Time allocation, market work, and changes in female health.

The conceptual model for analyzing the effect of time allocation on changes in health status is described, and empirical estimates of the relevant parameters of this model are presented, based on panel data and techniques to deal with initial and changing health heterogeniety.

Abstract

The past two decades have seen remarkable time reallocations of working-age women among child care, market work, housework, and leisure activities; indeed, the labor force participation rate of women aged 25-34 increased from 36 to 65 over this period. These reallocations have sparked speculation regarding their effects on female health status. One view suggests that increased female market work will contribute to health status, via its contributions to self-esteem and social contacts and support. A second view emphasizes the mortality differences between men and women, and suggests that female morbidity and mortality patterns will resemble those of men as the two sexes become increasingly similar in their exposure to the hazards and stresses of market work. A third view notes that the increasing market work of women is not associated with equivalent decreased time spent in child care and housework, implying an increase in the total demands on the time of women who work. It is suggested that the resulting decrease (increase) in time allocated to leisure (demanding and stressful work activities) will have deleterious effects on women's health status. For several reasons, it is difficult to sort out the determinants of changes in health status over a period of time. Clearly such changes depend on the age of the individual and especially her health at the beginning of the period over which measurement occurs. In addition, the individual's access to and utilization of health care and other resources during the period are likely determinants. And, while activities engaged in during a period (for example, hours of market work, work in hazardous or demanding circumstances, demands of dual roles) are likely to have an effect on changes in health status over the period, the extent of these activities is likely to be endogenous, depending on initial health status. Data limitations add to the difficulty of obtaining reliable estimates of the determinants of health status changes. Most basically, longitudinal data appear necessary to identify the determinants of changes in health status or the role of activities on health at a point in time. While certain activities and habits (for example, exercise, diet, cigarette and alcohol use) are likely to affect health, information on these variables is often not available in micro data. Detailed time allocations among child care, housework, and leisure activities are typically not available, and even if available, the composition of activities within each class are unknown. Finally, the inadequacy of reliable health status indicators plagues this question at least as much as any in the health economics area. In this paper, we first briefly relate what is known about the determinants of women's health and the impact of time allocations on changes in health. We then describe our conceptual model for analyzing the effect of time allocation on changes in health status. Finally, we present our empirical estimates of the relevant parameters of this model, based on panel data and techniques to deal with initial and changing health heterogeniety.