations within sociology. First, however, we will examine Hakim's arguments. WORK ORIENTATIONS AND WOMEN'S EMPLOYMENT CHOICES Hakim argues that, when measured by hours of work, rather than numbers of jobs, the extent of women's employment has not risen since the Second World War. She argues that childcare problems are not a barrier to women's employment, and that part-time workers are not exploited. Rather, as 'uncommitted' workers, they have chosen the flexibility of hours etc. associated with part-time work and express considerable satisfaction with their employment arrangements. Finally, women are more unstable employees than men, a fact which has been deliberately obscured by the practitioners of the feminist orthodoxy. The thrust of Hakim's argument is that the nature and pattern of women's labour force participation is largely a consequence of women's choices, and that the heterogeneity of women's employment statuses reflects the heterogeneity of female choice. 'Feminists', she argues, have claimed that women's apparent lack of commitment, job stability and so on is a reflection of the jobs they have been forced into because of patriarchal pressures. These jobs tend to be insecure, characterized by high rates of turnover, with few career prospects, and so on, and these characteristics are reflected in women's employment behaviour. Far from it, Hakim argues, women's lack of commitment, job instability, etc., accurately reflects the labour market behaviour of women whose employment is secondary to their domestic involvement. 120 Rosemary Crompton and Fiona Harris The deliberately provocative tenor of Hakim's recent commentaries has generated a number of critical replies, with which we would broadly concur (Ginn et al. 1996; Bruegel 1996). We would agree that she has set up straw feminists in constructing her arguments, and that it would be difficult to find academic 'feminists' (none are actually identified by Hakim) who have claimed either that 'women' are an undifferentiated mass in respect of their employment preferences, or that there are no differences at all between men and women as employees. We would also agree that some of her empirical procedures are highly questionable - for example, treating hours worked as being more significant than numbers of jobs, and including a population group that was not actually asked a question in calculating an averaged 'response' to it.1 Our purpose here, however, is not to further extend these criticisms of the details of Hakim's arguments, but to challenge, sociologically, the grounds upon which she makes her case. Hakim's argument moves directly from the macro to the micro level. Women in part-time employment, we are told, have 'chosen' to give priority to a marriage career, and no account is given of the mechanisms whereby this 'choice' was arrived at.2 Whilst asserting that 'Some women choose to be home-centred, with work a secondary activity', and 'Some women choose to be career-centred, with domestic activities a secondary consideration' (1996: 186), Hakim simultaneously holds that 'some women will switch between groups over their lifetime', thus contriving to have the argument all ways at once. Again, no suggestion is given as to why 'some women' might choose to change categories rather than remain in one or the other. The existence of these two types of women lends support to both rational choice and human capital theories, argues Hakim. 'Uncommitted' women make a rational decision to economize on the effort invested in employment, as this is not their main priority. In contrast, 'committed' women, in line with the prescriptions of human capital theory, will choose to invest in their employment careers. This fact of heterogeneous female preferences provides a link between psychological theories of male dominance and the concept of patriarchy (1996: 212). Goldberg (1973; cited in Hakim 1996: 5) argues that hormonal differences between men and women make men more 'self-assertive, aggressive, dominant and competitive'. The fact that women are fundamentally divided within themselves, Hakim argues, serves to amplify the effect of these 'natural' masculine characteristics and men are, as a consequence, disproportionately successful. We have said that Hakim provides no reasons for the existence of these two types of women, or why they might switch from one category to another.3 WTe would suggest that the reasons for these choices and changes lie in the exigencies of context and structural constraint which Hakim effectively disregards in her embrace of voluntaristic, rational-choice explanations of women's economic behaviour. To be sure, women can and do make choices - although in aggregate, their relative lack of power and resources relative to men means that both today and in the past, they have been less able to do so than the opposite sex. Women - and men - can Explaining women's employment patterns 121 choose but are also constrained, a fact which lies at the root of sociological explanations of human behaviour. However, Hakim (1991: 114) has argued that: '. . . theory and research on women's employment seems particularly prone to an over-socialised view of women, or with structural factors so weighted that choice flies out of the window'. This suggestion of structural over-determinism is somewhat paradoxical given the recent turn to 'discourse' in feminist debates. This trend has led some theorists to suggest that recent feminist analyses run the danger of disregarding structure altogether (Maynard 1995). Our own position is similar to that of Marshall (1994: 115), who has argued that although '. . . the content of gender is infinitely variable and continually in flux, . . . the salience of gender (i.e., 'male' and 'female') categories is persistent'. Thus gendered structures and categories - in employment, in families, in state institutions - play a major part in reproducing the gender order, but these structures are negotiated and interpreted by changing and flexible gendered subjects. Both the structures, and the manner in which they are interpreted, may be investigated empirically. One structure which has been argued to have been an important mechanism through which women have been subordinated is that of employment. Walby (1986) and Hartmann (1982) have argued that women have been deliberately denied access to jobs which would allow them to live independendy. In contrast, Hakim argues that Occupational segregation has been reconstructed in the late twentieth century to provide separate occupations and jobs for women following the marriage career, which allows only non-committed contingent work and non-career