HMOs and Fee-For-Service Health Care Expenditures: Evidence from Medicare

Increasing levels of HMO activity may influence health expenditures in other sectors of the market. Medicare provides FFS coverage to the majority of its beneficiaries and may thus provide a way of examining these so-called spillover effects. This paper examines 1986-1990 Medicare FFS expenditures at the county- and MSA- levels, coupled with county- and MSA-level measures of HMO market share. Fixed-effects and IV estimates of the relationship between market share and expenditures are presented. All of the models imply that FFS expenditures are concave in market share and that expenditures are decreasing in market share for market shares above about 18%. Many of the estimates suggest that expenditures become decreasing in market share at much lower levels (between 0% and 10%). Fixed-effects estimates imply that increases in market share from 20 to 30 percent would be associated with expenditure reductions of 3.4% -6.6% in Part A expenditures and 2.5% - 5.6% in Part B expenditures. IV estimates imply larger responses. The results are consistent with the hypothesis that managed care can affect non-managed-care expenditures.