The Health Effects of Medicaid Coverage for Children are Unclear

"The expansion of insurance coverage by Medicaid and private health insurance was associated with better health ratings by mothers of low-income black and Hispanic children, but it had no effect on the reported health status of poor white children."

Between 1986 and 1994, the number of children from low-income families enrolled in the Medicaid program, the government program that finances medical care for the nation's poor and near-poor families, rose by 71 percent. Medicaid spending on the health care of these children climbed 107 percent in real terms during that period. This was the largest expansion of public health insurance coverage for children since the original introduction of the Medicaid program. In addition, the Balanced Budget Act of 1997 included a $24.3 billion block grant to states to extend health insurance coverage to uninsured children from families with incomes up to 200 percent of the federal government's official poverty level. These spending increases were based on the assumption that increased utilization of medical services would improve the health of children from low-income families. Yet an examination of the data by NBER Research Associates Theodore Joyce, Robert Kaestner, and Andrew Racine finds only weak support for this belief.

While much previous research evaluated the merits of health insurance solely on the basis of whether it expanded the utilization of medical care by children, in Does Publicly Provided Health Insurance Improve the Health of Low-Income Children in the United States? (NBER Working Paper No. 6887), the authors concentrate on the health consequences of increased insurance coverage. They study a large, nationally representative sample of children from the National Health Interview Surveys of 1989 and 1992. They limit their analysis to children between the ages of two and nine who come from families with incomes below $25,000, as these are the families targeted by the legislation. These surveys include information about health insurance coverage, the health of the children as reported by their mothers, and the social and demographic characteristics of the families. Mothers rated the health of their children (excellent to poor) and provided the number of days a child was in bed sick (morbidity) in the previous 12 months. Both of these measures are correlated with acute and chronic health conditions. The authors also analyze hospital discharge data for a sample of 544 hospitals in 11 states to investigate whether expansion of Medicaid eligibility was associated with a decrease in the incidence of ambulatory care sensitive admissions for children from low-income areas.

Kaestner, Joyce, and Racine find that the expansion of insurance coverage by Medicaid and private health insurance was associated with better health ratings by mothers of low-income black and Hispanic children, but it had no effect on the reported health status of poor white children. In fact, with Medicaid coverage, the health status of the white children was slightly worse. Nor did insurance coverage reduce days sick in bed, which were actually greater for black and Hispanic children covered by Medicaid than for uninsured children.

"Perhaps," the authors speculate, "Medicaid provides greater access to care for poor black and Hispanic children than it does for white children, but this hypothesis is only conjecture at this point." As for the increased bed days, one hypothesis is that they are a result of greater access to primary care and of physicians recommending longer periods of rest or time in bed for treatment.

The authors do find that the Medicaid expansion decreased the incidence of ambulatory care sensitive discharges - mostly for children admitted to the hospital with convulsions, asthma, dehydration, and pneumonia -- by 10 to 20 percent among young children in low-income families. But there are some "anomalous results" in this data that raise questions about the robustness of their findings, they caution.

Reviewing possible explanations for their inconsistent findings, the authors note that uninsured children may already be receiving adequate health care, with parents paying for essential treatment out-of-pocket or getting it free at hospitals and clinics. But they do not find this explanation compelling, since health insurance does increase utilization of medical facilities. Another explanation, they note, could be that their measures of child health do not adequately reflect the benefits of having insurance. Certain unmeasured types of child illnesses may benefit from the extra health care.

They conclude that the proposition that the provision of health insurance through Medicaid to remedy the relatively poor health condition of children from poor and near-poor families "has not been adequately demonstrated."

-- David R. Francis

The Digest is not copyrighted and may be reproduced freely with appropriate attribution of source.
NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us