jobs which are always subordinate to non-market activities. (1995: 450) However, this explanation does not address the problem of disparities in power and resources between groups in the labour market. Rather, the status quo is described as being a reflection of the requirements of a population differentiated by 'choice' alone, rather than by any variations in their initial social and material endowments (e.g. material and social capital contributing to employment opportunities). As far as men and women are concerned, it would be difficult to argue that this has been or is the case. Hakim (1995: 450) also asserts that'. . . treating the workforce as a homogeneous group may work well for research on male employment'. This statement leaves out of account a body of research and theory in industrial sociology whose major conclusions were to demonstrate the heterogeneity of the male workforce. This was the 'orientations to work' debate in the Industrial Sociology of the 1960s and 70s. In the next section of this paper, therefore, we will first briefly review this debate, with the purpose of drawing out elements relevant to Hakim's discussion of women's employment. We will then present some findings from a cross-national study in order to illustrate the interaction between choice and constraint in the shaping of women's decisions relating to the employment/family interface. 122 Rosemary Crompton and Fiona Harris 'ORIENTATIONS TO WORK' The conceptualization of 'orientations to work' in Industrial Sociology arose out of an apparent empirical paradox. In the 1960s, assembly line work (particularly in the car industry) was widely regarded as boring, stressful and as giving rise to high levels of industrial conflict. However, a study of car workers in Luton (undertaken as part of the 'Affluent Worker' study, see Goldthorpe 1966; Goldthorpe et al. 1968) found that car assembly workers expressed considerable satisfaction with their employment, and manifested only low levels of industrial conflict. Goldthorpe et al. explained these surprising results by emphasizing the significance of the worker's 'prior orientations' to employment. The assembly line workers, they argued, had markedly 'instrumental' orientations to work. Putting an overwhelming emphasis upon material returns (with which they were content), their calculative involvement in employment left no room for individual self-realization and their major satisfactions were to be found outside of, rather than within, the workplace. Like Hakim's part-time workers (or 'grateful slaves' (1991)) the explanation of their apparent satisfaction with 'objectively' unsatisfactory employment was to be found in their orientations to work rather than in their experience of work itself: '. . .the orientation which workers have to their employment and the manner, thus, in which they define their work situation can be regarded as mediatingbetween features of the work situation objectively considered and the nature of workers' response' (Goldthorpe et al. 1968: 182). The Luton study was influential in developing the 'social action' approach within industrial sociology, in which the actors' . . own definitions of the situations in which they are engaged are taken as an initial basis for the explanation of their social behaviour and relationships' (Goldthorpe et al. 1968: 184). This approach was developed in contrast to what were perceived to be the universalist and over-determined assumptions of 1960s industrial sociology; in particular, the 'technological determinism' and 'human relations' approaches. In respect of 'technological determinism', authors such as Blauner (1964) had argued that worker satisfaction and associated behaviour varied according to different levels of technological development, and that assembly line technology, such as in the car industry, was the most 'alienating' form of work.4 'Human relations' theorists had argued that workers sought primarily social satisfactions in the workplace, and that the key to workplace satisfaction, therefore, was the development of cohesive work groups and supportive supervisor-worker relationships (Roethlisberger and Dickson 1947). In contrast, Goldthorpe et al., as we have seen, argued that to the extent that workers could choose the nature of their employment, then workers with particular orientations would tend to cluster in particular workplace situations, and that therefore 'orientations to work' should be treated as an . . important independent variable relative to the in-plant situation' (1968: 183). Explaining women's employment patterns 123 These arguments were developed in respect of the male labour force, but the parallels with Hakim's reasoning as to the existence of (at least) two types of wo men worker, characterized by different 'orientations to work', are very apparent. Goldthorpe et al.'s work generated an extensive debate, as well as further empirical studies, which are simply too numerous to review and summarize at any length here. The Luton studies, together with the development of the 'action' approach, also made an important contribution to the development of sociological theory.5 A major factor giving rise to the 'instrumental' orientation to work, according to Goldthorpe et al., was stage in the family life cycle. Men with young families, and non-working wives, were most likely to give priority to extrinsic returns from employment. However, the 'orientations' debate did not pursue this topic of the de facto intertwining of market work and non-market responsibilities and their relationship to the family life cycle, but rather, polarized into a dispute between protagonists arguing about the relative significance of workplace ('structural') and non-workplace ('action') factors to explanations of attitudes to and behaviour in work (see in particular Goldthorpe (1972) and Daniel 1969; 1971).6 The notion of 'orientations' was held up to close scrutiny. It was argued that orientations were complex and multi-stranded, rather than single-stranded. Many workers were found to desire both extrinsic and intrinsic rewards from employment and thus no single 'orientation' could be identified (Hill 1976; Blackburn and Mann 1979). Second, it was emphasized that individual orientations to work were variable and subject to change, both by context and over the life cycle. Daniel (1969), in particular, emphasized the contextual significance of orientations: in a (wage) 'bargaining' context, workers manifested highly 'instrumental' orientations, whereas in a 'work' (i.e., day to day) context, workers' orientations were more focused upon 'intrinsic' factors (see also Wedderburn