Despite my disclaimers in the main text I do need to draw some clear limits around the groups of families represented and be clear about who they are. I need to specify what I mean by middle class in this study, if it is to speak to debates and developments in class theory and contribute to the burgeoning scholarship on the middle class and their social reproduction.
Very simply the families included in the study are members of what is sometimes called the 'saliariat' or more commonly the 'service class'. As Goldthorpe (1995:314) makes very clear: 'the service class is a class of employees'. The main problem of demarcation is that of distinguishing them from other sorts of employees. There are two elements to this; first, benefits of employment over and above salary; that is, pension rights, increments, employment security and career opportunities - although as I have suggested these things are changing and are part of the threat to social reproduction (see Savage 2000, Chapter 8); second, some degree of professional autonomy and managerial or administrative authority - again professional autonomy is under threat for many public and private professionals, and there is a convergence of work practice between professional and managerial jobs. Nonetheless, these criteria do still provide a fairly robust basis for distinguishing service class employees from other middle-class groups, specifically the intermediate middle class; that is, those employed in routine, low-autonomy, white-collar jobs. I can add two further non-work criteria to this definition of the service class. One is education; the parents included in this study, with the exception of one couple and three other women, all experienced some form of higher education (the exceptions being a couple both of whom had attended private school and who were both employed in professional jobs, but who had not done formal higher education, but had done professional qualification courses, and three women, all with A levels). The other criterion is that they are house-owners.
On this basis, drawing on the four studies outlined below, I set about selecting the families with which I would engage and whose interviews I would analyse or re-analyse, by a process of exclusion. That is, I excluded all those families where one or both partners were working in routine, non-manual, white-collar jobs. Again one exception was made, a couple both of