Some pitfalls in making cost estimates of state health insurance coverage expansions.

This paper cautions state policy analysts who estimate expected costs of state health reforms using state-specific data from general purpose, national surveys. It compares cost estimates of subsidized insurance programs for low-income uninsured persons in 10 states using the Current Population Survey (CPS) and the Robert Wood Johnson Foundation state surveys. Depending on the measure of insurance and concept of the family that are used, costs of state programs could exceed the estimates by more than 50%. State policy analysts need to be aware of such pitfalls and make appropriate adjustments when using CPS data to estimate program costs.