and Crompton 1976, Cot-grove et al. 1971). Finally, it was argued more generally that studies of worker's attitudes had demonstrated that people have a tendency to adapt to what is realistically available for them and adjust to the realities of their employment situation (Blackburn and Mann 1979). This argument may be extended to cross-national comparisons, which have similarly demonstrated the significance of particular national contexts for worker attitudes and behaviour (Ingham 1974; Gallie 1978). Hakim's arguments, as we have seen, focus on the significance of 'prior orientations' for women's employment patterns, and may be criticized in a similar fashion. Much as male employees were found to desire both 'intrinsic' and 'extrinsic' rewards from their work, so women may be shown to desire both 'employment' and 'family' careers.7 Women's (and men's) 'work commitment' will vary by Hfe cycle stage, rather as men's 'orientations' were shown, in the 1970s, to vary depending on workplace context. And women's, as well as men's, employment-related attitudes and behaviour will vary depending on occupation-specific and national contexts. In 124 Rosemary Crompton and Fiona Harris the discussion which follows, we will illustrate these themes by drawing upon cross-national research which will demonstrate not only the manner in which women's work commitment is constructed over the family/employment life cycle, but also how significant external changes have reshaped orientations and commitment to fit changing circumstances. COMMITMENT AND CONTEXT The research reported in this paper was designed in order to explore the complex linkages between the changing system of gender relations and the structuring of women's employment through a comparative analysis which included five countries (Britain, Norway, France, Russia and the Czech Republic).8 We have sought to identify and describe relevant structural factors which, so to speak, offer a 'gendered template' to actors. It has been demonstrated that at the macro level, national differences in respect of important institutions - in particular, the development of the welfare state, family (reproductive) policies, and the approach to the liberal 'equality agenda' instantiated by 'first-wave' feminism - have had a significant and enduring impact on attitudes to gender roles and women's employment (Crompton and Harris 1997a). We have also demonstrated that these national variations in attitude are linked to behavioural differences in respect of the domestic division of labour (Crompton and Harris 1997b). Through an analysis of two 'feminizing' occupations, medicine and banking, we have also demonstrated that the occupational structure also plays an important role in shaping relatively more, or less, stereotyped gender identities. As we shall see, women doctors, in contrast to bankers, had tended to be more systematic in the construction of their work-life biographies, and this was reflected in their domestic lives (Crompton and Harris 1998). At the same time, we have, through our biographical interviews9 (fifteen with women in each occupation in each country), also focused upon the actors accounts and experiences of gendered structures - particularly in regard to paid employment and family life. Our interviews demonstrate that the continuities revealed in our 'structural' analyses are complemented by extensive difference at the level of the individual, indicating that women do, indeed, 'work on' their lives and shape their biographies in relation to their perceived possibilities. These possibilities vary relationally as well as culturally and historically (e.g. between East and West; Scandinavian and liberal democratic. See Crompton 1996; Crompton and Harris 1997a). Table I demonstrates that there were important differences in the employment/family biographies of the women in the two occupations studied. We may link these occupational differences to an ideal-typical contrast between 'professional' and 'managerial' career trajectories, and their interaction with the family life cycle. It can be seen that doctors are more likely Explaining women 's employment patterns 125 TABLE I: Occupational variations in career intentions, children and the domestic division of labour Doctors Bankers Early career/family intentions:** Career first/drifted 51% 78% Shaped by anticipated family responsibilities 49% 22% N 75 77 Children:** None/one child 42% 65% More than one child 58% 35% N 76 78 Domestic division of labour:* Traditional 75% 55% Other- 25% 45% N 69 68 Childcare: Respondent main responsibility 55% 36% Other- 45% 64% N 56 53 Notes: ** significant @ 99% * significant @ 95% ~ includes shared with partner, partner does most, moderate to extensive help from others (paid/relatives) to make career choices (i.e. of specialty) enabling them to combine employment and family life (e.g. General Practice, which offers regular hours. See Crompton and Le Feuvre 1997). Professionally-trained women often adopt 'family friendly' strategies which enable them to continue in professional practice (Crompton and Sanderson 1990). However, at the individual level, these kinds of biographies reproduce a relatively conventional gender division of labour, which feeds through to the occupational level, as manifest in gender segregation within the medical profession. In contrast, bankers (managers) are more likely to have made their career decisions at a later stage: A characteristic story is told by this Norwegian banker I was the youngest of three children and my parents were most concerned with my brother getting an education, that was (seen as) more natural ... I was young when I began at the bank. I would have done it differently now. I would never have begun at the bank before I was finished with my education ... it was a safe job ... at that time it was common to marry early and start working early (she had her children at age 21, 22) ... I was pretty unconscious when I chose the bank . . .