Women and Class Analysis – a debate


Goldthorpe, J (1983) ‘Women and Class Analysis’, Sociology 17 (4).

This article was partly based on a follow-up of a subsample of wives of the 1972 study, N = 578.

To argue that the male is best seen as the head of household is justifiable.  It is not necessarily a functionalist argument supporting the subordination of women in functionalist terms.  It stems rather from class analysis.  Women’s work is class conditioned, women’s job opportunities are less than males as a result of family demands, which themselves depend on social class.  Focusing on males is therefore not necessarily sexist.

There is an increasing level of married female employment, and an increasing incidence of female heads of families.  Nevertheless, none of this makes a difference in terms of the basics of class analysis, the classes vary according to economic positions.  There would be a problem for class analysis only if female labour was very different from that of males, if they occupied a different class, or if they were separate enough to warrant joint classification, as another head of family in class terms.

The data show that married women’s work is intermittent, limited and conditional, depending on social class.  Census data only gives general averages and cross sections here [which have misled the critics].  Class conditioning shows itself especially in that different classes have different childbearing characteristics, different economic requirements, different possibilities of choice and opportunity, and different chances of mobility.  Upwardly mobile wives stay in work longer, especially if they have been indirectly mobile [that is mobile through working their way up in occupation]. 

There are few ‘cross–class families’.  It is reasonable to assume that a female non- manual job carries the same status as a male non-manual job [and nothing more, such as more of a route to a top job].  Where there are differences, the most common pattern is where the wife has an inferior occupational status—in cases where wives occupy a superior status, this tends to be unstable [a short lived occurrence, soon to be restored to a more equal position].

Such instability means there is no particular need for new joint classifications for class analysis, even where women have a superior occupational status [these sorts of cross class marriages were seen as particularly important for social mobility by Goldthorpe’s critics, including Stanworth, since marriage itself might be an opportunity for upward social mobility for men, or at least for the children in the family].  Analysing cross class families like this might be useful for other purposes, for example in showing different fertility patterns, but purposes like that might as well have special socio cultural classifications to permit study.

We need to consider rates of instability.  If we just take snapshots, this would lead to a misleading picture of unusually high rates of social mobility—men would change their social class each time their women moved backwards and forwards in and out of different labour markets.  Classes would then be subject to higher rates of decomposition [precisely what some recent critics have suggested, of course, but ruled out by Goldthorpe here as inconsistent with his other findings?]


Stanworth, M.  (1984) ‘Women and Class Analysis: a reply to Goldthorpe’, Sociology 18 (2)

The concept of a head of family really turns on the idea of both length and commitment to the labour market, so it is understandable that researchers normally take it to be the male. This will be sexist only if (A) females were said to be necessarily unequal; (B) empirical evidence was denied.  Goldthorpe’s main effort is clearly directed at the second one, and he seems to be making three claims:

(1)    That male employment is more continuous than female employment, so that it is rare for females to be employed for ¾ of their married lives.  The problem with this argument is that this will vary according to the cohorts being studied, and the educational qualifications of the women.  Even if Goldthorpe is right, it would be misleading to disregard the position of women altogether: to do so would assume that the family is some sort of homogenous unit.  In practice, only 19% of men are now employed as the sole source of income.  Women’s work obviously now has an important effect on matters such as home owning, amount of savings, and the general stability of the working class.  Females will have their own sense of the standard of living, and possibly their own version of class politics.

(2)    That women’s employment is somehow conditioned by the work of the male breadwinner.  Again this assumes that the social class of the husband can somehow outweigh their own class experiences.  As an example, working class women seem to withdraw from work earlier than men because they are tired of their own work, not that of their husbands.

(3)    That marriages are homogenous in terms of social class, and that there are few cross class marriages.  But this will depend on how occupations are classified: the issue turns on whether women’s work can be classified in male terms as middle class or working-class [this is a genuine problem, as in minor clerical or shop work: in some ways they are white collar work which is characteristically a mixture of both working class wages and middle class work conditions].  Most women’s work is probably proletarian in practice. New classifications would produce lots more cross-class families.

The important category for Stanworth would be a cross class family which mixed working class men with middle class women.  This would raise all sorts of new possibilities for class consciousness.  It would be no more complex to consider these families.  Goldthorpe does admits that it might be possible to take working women as the head of the family, for example with divorced women who are the main breadwinners—currently 1/3 population.  Such women can achieve mobility in Goldthorpe’s schema only by remarriage [?] [If a man marries a woman from another class, and if this does make a difference, both to living standards and ambitions, say, then men as well can achieve mobility through marriage].

Thus Goldthorpe does not fully recognise sex inequalities as significant. Sex inequalities are related to the social class of women, say when they are forced to take proletarian jobs [because their husbands have moved somewhere where middle class work is not available?].  Class struggle becomes a complex matter located both in the labour market and between the sexes in the labour market.  Gender and class are interlinked. 

Goldthorpe, J.  (1984) ‘Women and Class Analysis –a reply to the replies’, Sociology 18 (4)

The employment of married women might indeed make a difference to family life, and matters like ambition – that has not been denied.  However, Goldthorpe is mostly interested in class analysis how classes form and are stabilised, and how they are differentiated.  Whether class analysis can then go on to explain all subsequent behaviour is a matter of empirical test.  The issue is --  are the differences mentioned by Stanworth sufficient to abandon the conventional view of social class.  There will be problems if we do, including a misleading view of the amount of class mobility.  There also problems with Stanworth’s argument.

Stanworth says that families are typically not unitary in terms of social class, but this implies an odd view of class as something determined by individual occupations rather than individual class positions.  Stanworth’s classification of female jobs as largely proletarian is too crude: there are relations of differentiation and association between women’s jobs and their husbands’ class.

If women are mostly treated as proletarians, what difference has it made?  Conventional class analysis still predicts differences in lifestyle and political orientations.  Women’s voting preference, for example, classically follows the social class of their husband—thus 45% of wives in the service class voted conservative regardless of any ‘proletarian’ work. 

Individual asymmetry, such as in different mortality rates, influences the work careers of both men and women.  Stanworth appears not to have described this asymmetry.

Now see Payne and Abbott on the position in 1990