Showing papers by "Robert Haveman published in 1991"

Childhood events and circumstances influencing high school completion.

Abstract: This paper is an empirical exploration of the effects of a variety of family and economic circumstances experienced during childhood on one indicator of success in young adulthood-high school completion. The estimates suggest that parental education and mother’s work are positive and significant determinants of high school completion, whereas growing up in a family with more children (who compete for resources), being persistently poor and on welfare, and moving one’s residence as a child have significant negative impacts on high school completion. The effects of some family stress and economic events differ depending on the age of the child when they occur. The results support the economic model of investment in children, as well as the welfare culture and socialization models.

Disability Transfers and the Work Decision of Older Men

Abstract: The potential causal relationship between the availability and generosity of disability transfers and the nonparticipation rates of older men is often asserted and much studied [Parsons, 1980; Haveman and Wolfe, 1984; Leonard, 1986]. However, estimates of the magnitude of this linkage vary widely: the growth in transfer availability and generosity has been credited with from 30 to 100 percent of the significant post-1960 fall in older male participation rates. The disparity in these research results stems primarily from (1) alternative sources of data that differentially constrain estimation of relevant expected wage and transfer incomes, (2) specifications that are unable to distinguish income transfer from labor market effects, and (3) reliance on measures of health status that are endogenous to the work-leisure decision. This study seeks to narrow the range of estimates by addressing each of these issues. Our utility maximization model of work status choice assumes that workers who differ in health status choose between the mutually exclusive options of working at the wage offered for their characteristics and accepting public or private disability transfers. Because the income flows in the two options are censored, they are estimated using switching regression techniques as expectations conditional on exogenous individual characteristics, hence accounting for the potentially endogenous process that separates wage earners from transfer recipients. The marginal utilities of income from the two sources are allowed to differ because of stigma effects. In the estimation we adjust the choice-based nature of our sample, employ longitudinal earnings data to measure expected changes in (as distinct from the expected level of) income in the two options, and use a separately estimated latent health status measure largely purged of its endogeneity with the participation decision.