(3/29). Women in banking often find it difficult to make 'family friendly' working arrangements, and Table I shows that they have fewer children, and are more likely to have a less stereotyped gender division of labour. At the individual level, therefore, their biographies are more likely to generate change 126 Rosemary Crompton and Fiona Harris TABLE II: Occupational variations in the interface between work and family life (percentages) Doctors Bankers Domestic life first 24 15 Satisficer 37 26 Maximize r 17 15 Careerist (by choice) 16 19 Careerist (by necessity) 15 Unclear/undecided 5 9 Number: 75 78 - although men continue to dominate within the organizational structure of banking. However, the biographies also revealed considerable heterogeneity, as described in Table II which summarizes the current work/life combinations of the women interviewed.10 In Hakim's terms, all of the women were committed to an employment, rather than a domestic or marriage, career. All had worked continuously, with only short breaks for childrearing. Nevertheless, despite the considerable sacrifices and investments which these women had made in developing their employment careers, most were explicit that they wanted to combine employment with family life. The 'orientations to work' debate emphasized that men sought both 'intrinsic' and 'extrinsic' satisfactions from employment, and in a parallel fashion, the majority of the women interviewed indicated that they, too, had multi- rather than single-stranded orientations. The concept of 'satisficing' has been developed by Chafetz and Hagan (1996) in their analysis of the growing impact of women's employment on family life. Increasingly, they argue, women will attempt to achieve success in both employment and family life goals without maximizing either. Table II shows that 'Satisficing' represents the largest single category of women interviewed. However, other women, whilst realizing both goals, had definitely given priority to their domestic lives ('Domestic lifefirst') I decided after I was married and doing house jobs that I didn't want to stay in the rat race that was hospital medicine . . .'cos my husband was doing that ... so I decided that as I wanted to have a family I would become a GP so that I could work part-time (Britain: 2/02). Others, rather than 'satisficing', had refused to compromise and had sought to maximize their goals in respect of both employment careers and family lives ('Maximizer'). They included a Norwegian banker (3/33) who scrupulously shared domestic labour with her husband (for example, they had each taken six months maternity leave for each of their three children), and had risen to a Directoral position by her early thirties. 'Domestic', Explaining women 's employment patterns 127 TABLE III: Occupational changes in work-life biographies (percentages) Doctors Bankers Family life course change External or organizational Shock/change 43 4 15 35 Personal life (crisis) change 26 3 25 14 17 19 Self-development change No change Number: 73 78 'Satisficer' and 'Maximizer' biographies all reflected multi-stranded work-life objectives, in contrast to 'careerists' - that is, women who consciously put their employment careers before their domestic lives, as in the traditional 'career woman' stereotype. Many of these women did not have children, although in some cases, children had been brought up by other relatives. Other careerists (who have been coded separately), had adopted an 'employment first' strategy more out of necessity than choice - after divorce or other personal and economic crises such as the collapse of state socialism in Eastern Europe. As can be seen from Table II, all 'careerists by necessity' were bankers (it is simply not possible to decide that a career in medicine is the answer to pressing financial difficulties!). Even in similar circumstances, therefore, it is evident that women make very different choices in the way in which they approach their employment and family lives. The women's biographies, therefore, demonstrate that their orientations to employment and family life were complex and variable. Some women choose a career with the clear intention of combining employment and family life, others do not. We have shown how orientations can change and develop over time - as in the examples of the bankers who went into employment with no clear intentions but nevertheless built successful careers. Table III summarizes the 'switching points' which had changed the work-life careers of individual women. Doctors are far more likely to have altered their work-life biographies as a response to changes over the family life cycle - for example, taking up part-time work when children are young. Bankers, however, are more likely to have changed in response to external 'shocks' - such as organizational restructuring, or economic crisis (personal life crises have been classified separately). In a similar vein, our interviews demonstrated that the experience of work itself had played an important part in stimulating changes in both paid employment and the domestic division of labour. For example, a British doctor, married and working overseas, had returned to train in a 'woman-friendly' area but decided to enter a highly competitive specialty because she 'just got interested' whilst working on a project to earn extra money. A banker had spent three years spent at home with her children 128 Rosemary Crompton and Fiona Harris My focus changed completely then to being a mum. Probably my mum's early indoctrination of me as a mother came through in that I'd never had any intentions of returning from maternity leave immediately. I'd always got it set in my mind that I wanted to be at home with my children . . . [And planned to return to work part-time. However, with bank reorganization, a new job was offered.] . . . Do I take the risk of forfeiting the extra two years home with the family, (or) to take this full time job which was a very good promotion opportunity? (she did) (Britain: 2/26). Not only do employment orientations and commitments, but also orientations to the domestic division of labour, change over the life cycle. Many women had renegotiated the domestic division of labour as their careers had developed - or had changed their partners for something more 'career-friendly'.11 In the case of the bankers in particular, 40 per cent reported problems relating to the domestic division of labour which had resulted in the end of a relationship. Often (but not always) there had been a change to a new and more 'domesticated' partner as a consequence . . . my first managerial job demanded a lot from me so I started putting more into the job ... I got more confidence - I took that home. I wasn't the wife he married ... It (domestic work) was all my responsibility, I did everything. Initially I didn't mind much . . . but as I got more mature and more confident I started to question why was all this my responsibility . . . I always said 'when I get married again I'm going to do it differently' (divorced; remarried, second marriage egalitarian, one child) (Britain: 3/21). Changes in the domestic division of labour were reflected in occupational differences. Women in banking (who were more likely to have experienced status changes over the employment life cycle) were more likely to have