Using data from the ninth wave of the Panel Study of Income Dynamics, changes in the wage effects of marital status and number-of-children for workers of the same race and sex are analyzed as more refined measures of work experience, training, and labor force attachment are substituted for conventional measures of these factors. The results indicate that number-ofchildren is a good proxy variable for differential work history and labor market attachment among white women, and that marital status is not a proxy for such differences among any of the four major race/sex subgroups of workers, including white women. Overall, the findings suggest that, controlling for numerous aspects of worker qualifications, workers with greater financial responsibilities to their families receive higher wages. I. INTRODUCTION Much of the recent literature on male/female wage differentials has directly or indirectly attributed part of the wage gap to division of labor in the home. These arguments have advanced from somewhat vague notions of sexspecific role differentiation (Fuchs [3]) to more rigorous mathematical models showing how sex differences in the allocation of human resources between the labor market and the home could produce wage differences between husbands and wives (Polachek [6]). But regardless of the explanatory framework, the central idea has been that women's labor market The author is Study Director,

No abstract

This paper examines a wide variety of forms, and full histories, of family structure to test existing theories of family in¯uences and identify needs for new theories. The focus is on links between childhood family structure and both completed schooling and risk of a nonmarital birth. Using a 27-year span of panel (PSID) data for U.S. children, we ®nd that: (a) change is stressful, (b) timing during childhood is relevant, (c) adults other than parents are important, and (d) two more recently studied family structures (motherwith-grandparent(s) and mother-with-stepfather) do not ®t the molds of existing theories. The ®ndings suggest that new theories should consider allocation of resources and reasons people group into family structures.

Like others before us using different data, we find significant effects of parental family income on the completed schooling and wage rates of adult children using intergenerational data from the Panel Study of Income Dynamics. We explore various hypotheses regarding these effects, finding substantial support for the economic hypothesis that income, regardless of its source, is invested by parents in their children; mixed support for the hypothesis that fathers serve as role models for their sons; and no support for the welfare dependency hypothesis. Rather than serving as positive role models, working mothers appear to have SigIIifiCRntly less successful sons. 0 1!%7 Academic press, Inc.

Given the prominence of marital dissolution in American life in recent decades, it is important to understand what contributes to or deters it. This article focuses on spouses' shared leisure activities as a possible deterrent. An “attachment hypothesis”— that spouses' shared leisure time is a form of pleasurable interaction that strengthens the attachment between them and helps prevent marital break-up at the time and into the future—is tested in the context of controls for a variety of hypotheses. The empirical tests are supportive of the attachment hypothesis and suggest that, because couples with children have less shared leisure time, children can contribute to marital break-up as well as help prevent it.