Racial Difference in Family Lineages in the United States

Rongjun Sun, Cleveland State University
Sarah Matthews, Cleveland State University

This paper uses the National Survey of Families and Households to present the distribution of three-generation lineage structures in the United States population to ascertain the accuracy of assertions made by gerontologists and demographers about the prevalence of age-gapped, age-condensed, and truncated lineages. Current descriptions of lineage structures in the population are of hypothetical lineages that are constructed by combining demographic data about age cohorts. No research has investigated how prevalent these various structures are in an actual population. This paper presents the distribution into lineage structures of a representative sample of the population in which persons aged 50+ (N = 3,550) are designated as the oldest generation of three. Specifically, the paper addresses the question, what is the distribution of three-generation lineages into the 10 types that result from combining, for the two older generations, age of parent at the birth of first child (young, punctual, and old) and childless (truncated)? In addition, whether there is a relationship between lineage type, and age of oldest generation, social class, and race/ethnicity is investigated. Rationale Bengtson, Rosenthal, and Burton (1990) describe a "demographic revolution" that has affected aging families. They argue that declining fertility and mortality have led to an increase in "diversity in family structure" and categorize lineages: age-condensed, age-gapped, and truncated. Age-condensed families are characterized by "teenage pregnancy across multiple generations" (p. 267). In contrast, age-gapped families are those in which childbearing is delayed across multiple generations. Truncated refers to lineages that end because the youngest generation does not reproduce. Most research on demographic aspects of lineages focuses on trends and social change (Bengtson and Silverstein, 1993; Easterlin, 1996; Uhlenberg, 1996a, 1996b). The logically prior question about prevalence of various lineages in the population has been ignored. Rossi and Rossi (1990:89) note that demographic trends are "widely known." Less widespread is the application of such demographic knowledge to the probable structure of family lineages or intergenerational relations. There has been, of course, much speculation about potential effects of demographic change, but much of it is not firmly grounded in actual data." Most research on family structures actually uses information about individuals in cohorts to simulate intergenerational ties (Uhlenberg, 1993) The specific research questions addressed in this paper are: (1) What is the distribution of lineages that results from combining the age of parent at the birth of first child (young, punctual, and old) and childless (truncated) for the two older generations? (2) How does the distribution of lineages vary by race? (3) How does social class affect the distribution of lineages? Can racial differentials be explained by social class? Research Design Generational age gap is defined as the interval between the respondent and his/her oldest child and the age interval between the respondent's oldest child and oldest grandchild. Family lineage types are constructed using information at the individual level. Intervals between generations are identified as condensed (19 and younger at birth of first child), punctual (between 19 and 30 at birth of first child) and gapped (30 and older at birth of first child). Specifying 19 and younger as "condensed" is consistent with Burton (1996); 30 and older as "gapped" is consistent with Chen and Morgan (1991). The lineage categories are defined as 1) punctual-punctual; 2) condensed-condensed; 3) gapped-gapped. Because children do not necessarily follow in their parents' footsteps, to be exhaustive four additional categories are necessary: 4) punctual or gapped to condensed; 5) punctual or condensed to gapped; 6) condensed to punctual; and 7) gapped to punctual. For respondents whose children had not (yet) reproduced, lineages are classified into two categories: 8) early censored (oldest child younger than 20); and 9) punctual censored (oldest child aged 20-29). (No cases of oldest child aged 30+ who had not reproduced were found in the sample). The last category is 10) respondent did not have children. Preliminary Results Punctual-punctual lineages (both generations reproduced in their 20s) comprise one-third of the population. The proportions for the other two lineage types for which children follow their parents' example are very small: 3.4% for condensed-condensed and 1.4% for gapped-gapped. The next most common lineage types are ones for which a third generation is absent, 15.3% because the oldest child who is in his her twenties has not (yet) reproduced and 13.9% because the respondent had no children. Almost one-third of the lineages fall into categories in which the second generation does not follow the first generation's example, slightly less than half (14.4%) because the second generation is "punctual," slightly more than half (17.1%) because someone in the second generation is either under 20 or over 30 when the third generation is started. In almost two-fifths of the lineage types, children follow in their parents' footsteps. In more than three-fifths, children do not. With respect to differences in lineage types among racial/ethnic groups, Blacks and Hispanics are ore likely to be in condensed families while non-Hispanic white are more likely to be in normal or gapped family types. Social status, measured by receipt of a high school diploma, has an impact on family lineages as well. Racial differences persist even after social status is controlled. Further analysis will be carried out to take into account additional socio-economic characteristics such as parent's education and occupation. A multinomial model will be used after the merging of some of the categories to assess the independent and interaction effects of race and socioeconomic status.