Date of Award

Fall 1999

Project Type


Program or Major


Degree Name

Doctor of Philosophy

First Advisor

Karen Smith Conway


While in recent years, the infant health production function literature has expanded to incorporate behavioral inputs into the production of infant health current research fails to incorporate the Medicaid-private insurance choice into the mother's decision making process. This dissertation seeks to address this by treating private insurance and Medicaid as endogenous while considering the effects that the Medicaid eligibility rules have on both kinds of coverage. Medicaid and private insurance are entered directly into the health production function so that quality effects may be captured. In addition, a woman's state of mental health is also incorporated into the health production process. Depression may have direct and indirect effects on infant health, as it is a potential negative input into the production process but also a factor that may influence the choice and productivity of other inputs (such as prenatal care).

A traditional infant health production model is extended to include the Medicaid-private insurance choice and guides the formulation of the two-stage empirical model. Two reduced form, univariate probits are estimated in the first stage on the Medicaid-private insurance choice. Then, a treatment effects model simultaneously estimates the interaction between prenatal care and birth weight as well as the effects of depression, Medicaid, and private insurance. Finally, utilizing the parameters estimated with my original (1988) data, the effects of expanding Medicaid eligibility according to the 1997 rules are simulated. Additionally, a similar simulation predicts the effect of mechanically "treating" depressed women or reducing their depressive symptoms (by lowering their CES-D depression score) on the key variables.

Results indicate that both depression and Medicaid participation lead to a reduction in birth weight. Prenatal care has no significant effect on birth weight but both smoking and anthropometry do. Simulations reveal that expanding Medicaid eligibility is predicted to improve health outcomes particularly for low-income women and black women. Treating depressed women is also predicted to increase birth weights.