renegotiated the domestic division of labour (18 per cent of bankers had moved from traditional to egalitarian relationships with the same partner, as compared to 9 per cent of doctors). Exogenous changes - such as, most dramatically, the collapse of state socialism in Eastern Europe - may lead to the renegotiation of the domestic division of labour, as in this case of a Russian bank employee12 Now my husband helps me more about the house. The thing is that over the last four years I am the only supporter to the family. . . . when I began working at the bank I told him that (the) work was new to me, and I needed a lot of time. ... in order to acquire the necessary professional skills. I also told him that since he did not have much professional work to do at the time he would have to perform most of the family duties. He understood it and he accepted it. Well, at first we had serious conflicts about it, but now his attitude to the problem is much more quiet (4/28, one child). Explaining women's employment patterns 129 We would suggest that this evidence of change in orientations amongst women, both as a consequence of employment experiences as well as over the employment/family life cycle, raises problems for Hakim's arguments. If orientations and attitudes to employment and the domestic division of labour change over time and by context, how can the women who hold them be described as 'qualitatively different' from each other? Can part-timers usefully be characterized as 'uncommitted' if they regard their status as a temporary one? (the latter circumstance, of course, would not preclude a female employee from expressing her 'satisfaction' with part-time work during her tenure of it). Although we have stressed the fluidity of women's employment orientations, we do not wish to argue that women's attitudes and employment experiences are identical to those of men, nor that women, as women, do not share many circumstances in common. Our cross-national occupational comparisons demonstrated very similar patterns of gender segregation within these occupations, despite the extent of national variability.13 The situation with regard to medicine has been demonstrated beyond empirical dispute (Allen 1994). Some specialties, in particular, surgery, are known as male preserves which are distinctly unfriendly to women, and masculine exclusionary practices are the major reason for the very low female presence in these specialties in all countries. For example, in the Czech Republic, where women are 51 per cent of doctors but only 14 per cent of surgeons, one of our respondents, a surgeon, described the situation as follows . . . keeping these disciplines 'male' is a kind of male tradition ... I almost intuitively understood from the very beginning of my active involvement in the medical world that I must not make any mistake because it would be, for many people around, evidence that women cannot be good surgeons (1/06 divorced, one child). The situation is broadly similar in banking, although there are important East-West contrasts. Whereas retail banking is feminizing in the West, in the East, men are rapidly moving into this previously female-dominated sector as employment opportunities expand. In broad outiine, the patterns of occupational segregation in the financial sector in the East are the same as in the West, in that men are over-represented in managerial positions, and women carry out low-level clerical work - which in the East is universally seen as a woman's job.14 Women in banking in Eastern Europe, even when in managerial positions, are conscious of the realities of masculine exclusionary practices and occupational sex segregation In the Czech Commercial Bank (i.e., pre-1989) all lower positions were occupied mostly by women, but when a man came, however stupid, he had his career carved out for him (1/21 divorced, one child, new relationship) When I applied for a higher post at the bank . . . though I was a strong candidate I was rejected ... it was said that the job involved dealing with 130 Rosemary Crompton and Fiona Harris the security forces and therefore it would be better if it was done by a man. I tried to argue with the head of the selection board but I didn't get very far (Czech Republic: 1/28 divorced (x2), one child). The situation was similar in Russia ... we are being discriminated according to sex, when we are admitted to work and when we are being promoted and in other ways too . . . Men are in securities, the credit department, women are in the accountancy department (4/24 divorced, one child). In Britain and Norway, there have been considerable, and recent changes, in gender structuring within the Finance sector. From being amongst the most discriminatory of employers, retail banks are now to be found at the forefront of Equal Opportunities policies (Crompton 1989). Thus reports of gender discrimination came mainly from older interviewees. However, in France, gender discrimination was often explicit. One woman applied for a Branch manager post (for the second time) . . . the most misogynist in our head office was the boss, who told me that he would never have a woman in a position of power, he actually said that to me. The others said it wasn't a question of sex but the cost of training, they were hypocrites . . . (5/28). Discouraged at being constantly turned down, she eventually decided to take advantage of her statutory right to work 80 per cent time after the birth of her second child in 1993. Although the circumstances of women in different countries are diverse, women in the same occupations share in important experiences relating to masculine exclusionary practices. Thus there are important cross-national continuities in intra-occupational segregation, as well as at the aggregate level. This kind of evidence suggests that universalistic, monocausal explanations such as Hakim's, which rest upon the assumption that the gender division of labour in employment as a whole can be explained as a consequence of 'qualitatively different' types of women exercising specific choices, are not very useful on their own. Many women will 'choose' low level clerical work in banking, for example, but as a wide range of empirical work has demonstrated, this does not fully explain the concentration of women in lower grades in retail banking. Gender segregation operates within particular occupations, as well as between different occupations, and, moreover, there would seem to be considerable cross-national continuity between the same occupations in different countries (Crompton and Le Feuvre 1992; 1996). Hakim's argument specifically allows for the impact of