Unequal Childhoods - Annette Lareau [207]
MEASURING SOCIAL CLASS IN A SMALL SAMPLE
Social scientists disagree over the proper way to measure inequality in the real world. Some take a gradational approach: on the basis of the key elements of inequality—especially occupational prestige, education, and income—they rank individuals or families in a relatively seamless hierarchy. Yet occupations differ greatly, particularly in the amount of autonomy workers enjoy, the degree to which some people supervise others, the pay, the cleanliness or dirtiness of the work performed, and the amount of prestige that the job commands. I think of these differences in nongradational terms.
There is also disagreement over how to conceptualize classes (i.e., whether to use a Marxian and Weberian approach).2 Regardless of approach, however, most current conceptualizations deploy a relatively large number of class categories in order to attain a fine-grained differentiation of economic positions. It was impossible for me to approximate such an approach in this study. Since my purpose was to develop an intensive, realistic portrait of family life, I was able to analyze only a small number of families. Indeed, with a small sample, and with a desire to compare children across gender and race lines, adopting the fine-grained differentiation of categories characteristic of neo-Marxist and neo-Weberian studies was untenable and unreasonable.
Initially, I settled on two class categories, guided by the populations represented in the town where I was observing. Since employers or self-employed workers were less common than others in the population as a whole, I decided to concentrate exclusively on parents who were employees rather than employers or self-employed. The question then became how to differentiate within this heterogeneous group. Various criteria have been proposed for this purpose, but authority in the workplace and “credential barriers” are the two most commonly used. The former entails the differentiation of those with supervisory or managerial authority over other workers from those with neither. The latter criterion entails separating occupations with stringent educational requirements from those with less demanding ones. On the basis of these considerations, coupled with a pragmatic assessment of what was realistically feasible, I settled on an approach that differentiated a working class and a middle class, each broadly construed. I planned to assign the families to one of these categories on the basis of discussions with each of the employed adults in which they would provide extensive information about the work they did, the nature of the organization that employed them (if there was one), and their place in it. If a family included two full-time workers with divergent class designations, I would assign the family to the higher category (i.e., “middle class”), irrespective of which family member had the defining job.
This plan was adjusted when I discovered that in Lawrenceville schools a substantial number of children were from households supported by public assistance. To ignore them would have been to restrict the scope of the study in a somewhat arbitrary fashion. As a result, I added a group of poor families not involved in the labor market—families that are traditionally excluded in social class groupings. In the end, I worked with three categories: middle-class families, working-class families, and poor families (see Table C1 for criteria for inclusion).
These social class categories conceal important internal variations. Both the Williams family (Black) and the Tallinger family (white) have very high incomes (i.e., annual incomes of more than $175,000). The differences in income among middle-class families, while real, did not