In lieu of an abstract, here is a brief excerpt of the content:

89 4 Other Correlates of Household Structure and Their Effects on Outcomes The previous chapter showed strong statistical relationships between household structure and a range of employment, educational, and behavioral outcomes of young adults—both for the full sample and for the subgroup of blacks. While family income accounted for a considerable portion (up to 40 percent) of the effects of household structure on outcomes, significant portions remained, both statistically and substantively . Results from fixed-effects models suggested some causal role for household structure on outcomes, as well. But how and why do household structures affect these outcomes? What are the mechanisms that account for the weaker performance of youth who have lived in single-parent households? Are these mechanisms themselves causal, and do they reflect causal effects of household structure? Or are they just spuriously related to household structure and to the outcomes themselves? In this chapter, we further explore three types of household characteristics that are likely to be correlated both with household structure and with the employment, educational, and behavioral outcomes we examine. They are measures of 1) human capital enrichment, 2) parenting and home environment, and 3) neighborhood characteristics. Using information from a subset of the NLSY97, we first show how measures in each of the three categories are associated with household structure. Next, we present regression models similar to those shown in Chapter 3, but now with these three types of household characteristics having been added. We show how the estimated effects of household structure differ once these characteristics are included in the models. We also show the joint influence of each of these three categories of variables on the outcomes. The evidence presented in this chapter indicates that the three sets of household characteristics we examine do account for some of the 90 Hill, Holzer, and Chen statistical associations between household structure and outcomes. Furthermore, these characteristics themselves are associated statistically , and in some cases substantively, with the outcomes we examine. Thus, they help us better understand why the household structures in which young people grow up might affect their later outcomes in life, and they suggest how these effects might be addressed through policy interventions. SAMPLE AND MEASURES The analysis in this chapter uses a subsample of NLSY97 respondents born from 1982 to 1984, who were mostly ages 20 to 22 at the time of the Round 8 interview in 2004–2005. This sample restriction is necessary because some of the additional measures we analyze were collected (by survey design) only for these younger members of the cohort. The NLSY97 collects a rich set of information about sample members ’ home and neighborhood environments and relationships with parents and peers.1 We select a relatively small subset of 11 of these variables for further investigation in this chapter. These reflect the three overarching constructs of 1) human capital enrichment, 2) parenting and home environment, and 3) neighborhood characteristics. We examine the extent to which the 11 variables reduce the estimated associations between household structure and the various outcomes , as well as the extent to which they themselves provide explanatory power for these outcomes. There are good theoretical reasons for believing that these three sets of factors at least partly account for the observed effects of household structure on youth outcomes, as we note below. But, within each construct , we also had to choose from among a wide variety of variables in the NLSY that were conceptually similar and often fairly highly correlated with one another. As described further below, we selected 11 variables in all that had face validity for representing each construct, were not too strongly correlated with each other, and were related to the outcomes we examined (individually and as a group). Other Correlates of Household Structure and Their Effects 91 Our intent was not, as has been successfully done elsewhere (Child Trends 1999), to develop or use a composite index for different constructs , but instead to select a few representative measures in each area that would be reasonable and readily interpretable. We acknowledge the limitations of some of these measures and encourage future research that would refine the measures and further investigate their relationships with household structure and the range of outcomes presented here. Our work should thus be viewed as exploratory, rather than definitive , in some ways. Why should these three sets of measures be related both to household structure and to youth outcomes? Regarding human capital, it appears that access to enriching and material resources early in life may...


Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.