Now that the attempted overhaul of the health care system is dead, piecemeal federal reforms and state initiated reforms will take center stage. An area of major concern is the explosive growth of the Medicaid program. Reform efforts are focused on the two major aspects of Medicaid, the acute care component and long-term care for the elderly. This research looks at (1) the effect on medical service utilization when Medicaid eligibles are switched from fee-for-service (FFS) medical care to capitated health plans, and (2) the effect of Medicaid subsidies on the decision of the frail elderly to enter a nursing home, and the role of their adult children in the decision. The goal is to explore six specific issues in order to inform policy makers of how the incentive schemes and delivery systems constructed by the government affect utilization of medical care services and therefore program cost. The first issue is estimating the program effect on usage of services such as doctor and clinic visits. The data set uniquely features randomized enrollment and observations of the same people enrolled in FFS and prepaid Medicaid plans. Count data regression models are used to explicitly account for the nonnegative integer aspect of the usage variables. Preliminary estimates show no significant change in doctor office visits, but the relatively more costly use of clinics, on average, falls by roughly one-half a visit per person in a year. Second, data for people in both FFS and prepaid Medicaid allow for estimation of the distribution of the program effect rather than the simple mean program effect across people with different levels of health care use. A distributional perspective informs policy makers of how managed care affects high versus low users of Medicaid services. Third, results are compared to reveal if significant differences in the program effect emerge when the same people are viewed over time versus viewing different people at one point in time. The primary issue concerns if there is enough latent heterogeneity among people to weaken cross-sectional results. The fourth issue involves estimating the effect of Medicaid subsidies on the decision of the elderly to enter a nursing home versus remaining independent or living with others. The data come from the Panel Study of Income Dynamics 1991 Parent Health Supplement. Preliminary results do not yet include policy variables, but they conform to expectations in that women and people with more disabilities are more likely to enter a nursing home. Fifth, the availability of data for parents and their adult children makes possible estimation of the relationship between living arrangements and the labor supply of the adult children of the elderly. The estimates will determine the extent to which children respond when their parents are sick, and the welfare costs of such response. The sixth issue concerns estimating the extent to which informal care serves as a substitute for formal long-term care. The benefit of the estimates is in gauging the need for inexpensive forms of care such as a periodic visit from a nurse instead of expensive institutionalization.