masculine exclusionary practices at the higher levels of the occupational structure (1996: 182ff). She argues that Goldberg's theory of male dominance suggests that women will tend to lose out when they attempt to compete on equal terms with men. Some of our evidence might be used to support such arguments. However, we have also argued that, even amongst this highly selected group Explaining women's employment patterns 131 of 'self-made women', work orientations are complex and variable, and do not correspond to a priori female 'types'. Many women's work orientations are multi-stranded, rather than single-stranded, and we have also shown how 'orientations' to both domestic and market work vary over the domestic life cycle, and can be transformed by major structural and organizational shocks and barriers. However, women (and men) are not structural 'dopes', and the biographical interviews also provide ample evidence of conscious domestic or market work choices, made within particular occupational or national constraints.15 CONCLUSIONS Our aim in this paper has been to illustrate the complex structuring of the gender division of labour in respect of both market and domestic work. Inter alia, we have also sought to demonstrate that one-sidedly voluntaristic explanations of women's (and men's) economic behaviour, in which 'orientations to work' and corresponding choice of economic activity are regarded as the major explanatory variables in respect of women's (and men's) economic behaviour, are inadequate and potentially misleading. We have seen that whilst women do indeed make choices, these choices are not necessarily between the alternatives of home-centredness and career-centredness. Some women want both - that is, their work orientations are not single-stranded - and they choose accordingly. We have also seen how contexts structure choices - a fact which should make us wary of assumptions that there exist identifiable 'types' of women. Some women go into employment and family life without the conscious exercise of choice -but this does not preclude their subsequently becoming highly committed to an employment career. As has been well-established empirically, direct male exclusionary practices have had a substantial impact on women's careers and occupational choices, and it would seem that there is a considerable amount of cross-national continuity in these processes. This kind of evidence demonstrates that occupational segregation by sex cannot be explained as being a consequence of women's choices alone. We have sought to emphasize that sociological explanations relating to women's employment patterns cannot rest upon a simplistic reduction to the argument that they are due to the fact that there are different 'types' of women. Merton (1957: 121) has argued that a sociological approach seeks to '. . . abandon (s) the position held by various individualistic theories that different rates of deviant behaviour in diverse groups and social strata are the accidental result of varying proportions of pathological personalities found in these groups or strata'.16 In a similar fashion, we would argue that the concentration of women in particular occupations and employment statuses cannot be 'read off' from the assumption that these correspond to different 'types' of women. Preferences may shape choices, but the do not, contrary to Hakim's assertions, determine them (1996: 214). 132 Rosemary Crompton and Fiona Harris The 'male breadwinner' model of the division of labour, in which women undertook domestic or non-market work and men 'went out' to work (employment) (Davidoff and Hall 1987) has had important consequences for the structuring of women's employment in many Western countries. This model also provided the initial framework for the constitution of major institutions; including welfare regimes, education systems, social security systems, etc. It has also shaped ideas of masculinity and femininity, and of the 'proper' kinds of work for men and women. Nevertheless, gender divisions of labour have varied considerably from country to country. In Finland, for example, the persistence of small farming meant that women never withdrew from market work to any great extent (Pfau-Effinger 1993). Different kinds of welfare state development have also affected the level and nature of women's employment - for example, in Germany there is a strong commitment to preserving the traditional caring functions of the family, and the tax system discourages female labour force participation (Esping-Andersen 1993: 35. See also Lewis 1992). In state socialist Eastern Europe, highly traditional gender roles persisted despite near-universal employment for women - a fact which Watson (1993) attributes (in part) to men's lack of standing in civil society under state socialism. At the aggregate national level, socio-historical and institutional differences have been reflected in systematic variations in the level of 'traditionalism' in gender role attitudes, which are carried over into the domestic division of labour between men and women (Crompton and Harris 1997a). This kind of evidence suggests that gender relations, and the choices associated with them, are shaped as much by context as by individual preferences, and we should expect these contextual variations to persist even as the male breadwinner model is modified. This heterogeneity of national patterns should also make us wary of assuming that a particular national compromise indicates a universal 'solution' as far as the gender division of labour is concerned, as would seem to be implied by Hakim's arguments relating to fundamentally different 'types' of women. It is likely that some degree of occupational segregation, reflecting cultural and psychological notions of masculinity and femininity as well as the organization of work and family life, would persist even if all gendered constraints on labour force participation were removed. However, to assert that patterns of occupational segregation in Britain represent the outcome of women's choices neglects important factors relevant to the British case. In particular, we would point to the weakness of UK employment protection and the recent promotion of labour market 'flexibility' (Beatson 1995). 