This paper examines whether public insurance (Medicaid) crowds out private insurance coverage in the United States. The authors use data from the 1987-1992 period, when Medicaid eligibility for children and pregnant women increased significantly. They estimate that between 50 and 75 percent of the increase in Medicaid coverage was associated with a reduction in private insurance coverage. This occurred largely because employees took up employer-based insurance less frequently, although employers may have encouraged them to do so by contributing less for insurance. There is some evidence that workers dropped coverage for their family and switched into individual policies. The authors find that public insurance expansions may reduce the "bang for the buck" from public insurance coverage, as the costs of the program rise without commensurate increase in insurance coverage. They also find that the effect of public insurance on private insurance coverage depends on the composition of coverage within the family. For example, if families value coverage of all members and some members cannot qualify for public coverage, crowdout is likely to be smaller than an individual-by-individual calculation would suggest. On the other hand, if the Medicaid subsidy is large, families may drop coverage of all members, even those who do not qualify for public insurance directly. The authors also find that the link between health insurance and employment may increase or decrease the amount of crowdout. If worker-specific shifting is not possible, then crowdout may be reduced, as employers do not want to impose extra costs on those workers who are not eligible for Medicaid. If employers increase cost sharing or reduce coverage for all workers, however, the decline in private coverage may be greater than the increase in Medicaid eligibility. The authors use data from the March Current Population Surveys (CPS) for 1987 through 1992 to estimate the effect of public insurance on private insurance coverage. They find that the expansion of Medicaid eligibility for children and pregnant women led to a significant reduction in private insurance coverage. For children, the estimated takeup rate of Medicaid was 24 percent, which is substantially below the average takeup rates for other public programs. For women, there was no statistically significant increase in Medicaid coverage as eligibility expands. The authors also find that the expansion of Medicaid eligibility leads to a significant offset in private insurance coverage for both children and women of child-bearing age. For children, each 10 percentage point increase in Medicaid eligibility results in a .74 percentage point decline in private insurance coverage. For women, each 10 percentage point increase in Medicaid eligibility results in a .45 percentage point decline in private insurance coverage and a .46 percentage point increase in uninsurance.This paper examines whether public insurance (Medicaid) crowds out private insurance coverage in the United States. The authors use data from the 1987-1992 period, when Medicaid eligibility for children and pregnant women increased significantly. They estimate that between 50 and 75 percent of the increase in Medicaid coverage was associated with a reduction in private insurance coverage. This occurred largely because employees took up employer-based insurance less frequently, although employers may have encouraged them to do so by contributing less for insurance. There is some evidence that workers dropped coverage for their family and switched into individual policies. The authors find that public insurance expansions may reduce the "bang for the buck" from public insurance coverage, as the costs of the program rise without commensurate increase in insurance coverage. They also find that the effect of public insurance on private insurance coverage depends on the composition of coverage within the family. For example, if families value coverage of all members and some members cannot qualify for public coverage, crowdout is likely to be smaller than an individual-by-individual calculation would suggest. On the other hand, if the Medicaid subsidy is large, families may drop coverage of all members, even those who do not qualify for public insurance directly. The authors also find that the link between health insurance and employment may increase or decrease the amount of crowdout. If worker-specific shifting is not possible, then crowdout may be reduced, as employers do not want to impose extra costs on those workers who are not eligible for Medicaid. If employers increase cost sharing or reduce coverage for all workers, however, the decline in private coverage may be greater than the increase in Medicaid eligibility. The authors use data from the March Current Population Surveys (CPS) for 1987 through 1992 to estimate the effect of public insurance on private insurance coverage. They find that the expansion of Medicaid eligibility for children and pregnant women led to a significant reduction in private insurance coverage. For children, the estimated takeup rate of Medicaid was 24 percent, which is substantially below the average takeup rates for other public programs. For women, there was no statistically significant increase in Medicaid coverage as eligibility expands. The authors also find that the expansion of Medicaid eligibility leads to a significant offset in private insurance coverage for both children and women of child-bearing age. For children, each 10 percentage point increase in Medicaid eligibility results in a .74 percentage point decline in private insurance coverage. For women, each 10 percentage point increase in Medicaid eligibility results in a .45 percentage point decline in private insurance coverage and a .46 percentage point increase in uninsurance.