It is not lack of opportunity that stops women climbing the career ladder, says Catherine Hakim, but their choice of marriages of unequals.

Women have flooded into ÖйúAƬ over the past four decades, not just in Britain, but throughout Europe. The university with a balanced sex ratio has ceased to be a rarity. The sex-segregation of occupations and the pay gap have fallen dramatically as a consequence.

Women's representation in the higher professional occupations rose steadily after 1971, following the introduction of equal opportunities legislation. For example, women's share of jobs in the legal professions (solicitors, barristers, advocates and judges) rose from 4 per cent in 1971 to per cent by 1990. It will, no doubt, have reached well over one-third by the forthcoming 2001 census. Nonetheless, commentators regularly complain about women's failure to break through the "glass ceiling" into the highest grades of professional and managerial occupations.

The gap between men's and women's average hourly earnings in full-time jobs shrank from 36 per cent in 1970 to 20 per cent by 1995. Since then, it has hovered around the 20 per cent level with little sign of any further major change. This has prompted calls for a strengthening of equal opportunities legislation and employer policies to promote women at work.

ÖйúAƬ

However, the real cause for this steady state in the position of women may lie outside the labour market. My study of marriage patterns and work orientations, Work-Lifestyle Choices in the 21st Century , found that a minority of women are aiming for the top jobs.

We tend to assume that women who have sought, and obtained, ÖйúAƬ qualifications are more career-minded than other women. This is not the case. Graduate women are not necessarily career-oriented, even if they welcome interesting jobs and good salaries in the period before they marry and have children.

ÖйúAƬ

As well as providing training for better jobs, colleges and universities give access to superior marriage markets. In Britain, for example, only 6 per cent of women marry a university graduate. Attending university hugely increases the chances that a woman will marry a graduate: in the 1990s, two-thirds of graduate women married a man with equal or better qualifications. For men, the exact opposite is found. Two-thirds of graduate men marry a woman who is significantly less educated than themselves.

In most cases, these marriages of unequals form the basis for a substantial sexual division of labour in the household, with the wife becoming a full-time homemaker in support of her husband, who may work long hours in his career.

Overall, it is wrong to assume that all graduate women will be career oriented. On the contrary, studies in Britain and the United States show that educational qualifications have a small impact on women's sex-role attitudes and interest in a career. The majority of women hold ambivalent attitudes on the relative importance of career versus family life.

Another hidden factor in a woman's employment decisions and choice of occupation is her spouse - more specifically her husband's education level. Women typically seek partners who are taller, older, better-educated, higher status and higher earning than themselves. This bias does not seem to have changed at all as a result of women's increased education; if anything, it has increased. In Britain, the proportion of marriages in which the husband is significantly better educated than his wife doubled between 1950 and 2000, while the proportion of marriages with equally educated spouses halved in the same period. Similar trends are found across Europe, Australia and the US. Only in Germany has the proportion of women marrying up the educational ladder declined over the past century.

As a result, there has been little change in the financial dependence of wives in recent decades. Across Western Europe, husbands typically earn at least twice as much as their wives across the life cycle, and very often three times as much. Even in marriages where both spouses are in employment, husbands typically earn twice as much as wives. In Britain, four-fifths of husbands earn more than their wives.

ÖйúAƬ

Marriage to a high-earning man is still an alternative to the full-time career, and one many women take up readily if the opportunity arises. This remains as true of graduate women as it does of less educated women. These women do sometimes work, full time or part time, but they are taking jobs, not pursuing careers on the same basis as their husbands, the majority of whom continue to view themselves as that family's main breadwinner - even when unemployed. It is these hidden attitudes and marriage patterns that probably explain the steady state in women's position in the workforce.

Only a minority of couples choose a dual-career marriage, Tony Blair and Cherie Booth and Bill and Hillary Clinton, for example. Most graduate couples choose a dual-earner partnership in which he pursues a career and she takes jobs that fit in with her family role. If they live in a small town or suburban area rather than a big city, her choice of jobs can become restricted. This explains the recurrent survey finding of graduate women who are "overqualified" for their jobs, or who are "underemployed" relative to their qualifications. Such results are frequently interpreted as the consequence of employer discrimination. More often, it is due to a lifestyle choice that couples make early on.

Although we live in a century of constant and accelerating change, the social sciences have not been very successful in theorising social change. The theory I have developed, preference theory, is an empirically based, multidisciplinary theory for explaining and predicting current and future patterns of women's choices between family and market work or equivalent activities in the public sphere.

ÖйúAƬ

Alternatively, it provides a theory of changing gender relations. In the context of prosperous modern societies, preference theory identifies five major social and economic changes that create a new scenario of opportunities for women. In particular, it underlines the contraceptive revolution, which, for women, is more important than recent technological change.

It is also important for theories of social change because the original cause of patriarchy was men's need to control women's reproductive activities, not their productive activities.

Social scientists must deliver theories of social change that address the social relations of reproduction as well as the social relations of productive work. Preference theory is a start.

Catherine Hakim is a senior research fellow at the London School of Economics. Her book, Work-Lifestyle Choices in the 21st Century: Preference Theory , was published by Oxford University Press this week.

ÖйúAƬ