'Non-standard', poorly paid jobs, in which women predominate, have increased as a consequence. In a competitive labour market, the weakest workers will get the worst jobs. Hakim argues that the lack of regulation of the British labour market means that it represents a 'natural experiment' in which gender preferences will find their true expression. As we have argued in this paper, this stance assumes a level playing field, in that there are no differences in material and power Explaining women's employment patterns 133 resources between men and women. We would venture to suggest otherwise. In conclusion, therefore, we would suggest that the position assumed by Hakim has significant, and negative, implications for the explanation and understanding of inequality in Britain as a whole. The inferior nature of much part-time, non-standard employment has been identified as a major factor contributing to increasing poverty and social polarization in Britain (Rowntree Foundation 1995). We should not, therefore, be content with explanations of this phenomenon which attributes its female component largely to 'choice', and fails to acknowledge its more negative aspects. (Date accepted: October 1996) Rosemary Crompton and Fiona Harris Department of Sociology University of Leicester NOTES * The authors wish to thank Gunn Birkelund, Kate Purcell, James Fulcher, Irene Bruegel, Nicky Le Feuvre, and the two anonymous BJS referees, for their advice and helpful comments on earlier drafts of this paper. 1. See Hakim 1991: 106, and Ginn et al. 1996: 168. 2. It may also be noted that Hakim's characterization of the two choices open to women fails to take account of the increasing numbers of women living in families but without men - i.e., lone parents. However, Hakim would seem to assume that all women who are not in work have access to a male supporter. For example, in adjusting the data on levels of work commitment in order to incorporate non-working women (who were not actually asked the question), she states that: '. . . by definition non-working women are choosing not to work given even a moderate income supplied by their husband' (1996: 105). However, the data in question gives no evidence of marital or partnership status. 3. Hakim also identifies a third category, the 'drifters', who are neither one thing nor another, but their activities 'probably defy explanation' (1996: 213). 4. The Luton study also included two other, technologically contrasting, workplace organizations: Laporte chemicals and Skefco ball bearings. 5. The 'action' approach may be seen as the precursor of the phenomenological and ethnomethodological 'turn' in British sociology during the 1970s, a development which was vehemently rejected by Goldthorpe. However, this research also demonstrated the manner in which attitudes to employment were and are shaped by perceptions of the opportunities available. See in particular Willis 1977. 6. The original debate may be criticized for this failure to pursue the question of the structuring of the employment/family interface (Dex 1985). Our discussion will incorporate this dimension. 7. Men, also, may have such a dual orientation, it is not intended to suggest that the possibility is gender-specific. 8. The cross-national project, 'Gender Relations and Employment' is funded by the ESRC (R000235617), the British Council, and the University of Bergen. In the cases of Norway, the Czech Republic and Britain we also have access to the core questions of the International Social 134 Rosemary Crompton and Fiona Harris Survey Programme (ISSP) Family and Gender Relations II module. Access to the data has been given by the Institute of Gender Studies, Prague, Social and Community Planning Research, London, and ISD, Norway. 9. Interviews have been carried out by Elena Mezentseva, Irina Aristarkheva, Professor Marie Cermakova, Dr Irena Hradecka, Dr Jaroslava Stastna, Dr Gunn Birkelund, Merete Helle, Rosemary Crompton and Fiona Harris. A common recording document has been used for part-transcriptions of interviews. All of the women interviewed were in employment, and aged between thirty and fifty. All of the doctors were fully qualified (i.e., were beyond the UK House Officer level or equivalent and were entitled to practice independently as doctors), and all of the bankers were in managerial positions. 10. These categories were developed following an exercise in which biographical summaries had been made of all of the interviews. 11. Research on changes in the domestic division of labour following women's entry into the labour force has tended to be rather pessimistic as to the possibilities of change - e.g. Hochshild (1990) - although Gershuny et al. (1994) do suggest that a process of 'lagged adaptation' is in train. However, a feature of these researches is that they have focused upon surviving two-couple households. Our evidence suggests that replacing or removing a partner might also be a common response amongst economically independent women. 12. In any case, empirical research (Pahl 1984) has shown that the domestic division of labour changes considerably over the domestic life cycle, as might be expected. 13. In Britain, women are 54 per cent of employees in 'financial intermediation' (1995), and 29 per cent of doctors (1994). In the Czech Republic, women are 70 per cent of employees in 'financial intermediation' (1994) and 51 per cent of doctors (1994). In Norway, women are 59 per cent of bank employees (1991), and 27 per cent of doctors (1994). In Russia, women are 76 per cent of bank employees (1994) and 74 per cent of doctors. 14. In the West, the situation in the banks is in fact in the process of transformation and gendered restructuring, due to both changes in the industry as well as the successful pursuit of an 'equality agenda' strategy by the EOC (Crompton and Sanderson 1994; Halford and Savage 1995). The situation of women in banking Eastern Europe has many parallels with the 'western model' of the 1960s and 70s. See Crompton 1996. 15. As our study has gathered occupational data only on the highly qualified, we are not in a position to present any new evidence relating to women in lower level jobs (Hakim's 'grateful slaves'). A recent study of homeworking, however, suggests that for significant subgroups of women, their 'choices' are, in fact, massively constrained by their economic circumstances (Phizacklea and Wolkowitz 1995). Their study of homeworking in Coventry showed that the homeworking labour force was differentiated along racial lines which reflected the divisions in non-homeworking employment. All of the clerical jobs in the homeworking sample were held by white women, ethnic minority women were concentrated in manual homeworking, with a heavy representation in clothing assembly, all Asian women. Nearly 60 per cent of the Asian women worked 45 hours or more and one third 60 hours or more (p. 57). These very long hours reflect economic need - a high proportion of households were on income support. Only 10 per cent of Asian women said that they preferred to work at home - they did the work because it was the only work that they could get. In the case of this section of the 'uncommitted' female labour force, therefore, it would be difficult to argue that their employment patterns had been freely 'chosen'. It would also seem to be the case that the extent of homeworking amongst ethnic minority women has been under-estimated in recent surveys, and that it is on the increase. See Felstead andjewson 1996. 