3. The PSID is a nationally representative longitudinal survey begun in 1968 with a focus on issues relating to household finances and employment. Annual or bi-annual waves of the survey have followed both the original sample families and the “breakaway” families formed by children of initial sample members. A weighting system has been devised to account for the effects of both the initial probability of being sampled and attrition (which is generally low) over time. Each wave of the PSID collects detailed information on the employment status, income, and finances of household members, as well as on related matters such as housing.
The CDS was first administered in 1997 to a subsample of PSID families with children between the ages of 0 and 12 years old. Data were collected on 3,563 children in 2,380 families. (Thus, two-thirds of these children are siblings of another child in the subsample, but no one family has more than two children in the study.) A set of child-level weights was created for use with these data by modifying the PSID weights to account for each child’s within-family probability of being sampled. These weights are used in our study. In order to maintain as much comparability as possible to the ethnography while simultaneously maximizing the number of cases available for analysis, we have restricted the subsample to children who were between the ages of 6 and 12 years old at the time of the data collection.
The PSID and CDS are collected and disseminated by the Survey Research Center, part of the Institute for Social Research at the University of Michigan.
4. Sandra L. Hofferth, “Response Bias in a Popular Indicator of Reading to Children,” for example, finds that in survey responses, highly educated parents (but not those with less education) appear to exaggerate the amount of time that their children spend reading, compared to time diary data on children’s reading patterns.
5. For the full results of these analyses, see the page for Unequal Childhoods at www.ucpress.edu.
6. These analyses include controls for the presence of relatives who reside in the focal child’s household and in his or her neighborhood.
AFTERWORD
1. Although I was able to contact most of the young adults, I did not connect with them all. For Harold, the information is based on my last contact in 2005. I did not attempt to reach Alexander; I confirmed his status via a website. As for the three young people in the intensive study but not featured in the book, Jessica Irwin graduated from college, married a policeman, and is now going to college to become an art therapist. Tara Carroll still hopes to return to community college, but she is working full-time as a caregiver in a home for disabled adults. I was unable to reach Karl Greeley; last I heard, he was working in a grocery store. He hoped to get his GED one day.
APPENDIX A
1. In retrospect, the decision to forgo interviewing the children was a serious mistake. I did, however, carry out “exit interviews” with children in the observation study.
2. See the work of Erik Olin Wright, especially his essay in the edited collection by John Hall, Reworking Class, as well as the work of Robert Erikson and John Goldthorpe, including The Constant Flux. Without artificially minimizing the divergences between Goldthorpe and Wright (the foremost of which is undoubtedly the latter’s insistence on retaining a capitalist class within his schema), it can be said that both use similar criteria (skills or credentials and authority) in drawing distinctions between categories of employees. For assistance with this discussion, I am grateful to Elliot Weininger.
3. See Frances Goldscheider and Linda Waite, New Families, No Families?
4. See Elliot Weininger and Annette Lareau, “Children’s Participation in Organized Activities and the Gender Dynamics of the ‘Time Bind.’”
5. Douglas Massey and Nancy Denton, American Apartheid.
6. Although I still see this choice as reasonable, it has drawbacks. It is hard to know whether poor or working-class families living in middle-class neighborhoods would adopt the same cultural logic of child rearing had they been living in poor or working-class neighborhoods. This problem is compounded by the small samples necessitated by intensive field research.
7. In addition, I made a donation of $100 to the school. In Lower Richmond the principal directed it to Ms. Green; at Swan it went to the Parent Association.
8. I began with Lower Richmond, but the method I used for selecting and recruiting interviewees was the same at both school sites. The content of the letters sent to parents at the two sites differed somewhat; for Lower Richmond parents, for example, I enclosed a photograph of their child taken (by me) during third grade.
9. Ten white poor families were recruited from welfare offices and other community programs. I paid these participants $25 per interview; none of the other families was paid.
10. We resisted the temptation to ask only those families who we thought, on the basis of the rapport established during the interview phase, would agree to be observed. We stuck by our first priority of recruiting the most representative families.
11. One white working-class and two Black poor families on welfare declined. The mother in the white family explained, “We’re not the perfect family.” One of the Black families objected to my request to have access to welfare records and declined for that reason. (I dropped the request after that.) The other Black family agreed but then dropped out after only a couple of days. The mother’s schedule changed frequently and there were indications of a possible problem with drugs within the family.
12. In some respects the Greeleys were not a typical poor family. They owned a car, for example. Still, they met enough of my basic criteria (e.g., they received various forms of public assistance, including medical coverage) to be included. Although the family lived in the same Lower Richmond neighborhood, they lived across a school boundary line and the son attended a different school.
13. This decision turned out to be problematic, however. A grade level can make a significant difference; Stacey seemed much more preteen than the other children we observed.
14. The payment was a lump sum (in cash), usually at the very end, when the intensive three weeks of field visits had been completed. In addition to offsetting some of the inconvenience to the families, the money was intended to compensate for expenses such as feeding the field-workers dinner. The amount offered meant something different to each family, depending on their income level.
15. A drawback to my having made friends with the children in their classroom was that some working-class and poor parents then assumed that I was affiliated with the school district. I worried this would increase their sense of distrust. In addition to assuring parents that all information was confidential, I made repeated efforts to clarify the fact that I worked at a college and was not in any way associated with the school district.
16. See Annette Lareau, “My Wife Can Tell Me Who I Know” for a discussion of problems in interviewing fathers.
17. My parents’ marriage was loving but cantankerous; there was a lot of yelling. There were other quirks, as well. My father worked for many years as a tutor, but there were also years when he did not work. This left my mother, who was a teacher, as the sole source of support for the family of four children. There were other ways that my family was seen as unusuaclass="underline" both my parents were atheists, my mother swore like a sailor, and my father had an unending series of broken-down cars he was always trying to fix.
18. I recorded similar feelings in my journal, noting that “. . . A two-site day is too much; it wears you out and doubles your field notes and makes your head spin with the contrast of [lack of] safety and opulence . . . [but to sustain the comparative character of the study] you need to have a two site (or really three site) day all of the time . . .”