Uses and abuses of long-term Medicare cost estimates.

Long-term cost estimates are central to the current Medicare policy debate. This paper uses Congressional Budget Office (CBO) and Health Care Financing Administration (HCFA) data, and comparisons to international trends, to make some distinctions and warnings about the use of those estimates. Projected spending increases related to a larger enrollee population may be more easily justified than increases related to higher prices or volume of services per enrollee. Projections of demographics as if policy will not change are at least plausible; projections as if payment policies will not change for the next seventy-five years absurd. The demand for solutions that would "fix" the problem for seventy-five years biases options in the direction of radical reforms, to the exclusion of more incremental and careful approaches.