16. As in, for example, Lombroso's explanation of crime via the identification of criminal 'types'. BIBLIOGRAPHY Allen, I. 1994 Doctors and their Careers, London: Policy Studies Institute. Explaining women's employment patterns 135 Barrett, M. 1987 'The Concept of "Difference" ', Feminist Review 26: 35-47. Beatson, M. 1995 'Progress towards a flexible labour market', Employment Gazette February: 55-66. Blackburn, R. and Mann, M. 1979 The Working Class in the Labour Market, London: Macmillan. Blauner, R. 1964 Alienation and Freedom, Chicago: University of Chicago Press. —t Bruegel, I. 1996 'Whose Myths are they anyway?' British Journal of Sociology, 47(1): 175-77. Chafetz, J. S. and Hagan, J. 1996 The Gender Division of Labour and Family Change in Industrial Societies: A Theoretical Accounting', Journal of Comparative Family Studies, 27: 187-219. Cotgrove, S., Dunham, J. and Vamplew, C. 1971 The Nylon Spinners, London: George Allen and Unwin. Crompton, R. 1989 'Women in Banking', Work, Employment and Society 3(2): 141-56. -1996 'Gender Relations and Employment: Retail Banking in Eastern and Western Europe', Paper presented at: 'The Globalization of Production and the Regulation of Labour', University of Warwick, September 1996. Crompton, R. and Harris, F. 1996 'Women's Employment, Gender Attitudes and the Household in the UK, Norway and the Czech Republic', University of Leicester Discussion Papers in Sociology, no. S96/2. Crompton, R. and Harris, F. 1997a 'Women's Employment and Gender Attitudes: A Comparative Analysis of Britain, Norway and the Czech Republic' Acta Soci-ologica 40: 183-202. - 1997b 'Attitudes to Women's Employment and the Changing Domestic Division of Labour: a Cross-National Analysis' (University of Bergen, April). -1998 'Gender Relations and Employment: the impact of occupation', Work am-Society (forthcoming). Crompton, R. and Le Feuvre, N. 1992 'Gender and bureaucracy: women in finance in Britain and France' in M. Savage and A. Witz (ed.) Gender and Bureaucracy, Oxford: Blackwell. - 1996 'Paid Employment and the Changing System of Gender Relations: A Cross-national Comparison', Sociology 30(3): 427-45. - 1997 'The Feminisation of the Professions Liberates, Theoretical and Empirical Implications' (University of Bergen, April). Crompton, R. and Sanderson, K. 1990 GenderedJobs and Social Change, London: Unwin Hyman. - 1994 'The gendered restructuring of employment in the finance sector', in A. MacEwen Scott (ed.) Gender Segregation and Social Change, Oxford: Oxford University Press. Daniel, W. 1969 'Industrial Behaviour and Orientations to Work', Journal of Management Studies 6: 152-69. - 1971 'Productivity Bargaining and Orientation to Work: A Rejoinder to Go\dihorpe\ Journal of Management Studies 8: 329-35. Davidoff, L. and Hall, C. 1987 Family Fortunes, London: Hutchinson. Dex, S. 1985 The Sexual Division of Work, Brighton: Wheatsheaf. Esping-Andersen, G. (ed.) 1993 Changing Classes, London: Sage. Felstead and Jewson 1996 Researching a Problematic Concept: Homeworkers in Britain, University of Leicester Discussion Papers in Sociology, no. S96/4. Gallie, D. 1978 In Search of the New Working Class, Cambridge: Cambridge University Press. Gershuny, J. Godwin, M. and Jones, S. 1994 'The Domestic Labour Revolution: a Process of Lagged Adaptation' in M. Anderson, F. Bechhofer and J. Gershuny (eds) The Social and Political Economy of the Household, Oxford: Oxford University Press. Ginn, J., Arber, S., Brannen, J., Dale, A, Dex, S., Elias, P., Moss, R, Pahl, J., Roberts, C. and Rubery, J. 1996 'Feminist Fallacies: a reply to Hakim on women's employment', British Journal of Sociology 7(1): 167-74. Goldthorpe, J. H. 1966 'Attitudes and Behaviour of Car Assembly Workers: A Deviant Case and a Theoretical Critique', British Journal of Sociology 17(3): 227-44. - 1972 'Daniel on Orientations to Work: a final Comment', Journal of 'Management Studies 9: 266-73. Goldthorpe, J. H., Lockwood, D., Bechhofer, F. and Piatt, J. 1968 The Affluent Worker: Industrial Attitudes and Behaviour, Cambridge: Cambridge University Press. 136 Rosemary Crompton and Fiona Harris Hakim, C. 1991 'Grateful slaves and self-made women: fact and fantasy in women's work orientations', European Sociological Review! {2): 101-21. —*- 1992 'Explaining Trends in Occu- pational Segregation: The Measurement, Causes, and Consequences of the Sexual Division of Labour', European Sociological Reviews): 127-52. —* - 1995 'Five feminist myths about women's employment', British Journal of Sociology 46(3): 429-55. - 1996 Key Issues in Women's Work, London: The Athlone Press Ltd. Halford, S. and Savage, M. 1995 Restructuring Organisations, Changing People', Work, Employment and Society 9(1): 97-122. Hartmann, H. 1982 'Capitalism, Patriarchy and Job Segregation by Sex' reprinted in A. Giddens and D. Held (eds) Classes, Power and Conflict, London and Basingstoke: Macmillan. Hill, S. 1976 The Dockers, London: Heinemann. Hochschild, A. 1990 The Second Shift, London: Piatkus. Ingham, G. K. 1974 Strikes and Industrial Conflict, London: Macmillan. Lewis, J. 1992 'Gender and the Development of Welfare Regimes', Journal of European SocialPolicy 2(3): 159-73. Marshall, B. L. 1994 Engendering Modernity, Cambridge: Polity. Maynard, M. 1995 'Beyond the 'Big Three': the development of feminist theory into the 1990s', Women's History Review 4:3; 259-81. Merton, R. K. 1957 Social Theory and Social Structure, New York: The Free Press. Pahl, R. E. 1984 Divisions of Labour, Oxford: Blackwell. Pfau-Effinger, B. 1993 'Modernisation, Culture and Part-time Employment', Work, Employment and Society 7(3): 383-410. Phizacklea, A. and Wolkowitz, C. 1995 Homeworking Women, London: Sage. Roethlisberger, F. J. and Dickson, W. J. 1947 Management and the Worker, Cambridge Mass.: Harvard University Press. Rowntree Foundation 1995 Inquiry into Income and Wealth, York. Walby, S. 1986 Patriarchy at Work, Cambridge: Polity. Watson, P. 1993 'Eastern Europe's Silent Revolution: Gender', Sociology 27(3): 471-87. Wedderburn, D. and Crompton, R. 1976 Workers Attitudes and Technology, Cambridge: Cambridge University Press. Willis, P. 1977 Learning to Labour, Farn-borough: Saxon House.