NSF Award Abstract - #0241861

AWSFL008-DS3

Human Capital, Social Capital and Infant Health: Evidence from New Panel Data

Abstract

Infant health is a key determinant of an individual's future success in life. For example, surviving infants of low birth weight are more likely to suffer from learning disabilities and other cognitive problems. Moreover, children of lower socioeconomic status have much poorer birth outcomes on average. To the extent that infant health is compromised by the mother's lack of opportunity, it represents a mechanism for the intergenerational perpetuation of poverty and inequality. This project assembles a new panel of data on infant health using California's Vital Statistics Natality Records for 1982 to 2001. This panel allows them to follow mothers between births in order to answer three outstanding questions about the relationships between human capital, social capital, and the determinants of infant health. First, how does maternal education affect infant health in a rich country such as the United States? Can spillovers from maternal education to infant health be considered as an additional return to education? Do estimated effects of maternal education differ with socioeconomic status, immigrant status, or race/ethnicity? In order to answer these questions this project compares the infant health outcomes of siblings whose mothers increased their education between births to those whose mothers did not. The project also examines the channels through which increased education improves infant health. Second, can "network effects" explain geographical/ethnic correlations in the use of publicly funded maternity care services. Many researchers have noted strong correlations between behaviors of women in networks defined using geographical and ethnic/racial affiliations, but it is difficult to prove that there is a causal link. A panel of data allows the researchers to control for many unobservable determinants of infant health which have confounded previous non-experimental studies. Moreover this project tests the prevalent hypothesis that estimated network effects reflect shared information. If networks act by transmitting information, then the estimated effects should be larger among mothers who have not previously used public maternity care services, since mothers who have already experienced the services will know how to access them and do not need to gain this information from networks. The panel permits comparisons of the estimated effects of networks on women who did and did not previously receive the services in order to conduct this test. The third question is how the implementation of Medicaid managed care in California has affected access to maternity care services. The construction of the panel enables the project to ask, for example, whether Medicaid mothers who are constrained to join Medicaid managed care organizations become less likely to enroll in Medicaid, use certain hospitals, to receive services, and ultimately to have healthy infants. This project assesses the extent to which an important change in access to health care for low income mothers has played a role in ameliorating or acerbating inequalities due to differentials in health at birth.